HCQ for COVID-19: real-time analysis of all 237 studies

Hussein et al., Journal of Molecular Structure, doi:10.1016/j.molstruc.2021.129979 (Peer Reviewed)

Molecular Docking Identification for the efficacy of Some Zinc Complexes with Chloroquine and Hydroxychloroquine against Main Protease of COVID-19

Molecular dynamics analysis recommending Zn (CQ) Cl2(H2O) and Zn (HCQ) Cl2(H2O) as potential inhibitors for COVID-19 Mpro. Zn (HCQ) Cl2(H2O) exhibited a strong binding to the main protease receptor, forming eight hydrogen bonds.

Zelenko, Z., Preprint (Preprint)

Nebulized Hydroxychloroquine for COVID-19 Treatment: 80x Improvement in Breathing

Report on the use of nebulized HCQ showing much more rapid improvement compared to tablets, with 95% of patients experiencing improved breathing within 1 hour. Author notes that the effectiveness of HCQ is time and dose dependent, with a primary issue being the time it takes for a therapeutic dosage to reach the lungs, and that treatment with tablets may take an average of 80 hours to achieve significant clinical improvement.

Author notes that it may take 3-7 days to achieve optimal alveolar concentrations with tablets, whereas nebulized HCQ administered as microdroplets directly to the lungs achieves optimal alveolar concentration in approximately one hour and is associated with faster clinical improvement, reduction in pulmonary complications, and a reduction in medical costs.

Covid Analysis (Preprint) (meta analysis - not included in study count)

HCQ is effective for COVID-19 when used early: real-time meta analysis of 195 studies

• HCQ is effective for COVID-19. The probability that an ineffective treatment generated results as positive as the 195 studies to date is estimated to be 1 in 1 quadrillion (p = 0.0000000000000009).

• Early treatment is most successful, with 100% of studies reporting a positive effect and an estimated reduction of 67% in the effect measured (death, hospitalization, etc.) using a random effects meta-analysis, RR 0.33 [0.25-0.43].

• 91% of Randomized Controlled Trials (RCTs) for early, PrEP, or PEP treatment report positive effects, the probability of this happening for an ineffective treatment is 0.0059.

• There is evidence of bias towards publishing negative results. 88% of prospective studies report positive effects, and only 75% of retrospective studies do.

• Studies from North America are 4.0 times more likely to report negative results than studies from the rest of the world combined, p = 0.00000005.

Total	195 studies	3,074 authors	166,999 patients
Positive effects	152 studies	2,394 authors	123,307 patients

Early treatment	67% improvement	RR 0.33 [0.25-0.43]
Late treatment	25% improvement	RR 0.75 [0.69-0.81]

Li et al., Science China Life Sciences, doi:10.1007/s11427-020-1871-4 (Peer Reviewed)

Evaluation of the efficacy and safety of hydroxychloroquine in comparison with chloroquine in moderate and severe patients with COVID-19

Small RCT comparing HCQ and CQ in China with 88 very late stage (17.6 days from onset to hospitalization and ~10 days to randomization) patients. The primary clinical outcomes (TTCR and TTCI) were not significantly different. Authors note that HCQ may have more promising efficacy in immune system modulation, indicated by ferritin reduction in the moderate cases and improvement of CT scores and lymphocyte counts in the severe cases. HCQ and CQ were well tolerated.

Authors also compare RCT patients to a matched sample of non-RCT patients in the same hospital, showing shorter time to discharge with CQ/HCQ, but not statistically significant due to the small size.

no hospital discharge, ↓50.0%, p=0.08, RCT patients vs. matched sample of non-treated patients

Li et al., Research Square, doi:10.21203/rs.3.rs-119202/v1 (Preprint)

Treatment of COVID-19 patients with hydroxychloroquine or chloroquine: A retrospective analysis

Small retrospective database analysis of 37 late stage patients hospitalized in an intensive care center in China, not finding a significant difference in viral shedding. Pateints were all in serious condition. There was only one death however the group is not specified. Confounding by indication is likely.

time to viral-, ↑40.0%, p=0.06

Rangel et al., Journal of the American Academy of Dermatology, doi:10.1016/j.jaad.2020.10.098 (Peer Reviewed)

Chronic Hydroxychloroquine Therapy and COVID-19 Outcomes: A Retrospective Case-Control Analysis

Retrospective 50 COVID-19 patients that take chronic HCQ, compared to a matched sample of patients not taking chronic HCQ, showing lower mortality and ICU admission, and shorter hospitalization for HCQ patients, but not statistically significant due to the small number of events.

The actual benefit for HCQ could be much larger. The study does not address the risk of being sick enough to visit the hospital. HCQ users are likely systemic autoimmune disease patients and authors do not adjust for the very different baseline risk for these patients. Other research shows that the risk of COVID-19 for systemic autoimmune disease patients is much higher overall, Ferri et al. show OR 4.42, p<0.001 [1].

[1] c19study.com/ferri.html

death, ↓25.1%, p=0.77, from all patients
hospitalization, ↓22.2%, p=0.29
hospitalization time, ↓41.2%, p=0.12

Yegerov et al., medRxiv, doi:10.1101/2021.01.06.20249091 (Peer Reviewed)

Epidemiological and Clinical Characteristics, and Virologic Features of COVID-19 Patients in Kazakhstan: a Nation-Wide, Retrospective, Cohort Study

Retrospective 1,072 hospitalized patients in Kazakhstan showing no mortality for HCQ treated patients, however only 23 patients received treatment - this result is not statistically significant.

death, ↓95.3%, p=1.00
(adjusted for zero (with reciprocal of the contrasting arm))

Baildya et al., Journal of Molecular Structure, doi:10.1016/j.molstruc.2021.129891 (Peer Reviewed) (not included in the study count)

Inhibitory capacity of Chloroquine against SARS-COV-2 by effective binding with Angiotensin converting enzyme-2 receptor: An insight from molecular docking and MD-simulation studies

Molecular docking study of 16 drugs showing CQ had the highest binding affinity with ACE2, and molecular dynamics study of the docked CQ-ACE2 structure. Authors conclude that CQ binds reasonably strongly with ACE2 and the stable ACE2-CQ may prevent further binding of ACE2 with the SARS-CoV-2 spike protein.

Noureddine et al., Journal of King Saud University - Science, doi:10.1016/j.jksus.2020.101334 (Peer Reviewed) (Theory)

Quantum chemical studies on molecular structure, AIM, ELF, RDG and antiviral activities of hybrid hydroxychloroquine in the treatment of COVID-19: molecular docking and DFT calculations

In silico analysis of hydroxychloroquine and hydroxychloroquine sulfate predicting that hydroxychloroquine sulfate is more stable and effective for COVID-19.

Gautret et al., International Journal of Antimicrobial Agents, doi:10.1016/j.ijantimicag.2020.106236 (Peer Reviewed) (not included in the study count)

Safety profile of hydroxychloroquine and azithromycin combined treatment in COVID-19 patients

Report on the safety of HCQ+AZ with 3,737 COVID-19 patients. 138 had contraindications and treatment was discontinued in 12 cases due to QTc prolongation. There were no cases of torsade de pointe or sudden death.

Sarfaraz et al., medRxiv, doi:10.1101/2020.12.28.20248920 (Preprint)

Determinants of in-hospital mortality in COVID-19; a prospective cohort study from Pakistan

Retrospective 186 hospitalized patients in Pakistan showing unadjusted HCQ mortality RR 1.45, p = 0.07. Confounding by indication is likely.

death, ↑45.0%, p=0.07

Sands et al., International Journal of Infectious Diseases, doi:/10.1016/j.ijid.2020.12.060 (Peer Reviewed)

No clinical benefit in mortality associated with hydroxychloroquine treatment in patients with COVID-19

Retrospective database analysis of 1,669 patients in the US showing OR 1.81, p = 0.01. Confounding by indication is likely.

COVID-19 was determined via PCR+ results, therefore authors include patients asymptomatic for COVID-19, but in the hospital for other reasons. While authors adjust for severity, the method used is very poor. 93.5% of patients are classified as "mild", which is patients with no documented care in a critical care unit within 8 hours of admission. Therefore almost all patients are in the same category, and those in a different category may be due to symptoms unrelated to COVID-19. Lower bias toward male patients in the control group also agrees with the hypothesis that the control group is made up of more people that were in hospital for another reason.

Since the analysis covers the initial period of the pandemic in the USA, it is likely that HCQ was used more often earlier in the analysis period when treatment protocols were considerably worse. It's unclear why the analysis only considers patients up to April 27, when the manuscript was submitted in October.

death, ↑69.9%, p=0.01
(odds ratio converted to relative risk)

Matada et al., Bioorganic & Medicinal Chemistry, doi:10.1016/j.bmc.2020.115973 (Review) (Peer Reviewed)

A comprehensive review on the biological interest of quinoline and its derivatives

Review of quinolone and derivatives, natural and drug sources, and biological activity.

Psevdos et al., Open Forum Infectious Diseases, doi:10.1093/ofid/ofaa439.721 (Peer Reviewed)

Corona Virus Disease-19 (COVID-19) in a Veterans Affairs Hospital at Suffolk County, Long Island, New York

Retrospective 67 hospitalized patients in the USA showing non-statistically significant unadjusted increased mortality with HCQ. Confounding by indication is likely.

Time varying confounding is likely. HCQ became controversial and was suspended during the end of the period studied, therefore HCQ use was likely more frequent toward the beginning of the study period, a time when overall treatment protocols were significantly worse.

death, ↑63.5%, p=0.52

Texeira et al., Open Forum Infectious Diseases, doi:10.1093/ofid/ofaa439.560 (Peer Reviewed)

Characteristics and outcomes of COVID-19 patients admitted to a regional health system in the southeast

Retrospective 161 hospitalized patients in the USA showing non-statistically significant unadjusted increased mortality with HCQ. Confounding by indication is likely.

Time varying confounding is likely. HCQ became controversial and was suspended towards the end of the period studied, therefore HCQ use was likely more frequent toward the beginning of the study period, a time when overall treatment protocols were significantly worse.

death, ↑79.3%, p=0.10

Vernaz et al., Swiss Medical Weekly, doi:10.4414/smw.2020.20446 (Peer Reviewed)

Early experimental COVID-19 therapies: associations with length of hospital stay, mortality and related costs

Retrospective 840 hospitalized patients in Switzerland showing non-statistically significant lower mortality with HCQ but significantly longer hospitalization times. Confounding by indication is likely. PSM fails to adjust for severity with a 16% higher mNEWS score for HCQ vs. SOC in the matched cohort.

Time varying confounding is likely. HCQ became controversial and was suspended towards the end of the period studied, therefore HCQ use was likely more frequent toward the beginning of the study period, a time when overall treatment protocols were significantly worse.

Authors note: "overall, there was an indication bias, with the reason of prescription being associated with the outcome of interest. Indeed, patients with more severe COVID-19 were more likely to receive experimental therapies."

death, ↓15.3%, p=0.71, HCQ vs. SOC
hospitalization time, ↑49.0%, p=0.002, HCQ vs. SOC

McCullough et al., Reviews in Cardiovascular Medicine, doi:10.31083/j.rcm.2020.04.264 (Review) (Peer Reviewed) (not included in the study count)

Multifaceted highly targeted sequential multidrug treatment of early ambulatory high-risk SARS-CoV-2 infection (COVID-19)

Review urging early treatment of COVID-19 with sequential multidrug treatment that has been shown to be safe and effective. Proposed treatment includes zinc, vitamin D & C, quercetin, and depending on age, comorbidities, and symptoms may include >=2 of HCQ, ivermectin, favipiravir; AZM/DOXY; corticosteroids; colchicine; bamlanivimab; aspirin; LMWH; and supplemental oxygen.

Procter et al., Reviews in Cardiovascular Medicine, doi:10.31083/j.rcm.2020.04.260 (Peer Reviewed)

Clinical outcomes after early ambulatory multidrug therapy for high-risk SARS-CoV-2 (COVID-19) infection

Retrospective 922 outpatients, with 320 treated early due to age>50 or comorbidities, showing 2.2% hospitalization and 0.3% death, which authors note is considerably lower than reported in other studies in their region.

At least two of zinc, HCQ, and ivermectin were used, along with one antibiotic, and budesonide and/or dexamethasone.

Güner et al., Journal of Infection and Public Health, doi:10.1016/j.jiph.2020.12.017 (Peer Reviewed)

Comparing ICU Admission Rates of Mild/Moderate COVID-19 Patients Treated with Hydroxychloroquine, Favipiravir, and Hydroxychloroquine plus Favipiravir

Retrospective 824 hospitalized patients in Turkey showing lower ICU admission for HCQ vs. favipiravir.

ICU admission, ↓77.3%, p=0.16, IPTW multivariate analysis

Cordtz et al., Rheumatology, doi:10.1093/rheumatology/keaa897 (Peer Reviewed)

Incidence and severeness of COVID-19 hospitalisation in patients with inflammatory rheumatic disease: a nationwide cohort study from Denmark

Retrospective 58,052 rheumatic disease patients in Denmark showing that RA patients have a higher risk of COVID-19 hospitalization in general. HCQ treated patients show lower risk, although this is not statistically significant with only 3 hospitalizations for HCQ treated patients.

HR 0.76 [0.23-2.52] time-dependent exposure model
HR 0.45 [0.11-1.92] time-fixed exposure model

hospitalization, ↓24.0%, p=0.67, time-dependent exposure model
hospitalization, ↓55.0%, p=0.28, time-fixed exposure model

Chari et al., Blood, doi:10.1182/blood.2020008150 (Peer Reviewed)

Clinical features associated with COVID-19 outcome in multiple myeloma: first results from the International Myeloma Society data set

Retrospective multiple myeloma patients showing lower mortality with HCQ treatment, unadjusted RR 0.67, p = 0.17 (data is in the supplementary material).

death, ↓33.1%, p=0.17

Su et al., BioScience Trends, doi:10.5582/bst.2020.03340 (Peer Reviewed)

Efficacy of early hydroxychloroquine treatment in preventing COVID-19 pneumonia aggravation, the experience from Shanghai, China

85% lower disease progression with early use of HCQ. Retrospective 616 patients in China showing adjusted progression HR 0.15, p = 0.006.

disease progression, ↓84.9%, p=0.006, binary logistic regression
improvement time, ↓24.0%, p=0.02, Cox proportional hazards regression

Taccone et al., The Lancet Regional Health - Europe, doi:10.1016/j.lanepe.2020.100019 (Peer Reviewed)

The role of organizational characteristics on the outcome of COVID-19 patients admitted to the ICU in Belgium

Retrospective 1,747 ICU patients in Belgium showing lower mortality with HCQ, multivariate mixed effects analysis HCQ aOR 0.64 [0.45-0.92].

death, ↓24.7%, p=0.0004
(odds ratio converted to relative risk)

Cangiano et al., Aging, doi:10.18632/aging.202307 (Peer Reviewed)

Mortality in an Italian nursing home during COVID-19 pandemic: correlation with gender, age, ADL, vitamin D supplementation, and limitations of the diagnostic tests

73% lower mortality with HCQ. Analysis of 98 PCR+ nursing home residents in Italy, mean age 90, showing HCQ mortality RR 0.27, p = 0.03. Subject to confounding by contraindication. The paper provides the p value for regression but not the effect size.

death, ↓73.4%, p=0.03

Huh et al., International Journal of Infectious Diseases, doi:10.1016/j.ijid.2020.12.041 (Peer Reviewed)

Association of prescribed medications with the risk of COVID-19 infection and severity among adults in South Korea

Retrospective database analysis with 17 existing users of HCQ and 5 severe cases, showing no significant difference for cases and higher risk for severe cases.

However, HCQ users are likely systemic autoimmune disease patients and authors do not adjust for the very different baseline risk for these patients. Other research shows that the risk of COVID-19 for systemic autoimmune disease patients is much higher overall, Ferri et al. show OR 4.42, p<0.001 [1].

[1] c19study.com/ferri.html

disease progression, ↑251.0%, p=0.11
COVID-19 case, ↓6.0%, p=0.82

Matangila et al., PLOS ONE, doi:10.1371/journal.pone.0244272 (Peer Reviewed)

Clinical characteristics of COVID-19 patients hospitalized at Clinique Ngaliema, a public hospital in Kinshasa, in the Democratic Republic of Congo: A retrospective cohort study

55% lower death with HCQ+AZ. Retrospective 160 hospitalized patients in the Democratic Republic of Congo, 92% receiving HCQ+AZ, showing adjusted OR 0.24 [0.03-2.2].

death, ↓54.9%, p=0.21
(odds ratio converted to relative risk)

Signes-Costa et al., Archivos de Bronconeumología, doi:10.1016/j.arbres.2020.11.012 (Peer Reviewed)

Prevalence and 30-day mortality in hospitalized patients with COVID-19 and prior lung diseases

47% lower mortality with HCQ/CQ. Retrospective 1,271 patients with lung disease in Canada, China, Cuba, Ecuador, Germany, Italy and Spain, 83% treated with HCQ/CQ.

Multivariable Cox regression HCQ/CQ mortality hazard ratio HR 0.53, p < 0.001.

death, ↓47.0%, p=0.0005

Gönenli et al., Research Square, doi:0.21203/rs.3.rs-107937/v1 (Preprint)

Prophylactic use of Hydroxychloroquine among Physicians working in Pandemic Hospitals

Small prophylaxis survey showing lower, but not statistically significant, progression to pneumonia (3 of 148 HCQ, 12 of 416 control), RR 0.70, p = 0.77. There was a higher incidence of cases with HCQ, OR 1.19, p = 0.58, which may be due to survey bias, treatment self-selection, and inconsistent regimens. Improvement on severity may be related to the much higher HCQ concentration in lung tissue, and also reflect that binary PCR does not distinguish replication-competence. Details of the pneumonia numbers for treatment/control are from the author.

risk of pneumonia, ↓29.7%, p=0.77
COVID-19 case, ↑18.9%, p=0.58
(odds ratio converted to relative risk)

Sofian et al., Wiener Medizinische Wochenschrift, doi:10.1007/s10354-020-00793-8 (Peer Reviewed)

SARS-CoV‑2, a virus with many faces: a series of cases with prolonged persistence of COVID-19 symptoms

Report on a series of 10 patients experiencing prolonged COVID-19 symptoms that were given HCQ 250mg bid for 5 days, with resolution of symptoms in all cases, and patients reporting they felt much better 2 days after treatment initiation.

Orioli et al., Diabetes & Metabolic Syndrome: Clinical Research & Reviews, doi:10.1016/j.dsx.2020.12.020 (Peer Reviewed)

Clinical characteristics and short-term prognosis of in-patients with diabetes and COVID-19: A retrospective study from an academic center in Belgium

Small retrospective study of 73 diabetic patients in Belgium, 55 HCQ patients, showing HCQ RR 0.87, p = 1.0.

death, ↓12.7%, p=1.00

Naseem et al., medRxiv, doi:10.1101/2020.12.13.20247254 (Preprint)

Predicting mortality in SARS-COV-2 (COVID-19) positive patients in the inpatient setting using a Novel Deep Neural Network

Retrospective 1,214 hospitalized patients in Pakistan, 77 HCQ patients, showing 33% lower mortality with HCQ, multivariate Cox HR 0.67, p = 0.34.

death, ↓33.3%, p=0.34, multivariate Cox

Tan et al., Virus Research, doi:10.1016/j.virusres.2020.198262 (Peer Reviewed)

A retrospective comparison of drugs against COVID-19

Retrospective 333 patients in China, with only 8 HCQ patients, showing shorter duration of hospitalization with HCQ.

hospitalization time, ↓35.2%, p=0.04

Bielza et al., Journal of the American Medical Directors Association, doi:10.1016/j.jamda.2020.12.003 (Peer Reviewed)

Clinical characteristics, frailty and mortality of residents with COVID-19 in nursing homes of a region of Madrid

Retrospective 630 elderly patients in Spain showing lower mortality with HCQ treatment, unadjusted relative risk RR 0.78, p = 0.09. HCQ was used more often with patients that were hospitalized (24% versus 3% use in the nursing homes). Median age 87.

death, ↓21.5%, p=0.09

Sogut et al., The American Journal of Emergency Medicine, doi:10.1016/j.ajem.2020.12.014 (Peer Reviewed) (not included in the study count)

Safety and efficacy of hydroxychloroquine in 152 outpatients with confirmed COVID-19: A pilot observational study

Safety study of 152 outpatients concluding that HCQ is safe for COVID-19, was well tolerated, and was not associated with a risk of ventricular arrhythmia due to drug-induced QTc interval prolongation.

Jung et al., Clinical Microbiology and Infection, doi:10.1016/j.cmi.2020.12.003 (Peer Reviewed)

Effect of hydroxychloroquine pre-exposure on infection with SARS-CoV-2 in rheumatic disease patients: A population-based cohort study

Retrospective cohort study of RA and SLE patients not showing a significant difference in PCR+ cases. PCR+ does not distinguish asymptomatic cases or severity. There was only one death which was in the control group. No other information on severity is provided.

33% of the control group used HCQ within the last year. Remaining confounding by differences in the nature and severity of rheumatic disease is likely.

death, ↓59.3%, p=1.00
COVID-19 case, ↑13.1%, p=0.86
(adjusted for zero (with reciprocal of the contrasting arm))

Alqassieh et al., F1000Research, Preprint (Preprint)

Clinical characteristics and predictors of the duration of hospital stay in COVID-19 patients in Jordan

Prospective observational study of 131 COVID-19 patients in Jordan, showing 18% shorter hospital stay with HCQ, p = 0.11.

hospitalization time, ↓18.2%, p=0.11

Italian Council of State (News) (not included in the study count)

Consiglio di Stato, sì all'uso dell'idrossiclorachina per la cura del Covid

Consiglio di Stato ruling in Italy re-establishes the right of Italian MDs to prescribe HCQ, which was suspended after the retracted Lancet study.

Agusti et al., Enfermedades Infecciosas y Microbiología Clínica, doi:10.1016/j.eimc.2020.10.023 (Peer Reviewed)

Efficacy and safety of hydroxychloroquine in healthcare professionals with mild SARS-CoV-2 infection: prospective, non-randomized trial

Small trial of low dose HCQ for healthcare workers with mild SARS-CoV-2 showing 68% lower progression to pneumonia, p = 0.21, and faster, but not statistically significant viral clearance.

There were no ICU admissions or deaths. Prospective non-randomized study. The figures and supplementary data are not currently available in the pre-proof edition.

disease progression, ↓68.4%, p=0.21, pneumonia
time to viral-, ↓31.8%

Guglielmetti et al., Journal of Infection and Public Health, doi:10.1016/j.jiph.2020.11.012 (Peer Reviewed)

Severe COVID-19 pneumonia in Piacenza, Italy – a cohort study of the first pandemic wave

Retrospective 218 hospitalized patients in Italy showing non-statistically significant 35% lower mortality with HCQ, hazard ratio aHR 0.65 [0.33–1.30].

death, ↓35.0%, p=0.22, multivariable Cox

Barnabas et al., Annals of Internal Medicine, doi:10.7326/M20-6519 (Peer Reviewed)

Hydroxychloroquine for Post-exposure Prophylaxis to Prevent Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Infection: A Randomized Trial

Early terminated PEP RCT comparing HCQ and vitamin C with 781 low-risk patients (83% household contacts), reporting no significant differences.

Different results were reported at IDWeek from the AIM results.

The study enrolled people with their last exposure within 4 days, i.e., if someone was exposed for 30 days in a row, they could be enrolled anywhere from day 1 to day 34. Therefore many were likely infected earlier than the enrollment date. Note that PCR has a very high false negative rates, e.g., 100% on day 1 and 67% on day 4 here [1].

50% of infections were detected by day 4. With the PCR false negatives and treatment delays it is likely that a majority of infections happened before enrollment or before HCQ can reach therapeutic levels.

Significantly more cases were caught at baseline in the control group (54 vs. 29 for HCQ) and excluded from analysis.

The early presentation stated that therapy started one day after enrollment and study supplies were sent to the participant "either by courier or mail". The published paper changes this to "courier delivery within 48 hours".

Overall delays are unclear but may be:

time since first exposure - unlimited
time from last exposure to enrollment - 10% reported as >= 5 days
time to telehealth meeting - 1 day (3 days if Friday enrollment?)
time to receive medication - <48 hours (including weekends?)

Symptomatic in this study was based on CDC-defined symptoms which contain symptoms that may be due to HCQ side effects.

Some results have not been reported, including symptomatic @28 days. The study uses a low dosage over an extended period, therapeutic levels may only be reached nearer to day 14, if at all, so day 28 results should be more informative when available (although labeled a PEP trial, with the low dosage and continuous exposure for most participants it is more of a PrEP/PEP trial where benefit might be seen later as HCQ levels increase).

Endpoints were:

Primary outcomes:
PCR+ @28 days mITT - aHR 1.16 [0.77-1.73]
PCR+ @14 days mITT - aHR 1.10 [0.73-1.66] IDWeek report was different: aHR 0.99 [0.64-1.52]
PCR+ @14 days ITT - aHR 0.81 [0.57-1.14]

Secondary outcomes:
PCR+ symptomatic @28 days - NOT REPORTED YET
duration of shedding - NOT REPORTED YET

Not in study protocol:
PCR+ cumulative symptomatic @14 days - aHR 1.23 [0.76-1.99].

Dose in first 24 hours - 0.8g (compare with Boulware et al. 2g)
Dose in first 5 days - 1.6g (compare with Boulware et al. 3.8g)

Other research suggests vitamin C may be beneficial for COVID-19, e.g. [2]. No information on severity of cases is provided. Binary PCR does not distinguish replication-competence. There were 2 COVID-19 hospitalizations, one in each group. Side effects were similar for HCQ and placebo. 83% medication adherence at day 14.

NCT04328961 [3]

[1] ncbi.nlm.nih.gov/pmc/articles/PMC7240870/

[2] researchsquare.com/article/rs-52778/v2

[3] clinicaltrials.gov/ct2/show/NCT04328961

hospitalization, ↑3.7%, p=1.00
COVID-19 case, ↑27.0%, p=0.33, day 14 symptomatic mITT PCR+ AIM
COVID-19 case, ↑23.0%, p=0.41, day 14 symptomatic mITT PCR+ IDWeek
COVID-19 case, ↑10.0%, p=0.66, day 14 PCR+ mITT AIM
COVID-19 case, ↓1.0%, p=0.97, day 14 PCR+ mITT IDWeek
COVID-19 case, ↓19.0%, p=0.23, day 14 PCR+ ITT AIM

Ozturk et al., Nephrology Dialysis Transplantation, doi:10.1093/ndt/gfaa271 (Peer Reviewed)

Mortality analysis of COVID-19 infection in chronic kidney disease, haemodialysis and renal transplant patients compared with patients without kidney disease: a nationwide analysis from Turkey

Retrospective 1210 hospitalized patients in Turkey focused on chronic kidney disease, haemodialysis and renal transplant patients, but also showing lower mortality with HCQ. Subject to confounding by indication.

death, ↓43.9%, p=0.14, CQ/HCQ

Modrák et al., medRxiv, doi:10.1101/2020.12.03.20239863 (Preprint)

Detailed disease progression of 213 patients hospitalized with Covid-19 in the Czech Republic: An exploratory analysis

Retrospective 213 hospitalized patients in Czech Republic showing lower mortality with HCQ. Subject to confounding by indication.

death, ↓59.0%, p=0.04, Cox (single)

Peng et al., Nephrology Dialysis Transplantation, doi:10.1093/ndt/gfaa288 (Peer Reviewed)

Early versus late acute kidney injury among patients with COVID-19—a multicenter study from Wuhan, China

Retrospective 4020 hospitalized patients in China showing non-statistically significant lower risk of acute kidney injury with HCQ.

disease progression, ↓10.8%, p=0.63, CQ/HCQ risk of AKI

Wiseman et al., medRxiv, doi:10.1101/2020.11.29.20235218 (Preprint) (meta analysis - not included in study count)

Effective post-exposure prophylaxis of Covid-19 is associated with use of hydroxychloroquine: Prospective re-analysis of a public dataset incorporating novel data

6th independent analysis showing efficacy from the Boulware PEP trial.

This prospective analysis corrects an error in the NEJM paper where shipping delays are omitted (still not corrected).

42% reduction in COVID-19 (9.6% vs. 16.5%), RR 0.58 [0.35 - 0.97], p=0.044, for the original trial 3 day specification.

COVID-19 case, ↓42.0%, p=0.04, <= 3 days after exposure

Capsoni et al., Research Square, doi:10.21203/rs.3.rs-113418/v1 (Preprint)

CPAP Treatment In COVID-19 Patients: A Retrospective Observational Study In The Emergency Department

Small 52 patient retrospective study of patients with acute respiratory failure showing lower rates of intubation with HCQ.

ventilation, ↓40.0%, p=0.30

Abdulrahman et al., medRxiv, doi:10.1101/2020.11.25.20234914 (Preprint)

The efficacy and safety of hydroxychloroquine in COVID19 patients : a multicenter national retrospective cohort

Retrospective medical record analysis of acute care patients in Bahrain not showing a significant effect of HCQ.

Confounding by indication is likely. Matching appears not to have matched for baseline severity - 17.5% of HCQ patients required oxygen while only 12.6% of control patients did.

death, ↓16.7%, p=1.00, PSM
combined intubation/death, ↑75.0%, p=0.24, PSM

Abd-Elsalam et al., Biological Trace Element Research, doi:10.1007/s12011-020-02512-1 (Peer Reviewed)

Do Zinc Supplements Enhance the Clinical Efficacy of Hydroxychloroquine?: a Randomized, Multicenter Trial

191 patient RCT in Egypt comparing the addition of zinc to HCQ, not showing a significant difference.

Clinical recovery at 28 days was 79.2% in the zinc group and 77.9% control, p = 0.969.

Mechanical ventilation was used with 4 patients in the zinc group and 6 control. There were 5 deaths in each group.

No information on baseline zinc values was recorded. We note that Egypt has a low rate of zinc deficiency so supplementation is less likely to be helpful in Egypt [1, 2]

[1] ncbi.nlm.nih.gov/pmc/articles/PMC3510072/

[2] ncbi.nlm.nih.gov/pmc/articles/PMC3510072/bin/pone.0050568.s003.xls

Lambermont et al., Critical Care Explorations, doi:10.1097/CCE.0000000000000305 (Peer Reviewed)

Predictors of Mortality and Effect of Drug Therapies in Mechanically Ventilated Patients With Coronavirus Disease 2019: A Multicenter Cohort Study

Retrospective 247 mechanically ventilated patients showing lower mortality with HCQ, but not statistically significant on multiple Cox regression.

The paper gives the p value for multiple Cox (0.46) and simple Cox (0.02), but does not specify the adjusted risk values.

death, ↓32.3%, p=0.46

Ruiz et al., International Journal of Antimicrobial Agents, doi:10.1016/j.ijantimicag.2020.106247 (Peer Reviewed)

Hydroxychloroquine lung pharmacokinetics in critically ill patients infected with COVID-19

HCQ lung pharmacokinetic study confirming that lung concentrations can be much higher than plasma.

The median lung epithelial lining fluid concentration was 38 times higher than plasma concentrations. 22 COVID-19 patients, median age 59.5.

Dosage determinations in studies assuming equilibrium between epithelium and plasma concentrations may lead to overly high dosages.

Authors find the median ELF concentration of HCQ above the maximum EC₅₀ value for 400 mg x 1 /day and 200 mg x 3 /day, 7-12 days after treatment initiation.

Rodriguez-Gonzalez et al., International Journal of Antimicrobial Agents, doi:10.1016/j.ijantimicag.2020.106249 (Peer Reviewed)

COVID-19 in hospitalized patients in Spain: a cohort study in Madrid

Retrospective 1255 patients in Spain showing lower mortality with HCQ. Subject to confounding by indication.

death, ↓22.8%, p=0.26

van Halem et al., BMC Infect Dis., doi:10.1186/s12879-020-05605-3 (Peer Reviewed)

Risk factors for mortality in hospitalized patients with COVID-19 at the start of the pandemic in Belgium: a retrospective cohort study

Retrospective 319 hospitalized patients in Belgium showing lower mortality with HCQ, although not reported to be statistically significant.

death, ↓31.6%, p=0.05

Abbas et al., Int. J. Clin. Pract., doi:10.1111/ijcp.13856 (Peer Reviewed)

Assessment of COVID-19 Treatment containing both Hydroxychloroquine and Azithromycin: A Natural Clinical Trial

Prospective study of 161 hospitalized patients in Iraq showing HCQ+AZ appears to help recovery. Most mortality was in patients that were already in critical condition on admission and died before treatment could be effective.

Qin et al., Thrombosis Research, doi:10.1016/j.thromres.2020.11.020 (Peer Reviewed)

Low molecular weight heparin and 28-day mortality among patients with coronavirus disease 2019: A cohort study in the early epidemic era

Low molecular weight heparin study also showing results for HCQ treatment, unadjusted HCQ mortality relative risk RR 0.66, p = 0.61.

death, ↓34.3%, p=0.61

Revollo et al., Journal of Antimicrobial Chemotherapy, doi:10.1093/jac/dkaa477 (Peer Reviewed)

Hydroxychloroquine pre-exposure prophylaxis for COVID-19 in healthcare workers

Retrospective PrEP analysis with 69 healthcare workers on PrEP HCQ, and 418 control.

Authors report PCR and IgG results, with no baseline results for either. Authors note they "identified 69 HCWs receiving HCQ" while providing no information as to why or when they started HCQ.

No conclusions can be drawn from this study because many workers may have been positive before starting HCQ. Only 14% of workers chose to use HCQ and they may have been motivated to do so because they had an infection.

Authors perform several different adjustments, finding very different results. No information on death, hospitalization, symptoms, or severity is provided. Details on timing of serology and baseline serology status is not provided. Potential bias due to self-selection for risk.

25% of infections were detected before 7 days, indicating that they actually happened earlier (PCR false positive is very high initially). It is likely that many infections were before HCQ could reach therapeutic levels.

COVID-19 case, ↓23.0%, p=0.52, PSM risk of PCR+
COVID-19 case, ↑43.0%, p=0.42, PSM risk of IgG+

Omrani et al., EClinicalMedicine, doi:10.1016/j.eclinm.2020.100645 (Peer Reviewed)

Randomized double-blinded placebo-controlled trial of hydroxychloroquine with or without azithromycin for virologic cure of non-severe Covid-19

Low risk patient RCT for HCQ+AZ and HCQ vs. control, not showing any significant differences.

Authors note that the results are not applicable to higher risk patients; that positive PCR may simply reflect detection of inactive (non-infectious) viral remnants; that an alternative dosage regimen may be more effective; and that medication adherence was unknown.

HCQ dosing was 600mg/day for 1 week, therapeutic levels may not be reached for several days. There were no deaths or serious adverse events.

hospitalization, ↓12.5%, p=1.00, HCQ+AZ or HCQ vs. control
risk of symptomatic at day 21, ↓25.8%, p=0.58, HCQ+AZ or HCQ vs. control
risk of Ct<=40 at day 14, ↑10.3%, p=0.13, HCQ+AZ or HCQ vs. control

Falcone et al., Open Forum Infectious Diseases, doi:10.1093/ofid/ofaa563 (Peer Reviewed)

Role of low-molecular weight heparin in hospitalized patients with SARS-CoV-2 pneumonia: a prospective observational study

Prospective observational study of 315 hospitalized patients in Italy showing 65% lower mortality with HCQ. The median treatment delay was 6 days for survivors and 6.5 days for non-survivors. Mortality relative risk:

RR 0.35, p = 0.2, propensity score matched
RR 0.75, p = 0.36, multivariate Cox regression
RR 0.43, p < 0.001, univariate Cox regression

death, ↓65.0%, p=0.20, PSM
death, ↓25.0%, p=0.36, multivariate Cox regression
death, ↓57.0%, p<0.001, univariate Cox regression

Budhiraja et al., medRxiv, doi:10.1101/2020.11.16.20232223 (Preprint)

Clinical Profile of First 1000 COVID-19 Cases Admitted at Tertiary Care Hospitals and the Correlates of their Mortality: An Indian Experience

Retrospective 976 hospitalized patients with 834 treated with HCQ+AZ showing HCQ mortality relative risk RR 0.35, p < 0.0001. Note that in this case HCQ was recommended for mild/moderate cases, so more severe cases may not have received HCQ (which may also be why they became severe cases). We note that this is opposite to a common bias in HCQ studies - in many cases HCQ was more likely to be given to more severe cases.

death, ↓65.4%, p<0.0001

Boari et al, Biosci. Rep., doi:10.1042/BSR20203455 (Peer Reviewed)

Prognostic factors and predictors of outcome in patients with COVID-19 and related pneumonia: a retrospective cohort study

Retrospective 258 hospitalized patients in Italy showing lower mortality with HCQ treatment, unadjusted relative risk RR 0.455, p<0.001.

(data is in the supplementary appendix)

death, ↓54.5%, p<0.001

Sheshah et al., Diabetes Research and Clinical Practice, doi:10.1016/j.diabres.2020.108538 (Peer Reviewed)

Prevalence of Diabetes, Management and Outcomes among Covid-19 Adult Patients Admitted in a Specialized Tertiary Hospital in Riyadh, Saudi Arabia

Retrospective 300 hospitalized patients in Saudi Arabia showing HCQ adjusted odds ratio aOR 0.12, p < 0.001.

death, ↓80.0%, p<0.001
(odds ratio converted to relative risk)

Simova et al., New Microbes and New Infections, doi:10.1016/j.nmni.2020.100813 (Peer Reviewed)

Hydroxychloroquine for prophylaxis and treatment of COVID-19 in health care workers

100% reduction in hospitalization and cases with early treatment using HCQ+AZ+zinc. Brief report on healthcare workers in Bulgaria.

0 hospitalizations with treatment vs. 2 for control
0 PCR+ at day 14 with treatment vs. 3 for control

33 treatment patients and 5 control patients.

No serious adverse events. This paper reports on both PEP and early treatment, we have separated the two studies.

hospitalization, ↓93.8%, p=0.01
risk of viral+ at day 14, ↓95.8%, p=0.001
(adjusted for zero (with reciprocal of the contrasting arm))

Simova et al., New Microbes and New Infections, doi:10.1016/j.nmni.2020.100813 (Peer Reviewed)

Hydroxychloroquine for prophylaxis and treatment of COVID-19 in health care workers

100% reduction in cases with HCQ+zinc post-exposure prophylaxis. Brief report for healthcare workers in Bulgaria.

0 cases with treatment vs. 3 for control.

156 treatment patients and 48 control patients.

No serious adverse events. This paper reports on both PEP and early treatment, we have separated the two studies.

COVID-19 case, ↓92.7%, p=0.01
(adjusted for zero (with reciprocal of the contrasting arm))

Tchounga et al., Journal of Pharmaceutical and Biomedical Analysis, doi:10.1016/j.jpba.2020.113761 (Peer Reviewed) (not included in the study count)

Composition analysis of falsified chloroquine phosphate samples seized during the COVID-19 pandemic

Analysis of fake CQ tablets finding:

- no CQ in six samples, substituted with metronidazole (at sub-therapeutic levels) or paracetamol.
- trace levels of paracetamol and chloramphenicol in four and two samples respectively.
- CQ levels too low in two samples.

Águila-Gordo et al., Revista Española de Geriatría y Gerontología, doi:10.1016/j.regg.2020.09.006 (Peer Reviewed)

Mortality and associated prognostic factors in elderly and very elderly hospitalized patients with respiratory disease COVID-19

67% lower mortality with HCQ. Retrospective 416 elderly patients in Spain showing adjusted HCQ mortality hazard ratio HR 0.33, p = 0.1.

death, ↓67.0%, p=0.10

Khamis et al., International Journal of Infectious Diseases, doi:10.1016/j.ijid.2020.11.008 (Peer Reviewed)

Randomized Controlled Open Label Trial on the Use of Favipiravir Combined with Inhaled Interferon beta-1b in Hospitalized Patients with Moderate to Severe COVID-19 Pneumonia

Small 89 patient RCT comparing favipiravir and inhaled interferon with HCQ for moderate to severe COVID-19 pneumonia, not finding significant differences.

Self et al., JAMA, doi:10.1001/jama.2020.22240 (Peer Reviewed)

Effect of Hydroxychloroquine on Clinical Status at 14 Days in Hospitalized Patients With COVID-19: A Randomized Clinical Trial

Early terminated very late stage (65% on supplemental oxygen) RCT with 242 HCQ and 237 control patients not showing a significant difference, 28 day mortality adjusted odds ratio aOR 0.93 [0.48-1.85].

For the subgroup not on supplemental oxygen at baseline (relatively early treatment), the odds ratio for the 7 point outcome scale is:

aOR 0.61 [0.34-1.08]

Dosage may be too low:

Dose in first 24 hours - 1g (compare to Boulware et al. 2g)
Dose in 5 days - 2.4g (compare to Boulware et al. 3.8g)

Dosage note: Boulware 2g within 24 hours includes the second day dose. Note two important differences with the RECOVERY/SOLIDARITY dosage which is believed to be dangerously high - in RECOVERY/SOLIDARITY the total dose is much higher, which is problematic because the half-life of HCQ is very long (the excess mortality is seen later), and it is given to patients that are already in very serious condition.

Note the paper reports values with OR>1 favoring HCQ, we have converted to OR<1 favoring HCQ. Subgroup analysis is in the supplemental appendix.

death, ↓7.0%, p=0.84

Brown et al., Annals of the American Thoracic Society, doi:10.1513/AnnalsATS.202008-940OC (Peer Reviewed)

Hydroxychloroquine vs. Azithromycin for Hospitalized Patients with COVID-19 (HAHPS): Results of a Randomized, Active Comparator Trial

Small early terminated very late stage (86% on oxygen, 44% enrolled in the ICU) RCT comparing HCQ vs. AZ, not finding a significant difference between the two treatments. There is no comparison with a control group. HCQ patients not in the ICU at enrollment (slightly earlier treatment) did better, OR 0.95 vs. 1.13.

HCQ dosage is relatively low:

Dose in first 24 hours - 1g (compare to Boulware et al. 2g)
Dose in 5 days - 2.4g (compare to Boulware et al. 3.8g)

Núñez-Gil et al., Intern. Emerg. Med., doi:10.1007/s11739-020-02543-5 (Peer Reviewed)

Mortality risk assessment in Spain and Italy, insights of the HOPE COVID-19 registry

Retrospective database study of 1,021 patients in Ecuador, Germany, Italy, and Spain, showing HCQ propensity score adjusted mortality odds ratio aOR 0.88, p=0.005.

death, ↓7.9%, p=0.005
(odds ratio converted to relative risk)

Mathai et al., J. Marine Medical Society, doi:10.4103/jmms.jmms_115_20 (Peer Reviewed)

Hydroxychloroquine as pre-exposure prophylaxis against COVID-19 in health-care workers: A single-center experience

90% reduction in cases with HCQ pre-exposure prophylaxis. Retrospective 604 healthcare workers.

COVID-19 case, ↓89.5%, p<0.0001
COVID-19 case, ↓88.5%, p<0.0001, symptomatic

Datta et al., Journal of Vaccines & Vaccination, S6:1000002 (Peer Reviewed)

No Role of HCQ in COVID-19 Prophylaxis: A Survey amongst Indian Doctors

Survey of Indian doctors not finding a significant effect of HCQ prophylaxis, p = 0.54.

We do not know the actual results for this study - the numbers for prophylaxis in Table 1 appear to be incorrect. They do not match the total using HCQ, and they are identical to the numbers for the analysis by specialty.

Dhibar et al., International Journal of Antimicrobial Agents, doi:10.1016/j.ijantimicag.2020.106224 (Peer Reviewed)

Post Exposure Prophylaxis with Hydroxychloroquine (HCQ) for the Prevention of COVID-19, a Myth or a Reality? The PEP-CQ Study

41% reduction in cases with HCQ PEP. Prospective open label trial with 132 HCQ patients and 185 control patients showing RR 0.59, p=0.03.

50% reduction in PCR+ cases, 44% reduction in symptomatic cases.

No serious adverse events. Relatively low dosage, the PEP group received HCQ 800mg on day one followed by 400mg once weekly for 3 weeks. NCT04408456

COVID-19 case, ↓41.0%, p=0.03
COVID-19 case, ↓50.0%, p=0.04, PCR+
symptomatic case, ↓43.9%, p=0.21

Maldonado et al., Nefrología, doi:10.1016/j.nefro.2020.09.002 (Peer Reviewed)

COVID-19 incidence and outcomes in a home dialysis unit in Madrid (Spain) at the height of the pandemic

Small retrospective 12 dialysis patients, 1/11 deaths with HCQ and 1/1 without HCQ.

death, ↓90.9%, p=0.17

Rodriguez-Nava et al., Mayo Clinic Proceedings: Innovations, Quality & Outcomes (Peer Reviewed)

Clinical characteristics and risk factors for mortality of hospitalized patients with COVID-19 in a community hospital: A retrospective cohort study

Retrospective 313 patients, mostly critical stage and mostly requiring respiratory support, showing unadjusted RR 1.06, p = 0.77. Confounding by indication likely.

death, ↑6.3%, p=0.77, unadjusted

Salazar et al., The American Journal of Pathology, doi:10.1016/j.ajpath.2020.10.008 (Peer Reviewed)

Significantly Decreased Mortality in a Large Cohort of Coronavirus Disease 2019 (COVID-19) Patients Transfused Early with Convalescent Plasma Containing High-Titer Anti–Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Spike Protein IgG

Convalescent plasma study also showing mortality based on HCQ treatment, unadjusted hazard ratio uHR 1.37, p = 0.28. Confounding by indication is likely.

death, ↑37.0%, p=0.28

Cadegiani et al., medRxiv, doi:10.1101/2020.10.31.20223883 (Preprint)

Early COVID-19 Therapy with Azithromycin Plus Nitazoxanide, Ivermectin or Hydroxychloroquine in Outpatient Settings Significantly Reduced Symptoms Compared to Known Outcomes in Untreated Patients

Comparison of HCQ, nitazoxanide, and ivermectin showing similar effectiveness for overall clinical outcomes in COVID-19 when used before seven days of symptoms, and overwhelmingly superior compared to the untreated COVID-19 population, even for those outcomes not influenced by placebo effect, at least when combined with azithromycin, and vitamin C, D and zinc in the majority of the cases.

585 patients with mean treatment delay 2.9 days. There was no hospitalization, mechanical ventilation, or mortality.

death, ↓81.2%, p=0.21, control group 1
ventilation, ↓95.1%, p=0.0008, control group 1
hospitalization, ↓98.3%, p<0.0001, control group 1
(adjusted for zero (with reciprocal of the contrasting arm))

Behera et al., medRxiv, doi:10.1101/2020.10.29.20222661v1 (Preprint)

Role of ivermectin in the prevention of COVID-19 infection among healthcare workers in India: A matched case-control study

Retrospective prophylaxis study for HCQ, ivermectin, and vitamin C with 372 healthcare workers, showing lower COVID-19 incidence for all treatments, with statistical significance reached for ivermectin.

HCQ OR 0.56, p = 0.29
Ivermectin OR 0.27, p < 0.001
Vitamin C OR 0.82, p = 0.58

COVID-19 case, ↓27.9%, p=0.29
(odds ratio converted to relative risk)

López et al., Annals of Pediatrics, doi:10.1016/j.anpedi.2020.10.017 (Peer Reviewed)

Telemedicine follow-ups for COVID-19: experience in a tertiary hospital

Retrospective 72 pediatric patients showing HCQ associated with a shorter duration of fever (p=0.023), less progression (p=0.016), and fewer return visits to the ER (p=0.017).

disease progression, ↓64.3%, p=0.02

Fonseca et al., Travel Medicine and Infectious Disease, doi:10.1016/j.tmaid.2020.101906 (Peer Reviewed)

Risk of Hospitalization for Covid-19 Outpatients Treated with Various Drug Regimens in Brazil: Comparative Analysis

64% lower hospitalization with HCQ. Retrospective 717 patients in Brazil with early treatment, adjusted OR 0.32, p=0.00081, for HCQ versus no medication, and OR 0.45, p=0.0065, for HCQ vs. anything else.

hospitalization, ↓64.0%, p=0.0008, HCQ vs. nothing
hospitalization, ↓50.5%, p=0.006, HCQ vs. anything else
(odds ratio converted to relative risk)

Tehrani et al., International Journal of Infectious Diseases, doi:10.1016/j.ijid.2020.10.071 (Peer Reviewed)

Risk factors for mortality in adult COVID-19 patients; frailty predicts fatal outcome in older patients

Retrospective 255 hospitalized patients, 65 treated with HCQ, showing unadjusted RR 0.87, p=0.63. Confounding by indication is likely.

death, ↓13.4%, p=0.63

Arleo et al., medRxiv, doi:10.1101/2020.10.26.20219154 (Preprint)

Clinical Course and Outcomes of coronavirus disease 2019 (COVID-19) in Rheumatic Disease Patients on Immunosuppression: A case Cohort Study at a Single Center with a Significantly Diverse Population

Retrospective hospitalized rheumatic disease patients showing 50% lower mortality for patients on HCQ.

death, ↓50.0%, p=0.67, all patients
death, ↓52.0%, p=0.64, inpatients

Choi et al., International Journal of Infectious Diseases, doi:10.1016/j.ijid.2020.10.062 (Peer Reviewed)

Comparison of antiviral effect for mild-to-moderate COVID-19 cases between lopinavir/ritonavir versus hydroxychloroquine: A nationwide propensity score-matched cohort study

Health insurance database analysis failing to adjust for disease severity and not finding a significant difference in time to PCR- for LPV/r and HCQ.

There are large differences in severity across groups. Authors did PSM but chose not to prioritize severity, resulting in incomparable groups, e.g., baseline pneumonia of 44% in the HCQ group and 15% in the control group (after PSM).

Authors note this but offer no explanation for not correcting for severity: "However, the disease severity and proportion of accompanying pneumonia were still significantly higher in the LPV/r and HCQ-group".

median time to PCR-, ↑22.0%, p<0.0001

Frontera et al., Research Square, doi:10.21203/rs.3.rs-94509/v1 (Preprint)

Treatment with Zinc is Associated with Reduced In-Hospital Mortality Among COVID-19 Patients: A Multi-Center Cohort Study

Retrospective 3,473 hospitalized patients showing lower mortality with HCQ+zinc.

death, ↓37.0%, p=0.01, PSM
death, ↓24.0%, p=0.02, regression

Derwand et al., International Journal of Antimicrobial Agents, doi:10.1016/j.ijantimicag.2020.106214 (preprint 7/3) (Peer Reviewed)

COVID-19 Outpatients – Early Risk-Stratified Treatment with Zinc Plus Low Dose Hydroxychloroquine and Azithromycin: A Retrospective Case Series Study

79% lower mortality and 82% lower hospitalization with early HCQ+AZ+Z.

No cardiac side effects. Retrospective 518 patients (141 treated, 377 control).

death, ↓79.4%, p=0.12
hospitalization, ↓81.6%, p<0.001
(odds ratio converted to relative risk)

Goenka et al., SSRN, doi:10.2139/ssrn.3689618 (Preprint)

Seroprevalence of COVID-19 Amongst Health Care Workers in a Tertiary Care Hospital of a Metropolitan City from India

Study of SARS-CoV-2-IgG antibodies in 1122 health care workers in India finding 87% lower positives for adequate HCQ prophylaxis, 1.3% HCQ versus 12.3% for no HCQ prophylaxis.

Adequate prophylaxis is defined as 400mg 1/wk for >6 weeks.

IgG positive, ↓87.2%, p=0.03
(odds ratio converted to relative risk)

Coll et al., American Journal of Transplantation, doi:10.1111/ajt.16369 (Peer Reviewed)

Covid‐19 in transplant recipients: the spanish experience

Retrospective 652 transplant recipient patients in Spain showing 46% lower mortality for patients treated with HCQ, unadjusted relative risk RR 0.54, p<0.0001.

death, ↓45.6%, p<0.0001

Lano et al., Clinical Kidney Journal, 13:5, October 2020, 878–888, doi:10.1093/ckj/sfaa199 (Peer Reviewed)

Risk factors for severity of COVID-19 in chronic dialysis patients from a multicentre French cohort

33% lower mortality with HCQ+AZ, p=0.28. Retrospective 122 French dialysis patients.

69% lower combined mortality/ICU, p=0.11, for the subgroup not requiring O2 on diagnosis (slightly earlier treatment).

death, ↓33.1%, p=0.28
combined death/ICU, ↓38.9%, p=0.23
combined death/ICU, ↓68.7%, p=0.11, not requiring O2 on diagnosis (relatively early treatment)
(odds ratio converted to relative risk)

Dubee et al., medRxiv, doi:10.1101/2020.10.19.20214940 (Preprint)

A placebo-controlled double blind trial of hydroxychloroquine in mild-to-moderate COVID-19

Small early terminated late stage (60% on oxygen) RCT in France showing 46% lower mortality.

mortality at 28 days relative risk RR 0.54 [0.21-1.42]
combined mortality/intubation at 28 days relative risk RR 0.74 [0.33-1.70]

If not stopped early and the same trend continued, statistical significance would be reached on 28 day mortality after ~550 patients (1,300 patients were planned).

Mortality results are not provided for subgroups. For the subgroups receiving AZ:

HCQ+(AZ from day 0): combined mortailty/intubation RR 0.16, p = 0.21
(0/10 HCQ+AZ and 3/11 placebo, 0.5 added for calculations due to 0)
HCQ+(AZ later), combined mortality/intubation RR 0.42 [0.05-3.54]

No safety concerns were identified. This study has been presented as negative, however the results do not support that conclusion.

NCT04325893

risk of death at day 28, ↓46.0%, p=0.21
risk of combined intubation/death at day 28, ↓26.0%, p=0.82
combined intubation/death, ↓85.1%, p=0.21, HCQ+AZ from day 0 subgroup
(adjusted for zero (with reciprocal of the contrasting arm))

Ñamendys-Silva et al., Heart & Lung, doi:10.1016/j.hrtlng.2020.10.013 (Peer Reviewed)

Outcomes of patients with COVID-19 in the Intensive Care Unit in Mexico: A multicenter observational study

Retrospective 164 ICU patients in Mexico showing 32% lower mortality with HCQ+AZ and 37% lower with CQ.

HCQ+AZ vs. neither HCQ or CQ relative risk RR 0.68, p = 0.03
CQ vs. neither HCQ or CQ relative risk RR 0.63, p = 0.02
HCQ+AZ or CQ vs. neither relative risk RR 0.65, p = 0.006

death, ↓32.3%, p=0.18, HCQ+AZ vs. neither HCQ or CQ
death, ↓37.1%, p=0.09, CQ vs. neither HCQ or CQ
death, ↓34.5%, p=0.006, HCQ+AZ or CQ

IHU Marseille (Preprint) (meta analysis - not included in study count)

Meta-analysis on chloroquine derivatives and COVID-19 mortality

Updated meta analysis of 41 studies showing CQ/HCQ OR 0.57, p<0.0001 from clinical studies.

For big data studies authors find inconsistent results and OR 0.83, p=0.0014, and for all studies combined OR 0.72, p<0.0001.

Solh et al., medRxiv, doi:10.1101/2020.10.16.20214130 (Preprint)

Clinical course and outcome of COVID-19 acute respiratory distress syndrome: data from a national repository

Retrospective database analysis of 7,816 Veterans Affairs hospitalized patients analyzing progression to ARDS and 30-day mortality from ARDS. Confounding by indication is likely. Chronological bias is likely, with HCQ more likely to be used earlier on, before significant improvements in overall treatment.

No results are provided for HCQ for progression to ARDS.

death, ↑18.0%, p=0.17

Mohana et al., International Journal of Infectious Diseases, doi:10.1016/j.ijid.2020.10.031 (preprint 8/17) (Peer Reviewed) (not included in the study count)

Hydroxychloroquine Safety Outcome within Approved Therapeutic Protocol for COVID-19 Outpatients in Saudi Arabia

Safety study of 2,733 patients in Saudi Arabia showing HCQ in mild to moderate cases in an outpatient setting, within the protocol recommendation and inclusion/exclusion criteria, is safe, highly tolerable, and has minimal side effects. No ICU admission or deaths were reported.

Guisado-Vasco (Peer Reviewed)

Clinical characteristics and outcomes among hospitalized adults with severe COVID-19 admitted to a tertiary medical center and receiving antiviral, antimalarials, glucocorticoids, or immunomodulation with tocilizumab or cyclosporine: A retrospective observational study (COQUIMA cohort)

Retrospective 607 patients reporting results for early outpatient HCQ use with mortality odds ratio OR 0.092 [0.022-0.381], p = 0.001 (65 patients), and for hospital use, mortality odds ratio OR 0.737 [0.38-1.41], p = 0.36 (558 patients). Median age 69.

death, ↓20.3%, p=0.36
outpatient use, ↓88.0%, p=0.001
(odds ratio converted to relative risk)

SOLIDARITY Trial Consortium, NEJM, doi:10.1056/NEJMoa2023184 (preprint 10/15) (Peer Reviewed)

Repurposed antiviral drugs for COVID-19; interim WHO SOLIDARITY trial results

WHO SOLIDARITY open-label trial with 954 very late stage (64% on oxygen/ventilation) HCQ patients, mortality relative risk RR 1.19 [0.89-1.59], p=0.23.

HCQ dosage very high as in RECOVERY, 1.6g in the first 24 hours, 9.6g total over 10 days, only 25% less than the high dosage that Borba et al. show greatly increases risk (OR 2.8) [1].

Authors state they do not know the weight or obesity status of patients to analyze toxicity (since they do not adjust dosage based on patient weight, toxicity may be higher in patients of lower weight).

KM curves show a spike in HCQ mortality days 5-7, corresponding to ~90% of the total excess seen at day 28 (a similar spike is seen in the RECOVERY trial).

Almost all excess mortality is from ventilated patients.

Authors refer to a lack of excess mortality in the first few days to suggest a lack of toxicity, but they are ignoring the very long half-life of HCQ and the dosing regimen - much higher levels of HCQ will be reached later. Increased mortality in Borba et al. occurred after 2 days.

An unspecified percentage used the more toxic CQ. No placebo used.

[1] c19study.com/borba.html

death, ↑19.0%, p=0.23

Annie et al., Pharmacotherapy, doi:10.1002/phar.2467 (Peer Reviewed)

Hydroxychloroquine in hospitalized COVID‐19 patients: Real world experience assessing mortality

Retrospective database analysis with PSM not including COVID-19 severity, finding mortality OR 0.95 [0.62-1.46] for HCQ, and 1.24 [0.70-2.22] for HCQ+AZ. Confounding by indication likely.

death, ↓4.3%, p=0.83
death, ↑20.5%, p=0.46
(odds ratio converted to relative risk)

Sili et al., medRxiv, doi:10.1101/2020.10.09.20209775 (Preprint) (not included in the study count)

Factors associated with progression to critical illness in 28 days among COVID-19 patients: results from a tertiary care hospital in Istanbul, Turkey

Analysis of hospitalized patients in Turkey showing HCQ was given to 99.2% of patients and the incidence of critical illness was lower than most studies. Authors note "whether HCQ administration lowered the rates of critical illness development is beyond the scope of this study." There is no comparison with a control group.

Aparisi et al., medRxiv, doi:10.1101/2020.10.06.20207092 (Preprint)

Low-density lipoprotein cholesterol levels are associated with poor clinical outcomes in COVID-19

Retrospective 654 hospitalized patients focused on low-density lipoprotein cholesterol levels, also showing results for HCQ with 605 HCQ patients, unadjusted 30 day mortality relative risk RR 0.37, p = 0.008.

death, ↓63.0%, p=0.008

Soto-Becerra et al., medRxiv, doi:10.1101/2020.10.06.20208066 (Preprint)

Real-World Effectiveness of hydroxychloroquine, azithromycin, and ivermectin among hospitalized COVID-19 patients: Results of a target trial emulation using observational data from a nationwide Healthcare System in Peru

Retrospective database study of 5683 patients, 692 received HCQ/CQ+AZ, 200 received HCQ/CQ, 203 received ivermectin, 1600 received AZ, 358 received ivermectin+AZ, and 2630 received standard of care.

This study includes anyone with ICD-10 COVID-19 codes which includes asymptomatic PCR+ patients, therefore many patients in the control group are likely asymptomatic with regards to SARS-CoV-2, but in the hospital for another reason. For those that had symptomatic COVID-19, there is also likely significant confounding by indication.

In this study all medications show higher mortality at day 30, which is consistent with asymptomatic (for COVID-19) or mild condition patients being more common in the control group.

For ivermectin they show 30 day mortality aHR = 1.39 [0.88 - 2.22]. KM curves show that the treatment groups were in more serious condition, and also that after about day 35 survival became better with ivermectin. The last day available for ivermectin shows RR 0.83, p = 0.01. More than the total excess mortality happened on the first day. This is consistent with treated patients being in more serious condition, and with many of the control group patients being in hospital for something unrelated to COVID-19.

Authors use a machine learning based propensity scoring system that appears over-parameterized and likely to result in significant overfitting and inaccurate results. Essentially they test for all interactions between two and three covariates. The nature and large number of covariates means many random correlations may be found. COVID-19 severity is not used.

See also: [1].

Ivermectin dosage details: [2]

[1] twitter.com/Covid19Crusher/status/1315461049034907650

[2] cdn.www.gob.pe/uploads/document/file/830595/RM_375-2020-MINSA.PDF

death, ↓18.1%, p<0.0001, day 54 (last day available) weighted KM
death, ↑84.0%, p=0.02, day 30

Ader et al., medRxiv, doi:10.1101/2021.01.08.20248149 (news 10/6) (Preprint)

Antiviral drugs in hospitalized patients with COVID-19 - the DisCoVeRy trial

Early terminated very late stage (95% on oxygen at baseline) DISCOVERY trial. 4% more patients were on ventilation at baseline in the HCQ group.

risk of death at day 29, ↓6.4%, p=1.00

Mori et al., Journal of Microbiology, Immunology and Infection, doi:10.1016/j.jmii.2020.09.003 (Peer Reviewed)

Triple therapy with hydroxychloroquine, azithromycin, and ciclesonide for COVID-19 pneumonia

Small case study of 5 patients in Japan showing improvement with HCQ+AZ+ciclesonide.

Prodromos et al., New Microbes and New Infections, doi:10.1016/j.nmni.2020.100776 (Peer Reviewed) (meta analysis - not included in study count)

Hydroxychloroquine is effective, and consistently so used early, for Covid-19: A systematic review

Meta analysis of 43 studies: "HCQ was found consistently effective against COVID-19 when used early, in the outpatient setting. It was found overall effective also including inpatient studies. No unbiased study found worse outcomes with HCQ use. No mortality or serious safety adverse event was found.

Nachega et al., The American Journal of Tropical Medicine and Hygiene, doi:10.4269/ajtmh.20-1240 (Peer Reviewed)

Clinical Characteristics and Outcomes of Patients Hospitalized for COVID-19 in Africa: Early Insights from the Democratic Republic of the Congo

Retrospective 766 hospitalized patients in DRC showing mortality reduced from 29% to 11%, and improvement at 30 days increased from 65% to 84%.

Mortality cox regression adjusted hazard ratio aHR 0.26, p < 0.001
Risk of no improvement adjusted odds ratio aOR 0.28, p < 0.001

Using marginal structural model analysis these risks became:

Mortality MSM adjusted odds ratio aOR 0.65, p = 0.166
Risk of no improvement MSM adjusted odds ratio aOR = 0.65, p = 0.132

Median age 46, 630 treated with CQ+AZ.

death, ↓27.6%, p=0.17
no improvement, ↓25.8%, p=0.13
(odds ratio converted to relative risk)

Almazrou et al., Saudi Pharmaceutical Journal, doi:10.1016/j.jsps.2020.09.019 (Peer Reviewed)

Comparing the impact of Hydroxychloroquine based regimens and standard treatment on COVID-19 patient outcomes: A retrospective cohort study

Retrospective 161 hospitalized patients in Saudi Arabia showing lower ventilation and ICU admission with HCQ, but not statistically significant with the small sample sizes.

ventilation, ↓65.0%, p=0.16
ICU admission, ↓21.0%, p=0.78

Garcia-Albeniz et al., medRxiv, doi:10.1101/2020.09.29.20203869 (Preprint) (meta analysis - not included in study count)

Brief communication: A meta-analysis of randomized trials of hydroxychloroquine for the prevention of COVID-19

Combination of the four underpowered prophylaxis RCTs to date showing statistically significant results, RR 0.78 [0.61-0.99].

The actual effect of HCQ is likely to be higher for several reasons: the trials did not adjust for losses to follow-up or other deviations from protocol. There was very long treatment delays in the postexposure prophylaxis trials - in one trial, about a third of participants were enrolled 4 days after exposure with an additional shipping delay of ~46 hours on average, in the other, participants were enrolled up to 7 days after exposure, with an unknown additional delay before treatment, and results suggesting that exposure detection was delayed.

For other reasons see: [1, 2, 3, 4].

[1] c19study.com/boulware.html

[2] c19study.com/mitjapep.html

[3] c19study.com/rajasingham.html

[4] c19study.com/abella.html

COVID-19 case, ↓22.0%, p=0.04

Polat et al., Medical Journal of Bakirkoy, 16:3, 280-6, doi:10.5222/BMJ.2020.50469 (Peer Reviewed)

Hydroxychloroquine Use on Healthcare Workers Exposed to COVID-19 -A Pandemic Hospital Experience

Small prophylaxis study of 208 healthcare workers in Turkey, 138 with high risk exposure received HCQ, while 70 with low and medium risk exposure did not. COVID-19 cases were lower in the treatment group, relative risk RR 0.43, p = 0.026. Since the control group had lower risk, the actual benefit may be larger.

COVID-19 case, ↓57.0%, p=0.03

Ayerbe et al., Internal and Emergency Medicine, doi:0.1007/s11739-020-02505-x (Peer Reviewed)

The association of treatment with hydroxychloroquine and hospital mortality in COVID-19 patients

2075 hospital patients in Spain showing HCQ reduces mortality 52%, odds ratio OR 0.39, p<0.001, after adjustment for age, gender, temperature > 37 °C, and saturation of oxygen < 90% treatment with azithromycin, steroids, heparin, tocilizumab, a combination of lopinavir with ritonavir, and oseltamivir, and date of admission (model 4).

death, ↓52.2%, p<0.001
(odds ratio converted to relative risk)

Ladapo et al., medRxiv, doi:10.1101/2020.09.30.20204693 (Preprint) (meta analysis - not included in study count)

Randomized Controlled Trials of Early Ambulatory Hydroxychloroquine in the Prevention of COVID-19 Infection, Hospitalization, and Death: Meta-Analysis

Meta analysis of prophylactic and early treatment RCTs, 24% reduction in cases, hospitalization or death with HCQ, RR 0.76, p=0.025. No serious adverse cardiac events were reported. 5,577 patients.

The Boulware study provides a breakdown for treatment delay. For the case of < ~4 days (2 days enrollment, ~46 hours shipping), the result of the meta analysis becomes RR 0.68, p=0.0097.

The actual effect may be larger due to treatment delays, followup loss, protocol deviation, active placebos, no severity analysis for cases, and suboptimal regimens.

For the individual studies see [1, 2, 3, 4, 5].

[1] c19study.com/boulware.html

[2] c19study.com/mitjapep.html

[3] c19study.com/rajasingham.html

[4] c19study.com/skipper.html

[5] c19study.com/mitja.html

combined cases/death/hospitalization, ↓24.0%, p=0.03

Abella et al., JAMA Internal Medicine, doi:doi:10.1001/jamainternmed.2020.6319 (Peer Reviewed)

Efficacy and Safety of Hydroxychloroquine vs Placebo for Pre-exposure SARS-CoV-2 Prophylaxis Among Health Care Workers

Very small early-terminated underpowered PrEP RCT with 64/61 HCQ/control patients and only 8 infections, HCQ infection rate 6.3% versus control 6.6%, RR 0.95 [0.25 - 3.64].

There was no hospitalization or death, no significant difference in QTc, no severe adverse events, no cardiac events (e.g., syncope and arrhythmias) observed. Medication adherence was 81%. Therapeutic levels of HCQ may not have been reached by the time of the infection in the first week.

2 infections were reported to be after discontinuation of the medication, but the authors do not specify which arm these were in. Hypothetically, if these were both in the HCQ arm, the resulting RR for treatment would be much lower.

COVID-19 case, ↓5.0%, p=1.00

Lammers et al., Int. J. Infectious Diseases, doi:10.1016/j.ijid.2020.09.1460 (Peer Reviewed)

Early hydroxychloroquine but not chloroquine use reduces ICU admission in COVID-19 patients

Observational study 1,064 hospitalized patients in the Netherlands, 53% reduced risk of transfer to the ICU for mechanical ventilation with HCQ treatment starting on the first day of admission.

Weighted propensity score adjusted hazard ratio for transfer to the ICU with HCQ treatment, HR = 0.47, p = 0.008. For CQ, HR = 0.8, p = 0.207. Mortality results in this study are only for mortality before transfer to the ICU. The combined ICU/death HR was 0.68, p = 0.024 for HCQ, and 0.85, p = 0.224 for CQ.

Observational, multicenter, cohort study of hospitalized COVID-19 patients. 189 HCQ patients, 377 CQ, 498 control.

combined death/ICU, ↓32.0%, p=0.02

Dabbous et al., Research Square, doi:10.21203/rs.3.rs-83677/v1 (Preprint)

A Randomized Controlled Study Of Favipiravir Vs Hydroxychloroquine In COVID-19 Management: What Have We Learned So Far?

Small RCT comparing HCQ and favipiravir, with 50 patients in each arm, finding that 55.1% of HCQ patients were PCR negative on day 7 compared to 48% for favipiravir, p = 0.7. There was no comparison with a control group.

Luco, J., ResearchGate, doi:10.13140/RG.2.2.24214.98880 (Preprint) (meta analysis - not included in study count)

Hydroxychloroquine as Post-Exposure Prophylaxis for Covid-19: Why simple data analysis can lead to the wrong conclusions from well-designed studies

Reanalysis of Boulware et al. PEP trial data showing statistically significant improvements with HCQ.

Gasperetti et al., EP Europace, doi:10.1093/europace/euaa216 (Peer Reviewed) (not included in the study count)

Arrhythmic safety of hydroxychloroquine in COVID-19 patients from different clinical settings

Safety study of 649 patients finding that HCQ administration is safe for short-term treatment for patients with COVID-19 infection regardless of the clinical setting of delivery, causing only modest QTc prolongation and no directly attributable arrhythmic deaths.

Arrhythmic safety data from a large cohort of patients treated with HCQ alone or in combination with other QT-prolonging drugs.

Shoaibi et al., medRxiv, doi:10.1101/2020.09.23.20199463 (Preprint)

Comparative Effectiveness of Famotidine in Hospitalized COVID-19 Patients

Retrospective database analysis focused on Famotidine but also showing results for HCQ users, with unadjusted mortality RR 0.85, p<0.001 (13.6% vs. 16.1%).

death, ↓15.4%, p<0.001

Ulrich et al., Open Forum Infectious Diseases, doi:10.1093/ofid/ofaa446 (Peer Reviewed)

Treating Covid-19 With Hydroxychloroquine (TEACH): A Multicenter, Double-Blind, Randomized Controlled Trial in Hospitalized Patients

Small RCT on very late stage use of HCQ, with 48% on oxygen at baseline. 67 HCQ patients, 61 control.

Baseline states were not comparable - 82% more HCQ patients had the highest severity at baseline, there was 32% more male HCQ patients, and 44% more control patients used AZ. The HCQ group also had significantly more patients with cerebrovascular disease, cardiovascular disease (non-hypertension), renal disease (non-dialysis), and a history of organ transplants.

30 day mortality RR 1.06, p = 1.0.

death, ↑6.0%, p=1.00

Serrano et al., Ann. Oncol., 2020, Sep, 31, S1026, doi:10.1016/j.annonc.2020.08.1830 (Peer Reviewed)

COVID-19 and lung cancer: What do we know?

Small retrospective study of 22 lung cancer patients, 14 treated with HCQ+AZ, showing HCQ+AZ mortality relative risk RR 0.57, p = 0.145.

death, ↓43.0%, p=0.14

Gentry et al., Lancet Rheumatology, doi:10.1016/S2665-9913(20)30305-2 (Peer Reviewed)

Long-term hydroxychloroquine use in patients with rheumatic conditions and development of SARS-CoV-2 infection: a retrospective cohort study

Retrospective patients with rheumatologic conditions showing zero of 10,703 COVID-19 deaths for HCQ patients versus 7 of 21,406 propensity matched control patients (not statistically significant). The average age of HCQ patients is slightly lower 64.8 versus 65.4 control.

COVID-19 cases OR 0.79, p=0.27. There are several significant differences in the propensity matched patients that could affect results, e.g., 20.9% SLE versus 24.7%.

death, ↓91.3%, p=0.10, COVID-19 mortality within all patients
death, ↓90.7%, p=0.19, mortality for infected patients
COVID-19 case, ↓20.9%, p=0.27
(adjusted for zero (with reciprocal of the contrasting arm))
(odds ratio converted to relative risk)

Rajasingham et al., medRxiv, doi:10.1101/2020.09.18.20197327 (Peer Reviewed)

Hydroxychloroquine as pre-exposure prophylaxis for COVID-19 in healthcare workers: a randomized trial

PrEP RCT showing HR 0.73, p = 0.12.

Trial halted after 47% enrollment, p < 0.05 will be reached at ~75% enrollment if similar results continue. Trial halted due to low enrollment. US prescriptions only stopped briefly, resuming in June, so the authors can complete the study [1].

HR 0.66/0.68 for full medication adherence, 0.72/0.74, p = 0.18/0.22 overall (1x/2x dosing).

Efficacy for first responders was higher, OR 0.32, p = 0.01. First responders had a much higher incidence, allowing greater power, and reducing the effect of confounders such as misdiagnosis of other conditions or survey issues.

Performance is similar to placebo for the first 3 weeks. The effect may be greater with a dosage regimen that achieves therapeutic levels faster [2]. ~40% of participants suspected they might have had COVID-19 before the trial, the effect in people without prior COVID-19 may be higher.

Authors note:

- the trial was underpowered
- investigation into more frequent dosing may be warranted
- insufficient dosing with no participants achieving more than the in vitro EC₅₀

Internet survey RCT subject to survey bias. There was no death or ICU admission. Low risk healthcare workers, median age ~40. 494 1x/week dosing, 495 2x/week dosing, 494 control participants (1x and 2x participants received the same overall dosage).

[1] wattsupwiththat.com/2020/08/24/hydroxychloroquine-in-covid-19-treatment-actual-usage-in-the-usa/

[2] tandfonline.com/doi/full/10.1080/00498254.2020.1824301

hospitalization, ↓50.1%, p=1.00
COVID-19 case, ↓27.0%, p=0.12

Grau-Pujol et al., Research Square, doi:10.21203/rs.3.rs-72132/v1 (Preprint)

Pre-exposure prophylaxis with hydroxychloroquine for COVID-19: initial results of a double-blind, placebo-controlled randomized clinical trial

Small PrEP RCT showing that PrEP with HCQ is safe at the dosage used. No deaths, hospitalizations, or serious adverse events. With only one case (in the placebo arm), efficacy was not evaluated.

The abstract appears to have a typo and does not match the results section. We believe the results section is correct with one case in the control arm - the 0.8% matches the percentage for the control arm (there is no possible number of cases for the combined arms that matches 0.8%). Note that table 2 refers to risk factors, not outcomes.

COVID-19 case, ↓67.9%, p=0.47
(adjusted for zero (with reciprocal of the contrasting arm))

Lofgren et al., Open Forum Infectious Diseases, doi:10.1093/ofid/ofaa500 (preprint 9/21) (Peer Reviewed) (not included in the study count)

Safety of Hydroxychloroquine among Outpatient Clinical Trial Participants for COVID-19

Analysis of 2,795 outpatients not showing significant safety concerns with HCQ. No deaths were related to HCQ. There was one serious event requiring hospitalization, identical to the frequency with placebo.

Axfors et al., medRxiv, doi:10.1101/2020.09.16.20194571 (Preprint) (meta analysis - not included in study count)

Mortality outcomes with hydroxychloroquine and chloroquine in COVID-19: an international collaborative meta-analysis of randomized trials

Meta analysis assigning 87% weight to the RECOVERY trial, producing the same result. The RECOVERY trial used excessively high non-patient-customized dosage in very sick late stage patients and the result is not generalizable to typical dosage or earlier treatment.

Karatza et al., Xenobiotica (Peer Reviewed) (not included in the study count)

Optimization of hydroxychloroquine dosing scheme based on COVID-19 patients’ characteristics: a review of the literature and simulations

Analysis of HCQ dosing, suggesting that high initial doses followed by low and sparse doses may offer significant benefits to patients by decreasing the viral load without reaching levels considered to produce adverse effects.

For instance, the dosing scheme proposed for a 70kg adult with moderate COVID-19 symptoms would be 600mg upon diagnosis, 400mg after 12h, 300mg after 24h, 200mg after 36h, followed by 200mg BID for 4 days, followed by 200mg OD for 5 days.

Suboptimal dosing regimens that do not fully account for the long half-life of HCQ or the patient characteristics are likely contribute to either limited efficacy where therapeutic levels take too long to reach, or significant adverse effects due to excessive dosage.

Ashinyo et al., Pan African Medical Journal, 37:1, doi:10.11604/pamj.supp.2020.37.1.25718 (Peer Reviewed)

Clinical characteristics, treatment regimen and duration of hospitalization among COVID-19 patients in Ghana: a retrospective cohort study

Retrospective 307 hospital patients in Ghana showing 33% reduction in hospitalization time with HCQ, 29% reduction with HCQ+AZ, and 37% reduction with CQ+AZ.

hospitalization time, ↓33.0%, p=0.03

Lauriola et al., Clinical and Translational Science, doi:10.1111/cts.12860 (Peer Reviewed)

Effect of combination therapy of hydroxychloroquine and azithromycin on mortality in COVID‐19 patients

Retrospective 377 patients, 73% reduction in mortality with HCQ+AZ, adjusted hazard ratio HR 0.27 [0.17-0.41]. Mean age 71.8. No serious adverse events. Subject to incomplete adjustment for confounders.

death, ↓73.5%, p<0.001

Sulaiman et al., medRxiv, doi:10.1101/2020.09.09.20184143 (Preprint)

The Effect of Early Hydroxychloroquine-based Therapy in COVID-19 Patients in Ambulatory Care Settings: A Nationwide Prospective Cohort Study

Observational prospective 5,541 patients, adjusted HCQ mortality odds ratio OR 0.36, p = 0.012. Adjusted hospitalization OR 0.57, p < 0.001. Zinc supplementation was used in all cases. Early treatment in ambulatory fever clinics in Saudi Arabia.

death, ↓63.7%, p=0.01
hospitalization, ↓38.6%, p=0.001
(odds ratio converted to relative risk)

Heberto et al., IJC Heart & Vasculature, doi:10.1016/j.ijcha.2020.100638 (Peer Reviewed)

Implications of myocardial injury in Mexican hospitalized patients with coronavirus disease 2019 (COVID-19)

Observational prospective 254 hospitalized patients, HCQ+AZ mortality odds ratio OR 0.36, p = 0.04.

Ventilation OR 0.20, p = 0.008.

death, ↓53.6%, p=0.04
ventilation, ↓65.6%, p=0.008
(odds ratio converted to relative risk)

Alamdari et al., Tohoku J. Exp. Med., 2020, 252, 73-84, doi:10.1620/tjem.252.73 (Peer Reviewed)

Mortality Risk Factors among Hospitalized COVID-19 Patients in a Major Referral Center in Iran

Retrospective 459 patients in Iran 93% using HCQ, showing HCQ mortality RR 0.45, p = 0.028. HCQ was the only antiviral that showed a significant difference. Small number of control patients and subject to confounding by indication. Average admission delay 5.72 days.

death, ↓55.0%, p=0.03

Kirenga et al., BMJ Open Respiratory Research, doi:10.1136/bmjresp-2020-000646 (Peer Reviewed)

Characteristics and outcomes of admitted patients infected with SARS-CoV-2 in Uganda

Prospective 56 patients in Uganda, 29 HCQ and 27 control, showing 25.6% faster recovery with HCQ, 6.4 vs. 8.6 days (p = 0.20). There was no ICU admission, mechanical ventilation, or death.

Treatment delay is not specified but at least a portion of patients appear to have been treated early.

median time to recovery, ↓25.6%, p=0.20

Rentsch et al., The Lancet Rheumatology, doi:10.1016/S2665-9913(20)30378-7 (preprint 9/9, https://www.medrxiv.org/content/10.1101/2020.09.04.20187781v1) (Peer Reviewed)

Effect of pre-exposure use of hydroxychloroquine on COVID-19 mortality: a population-based cohort study in patients with rheumatoid arthritis or systemic lupus erythematosus using the OpenSAFELY platform

Observational database study of RA/SLE patients in the UK, 194,637 RA/SLE patients with 30,569 having >= 2 HCQ prescriptions in the prior 6 months, HCQ HR 1.03 [0.80-1.33] (HR 0.78 before adjustments).

70 patients with HCQ prescriptions died. One major problem is that there is no knowlege of medication adherence for these 70 - for example, it is possible that they were part of the expected percentage of patients that did not take the medication as prescribed, invalidating the result. Other limitations include confounding by use of bDMARDs and confounding by severity of rheumatological disease.

death, ↑3.0%, p=0.83

Laplana et al., PLOS ONE, doi:10.1371/journal.pone.0243598 (Peer Reviewed)

Lack of protective effect of chloroquine derivatives on COVID-19 disease in a Spanish sample of chronically treated patients

Survey of 319 autoimmune disease patients taking CQ/HCQ with 5.3% COVID-19 incidence, compared to a control group from the general population (matched on age, sex, and region, but not adjusted for autoimmune disease), with 3.4% incidence.

It not clear why authors did not compare with autoimmune patients not on CQ/HCQ. Other research shows that the risk of COVID-19 for systemic autoimmune disease patients is much higher overall, Ferri et al. show OR 4.42, p<0.001 [1], which is the observed real-world risk, taking into account factors such as these patients potentially being more careful to avoid exposure. If we adjust for the different baseline risk, the result becomes RR 0.36, p<0.001, suggesting a substantial benefit for HCQ/CQ treatment (as shown in other studies).

There may also be significant survey bias - those experiencing COVID-19 may be more likely to respond to the survey.

Authors note that they "could not eliminate completely the possibility of some bias due to the intrinsic condition of the individuals within the treatment group that are undergoing chloroquine or derivative drug treatment due to other diseases that alter their health status and may have different comorbidities", however they could account for one significant bias by comparing with matched autoimmune disease patients.

[1] c19study.com/ferri.html

COVID-19 case, ↑56.0%, p=0.24

IHU, Expert Review of Clinical Immunology (Review) (Peer Reviewed) (not included in the study count)

Natural history and therapeutic options for COVID-19

Review of the current state of knowledge regarding the natural history of and therapeutic options for COVID-19.

Treatment with an oral combination of hydroxychloroquine, azithromycin and zinc may represent the best current therapeutic option in relation to its antiviral and
immunomodulatory effects.

Synolaki et al., medRxiv, doi:10.1101/2020.09.05.20184655 (Preprint)

The Activin/Follistatin-axis is severely deregulated in COVID-19 and independently associated with in-hospital mortality

Retrospective 117 patients, 58 HCQ showing lower mortality for HCQ patients.

Version 1 of this paper stated: "HCQ, AZ, [and ...] were found to be independently associated with survival when treatment commenced at FACTCLINYCoD scores <3".

death, ↓23.6%, p=0.27

Furtado et al., The Lancet, doi:10.1016/S0140-6736(20)31862-6 (Peer Reviewed)

Azithromycin in addition to standard of care versus standard of care alone in the treatment of patients admitted to the hospital with severe COVID-19 in Brazil (COALITION II): a randomised clinical trial

Small RCT comparing the addition of AZ for very late stage patients on ventilation or oxygen. No significant difference was found, OR 1.36, p=0.11. One notable result is that even within this extremely late stage population, results suggest increased efficacy with the addition of AZ for patients with earlier use of AZ/HCQ, OR 0.71, p=0.28.

Patients received 8g of HCQ over 10 days, approaching the high levels used in the RECOVERY trial (9.2g over 10 days), showing significantly more adverse events than typical trials.

Since all patients were on HCQ, this study does not provide information on the efficacy of HCQ.

Wang et al., Phytomedicine, doi:10.1016/j.phymed.2020.153333 (Peer Reviewed) (In Vitro) (not included in the study count)

Chloroquine and hydroxychloroquine as ACE2 blockers to inhibit viropexis of 2019-nCoV Spike pseudotyped virus

In Vitro study providing novel insights into the molecular mechanism of CQ/HCQ treatment, showing that CQ and HCQ both inhibit the entrance of 2019-nCoV into cells by blocking the binding of the virus with ACE2.

Heras et al., European Geriatric Medicine, doi:10.1007/s41999-020-00432-w (preprint 9/2) (Peer Reviewed)

COVID-19 mortality risk factors in older people in a long-term care center

Retrospective 100 elderly nursing home patients, HCQ+AZ mortality 11.4% vs. control 61.9%, RR 0.18, p<0.001. Median age 85.

COVID-19 confirmed. 70% treated with HCQ+AZ. Details of differences between groups are not provided, and no adjustments are made. It is not clear how the groups were selected. Authors indicate treatment was early but do not specify the treatment delay.

death, ↓95.6%, p=0.004

de la Iglesia et al., medRxiv, doi:10.1101/2020.08.31.20185314 (Preprint)

Hydroxicloroquine for pre-exposure prophyylaxis for SARS-CoV-2

Analysis of autoimmune disease patients on HCQ, compared to a control group from the general population (matched on age and sex, but not adjusted for autoimmune disease), showing non-significant differences between groups.

Other research shows that the risk of COVID-19 for systemic autoimmune disease patients is much higher overall, Ferri et al. show OR 4.42, p<0.001 [1], which is the observed real-world risk, taking into account factors such as these patients potentially being more careful to avoid exposure.

If we adjust for the different baseline risk, the mortality result becomes RR 0.35, p=0.23, suggesting a substantial benefit for HCQ treatment (as shown in other studies).

[1] c19study.com/ferri.html

hospitalization, ↑50.0%, p=1.00
COVID-19 case, ↑42.6%, p=0.15, suspected COVID-19
COVID-19 case, ↓7.8%, p=0.84, confirmed COVID-19

Hecel et al., Pharmaceuticals, 13:9, 228, doi:10.3390/ph13090228 (Review) (Peer Reviewed) (not included in the study count)

Zinc(II)—The Overlooked Éminence Grise of Chloroquine’s Fight against COVID-19?

Review of zinc as an inhibitor of SARS-CoV-2′s RNA-dependent RNA polymerase, and zinc ionophores including CQ/HCQ, showing the latest evidence for zinc and CQ/HCQ having antiviral, and in particular anticoronaviral action.

Elbazidi et al., New Microbes and New Infections, doi:10.1016/j.nmni.2020.100749 (Peer Reviewed)

Pandemic and social changes, political fate

Analysis of US states and countries. Country analysis shows a significant correlation between the dates of decisions to adopt/decline HCQ, and corresponding trend changes in CFR. US state analysis shows a significant correlation between CFR and the level of acceptance of HCQ.

Castillo et al., Journal of Steroid Biochemistry and Molecular Biology, 203, October 2020, doi:10.1016/j.jsbmb.2020.105751 (Peer Reviewed)

Effect of calcifediol treatment and best available therapy versus best available therapy on intensive care unit admission and mortality among patients hospitalized for COVID-19: A pilot randomized clinical study

RCT on calcifediol (25-hydroxyvitamin D) treatment for hospitalized COVID-19 patients showing significantly reduced intensive care unit admissions. All patients received standard care including HCQ+AZ. Significantly lower ICU admission with the addition of calcifediol - adjusted odds ratio 0.03 [0.003-0.25]. No deaths for calcifediol (0/50), 2 deaths for SOC (2/26).

Fried et al., Clinical Infectious Disease, doi:10.1093/cid/ciaa1268 (Peer Reviewed)

Patient Characteristics and Outcomes of 11,721 Patients with COVID19 Hospitalized Across the United States

Database analysis of 11,721 hospitalized patients, 4,232 on HCQ. Strong evidence for confounding by indication and compassionate use of HCQ. 24.9% of HCQ patients were on mechanical ventilation versus 12.2% control. Ventilation mortality was 70.5% versus 11.6%.

This study does not adjust for the differences in comorbid conditions and disease severity, and therefore does not make a conclusion. Unadjusted HCQ mortality was 24.8% versus control 19.6%. Adjusting for ventilation only gives us 17.7% HCQ versus 19.6% control (adjusting the HCQ group to have the same proportion of ventilation patients), RR 0.90. Hopefully authors can do a full adjustment analysis. Comorbidities may favor control, while patients remaining in the hospital (5.3%) may favor HCQ (other studies show faster resolution for HCQ patients).

Data inconsistencies have been found in this study, for example 99.4% of patients treated with HCQ were treated in urban hospitals, compared to 65% of untreated patients (Supplemental Table 3), while patients are distributed in a more balanced manner between teaching or not-teaching hospitals, as well as in the most urbanized (Northeast) and less urbanized (Midwest) regions of the United States [1].

[1] academic.oup.com/cid/advance-article/doi/10.1093/cid/ciaa1618/5934822

death, ↑27.0%, p<0.001

Ferri at al., Clinical Rheumatology, doi:0.1007/s10067-020-05334-7 (Peer Reviewed)

COVID-19 and rheumatic autoimmune systemic diseases: report of a large Italian patients series

Analysis of 1641 systemic autoimmune disease patients showing csDMARD (HCQ etc.) RR 0.37, p=0.015.

csDMARDs include HCQ, CQ, and several other drugs, so the effect of HCQ/CQ alone could be higher.

This study also confirms that the risk of COVID-19 for systemic autoimmune disease patients is much higher overall, OR 4.42, p<0.001 (this is the observed real-world risk which takes into account factors such as these patients potentially being more careful to avoid exposure).

(results are for "definite + highly suspected" cases and the main result is presented in the paper as the OR for not taking csDMARDs, we have converted this to RR for taking csDMARDs).

risk of COVID-19 case, ↓63.0%, p=0.01

Fiolet et al., Clinical Microbiology and Infection (Peer Reviewed) (meta analysis - not included in study count)

Effect of hydroxychloroquine with or without azithromycin on the mortality of COVID-19 patients: a systematic review and meta-analysis

Meta analysis of late stage studies (and one early treatment study with only 2 deaths), showing HCQ RR 0.83 [0.65-1.06], before exclusions RR 0.80 [0.65-1.0].

Authors claim "HCQ alone is not effective", but the result directly contradicts this, RR 0.83 [0.65-1.06], i.e., inconclusive but much more likely to be effective than not.

There are many errors in this meta analysis which introduce critical bias, for example:

- Very biased sample of studies, including <4% of early treatment studies (only 1), and <30% of late treatment studies, focused on negative studies.

- Arshad et al. (propensity matched HR 0.49, p=0.009) was excluded because the authors claim a "critical" risk of confounding bias due to steroid use, however steroids were controlled for in the multivariate and propensity analyses [1].

- For Skipper et al., authors use an RR of 1.01, however the study had one hospitalized control death and one non-hospitalized HCQ death. Since the HCQ death was non-hospitalized, it may not be caused by COVID-19, or the patient did not receive standard care, therefore this should not be treated as equal to the control death. Further, medication adherence was only 77%, the HCQ patient may not have taken the medication (Skipper et al. neglects to answer this question). In any case, including a trial with only 1-2 deaths is likely to increase bias.

- Cavalcanti et al. received the lowest bias rating, despite having treatment delayed up to 14 days after symptoms, randomizing 14% of patients in the ICU, having significant protocol deviations, unusually low medication adherence, randomization that resulted in 64.3% male patients (HCQ) vs. 54.2% (control), and excluding patients already receiving longer and potentially therapeutic doses of the study treatments.

- Sbidian el al. received the lowest bias rating, however many more control patients are still in hospital at 28 days suggesting there will be a significant improvement when extending past 28 days.

- The RECOVERY trial received the lowest bias rating, despite using a very high dose likely responsible for the increased mortality. Results of this trial are not relevant to use at normal dosages.

- Inclusion criteria required RT-PCR confirmed cases, but this was disregarded when including Horby et al. (very negative, excessive dose) and Skipper et al.

- Authors do not consider different treatment delays, risk level of patients, differences in dosage, or usage of Zinc.

Also see many other reports of problems and fatal flaws: [2, 3, 4, 5, 6]

This analysis is also missing several recent studies, for a more up-to-date analysis see [7].

[1] ijidonline.com/article/S1201-9712(20)30604-4/fulltext

[2] clinicalmicrobiologyandinfection.com/article/S1198-743X(20)30649-2/fulltext

[3] mediterranee-infection.com/wp-content/uploads/2020/04/Response-to-Fiolet-et-al-Manuscript.pdf

[4] francesoir.fr/societe-sante/covid-19-les-anti-hydroxychloroquine-et-une-certaine-science-francaise..

[5] twitter.com/Covid19Crusher/status/1299746474478243841

[6] twitter.com/Covid19Crusher/status/1321426322049142791

[7] c19study.com/ihu.html

Ip et al., BMC Infectious Diseases, doi:10.1186/s12879-021-05773-w (preprint 8/25) (Peer Reviewed)

Hydroxychloroquine in the treatment of outpatients with mildly symptomatic COVID-19: A multi-center observational study

Retrospective 1,274 outpatients, 47% reduction in hospitalization with HCQ with propensity matching, HCQ OR 0.53 [0.29-0.95]. Sensitivity analyses revealed similar associations.

Adverse events were not increased (2% QTc prolongation events, 0% arrhythmias).

death, ↓54.5%, p=0.43
ICU admission, ↓28.6%, p=0.79
hospitalization, ↓37.3%, p=0.04
(odds ratio converted to relative risk)

Di Castelnuovo et al., European J. Internal Medicine, doi:10.1016/j.ejim.2020.08.019 (Peer Reviewed)

Use of hydroxychloroquine in hospitalised COVID-19 patients is associated with reduced mortality: Findings from the observational multicentre Italian CORIST study

Retrospective 3,451 hospitalized patients, 30% reduction in mortality with HCQ after propensity adjustment, HR 0.70 [0.59 - 0.84].

death, ↓30.0%, p<0.0001

Catteau et al., Int. J. Antimicrobial Agents, doi:10.1016/j.ijantimicag.2020.106144 (Peer Reviewed)

Low-dose Hydroxychloroquine Therapy and Mortality in Hospitalized Patients with COVID-19: A Nationwide Observational Study of 8075 Participants

Retrospective 8,075 hospitalized patients, 4,542 low-dose HCQ, 3,533 control. 35% lower mortality for HCQ (17.7% vs. 27.1%), adjusted HR 0.68 [0.62–0.76]. Low-dose HCQ monotherapy was independently associated with lower mortality in hospitalized patients.

Patients exposed to others therapies (TCZ, AZ, LPV/RTV) were excluded.

Statistical analysis was performed by an independent group. Calendar time of prescription and immortal time bias was taken into account. Corticosteroids prescriptions was low in both groups.

death, ↓32.0%, p<0.0001

Ly et al., International Journal of Antimicrobial Agents, doi:10.1016/j.ijantimicag.2020.106219 (preprint 8/21) (Peer Reviewed)

Pattern of SARS-CoV-2 infection among dependant elderly residents living in retirement homes in Marseille, France, March-June 2020

Retrospective analysis of retirement homes, HCQ+AZ >= 3 days mortality OR 0.37, p=0.02. 1690 elderly residents (mean age 83), 226 infected residents, 116 treated with HCQ+AZ >= 3 days.

Detection via mass screening also showed significant improvements (16.9% vs. 40.6%, OR 0.20, p=0.001), suggesting that earlier detection and treatment is more successful.

death, ↓55.6%, p=0.02
(odds ratio converted to relative risk)

Lane et al., The Lancet Rheumatology, doi:10.1016/S2665-9913(20)30276-9 (Peer Reviewed) (not included in the study count)

Risk of hydroxychloroquine alone and in combination with azithromycin in the treatment of rheumatoid arthritis: a multinational, retrospective study

Retrospective study of RA patients using HCQ vs. sulfasalazine (another DMARD). HCQ treatment showed no increased risk in the short term (up to 30 days) among patients with RA. Long term use was associated with excess cardiovascular mortality.

Addition of AZ increased the risk of cardiovascular mortality with combined use up to 30 days. This is several times longer than typical COVID-19 use. This result also comes from just 2 of the 14 databases, with the negative result from just one database (VA) and much lower statistically insignifant difference in mortality from the other database (Clinformatics).

Confounding by indication. Patients conditions vary, the severity of a patient's RA or other conditions was not taken into account. Results varied widely across different databases, and different subsets of databases were used in different analyses. Baseline risk of serious adverse events unknown. Health care database analysis subject to misclassification errors.

Gonzalez et al., medRxiv, doi:10.1101/2020.08.18.20172874 (Preprint)

The Prognostic Value of Eosinophil Recovery in COVID-19: A Multicentre, Retrospective Cohort Study on Patients Hospitalised in Spanish Hospitals

Retrospective study focused on eosinophil recovery with 9,644 hospitalized patients in Spain, showing lower mortality for HCQ (14.7% vs 29.2%, p<0.001), and AZ (15.3% vs. 18.4%, p<0.001). With a multivariate model including potential confounding factors, HCQ and AZ are associated with lower mortality, HCQ OR 0.662, p=0.057.

death, ↓26.6%, p=0.06
(odds ratio converted to relative risk)

Dubernet et al., J. Global Antimicrobial Resistance, doi:10.1016/j.jgar.2020.08.001 (Peer Reviewed)

A comprehensive strategy for the early treatment of COVID-19 with azithromycin/hydroxychloroquine and/or corticosteroids: results of a retrospective observational study in the French overseas department of Reunion Island

Retrospective analysis of 36 hospitalized patients showing HCQ/AZ associated with lower ICU admission, p=0.008. Median age 66, no mortality. Confounding by indication, however it was patients with hypoxemic pneumonia that were treated with HCQ/AZ, patients were not treated with HCQ/AZ if they didn't need oxygen therapy.

ICU admission, ↓87.6%, p=0.008

Prodromos, C., New Microbes and New Infections, doi:10.1016/j.nmni.2020.100747 (Peer Reviewed) (not included in the study count)

Hydroxychloroquine is protective to the heart, not harmful: A systematic review

Review concluding that HCQ/AZ does not cause Torsade de Pointes or related deaths, HCQ decreases cardiac events, and HCQ should not be restricted in use for COVID-19 patients because of fear of cardiac mortality.

Pinato et al., Cancer Discovery, doi:10.1158/2159-8290.CD-20-0773 (Peer Reviewed)

Clinical portrait of the SARS-CoV-2 epidemic in European cancer patients

Restrospective 890 cancer patients with COVID-19, adjusted mortality HR for HCQ/CQ 0.41, p<0.0001.

Confirmed SARS-CoV-2 infection was required, which may help focus on more severe cases. Analysis with Cox proportional hazard model. Potential unmeasured confounders.

death, ↓59.0%, p<0.0001

Peters et al., Clinical Microbiology and Infection, doi:10.1016/j.cmi.2020.10.004 (preprint 8/15) (Peer Reviewed)

Outcomes of Persons With COVID-19 in Hospitals With and Without Standard Treatment With (Hydroxy)chloroquine

Retrospective study of HCQ use in 9 hospitals in the Netherlands, showing no significant difference in mortality with HCQ/CQ or dexamethasone. Late stage (admitted to hospital with positive test or CT scan abnormalities). 4 of 7 hospitals started treatment only after further deterioration. Short cutoff (21 days) - other studies have shown treated patient cases resolved faster and more control patients remaining in hospital at this time.

In the preprint, 58 of 341 control patients died. In the journal version, 53 of 353 control patients died.

Significant differences between hospitals - HCQ hospitals had significantly older patients with significantly more comorbidities. Non-HCQ hospitals were "tertiary academic centres" whereas HCQ hospitals were "secondary care hospitals". Residual confounding likely. This study compares overcrowded regular hospitals with undercrowded academic hospitals.

A subset of patients were excluded due to transfer to other hospitals. This introduces bias because patients in critical condition are not transferred. For examples, patients benefiting from HCQ treatment may have been transferred to the tertiary centres and excluded from analysis, increasing the percentage of critical cases in the secondary hospitals.

Among the seven (H)CQ-hospitals, the timing of start of (H)CQ treatment differed; three hospitals started at the moment of COVID-19 diagnosis, four started after diagnosis but only when patients clinically deteriorated e.g., when there was an increase in respiratory rate or increase in use of supplemental oxygen.

Most patients received CQ instead of the safer HCQ, receiving late treatment with CQ. Patients were given an initial dose of 600mg CQ then every 12 hours, for 5 days a dose of 300 mg, for a total of 3600mg CQ. This dose is likely to be toxic, see for example [1].

Authors mention a subset of hospitals started treatment relatively earlier, which seems like the most important area to analyze, but no results are provided.

[1] apps.who.int/iris/bitstream/handle/10665/65773/WHO_MAL_79.906.pdf

death, ↑9.0%, p=0.57

Abd-Elsalam et al., American Journal of Tropical Medicine and Hygiene, 10.4269/ajtmh.20-0873 (Peer Reviewed)

Hydroxychloroquine in the Treatment of COVID-19: A Multicenter Randomized Controlled Study

Small RCT in Egypt with 97/97 HCQ/control patients, showing 58% more recovery @28days for HCQ (53.6% HCQ, 34% control), p=0.009 (0.06 in the paper refers to the 5 combined recovery/death/ICU values).

No significant difference in ventilation and mortality (<=6 examples in each case). Authors note the "sample size was not adequately powered for [the] survival endpoint".

Other studies have also shown treated patient cases resolved faster. Continuing analysis past 28 days would be useful. Group characteristics are given, with for example 36% vs. 26% smokers, but they do not identify which group is which. Group 1 and 2 have 97 patients but the total given is 175.

death, ↑20.0%, p=1.00
risk of no recovery at day 28, ↓30.0%, p=0.009

Roomi et al., J. Medical Internet Research, doi:10.2196/21758 (Peer Reviewed)

Efficacy of hydroxychloroquine and tocilizumab in patients with COVID-19: A single-center retrospective chart review

Retrospective 176 hospitalized patients (144 HCQ, 32 control) showing no significant differences with HCQ or TCZ. Confounding by indication.

death, ↑37.7%, p=0.54
(odds ratio converted to relative risk)

Bakhshaliyev et al., J. Electrocardiology, doi:10.1016/j.jelectrocard.2020.08.008 (Peer Reviewed) (not included in the study count)

The effect of 5-day course of hydroxychloroquine and azithromycin combination on QT interval in non-ICU COVID19(+) patient

Safety study of 109 patients showing 5 days of HCQ+AZ did not lead to clinically significant QT prolongation or other conduction delays compared to baseline ECG in non-ICU patients.

Saleemi et al., medRxiv, doi:10.1101/2020.08.05.20151027 (Preprint)

Time to negative PCR from symptom onset in COVID-19 patients on Hydroxychloroquine and Azithromycin - A real world experience

Retrospective 65 HCQ+AZ, 20 control patients, showing median time to negative PCR of 23 days for HCQ+AZ vs. 19 days for control. Confounding by indication. 100% of non-HCQ group had mild disease vs. 63% of the HCQ+AZ group. More comorbidities and symptoms in the HCQ+AZ group.

median time to PCR-, ↑21.0%, p<0.05

Lopez et al., Int. J. Antimicrob. Agents, doi:/j.ijantimicag.2020.106136 (Peer Reviewed)

Effects of Hydroxychloroquine on Covid-19 in Intensive Care Unit Patients: Preliminary Results

Small retrospective study of 29 ICU patients comparing those with HCQ plasma concentration within target to those with a concentration below the target value, with no significant differences found. Mortality in the on-target group was 0% versus 17% for the off-target group, mortality relative risk 0.14, p = 0.16.

McCullough et al., The American Journal of Medicine, doi:10.1016/j.amjmed.2020.07.003 (Review) (Peer Reviewed) (not included in the study count)

Pathophysiological Basis and Rationale for Early Outpatient Treatment of SARS-CoV-2 (COVID-19) Infection

Review of pathophysiological principles related to early outpatient treatment and therapeutic approaches including reduction of reinoculation, combination antiviral therapy, immunomodulation, antiplatelet/antithrombotic therapy, and administration of oxygen, monitoring, and telemedicine.

Proposes an algorithm based on age and comorbidities that allows for a large proportion to be monitored and treated at home during self-isolation with the aim of reducing the risks of hospitalization and death.

Watanabe et al., Open Letter (Letter) (meta analysis - not included in study count)

Concerns regarding the misinterpretation of statistical hypothesis testing in clinical trials for COVID-19

Open letter signed by 38 professors and doctors regarding misinterpretation of statistics in HCQ RCTs.

Authors note [1] that data from RCTs for early treatment in outpatients to date actually show favorable effects, especially in high-risk patients such as the elderly, where efficacy was up to three times higher than in young people. Because most samples were made up of young people without comorbidities, the studies were statistically inconclusive with the entire samples. Authors note that instead of the papers reporting this, they incorrectly claim that the treatment had no effect compared to the placebo. “This misinterpretation in statistical tests is well known and explained in most undergraduate books in the field,” says Watanabe. "An article published in Nature last year states that about 51% of the work on clinical trials with this type of result has incorrect conclusions."

[1] veja.abril.com.br/saude/especialistas-contestam-estudos-que-nao-viram-beneficios-na-cloroquina/amp/

Singer et al., Annals of the Rheumatic Diseases, doi:10.1136/annrheumdis-2020-218500 (Letter)

Hydroxychloroquine ineffective for COVID-19 prophylaxis in lupus and rheumatoid arthritis

Comparison of the percentage of SLE/RA patients on immunosuppressants that were taking HCQ, for COVID-19 diagnosis versus other infections or outpatient visits, finding a similar percentage in each case.

No mortality of severity information is provided to determine if HCQ treated patients fared better. No adjustment for concomitant medications or severity.

COVID-19 case, ↑9.0%, p=0.62

Kalligeros et al., Journal of Global Antimicrobial Resistance, doi:10.1016/j.jgar.2020.07.018 (Peer Reviewed)

Hydroxychloroquine use in hospitalised patients with COVID-19: An observational matched cohort study

Small retrospective database analysis of 36 patients receiving HCQ not showing significant differences. Confounding by indication is likely.

death, ↑67.0%, p=0.57

Kamran et al., medRxiv, doi:10.1101/2020.07.30.20165365 (Preprint)

Clearing the fog: Is HCQ effective in reducing COVID-19 progression: A randomized controlled trial

Study of 349 low-risk hospitalized patients with 151 non-consenting or ineligible patients used as controls. SOC included zinc, vitamin C and vitamin D. A statistically significant improvement in PCR negativity is shown at day 7 with HCQ treatment, 52.1% (HCQ) versus 35.7% (control), p=0.001, but no statistically significant difference at day 14, or in progression. Patients were relatively young and there was no mortality. Only 3% of patients had any disease progression and all patients recovered, so there is little if any room for treatment benefit. Progression among higher-risk patients with comorbidities was lower with treatment (12.9% versus 28.6%, p=0.3, very few cases).

Despite the title, this is not an RCT since patients self-selected the arm, or were chosen based on allergies/contraindications. The treatment group had about twice the number of patients with comorbidities. Treatment delay is unknown - it was recorded but not reported in the paper.

Viral load was not measured. As with other studies, PCR may detect non-replicable viral nucleic acid, this is more likely at day 14. Details on the test accuracy are not provided, authors note that RT-PCR sensitivity ranges from 34-80%.

disease progression, ↓5.0%, p=1.00
disease progression, ↓54.8%, p=0.30, with comorbidities
risk of viral+ at day 7, ↓25.5%, p=0.001
risk of viral+ at day 14, ↑10.0%, p=0.52

Berenguer et al., Clinical Microbiology and Infection, doi:10.1016/j.cmi.2020.07.024 (Peer Reviewed)

Characteristics and predictors of death among 4035 consecutively hospitalized patients with COVID-19 in Spain

Retrospective 4035 hospitalized patients in Spain showing reduced mortality with HCQ (data is in the supplementary appendix).

death, ↓61.9%, p<0.0001

Yu et al., Science China Life Sciences, 2020 Aug 3, doi:10.1007/s11427-020-1782-1 (Letter)

Beneficial effects exerted by hydroxychloroquine in treating COVID-19 patients via protecting multiple organs

Retrospective 2,882 patients in China, median age 62, 278 receiving HCQ, median 10 days post hospitalization, showing that HCQ treatment can reduce systemic inflammation and inhibit the cytokine storm, thus protecting multiple organs from inflammatory injuries, such as detoxification in the liver and attenuation of cardiac injury. IL-6 levels significantly reduced after HCQ treatment, p<0.05, and elevated after HCQ withdrawal. The significantly lower dose used here is potentially related to the different observations from the RECOVERY trial results. Authors suggest that treatment should be started as soon as possible.

The 550 patients that were critically ill at baseline are reported in a separate paper. For the non-critically-ill patients at baseline, the proportion of patients that became critically ill was significantly lower for those treated with HCQ.

For the subset of patients that started HCQ treatment early only 1.4% died versus 3.9% for HCQ started late and 9.1% for control patients.

risk of progression to critical, ↓82.5%, p=0.05, baseline critical cohort reported separately in Yu et al.
death, ↓85.0%, p=0.02, HCQ treatment started early vs. non-HCQ

Davido et al., Int. J. Antimicrobial Agents, 2020, doi:10.1016/j.ijantimicag.2020.106129 (Peer Reviewed)

Impact of medical care including anti-infective agents use on the prognosis of COVID-19 hospitalized patients over time

Retrospective of 132 hospitalized patients. HCQ+AZ(52)/AZ(28) significantly reduced death/ICU, HR=0.45, p=0.04. Adjusted for Charlson Comorbidity Index (including age), obesity, O2, lymphocyte count, and treatments. Mean delay from admission to treatment 0.7 days.

combined intubation/hospitalization, ↓55.0%, p=0.04

Sheaff, R., bioRxiv, doi:10.1101/2020.08.02.232892 (Preprint) (In Vitro) (not included in the study count)

A New Model of SARS-CoV-2 Infection Based on (Hydroxy)Chloroquine Activity

In vitro study presenting a new theory on SARS-CoV-2 infection and why HCQ/CQ provides benefits, which potentially explains the observed relationships with smoking, diabetes, obesity, age, and treatment delay, and confirms the importance of accurate dosing. Metabolic analysis revealed HCQ/CQ inhibit oxidative phosphorylation in mitochondria (likely by sequestering protons needed to drive ATP synthase), inhibiting infection and/or slowing replication.

Mazzitelli et al., Travel Medicine and Infectious Disease, 37, doi:10.1016/j.tmaid.2020.101826 (Letter) (not included in the study count)

Apparent inefficacy of hydroxychloroquine combined with azithromycin on SARS-CoV-2 clearance in an incident cohort of geriatric patients with COVID-19

Report on HCQ+AZ use in 41 elderly high-risk patients. 29 of 30 patients with treatment >= 5 days survived. Only 10% were PCR negative after one week, however the Ct value is not specified.

D'Arminio Monforte et al., Int. J. Infectious Diseases, doi:10.1016/j.ijid.2020.07.056 (Letter)

Effectiveness of Hydroxychloroquine in COVID-19 disease: A done and dusted situation?

HCQ+AZ adjusted death HR 0.44, p=0.009. Propensity scores include baseline COVID-19 disease severity, age, gender, number of comorbidities, cardio-vascular disease, duration of symptoms, date of admission, baseline plasma CRP. IPW censoring. Retrospective study of 539 COVID-19 hospitalized patients in Milan, with treatment a median of 1 day after admission. HCQ 197 patients, HCQ+AZ 94, control 92. Control group received various other treatments. Authors excluded people receiving other drugs which could have biased the effect of HCQ when used in combination. Residual confounding is possible (e.g., people with CVD were more frequent in control), however people in the control group were more likely to require mechanical ventilation.

death, ↓34.0%, p=0.12
HCQ+AZ, ↓56.0%, p=0.009

BaŞaran et al., Turk. J. Med. Sci., doi:10.3906/sag-2006-173 (Peer Reviewed)

Outcome of Non-Critical COVID-19 Patients with Early Hospitalization and Early Antiviral Treatment Outside the ICU

Observational study of 174 hospitalized patients in Turkey, median age 45.4, 23 treated with HCQ, 113 with HCQ+AZ, and 32 with regimens including favipiravir. 75% reduction in the median time to clinical improvement for HCQ+AZ vs. FAV, RR 0.25, p<0.001. 83% reduction for HCQ. However, there was significant confounding by indication.

There were no significant adverse events.

Mitjà et al., NEJM, doi:10.1056/NEJMoa2021801 (preprint 7/26) (Peer Reviewed)

A Cluster-Randomized Trial of Hydroxychloroquine as Prevention of Covid-19 Transmission and Disease

Death rate reduced from 0.6% to 0.4%, RR 0.68, not statistically significant due to low incidence (8 control cases, 5 treatment cases).

For positive symptomatic cases, a greater effect is seen for nursing home residents, RR=0.49 [0.21 - 1.17], vs. overall 0.89, possibly because the exposure events are identified faster in this context, versus home exposure where testing of the source may be more delayed. The trial is too small for significance here. If the trend continued this result would be significant at p<0.05 after about 25% more patients were added.

There are 2 groups in this study: PCR+ at baseline (n=314) and PCR- at baseline (n=2000), which should be separated as they are different populations (primary outcome rates 18.6% and 22.2% compared to 3.0% and 4.3%). PCR+ already have COVID-19, so PEP analysis should be for the 2,000 PCR-, showing symptomatic COVID-19 of 4.3% (control) and 3.0% (treatment), RR 0.7, p=0.154.

The paper has different RR values here, stating that they are adjusted for contact-level variables. It is not clear how they are computed - the adjusted RR for the overall sample is 4% lower, for PCR+ it is 20% lower, but for PCR- it is 107% higher, even though PCR- represents 86% of the sample.

Hopefully, supplementary data will provide a breakdown on cases in this PCR- @baseline sample by number of days since exposure, and also provide relevant hospitalisation and death results.

Enrollment was up to 7 days after exposure, median 4 days. Treatment delay is unclear. The exposure event timing is not detailed. It appears to be based on the date of a positive test for a contact, which is likely to be much later than the actual exposure time. 13.1% were already positive at baseline, which is consistent with the actual exposure time being significantly earlier. PCR testing has a very high false-negative rate in early stages (e.g., 100% on day 1, 67% on day 4, and 20% on day 8 [1]), hence it is likely that a much higher percentage were infected at an unknown time before enrollment. Medication administration is not detailed. Sensitivity and specificity of the tests is not provided.

Given the delay identifying index cases, PCR test delay, and PCR false negative rate at early stages, the treatment delay in general was very long and could be over 2 weeks.

The RR for non-PCR positive at baseline is 0.74. Including the PCR-positive at baseline patients reduced this to 0.89. This is also consistent with earlier treatment being more effective.

The paper does not mention zinc. Zinc deficiency in Spain has been reported at 83% [2], this may significantly reduce effectiveness. HCQ is a zinc ionophore which increases cellular uptake, facilitating significant intracellular concentrations of zinc, and zinc is known to inhibit SARS-CoV RNA-dependent RNA polymerase activity, and is widely thought to be important for effectiveness with SARS-CoV-2 [3].

This study focuses on the existence of symptoms or PCR-positive results, however severity of symptoms is more important. Research has shown HCQ concentrations can be much higher in the lung compared to plasma [4], which may help minimize the occurrence of severe cases and death.

There is a treatment-delay response relationship consistent with an effective treatment, however the authors only provide 3 ranges and do not break down the earliest treatment delay times.

The definition of COVID-19 symptoms is very broad - just existence of a headache alone or muscle pain alone was considered COVID-19. There was an overall very low incidence of confirmed COVID-19 (138 cases across both arms). There were no serious adverse events that were adjudicated as being treatment related. Authors exclude those with symptoms in the previous two weeks, however, those with symptoms up to several months before may still test PCR-positive even though there may be no viable virus.

There appears to be incorrect data. Table 2, secondary outcomes, control, hospital/vital records shows that 8 of 1042 is 9.7% (we get 0.8%).

Nasopharyngeal viral load analysis issues include test unreliability and temporo-spatial differences in viral shedding [5].

In summary, this study appears positive in the context of very delayed treatment and very small sample sizes, however we have classified it as inconclusive for now pending further analysis and feedback. Preliminary analysis. Supplementary Appendix is not currently available. Please submit any corrections or comments.

Data from this study has been used to show that viral load is the primary factor in transmission: [6].

[1] acpjournals.org/doi/10.7326/M20-1495

[2] mdpi.com/2072-6643/9/7/697

[3] infezmed.it/index.php/article?Anno=2020&numero=2&ArticoloDaVisualizzare=Vol_28_2_2020_192

[4] ncbi.nlm.nih.gov/pmc/articles/PMC7122276/

[5] journal.chestnet.org/article/S0012-3692(20)31718-9/fulltext

[6] medrxiv.org/content/10.1101/2020.10.27.20220277v1

death, ↓32.0%, p=0.58
baseline pcr- risk of cases, ↓32.0%, p=0.27

Khurana et al., medRxiv, doi:10.1101/2020.07.21.20159301 (Preprint)

Prevalence and clinical correlates of COVID-19 outbreak among healthcare workers in a tertiary level hospital

Study of hospital health care workers showing HCQ prophylaxis reduces COVID-19 significantly, OR 0.30, p=0.02. 94 positive health care workers with a matched sample of 87 testing negative. Full course prophylaxis was important in this study which used a low dose of 400mg/week HCQ (800mg for week 1), so it may take longer to reach therapeutic levels. Actual benefit of HCQ may be larger because severity of symptoms are not considered here but HCQ may also reduce severity.

COVID-19 case, ↓51.0%, p=0.02
(odds ratio converted to relative risk)

Cavalcanti et al., NEJM, July 23, 2020, doi:10.1056/NEJMoa2019014 (Peer Reviewed)

Hydroxychloroquine with or without Azithromycin in Mild-to-Moderate Covid-19

Late stage RCT of 667 hospitalized patients with up to 14 days of symptoms at enrollment and receiving up to 4 liters per minute supplemental oxygen, not finding a significant effect after 15 days.

Authors note: "the trial cannot definitively rule out either a substantial benefit of the trial drugs or a substantial harm", sample sizes are too small.

The paper uses the terms mild and moderate, however all patients had serious enough disease to be hospitalized, and 14% were actually randomized in the ICU.

The trial had significant protocol deviations and unusually low medication adherence. Randomization resulted in 64.3% male patients (HCQ) vs. 54.2% (control) which may significantly affect results due to the much higher risk for male patients.

Authors note: "our aim was to exclude patients already receiving longer and potentially therapeutic doses of the study treatments" in explanation for why the study protocol was changed to exclude patients with previous use of the medications >24hrs. Analyzing these patients rather than excluding them may have revealed effectiveness with early use as shown in other studies.

The trial initially required enrollment within 48 hours of admission and was changed to remove this requirement, this change is likely to reduce effectiveness because enrollment was moved later, compared to the time the disease became serious enough for hospitalization. Total HCQ dosage 5.6g.

A correction for 17 errors has been published: [1]

[1] nejm.org/doi/full/10.1056/NEJMx200021

death, ↓16.0%, p=0.77, HCQ+HCQ/AZ
hospitalization, ↑28.0%, p=0.30, HCQ+HCQ/AZ

Ou et al., PLOS Pathogens, doi:10.1371/journal.ppat.1009212 (preprint 7/22) (Peer Reviewed) (In Vitro)

Hydroxychloroquine-mediated inhibition of SARS-CoV-2 entry is attenuated by TMPRSS2

In Vitro analysis showing that HCQ efficiently blocks viral entry mediated by cathepsin L, but not by TMPRSS2, and that a combination of HCQ and a TMPRSS2 inhibitor prevents SARS-CoV-2 infection more potently than either drug alone.

Hoffmann et al., Nature, (2020), doi:10.1038/s41586-020-2575-3 (Peer Reviewed) (In Vitro) (not included in the study count)

Chloroquine does not inhibit infection of human lung cells with SARS-CoV-2

The title of this paper does not appear to match the results. Fig. 1b @100uM shows CQ results in a ~4.5 fold decrease (on a linear scale) in extracellular virus, p=0.05, after 24 hours (we do not see the supplementary data at this time so this is estimated from the graph). This decrease may continue if examined over longer time periods. Fig. 1a shows a ~45-50% entry inhibition @100uM for HCQ/CQ (p=0.0005/0.0045), ~10-30% @10uM (p=0.13/0.99). Inhibition is significantly better with Vero cells. Note that the safe concentration in practice for different cells is not well known, Keyaerts et al. find CC50 of 261uM [1].

In vitro study of CQ and HCQ inhibition of SARS-CoV-2 into Vero (kidney), Vero-TMPRSS2, and Calu-3 (derived from human lung carcinoma) cells.

Authors reportedly used sodium pyruvate which may inhibit CQ from entering cells [2].

Although there are several theories on how HCQ may help with COVID-19, authors do not consider the most common theory where HCQ functions as a zinc ionophore, facilitating significant intracellular concentrations of zinc. Zinc is known to inhibit SARS-CoV RNA-dependent RNA polymerase activity, and is widely thought to be important for effectiveness with SARS-CoV-2 [3].

Calu-3 is one of many cell lines derived from human lung carcinomas [4]. Calu-3 cells resemble serous gland cells. They do not express 15-lipoxygenase, an enzyme specifically localized to the surface epithelium, but they do express secretory component, secretory leukocyte protease inhibitor, lysozyme, and lactoferrin, all markers of serous gland cells. [5] note that the absence of systemic inflammation, circulatory factors, and other paracrine systemic influences is a potential limitation of the isolated cell system.

RT-PCR is used, we note that nucleic acid may persist even after the virus is no longer viable [6].

It is unclear how the authors conclude "CQ does not block SARS-CoV-2 infection of Calu-3" cells, when the results show statistically significant inhibition at higher concentrations.

Further, it is unclear how the authors go from these results in one specific type of pulmonary adenocarcinoma cells that resemble serous gland cells, in vitro, into the title of the paper which claims no inhibition in lung cells.

Further, it is unclear how another leap is made to "will not be effective against COVID-19" given the multiple theories of HCQ/CQ effectiveness.

[1] c19study.com/keyaerts.html

[2] twitter.com/JaclynHord/status/1302680394244947969

[3] infezmed.it/index.php/article?Anno=2020&numero=2&ArticoloDaVisualizzare=Vol_28_2_2020_192

[4] journals.physiology.org/doi/pdf/10.1152/ajplung.1994.266.5.L493

[5] journals.physiology.org/doi/pdf/10.1152/ajplung.1994.266.5.L493

[6] fda.gov/media/136472/download

Rivera et al., Cancer Discovery, doi:10.1158/2159-8290.CD-20-0941 (Peer Reviewed)

Utilization of COVID-19 Treatments and Clinical Outcomes among Patients with Cancer: A COVID-19 and Cancer Consortium (CCC19) Cohort Study

Retrospective cancer patients, showing adjusted OR 1.03 [0.62-1.73] for HCQ. The study reports the number of HCQ+AZ patients but they do not provide results for HCQ+AZ (only HCQ + any other treatment). Significant confounding by indication and compassionate use is likely.

death, ↑2.4%, p=0.90
(odds ratio converted to relative risk)

Kelly et al., British Journal of Clinical Pharmacology, doi:10.1111/bcp.14482 (Peer Reviewed)

Clinical outcomes and adverse events in patients hospitalised with COVID‐19, treated with off‐label hydroxychloroquine and azithromycin

Retrospective 82 hospitalized patients HCQ/AZ, 52 SOC, not finding statistically significant differences. Confounding by indication - authors note that the HCQ/AZ patients were more severely ill, and do not attempt to adjust for confounders.

death, ↑143.0%, p=0.03

Bernaola et al., medRxiv, doi:10.1101/2020.07.17.20155960 (Preprint)

Observational Study of the Efficiency of Treatments in Patients Hospitalized with Covid-19 in Madrid

HCQ HR 0.83 [0.77-0.89] based on propensity score matched retrospective analysis of 1,645 hospitalized patients. Prednisone HR 0.85 [0.82-0.88], 14 other medications showed either no signicant benefit or a negative effect.

death, ↓17.0%, p<0.0001

Desbois et al., Research Square, doi:10.21203/rs.3.rs-41653/v1 (Preprint)

Prevalence and clinical features of COVID-19 in a large cohort of 199 patients with sarcoidosis

Retrospective 199 sarcoidosis patients showing non-statistically significant HCQ RR 0.83, p=1.0.

COVID-19 case, ↓16.9%, p=1.00

Risch, H., American Journal of Epidemiology, July 20, 2020, doi:10.1093/aje/kwaa152 (Peer Reviewed) (meta analysis - not included in study count)

Response to: “Early Outpatient Treatment of Symptomatic, High-Risk Covid-19 Patients” and “Re: Early Outpatient Treatment of Symptomatic, High-Risk Covid-19 Patients that Should be Ramped-Up Immediately as Key to the Pandemic Crisis”

Updated meta analysis including 7 new studies of high-risk outpatients, for a total of 12 studies, all showing significant benefit.

Watanabe, M., arXiv.org, arXiv:2007.09477 (Preprint) (meta analysis - not included in study count)

Efficacy of Hydroxychloroquine as Prophylaxis for Covid-19

Secondary analysis of Boulware et al.'s PEP trial and treatment delay-response data, confirming that HCQ is effective when used early, p<0.01.

The effectiveness found is especially notable considering the limitations of the study. Treatment was relatively late, with enrollment up to
4 days after exposure, and an unspecified shipping delay. While the paper does not provide shipping details, the study protocol gives some detail allowing us to estimate the treatment delay as ~70 to 140 hours after exposure on average for the 1-4 days since enrollment specified in the paper (we will update this when authors respond to our request for details). There was only 75% medication adherence, including 16% who did not take the medication at all. The study relies on Internet surveys.

Some issues have been raised with this analysis. 1-tailed vs. 2-tailed tests - this is debatable, an argument can be made for both cases. However, it doesn't affect the conclusion in terms of the delay-response relationship showing statistically significant efficacy. Secondly, the paper projects the "1-4" day results to a day "0" result (in reality about 46 hours later in all cases), while the trend may well continue, we do not know this. However it doesn't change the outcome that the 1-4 day results show a statistically significant delay-response relationship.

McGrail et al., medRxiv, doi:10.1101/2020.07.17.20156521 (Preprint)

COVID-19 Case Series at UnityPoint Health St. Luke’s Hospital in Cedar Rapids, IA

HCQ+AZ early in the epidemic had a fairly good success rate with few complications, 86% of HCQ patients survived and 92% of HCQ+AZ patients. Patients not receiving either had 93% survival but were not considered comparable because the treated groups were significantly more ill (100% hypoxic at admission vs. 59%) and this study does not adjust for the differences.

Transition from an early intubation strategy to aggressive utilization of high flow nasal cannula and noninvasive ventilation (i.e, BiPAP) was successful in freeing up ICU resources.

death, ↑70.0%, p=0.69

Lyngbakken et al., Nature Communications, doi:10.1038/s41467-020-19056-6 (Peer Reviewed)

A pragmatic randomized controlled trial reports lack of efficacy of hydroxychloroquine on coronavirus disease 2019 viral kinetics

Small RCT of nasopharyngeal viral load not showing significant differences. The rate of reduction for HCQ was 0.24 [0.03-0.46] RNA copies/mL/24h, and 0.14 [-0.10-0.37] for the control group (71% faster with HCQ but not statistically significant with the small sample size of 27 HCQ and 26 control patients). Analysis only over 96 hours.

NCT04316377

death, ↓3.7%, p=1.00
improvement in viral load reduction rate, ↓71.0%, p=0.51

Hong et al., Infect. Chemother., 2020, doi:10.3947/ic.2020.52.e43 (Peer Reviewed)

Early Hydroxychloroquine Administration for Rapid Severe Acute Respiratory Syndrome Coronavirus 2 Eradication

HCQ 1-4 days from diagnosis was the only protective factor against prolonged viral shedding found, OR 0.111, p=0.001. 57.1% viral clearance with 1-4 days delay vs. 22.9% for 5+ days delayed treatment. Authors report that early administration of HCQ significantly ameliorates inflammatory cytokine secretion and that COVID-19 patients should be administrated HCQ as soon as possible. 42 patients with HCQ 1-4 days from diagnosis, 48 with HCQ 5+ days from diagnosis.

risk of prolonged viral shedding, ↓64.9%, p=0.001
(odds ratio converted to relative risk)

Skipper et al., Annals of Internal Medicine, doi:10.7326/M20-4207 (Peer Reviewed)

Hydroxychloroquine in Nonhospitalized Adults With Early COVID-19: A Randomized Trial

~70 to 140 hour (inc. shipping) delayed outpatient treatment with HCQ reduced combined hospitalization/death by 50%, p=0.29 (5 HCQ cases, 10 control cases), and reduced hospitalization by 60%, p=0.17. There was one hospitalized control death and one non-hospitalized HCQ death. It is unclear why there was a non-hospitalized death, external factors such as lack of standard care may be involved. Excluding that case results in one control death and zero HCQ deaths (not statistically significant but noted as reducing mortality is the most important outcome). Details for the hospitalizations and deaths such as medication adherence and treatment delay may be informative but are not provided.

The paper states the end point was changed from hospitalization/death to symptom severity because they would have required 6,000 participants. However, if the observed trend continued, they would hit 95% significance on the reduction in hospitalization at ~725 patients, and 95% on the reduction in combined hospitalization/death at ~1,145 patients, which is a lot less than 6,000, and also less than the original plan of 1,242 patients. We hope the trial can be continued for statistical significance.

Treatment is relatively late, ~70 to 140 hours after symptoms, including the shipping delay. The paper does not mention the shipping delay but partial details are provided in the study protocol. They are not clear but indicate no shipping on the weekends and a possible 12pm cutoff for same day dispensing and mailing. Assuming that enrollments were evenly distributed between 6am and 12am each day, we get an average of ~46 hours shipping delay. We have asked for shipping details and will update with more accurate values when available. In any case the treatment delay is quite long and there is no overlap with the more typical delays used such as 0 - 36 hours for oseltamivir.

The paper compares 0 - 36 hour delayed treatment with oseltamivir (influenza) and ~70 to 140 hour delayed treatment with HCQ (COVID-19), noting that oseltamivir seemed more effective. However, a more comparable study is McLean (2015) who showed that 48 - 119 hour delayed treatment with oseltamivir has no effect. This suggests that HCQ is more effective than oseltamivir, and that HCQ may still have significant effect for some amount of delay beyond the delay where oseltamivir is effective.

6 people were included that enrolled with >4d symptoms, although they do not match the study inclusion criteria. This reduces observed effectiveness. The paper says 56% (236) were enrolled within 1 day of symptoms, but results show only 40% for "<1d", 56% is possibly for <48hrs, we have asked for clarification.

Patients in this study are relatively young and most of them recover without assistance. This reduces the room for a treatment to make improvements. The maximum improvement of an effective treatment would be expected before all patients approach recovery, as shown in the figure below. Authors focus on the end result where most have recovered, but it is more informative to examine the curve and the point of maximum effectiveness. Authors did not collect data for every day but they do have interim results for days 3, 5, 10. The results are consistent with an effective treatment and show a statistically significant improvement, p = 0.05, at day 10 (other unreported days might show increased effectiveness).

Results also show a larger treatment effect for those >50, not statistically significant due to the small sample, but noted as COVID-19 risk dramatically increases with age. The effect may be more visible here because younger patients may on average have more mild cases with less room for improvement. In general patients in this study have relatively mild symptoms on average, limiting the chance to observe improvement.

The study relies on Internet surveys. Known fake surveys were submitted to the similar PEP trial and there could be an unknown number of undetected fake surveys in both trials. The study shows a high incidence of side effects in the placebo arm, which could be in part due to fake entries [1].

The granularity change in the histograms of Figure S4 raise concerns [2]. Data on increasing severity, less affected by the lower bound where everyone has recovered, also supports effectiveness [3].

Research shows the placebo used in the US may be protective for COVID-19 [4] so the true effectiveness of HCQ could be higher than observed.

Treatment delay reporting has changed from the companion PEP trial which reported results for enrollment delays 1, 2, 3, and 4 separately (and from which we can confirm a statistically significant delay-response relationship), while this trial combines 1-2 and 3-4, and adds <1. Since the two trials share reporting (some patients were moved between trials) it's not clear how the new category was added.

RCT of 423 patients with Internet surveys. Medication adherence was only 77% so the true effect of treatment is likely higher. Analysis of primarily low risk patients, authors note the results are not generalizable to the COVID high-risk population. We will update when hearing back on questions asked.

In summary, we believe the results of this study are positive for HCQ being an effective treatment, however we have classified this study as inconclusive for now pending feedback and further analysis.

Also see: [5] and [6] regarding flaws in this study.

[1] twitter.com/Covid19Crusher/status/1284515906375356416

[2] twitter.com/Covid19Crusher/status/1284515906375356416

[3] twitter.com/Covid19Crusher/status/1284515906375356416

[4] frontiersin.org/articles/10.3389/fphar.2020.01062/full

[5] drive.google.com/file/d/1NZOJ57fM0RTaHD1t_9w2iua7lUJhOgWT/view

[6] twitter.com/cnpaiva/status/1303324404630388738

hospitalization, ↓51.7%, p=0.19
risk of no recovery at day 14, ↓20.0%, p=0.21

Mitjà et al., Clinical Infectious Diseases, ciaa1009, doi:10.1093/cid/ciaa1009 (Peer Reviewed)

Hydroxychloroquine for Early Treatment of Adults with Mild Covid-19: A Randomized-Controlled Trial

This paper has inconsistent data - some of the values reported in Table 2 and the abstract correspond to 12 control hospitalizations, while others correspond to 11 control hospitalizations.

There was a 25% reduction in hospitalization and 16% reduction in the median time to symptom resolution for HCQ, without statistical significance due to small samples.

Treatment delay is unknown at this time. They report a delay of up to 120 hours after symptoms plus an additional unspecified delay where medication was provided to patients at the first home visit. We have asked for details of the treatment delay and will update when hearing back. They do not break down results by treatment delay.

The paper does not mention zinc. Zinc deficiency in Spain has been reported at 83% [1], this may significantly reduce effectiveness. HCQ is a zinc ionophore which increases cellular uptake, facilitating significant intracellular concentrations of zinc, and zinc is known to inhibit SARS-CoV RNA-dependent RNA polymerase activity, and is widely thought to be important for effectiveness with SARS-CoV-2 [2].

Undetectable viral load was changed to 3 log10 copies/mL potentially masking effectiveness. For viral load authors use nasopharyngeal swabs, we note that viral activity in the lung may be especially important for COVID-19, and that research has shown HCQ concentrations can be much higher in the lung compared to plasma [3]. We also note that viral detection by PCR does not equate to viable virus [4]. Accuracy of the tests is not provided.

Nasopharyngeal viral load analysis issues include test unreliability and temporo-spatial differences in viral shedding [5].

293 low-risk patients with no deaths. No serious adverse events. We have asked for more details on the treatment delay and viral load change and will update when hearing back.

Also see this open letter: [6]

[1] mdpi.com/2072-6643/9/7/697

[2] infezmed.it/index.php/article?Anno=2020&numero=2&ArticoloDaVisualizzare=Vol_28_2_2020_192

[3] ncbi.nlm.nih.gov/pmc/articles/PMC7122276/HCQ

[4] ams.edu.sg/view-pdf.aspx?file=media%5c5556_fi_331.pdf&ofile=Period+of+Infectivity+Position+Stateme..

[5] journal.chestnet.org/article/S0012-3692(20)31718-9/fulltext

[6] drive.google.com/file/d/1NZOJ57fM0RTaHD1t_9w2iua7lUJhOgWT/view

hospitalization, ↓25.0%, p=0.64
recovery time, ↓16.7%, p=0.38

Gupta et al., JAMA Intern. Med., doi:10.1001/jamainternmed.2020.3596 (Peer Reviewed)

Factors Associated With Death in Critically Ill Patients With Coronavirus Disease 2019 in the US

Analysis of 2215 intensive care unit patients showing no significant differences with this very late stage use of HCQ. HCQ+AZ mortality relative risk RR 0.96, p=0.53, HCQ and HCQ+AZ combined RR 1.06, p=0.409.

death, ↑6.0%, p=0.41

Kavanagh et al., Med. Hypotheses, doi:10.1016/j.mehy.2020.110110 (Peer Reviewed)

Inhaled hydroxychloroquine to improve efficacy and reduce harm in the treatment of COVID-19

Proposal to use an inhaled formulation of HCQ which has passed safety studies in clinical trials for the treatment of asthma. Authors advocate for early treatment or prophylaxis of COVID-19, using HCQ as an inhaled aerosol, to deliver the drug directly to the lungs at a lower dose than that required for oral systemic delivery.

Trullàs et al., Research Square, doi:10.21203/rs.3.rs-39421/v1 (Preprint)

High in-hospital mortality due to COVID-19 in a community hospital in Spain: a prospective observational study

Retrospective 100 hospitalized patients in Spain showing lower mortality with HCQ+AZ.

death, ↓35.6%, p=0.12

Chowdhury et al., Research Square, doi:10.21203/rs.3.rs-38896/v1 (Preprint)

A Randomized Trial of Ivermectin-Doxycycline and Hydroxychloroquine-Azithromycin therapy on COVID19 patients

Small 116 patient RCT comparing Ivermectin-Doxycycline and HCQ+AZ, not showing a significant difference in time to PCR negative or symptom resolution. Time to symptomatic recovery was 5.93 days for Ivermectin-Doxycycline vs. 6.99 days for HCQ+AZ. Given the long half-life of HCQ and the lack of a loading dose, it may take several days for HCQ to reach therapeutic levels. 10% of HCQ+AZ patients were lost to followup (2x Ivermectin-Doxycycline). There is no comparison with a control group.

Lecronier et al., Critical Care, 24:418, 2020, doi:10.1186/s13054-020-03117-9 (Peer Reviewed)

Comparison of hydroxychloroquine, lopinavir/ritonavir, and standard of care in critically ill patients with SARS-CoV-2 pneumonia: an opportunistic retrospective analysis

Retrospective 80 ICU patients, 22 SOC, 20 lopinavir/ritonavir, 38 HCQ. 28 day mortality 24% (HCQ) versus 41% (SOC), a 41% decrease, but not statistically significant due to very small sample sizes. No statistically significant differences found for treatment escalation, ventilator-free days, viral load, or mortality. Authors consider treatment escalation more important than mortality, for unknown reasons.

death, ↓42.0%, p=0.24, HCQ vs. control
treatment escalation, ↓6.0%, p=0.73, HCQ vs. control
risk of viral+ at day 7, ↓15.0%, p=0.61, HCQ vs. control

Cravedi et al., American Journal of Transplantation, doi:10.1111/ajt.16185 (Peer Reviewed)

COVID‐19 and kidney transplantation: Results from the TANGO International Transplant Consortium

Analysis of 144 hospitalized kidney transplant patients showing HCQ mortality HR 1.53, p = 0.17. Subject to confounding by indication.

death, ↑53.0%, p=0.17

Chen et al., PLOS ONE, doi:10.1371/journal.pone.0242763 (Peer Reviewed)

A Multicenter, randomized, open-label, controlled trial to evaluate the efficacy and tolerability of hydroxychloroquine and a retrospective study in adult patients with mild to moderate Coronavirus disease 2019 (COVID-19)

2 very small studies with hospitalized patients in Taiwan.

RCT with 21 treatment and 12 SOC patients. No mortality, or serious adverse effects. Median time to negative RNA 5 days versus 10 days SOC, p=0.4. Risk of PCR+ at day 14, RR 0.76, p = 0.71.

This paper also reports on a small retrospective study with 12 of 28 HCQ patients and 5 of 9 in the control group being PCR- at day 14, RR 1.29, p = 0.7.

no virological cure, ↓24.0%, p=0.71, day 14
median time to PCR-, ↓50.0%, p=0.40

Rivera-Izquierdo et al., Medicina Clínica, doi:10.1016/j.medcli.2020.06.025 (Peer Reviewed)

Agentes terapéuticos utilizados en 238 pacientes hospitalizados por COVID-19 y su relación con la mortalidad

Retrospective 238 hospitalized patients in Spain showing lower mortality with HCQ, adjusted hazard ratio aHR 0.81 [0.24-2.76].

death, ↓19.0%, p=0.75

Raoult et al., Preprint (Preprint) (meta analysis - not included in study count)

Hydroxychloroquine and Azithromycin as a Treatment of COVID-19: Results of an Open-Label Non-Randomized Clinical Trial: Response to David Spencer (Elsevier)

Updated meta analysis showing significant reductions in mortality and viral shedding. Mortality OR 0.53 [0.4-0.71] for clinical studies, 0.92 big data studies, 18,211 patients. Persistent viral shedding OR 0.47 [0.28-0.79], 4,540 patients.

Marzolini et al., Antimicrobial Agents and Chemotherapy, doi:10.1128/AAC.01177-20 (Peer Reviewed) (not included in the study count)

Effect of Systemic Inflammatory Response to SARS-CoV-2 on Lopinavir and Hydroxychloroquine Plasma Concentrations

Study of Lopinivar and HCQ plasma concentrations and CRP levels in late stage (treatment initiation median 8 days from onset) COVID-19 patients. The median HCQ plasma concentration was 171 ng/ml, which authors suggest indicates that HCQ levels achieved in vivo do not result in adequate clinical activity for COVID-19, however this is incorrect as tissue concentration can be many times higher [1].

[1] c19study.com/ruiz.html

Li et al., Cell Death & Disease volume 11, doi:10.1038/s41419-020-2721-8 (Review) (Peer Reviewed) (not included in the study count)

Is hydroxychloroquine beneficial for COVID-19 patients?

Review of the anti-inflammatory, antiviral, and protective vascular effects of CQ and HCQ, noting that HCQ may be preferable for COVID-19 due to fewer side effects.

Goldstein, L., Preprint, July 7, 2020 (Review) (Preprint) (not included in the study count)

Hydroxychloroquine-based COVID-19 Treatment, A Systematic Review of Clinical Evidence and Expert Opinion from Physicians’ Surveys

85% of globally surveyed physicians recognized HCQ as at least partially effective in treating COVID-19, according to Sermo W3. More than half of the surveyed US physicians would take the drug or give it to family members early or even before onset of symptoms, according to JC.

Aside from the rarely used plasma, HCQ / HCQ+AZ based treatments are preferred by physicians by wide margin over other drugs. HCQ / HCQ+AZ based treatments are the most used, most recommended, and most highly rated by physicians treating COVID-19 at an early stage.

An et al., medRxiv, doi:10.1101/2020.07.04.20146548 (Preprint)

Treatment Response to Hydroxychloroquine and Antibiotics for mild to moderate COVID-19: a retrospective cohort study from South Korea

Retrospective of hospitalized patients with 31 HCQ patients and 195 standard treatment patients, not showing a significant difference in terms of viral clearance or recovery. There was no mortality in either group.

"It is notable that HQ plus antibiotics group had worse baseline clinical profiles (i.e. higher percentage of moderate severity patients, more patients with fever >=37.5C, higher average body temperature) and prognostic indicators such as age, LDH, lymphocyte count, and CRP".

We note that propensity score matching removed almost all of the male patients in the control group (40% -> 5%) but increased the percentage of male patients in the treatment group. This provides a large advantage to the control group because there is a very large difference in severity and mortality based on gender [1].

In terms of viral RNA clearance we note that other research has found that "active viral replication drops quickly after the first week, and viable virus was not found after the second week of illness despite the persistence of PCR detection of RNA” [2].

[1] ncbi.nlm.nih.gov/pmc/articles/PMC7247289/

[2] ams.edu.sg/view-pdf.aspx?file=media%5c5556_fi_331.pdf&ofile=Period+of+Infectivity+Position+Stateme..

time to viral clearance, ↓3.0%, p=0.92

Zhong et al., Lancent Rheumatology, 10.1016/S2665-9913(20)30227-7 (Peer Reviewed)

COVID-19 in patients with rheumatic disease in Hubei province, China: a multicentre retrospective observational study

Rheumatic disease patients on HCQ had a lower risk of COVID-19 than those on other disease-modifying anti-rheumatic drugs, OR 0.09 (0.01–0.94), p=0.044 after adjusting for age, sex, smoking, systemic lupus erythematosus, infection in other family members, and comorbidities. 43 patients with rheumatic disease and COVID-19 exposure.

COVID-19 case, ↓91.0%, p=0.04

Arshad et al., Int. J. Infect. Dis., July 1 2020, doi:10.1016/j.ijid.2020.06.099 (Peer Reviewed)

Treatment with Hydroxychloroquine, Azithromycin, and Combination in Patients Hospitalized with COVID-19

HCQ decreases mortality from 26.4% to 13.5% (HCQ) or 20.1% (HCQ+AZ). Propensity matched HCQ HR 0.487, p=0.009. Michigan 2,541 patients retrospective. Before propensity matching the HCQ group average age is 5 years younger and the percentage of male patients is 4% higher which is likely to favor the treatment and the control respectively in the before-propensity matching results.

Some reported limtiations of this study are inaccurate [1]. Corticosteroids were controlled for in the multivariate and propensity analyses as were age and comorbidities including cardiac disease and severity of illness. Age was an independent risk factor associated with mortality. HCQ was independently associated with decreased mortality, distinct from the steroid effect. 91% of all patients began treatment within two days of admission. HCQ was used throughout the study period, limiting time bias. Patients assigned to HCQ group had moderate and severe illness at presentation, which would favor worse outcome with HCQ.

[1] ijidonline.com/article/S1201-9712(20)30604-4/fulltext

death, ↓51.3%, p=0.009

Samuel et al., Heart Rhythm, doi:10.1016/j.hrthm.2020.06.033 (Peer Reviewed) (not included in the study count)

Incidence of arrhythmias and electrocardiographic abnormalities in symptomatic pediatric patients with PCR positive SARS-CoV-2 infection including drug induced changes in the corrected QT interval (QTc)

In pediatric patients with PCR positive active COVID-19 infection, significant arrhythmias are infrequent, but occur at an incidence higher than expected in a general pediatric population. Comorbidities are not more common in patients with arrhythmias than in patients without arrhythmias. However, providers still need to be vigilant for comorbidities that may independently place patients at risk for arrhythmias. COVID-19 treatment using HCQ leads to significant QTc prolongation, but was not associated with arrhythmias in pediatric patients. The long term sequelae of arrhythmia development in this population and their impact on outcome needs to be studied.

Martinez-Lopez et al., Blood Cancer Journal, doi:10.1038/s41408-020-00372-5 (Peer Reviewed)

Multiple Myeloma and SARS-CoV-2 Infection: Clinical Characteristics and Prognostic Factors of Inpatient Mortality

Retrospective 167 multiple myeloma patients in Spain.

death, ↓33.0%, p=0.20

Mikami et al., J. Gen. Intern. Med., doi:10.1007/s11606-020-05983-z (Peer Reviewed)

Risk Factors for Mortality in Patients with COVID-19 in New York City

HCQ decreases mortality, HR 0.53 (CI 0.41–0.67). IPTW adjustment does not significantly change HR 0.53 (0.41-0.68). Retrospective 6,000 patients in New York City.

death, ↓47.0%, p<0.0001

Komissarov et al., medRxiv, doi:10.1101/2020.06.30.20143289 (Preprint)

Hydroxychloroquine has no effect on SARS-CoV-2 load in nasopharynx of patients with mild form of COVID-19

Small late stage (7-10 days post symptoms) study of nasal swab RNA with 12 control and 33 patients, showing no significant differences (significant reduction in viral load is seen in both groups).

The groups are not comparable, with significant differences seen between hospitalized and non-hospitalized patients. 9 of 10 hospitalized patients were in the HCQ group and only one in the control group. 2 additional control patients were added between the first and second version of this preprint (including the only hospitalized control patient).

viral load, ↑25.0%, p=0.45

Sosa-García et al., Cir Cir. 2020, 88:5, 569-575, doi:10.24875/CIRU.20000675 (Peer Reviewed)

Experience in the management of severe COVID-19 patients in an intensive care unit

Small retrospective study of 56 ICU patients in Mexico showing HCQ RR 1.1, p = 1.0.

death, ↑10.5%, p=1.00

Ferreira et al., J. Medical Virology, July 9, 2020, doi:10.1002/jmv.26286 (preprint 6/29) (Peer Reviewed)

Chronic treatment with hydroxychloroquine and SARS-CoV-2 infection

Chronic treatment with HCQ provides protection against COVID, odds ratio 0.51 (0.37-0.70).

The actual benefit is likely to be larger becasue research shows that the risk of COVID-19 for systemic autoimmune disease patients is much higher overall. Ferri et al. show OR 4.42, p<0.001 [1], which is the observed real-world risk, taking into account factors such as these patients potentially being more careful to avoid exposure.

[1] c19study.com/ferri.html

COVID-19 case, ↓47.1%, p<0.0001
(odds ratio converted to relative risk)

Mfeukeu-Kuate et al. (Preprint) (not included in the study count)

Electrocardiographic safety of daily Hydroxychloroquine 400mg plus Azithromycin 250mg as an ambulatory treatment for COVID-19 patients in Cameroon

No life-threatening modifications of the QT interval was observed in non-severe COVID-19 patients treated ambulatory with HCQ+AZ. 51 relatively young patients 39 +/- 11.

Gendebien et al., Annals of the Rheumatic Diseases, doi:10.1136/annrheumdis-2020-218244 (Letter)

Systematic analysis of COVID-19 infection and symptoms in a systemic lupus erythematosus population: correlation with disease characteristics, hydroxychloroquine use and immunosuppressive treatments

Small study of 152 SLE patients taking HCQ with a phone survey for COVID-19 suggestive symptoms. There was 2 hospitalizations (group not identified) and no ICU or death cases. A similar percentage of suspected infections were reported for HCQ users and non-HCQ users, RR 0.96, p = 0.93.

There was no mortality and severity was not analyzed to determine if HCQ treated patients fared better. No adjustment for concomitant medications or severity of SLE. Only 5 cases were PCR confirmed.

COVID-19 case, ↓3.9%, p=0.93

Lagier et al., Travel Med. Infect. Dis. 101791, Jun 25, 2020, doi:10.1016/j.tmaid.2020.101791 (Peer Reviewed)

Outcomes of 3,737 COVID-19 patients treated with hydroxychloroquine/azithromycin and other regimens in Marseille, France: A retrospective analysis

Early treatment leads to significantly better clinical outcome and faster viral load reduction. Matched sample mortality HR 0.41 p-value 0.048. Retrospective 3,737 patients.

death, ↓59.0%, p=0.05

Bousquet et al., Aging, 12:12, 11306-11313, doi:10.18632/aging.103583 (Peer Reviewed)

ADL-dependency, D-Dimers, LDH and absence of anticoagulation are independently associated with one-month mortality in older inpatients with Covid-19

Observational prospective 108 hospitalized patients 65 and older, showing HCQ mortality OR 0.49, p = 0.15.

death, ↓42.8%, p=0.15
(odds ratio converted to relative risk)

Isaksen et al., medRxiv, doi:10.1101/2020.06.19.20135475 (Preprint)

Chloroquine, but not hydroxychlorquine, prolongs the QT interval in a primary care population

Safety analysis in patients without COVID-19, finding a small increase in QTc associated with use of CQ, but not HCQ. Authors also study mortality (n=3,368), with HCQ hazard ratio 0.67 [0.43-1.05].

Fontana et al., Clinical Kidney Journal, 13:3, 334–339, doi:10.1093/ckj/sfaa084 (Peer Reviewed)

SARS-CoV-2 infection in dialysis patients in northern Italy: a single-centre experience

Very small observational study of 15 dialysis patients showing HCQ mortality RR 0.50, p = 0.53.

death, ↓50.0%, p=0.53

Chen et al., medRxiv, doi:10.1101/2020.06.19.20136093 (Preprint)

Efficacy and safety of chloroquine or hydroxychloroquine in moderate type of COVID-19: a prospective open-label randomized controlled study

Significantly faster clinical recovery and shorter time to RNA negative (from 7.0 days to 2.0 days (HCQ), p=0.01. 67 patients with mild/moderate cases.

median time to PCR-, ↓72.0%, p=0.01

Faíco-Filho et al., Braz J Microbiol, doi:10.1007/s42770-020-00395-x (preprint 6/21) (Peer Reviewed)

No benefit of hydroxychloroquine on SARS-CoV-2 viral load reduction in non-critical hospitalized patients with COVID-19

Viral load comparison for 34 HCQ and 32 control patients hospitalized with moderate COVID-19. All patients recovered limiting the room for beneficial effects.

While not achieving statistical significance, results show faster recovery with HCQ. The greatest benefit is seen mid-recovery as expected for an effective treatment:

Δt7-12: 81% improvement with HCQ
Δt<7: 24% improvement with HCQ

For Δt>12, everyone has recovered so there is no room for improvement. Since the HCQ group started slightly higher the improvement is slightly less. Most participants have also dropped out by this test, with only 6 HCQ and 9 control remaining (also suggesting HCQ patients recovered faster).

Δt7-12 ΔCt improvement, ↓80.8%, p=0.40
Δt<7 ΔCt improvement, ↓24.0%, p=0.36
Δt>12 ΔCt improvement, ↑15.0%, p=0.52

SMSH Sawai Man Singh Hospital, India (News) (not included in the study count)

HCQ beneficial as preventive drug: SMS doctors told ICMR

PrEP with 4,300 very high risk healthcare workers in a hospital with up to 500+ COVID patients at a time, only 1% cases, all recovered.

NIH, study not available yet (News) (not included in the study count)

NIH halts clinical trial of hydroxychloroquine

NIH halts late stage trial reporting no harm and no benefit. 470 patients.

Sbidian et al., medRxiv, doi:10.1101/2020.06.16.20132597 (Preprint)

Hydroxychloroquine with or without azithromycin and in-hospital mortality or discharge in patients hospitalized for COVID-19 infection: a cohort study of 4,642 in-patients in France

Retrospective of 4,642 hospitalized patients in France showing significantly faster discharge with HCQ and HCQ+AZ. No significant effect is seen on 28-day mortality, however many more control patients are still in hospital at 28 days. Other studies show faster resolution for HCQ, suggesting there will be a significant improvement when extending past 28 days. Hopefully authors will extend the analysis. Note that the median age is higher in the group not treated with HCQ or AZ.

For other issues with the adjustments see [1]. Also see the analysis here [2].

[1] medrxiv.org/content/10.1101/2020.06.16.20132597v1#disqus_thread

[2] twitter.com/Covid19Crusher/status/1274668697408471040

death, ↑5.0%, p=0.74, whole population HCQ AIPTW adjusted
no hospital discharge, ↓20.0%, p=0.002, whole population HCQ AIPTW adjusted

Kaptein et al., bioRxiv, doi:10.1101/2020.06.19.159053 (Peer Reviewed) (not included in the study count)

Favipiravir at high doses has potent antiviral activity in SARS-CoV-2−infected hamsters, whereas hydroxychloroquine lacks activity

Animal study with Syrian hamsters, showing treatment of SARS-CoV-2-infected hamsters with favipiravir or HCQ (with and without AZ). Treatment with HCQ alone resulted in a very modest reduction of 0.3 log10 viral RNA copies/mg lung, and no reduction in viral RNA load in the ileum or stool.

Therapeutic levels of HCQ may not have been reached. Cytosolic concentrations in the lung were far below the EC90 target.

A number of issues have reportedly been raised, including the following, for which the authors did not respond:

- HCQ was administered with DMSO and Cremophor - why were non-neutral carriers chosen? DMSO has anti-inflammatory properties and Cremophor has a range of side effects [1]: "use has been associated with severe anaphylactoid hypersensitivity reactions, hyperlipidemia, abnormal lipoprotein patterns, aggregation of erythrocytes and peripheral neuropathy". Why weren't the same amounts of solvent applied to the non-treated animals? Shouldn't favipiravir have been dissolved in the same way to avoid bias?

- Why is the method of administration different for both products (oral gavage/intraperitoneal injection)?

- One of the HCQ protocols used in France (IHU Marseille) prescribes 600mg of HCQ per day to patients. For a body weight of 60kg, this corresponds to a dose of 10mg/kg. How is a dose of 50 mg/kg in this study justified, i.e. 5 times more, with possible systemic effects on the animals and possible influence on the results?

- Why were the animals killed after 4/5 days and did the treatment not continue? Unfortunately, this does not allow us to know the real course of the disease or the mortality of the animals beyond day 4. Is it because the animals spontaneously improve without treatment as Professor Neyts seems to say in his webinar for the GVN? And if this is indeed the case, isn't the use of these hamsters exactly an objection as a model for COVID-19 research?

[1] sciencedirect.com/topics/pharmacology-toxicology-and-pharmaceutical-science/cremophor

Paccoud et al., Clinical Infectious Diseases, doi:10.1093/cid/ciaa791 (Peer Reviewed)

Compassionate use of hydroxychloroquine in clinical practice for patients with mild to severe Covid-19 in a French university hospital

Retrospective of 89 hospitalized patients, survival HR 0.89 [0.23-3.47], not statistically significant. Authors note that unmeasured confounders may have persisted and the study may be underpowered.

death, ↓11.0%, p=0.88

Capucci et al., J. Cardiovasc. Med. 21, 922–923, doi:10.2459/JCM.0000000000001061 (Peer Reviewed)

Low hospitalization rate without severe arrhythmias: a prospective survey on 350 patients early home treated with hydroxychloroquine during COVID-19 pandemic

Prospective analysis of early treatment of 350 patients in Italy (without waiting for PCR results), showing low hospitalization rates and no serious adverse events.

From 274 patients treated with HCQ, 16 required hospitalization (5.8%). Minor complications (mainly gastrointestinal, diarrhoea) were found in eight patients (2.9%), none of whom had to interrupt treatment. No major cardiac complications were found.

Of the 76 patients treated with HCQ+AZ, 4 were hospitalized (5.2%). Minor complications occurred in two patients (2.6%). There were no reports of any major arrhythmias, syncope or sudden death.

Xue et al., J. Med. Virology, June 17, 2020, doi:10.1002/jmv.26193 (Peer Reviewed)

Hydroxychloroquine treatment in COVID-19: a descriptive observational analysis of 30 cases from a single center in Wuhan, China

30 hospitalized patients. Early use of HCQ is more effective, 43% reduction in progression from moderate to severe. "Early" is relative here, within 7 days of hospitalization.

Luo et al., Annals of Oncology, 31:10, 1386-1396, doi:10.1016/j.annonc.2020.06.007 (Peer Reviewed)

COVID-19 in patients with lung cancer

Analysis of hospitalized lung cancer patients with 35 of 48 taking HCQ, mortality OR 1.03, p = 0.99.

death, ↑2.2%, p=0.99
(odds ratio converted to relative risk)

Kim et al., Korean J Intern Med, doi:10.3904/kjim.2020.224 (Peer Reviewed)

Lopinavir-ritonavir versus hydroxychloroquine for viral clearance and clinical improvement in patients with mild to moderate coronavirus disease 2019

Small retrospective study of hospitalized patients with 31 lopinavir-ritonavir and 34 HCQ patients, HCQ 400mg once per day, finding no significant difference in clinical response, but more rapid viral clearance with lopinavir-ritonavir.

WHIP COVID-19 (News) (not included in the study count)

Henry Ford Health System still moving forward with hydroxychloroquine study

Ongoing WHIP COVID-19 HCQ PrEP study reports analyzing their data and seeing a significantly improved outcome in a group of COVID-19 patients who received HCQ. For more details on the study see [1].

[1] henryford.com/whip-covid-19

Huang et al., Annals of the Rheumatic Diseases 2020:79, 1163-1169, doi:10.1136/annrheumdis-2020-217425 (Peer Reviewed)

Clinical characteristics of 17 patients with COVID-19 and systemic autoimmune diseases: a retrospective study

Analysis of 1255 COVID-19 patients in Wuhan Tongji Hospital finding 0.61% with systemic autoimmune diseases, much lower than authors expected (3%–10%). Authors hypothesise that protective factors, such as CQ/HCQ use, reduce hospitalization.

hospitalization, ↓80.0%, p<0.001

Scherrmann, AAPS J 22, 86 (2020), doi:10.1208/s12248-020-00465-w (Peer Reviewed) (Theory) (not included in the study count)

Intracellular ABCB1 as a Possible Mechanism to Explain the Synergistic Effect of Hydroxychloroquine-Azithromycin Combination in COVID-19 Therapy

Theory paper, not included in the study count or percentages. Proposes a new mechanism supporting the synergistic interaction between HCQ+AZ.

Giacomelli et al., Journal of Medical Virology, doi:10.1002/jmv.26407 (preprint 6/12) (Peer Reviewed)

Early administration of lopinavir/ritonavir plus hydroxychloroquine does not alter the clinical course of SARS-CoV-2 infection: a retrospective cohort study

Late stage study of hospitalized patients comparing treatment starting within 5 days versus later. Note that "early" here is only relative - all patients are hospitalized so this is "late" and "very late". The "early" treatment group is significantly older. Severe adverse events attributed by authors to concurrent administration of LPV, making it difficult to make conclusions about HCQ.

Wang et al., medRxiv, doi:10.1101/2020.06.11.20128926 (Peer Reviewed)

Comorbidity and Sociodemographic determinants in COVID-19 Mortality in an US Urban Healthcare System

Database analysis of 7,592 patients in NYC, showing adjusted HCQ mortality odds ratio OR 0.96, p = 0.82, and HCQ+AZ OR 0.94, p = 0.63

death, ↓5.8%, p=0.63
(odds ratio converted to relative risk)

Otea et al., medRxiv, doi:10.1101/2020.06.10.20101105 (Preprint)

A short therapeutic regimen based on hydroxychloroquine plus azithromycin for the treatment of COVID-19 in patients with non-severe disease. A strategy associated with a reduction in hospital admissions and complications.

80 moderate cases, HCQ+AZ appears to reduce serious complications and death. Moderate treated cases resulted in hospitalization at the same rate as mild untreated cases suggesting efficacy.

Pirnay et al., Hosp. Pharm. and Clinician, doi:10.1016/j.phclin.2020.06.001 (Peer Reviewed)

Beneficial effect of Hydroxychloroquine-Azithromycin combination in the treatment of elderly patients with Covid-19: results of an observational study

68 very high risk nursing home residents, median age 86, HCQ+AZ early treatment within 2.5 days onset, 2 stopped due to QTc. Only 7 died, significantly less than other nursing homes in France and the same as the median death for the same period in 2019/2018.

Bhattacharya et al., medRxix, doi:10.1101/2020.06.09.20116806 (Preprint)

Pre exposure Hydroxychloroquine use is associated with reduced COVID19 risk in healthcare workers

HCQ reduced cases from 38% to 7%. 106 people. No serious adverse effects.

COVID-19 case, ↓80.7%, p=0.001

Roussel et al., New Microbes and New Infections, Volume 38, doi:10.1016/j.nmni.2020.100710 (Review) (Peer Reviewed) (not included in the study count)

Influence of conflicts of interest on public positions in the COVID-19 era, the case of Gilead Sciences

Shows a correlation (Spearman test, p = 0.017) between the amount received from Gilead Sciences and public opposition to the use of HCQ in France.

Million et al., New Microbes and New Infections, doi:10.1016/j.nmni.2020.100709 (Peer Reviewed) (meta analysis - not included in study count)

Clinical Efficacy of Chloroquine derivatives in COVID-19 Infection: Comparative metaanalysis between the Big data and the real world

[H]CQ effective and reduces mortality by a factor 3. Meta analysis of 20 studies.

RECOVERY Collaborative Group, NEJM, doi:10.1056/NEJMoa2022926 (press release 6/5) (Preprint)

Effect of Hydroxychloroquine in Hospitalized Patients with COVID-19: Preliminary results from a multi-centre, randomized, controlled trial

RECOVERY trial finds no significant benefit for very late stage very sick patients. Results may be due to the unusually high dosage used (9.2g total over 10 days) [1, 2].

The overall dosage used is only 23% less than the high dosage that Borba et al. show greatly increases risk (OR 2.8) [3].

Authors do not report results based on weight, BMI, or related conditions such as diabetes, which may provide additional evidence of toxic dosages. Authors do not adjust dosage based on patient weight, so toxicity may be higher in patients of lower weight.

KM curves show a spike in HCQ mortality days 5-8, corresponding to ~85% of the total excess seen at day 28 (a similar spike is seen in the SOLIDARITY trial).

Authors will not release the data until Jan 8, 2021.

Authors note: "we did not observe excess mortality in the first 2 days of treatment ... when early effects of dose-dependent toxicity might be expected", but they are ignoring the very long half-life of HCQ and the dosing regimen - much higher levels of HCQ will be reached later. Increased mortality in Borba et al. occurred after 2 days.

Patients were extremely sick (median 9 days post symptoms, 60% requiring oxygen and an additional 17% requiring ventilation/ECMO), and an unusually high death rate was seen in both arms. 1,561 HCQ patients, 3,155 SOC.

A secondary analysis has found several inconsistencies in the data: [4]. Hypoxia may inhibit HCQ entering cells [5], making it less effective for late stage use.

[1] twitter.com/JamesTodaroMD/status/1272661099985481733

[2] twitter.com/JamesTodaroMD/status/1276245669372723200

[3] c19study.com/borba.html

[4] francesoir.fr/politique-monde/oxford-etude-recovery-ou-sont-les-morts

[5] twitter.com/JaclynHord/status/1302704076573077507

death, ↑9.0%, p=0.15

Boulware et al., NEJM, June 3 2020, doi:10.1056/NEJMoa2016638 (Peer Reviewed)

A Randomized Trial of Hydroxychloroquine as Postexposure Prophylaxis for Covid-19

COVID-19 cases are reduced by [49%, 29%, 16%] respectively when taken within ~[70, 94, 118] hours of exposure (including shipping delay). The treatment delay-response relationship is significant at p=0.002. PEP delayed treatment RCT.

For more detailed analysis, see [1].

6 independent analyses of the data in this study confirm efficacy: [2, 3, 4, 5, 6, 7, 8].

[1] c19study.com/boulware.html

[2] drive.google.com/file/d/1NZOJ57fM0RTaHD1t_9w2iua7lUJhOgWT/view

[3] arxiv.org/abs/2007.09477

[4] medrxiv.org/content/10.1101/2020.08.19.20178376v1

[5] researchgate.net/publication/344369617_Hydroxychloroquine_as_Post-Exposure_Prophylaxis_for_Covid-1..

[6] blog.philbirnbaum.com/2020/08/the-nejm-hydroxychloroquine-study-fails.html

[7] longdom.org/open-access/hydroxychloroquine-and-interferons-for-the-prophylaxis-and-early-treatment..

[8] osf.io/vz8a7/

COVID-19 case, ↓17.0%, p=0.35
COVID-19 case, ↓25.1%, p=0.22, probable COVID-19 cases

Al-Kofahi et al., Clin. Pharmacol. Ther., Jun 1, 2020, doi:10.1002/cpt.1874 (Peer Reviewed) (not included in the study count)

Finding the Dose for Hydroxychloroquine Prophylaxis for COVID‐19: The Desperate Search for Effectiveness

Analysis of HCQ dosing regimens, recommending:

PrEP: 800mg loading dose followed by 400mg 2 or 3 times weekly to maintain weekly troughs above EC₅₀ in >50% of patients at steady-state.

PEP: 800mg loading dose followed in 6 hours by 600mg, then 600mg daily for 4 more days to achieve daily troughs above EC₅₀ in >50% subjects.

Guérin et al., Asian J. Medicine and Health, July 15, 2020, doi:10.9734/ajmah/2020/v18i730224 (preprint 5/31) (Peer Reviewed)

Azithromycin and Hydroxychloroquine Accelerate Recovery of Outpatients with Mild/Moderate COVID-19

Mean clinical recovery time reduced from 26 days (SOC) to 9 days, p<0.0001 (HCQ+AZ) or 13 days, p<0.0001 (AZ). No cardiac toxicity. Small retrospective study of 88 patients with case control analysis with matched patients.

death, ↓61.4%, p=1.00
recovery time, ↓65.0%, p<0.0001
(adjusted for zero (with reciprocal of the contrasting arm))

Chamieh et al., medRxiv 2020.05.28.20114835, doi:10.1101/2020.05.28.20114835 (Preprint)

Viral Dynamics Matter in COVID-19 Pneumonia: the success of early treatment with hydroxychloroquine and azithromycin in Lebanon

HCQ+AZ potentially explains 94.7% success in treating a fairly complex cohort.

Chatterjee et al., Indian J. Med. Res., June 20, 2020, doi:10.4103/ijmr.IJMR_2234_20 (Peer Reviewed)

Healthcare workers & SARS-CoV-2 infection in India: A case-control investigation in the time of COVID-19

4+ doses of HCQ associated with a significant decline in the odds of getting infected, dose-response relationship exists.

COVID-19 case, ↓66.8%, p<0.001, full course vs. unused

Huang et al., National Science Review, nwaa113, doi:10.1093/nsr/nwaa113 (Peer Reviewed)

Preliminary evidence from a multicenter prospective observational study of the safety and efficacy of chloroquine for the treatment of COVID-19

197 CQ patients, 176 control. Mean time to undetectable viral RNA and duration of fever significantly reduced. No serious adverse events.

time to viral-, ↓67.0%, p<0.0001
time to viral-, ↓59.1%, p<0.0001, early treatment

Goldman et al., NEJM, doi:10.1056/NEJMoa2015301 (Peer Reviewed)

Remdesivir for 5 or 10 Days in Patients with Severe Covid-19

Study focused on remdesivir but with results for HCQ in the supplementary appendix, showing 9% death with HCQ versus 12% control, unadjusted relative risk uRR 0.78, p = 0.46.

death, ↓22.3%, p=0.46

Kuderer et al., Lancet, June 20, 2020, doi:10.1016/S0140-6736(20)31187-9 (preprint 5/28) (Peer Reviewed)

Clinical impact of COVID-19 on patients with cancer (CCC19): a cohort study

Retrospective 928 cancer patients, showing HCQ OR 1.06 [0.51-2.20]. HCQ+AZ OR 2.93 [1.79-4.79]. The relative risks of different therapies suggest that the results are overly affected by confounding by indication. Authors note: HCQ+AZ might not be the cause of increased mortality, but instead these were given to patients with more severe COVID-19.

death, ↑134.2%, p<0.0001, HCQ+AZ
(odds ratio converted to relative risk)

Gianfrancesco et al., Annals of the Rheumatic Diseases, 79:7, 859-866, doi:10.1136/annrheumdis-2020-217871 (Peer Reviewed)

Characteristics associated with hospitalisation for COVID-19 in people with rheumatic disease: data from the COVID-19 Global Rheumatology Alliance physician-reported registry

Analysis of rheumatic disease patients showing no significant association between antimalarial therapy and hospitalisation, OR=0.94 [0.57-1.57], p=0.82 after adjustments.

hospitalization, ↓3.3%, p=0.82
(odds ratio converted to relative risk)

Risch, American Journal of Epidemiology, kwaa093, 27 May 2020, doi:10.1093/aje/kwaa093 (Peer Reviewed) (meta analysis - not included in study count)

Early Outpatient Treatment of Symptomatic, High-Risk Covid-19 Patients that Should be Ramped-Up Immediately as Key to the Pandemic Crisis

Five studies, including two controlled clinical trials, have demonstrated significant outpatient treatment efficacy.

Ip et al., PLOS ONE, doi:10.1371/journal.pone.0237693 (Peer Reviewed)

Hydroxychloroquine and Tocilizumab Therapy in COVID-19 Patients - An Observational Study

Retrospective study of late stage use on 2,512 hospitalized patients showing no significant differences in associated mortality for patients receiving any HCQ during the hospitalization (HR, 0.99 [95% CI, 0.80-1.22]), HCQ alone (HR, 1.02 [95% CI, 0.83-1.27]), or HCQ+AZ (HR, 0.98 [95% CI, 0.75-1.28]). Misclassification is possible due to manual abstraction of EHR data. They observed a change in the prescribing patterns of HCQ during the study timeframe. Confounding by indication.

death, ↓1.0%, p=0.93

ICMR, Indian Council of Medical Research (Advisory) (not included in the study count)

Revised advisory on the use of Hydroxychloroquine (HCQ) as prophylaxis for SARS-CoV-2 infection

Healthcare workers on HCQ prophylaxis less likely to get COVID. Significant dose-response relationship. Extends recommended HCQ prophylaxis to asymptomatic household contacts of cases and frontline workers. Degree of benefit not quantified.

Mehra et al., The Lancet, May 22, 2020, doi: 10.1016/S0140-6736(20)31180-6 (Peer Reviewed) (not included in the study count)

Hydroxychloroquine or chloroquine with or without a macrolide for treatment of COVID-19: a multinational registry analysis

Incorrect at first read (implausible death, ventilation, and population numbers). This paper was retracted.

Singh et al., medRxiv, doi:10.1101/2020.05.12.20099028 (Preprint)

Outcomes of Hydroxychloroquine Treatment Among Hospitalized COVID-19 Patients in the United States- Real-World Evidence From a Federated Electronic Medical Record Network

EHR analysis of 3,372 hospitalized COVID-19 patients not showing a significant difference for mortality or the risk of mechanical ventilation. Subject to the limitations of EHR analysis. Misclassification is possible. Confounding by indication.

death, ↓5.0%, p=0.72
ventilation, ↓19.0%, p=0.26

Kim et al., medRxiv, doi:10.1101/2020.05.13.20094193 (Preprint)

Treatment Response to Hydroxychloroquine, Lopinavir/Ritonavir, and Antibiotics for Moderate COVID 19: A First Report on the Pharmacological Outcomes from South Korea

Retrospective of 97 moderate cases. Time to viral clearance significantly shorter for HCQ+antibiotic. Preprint withdrawn pending peer review.

hospitalization time, ↓51.0%, p=0.01
time to viral-, ↓56.0%, p=0.005

Ahmad et al., doi:10.1101/2020.05.18.20066902 (Preprint)

Doxycycline and Hydroxychloroquine as Treatment for High-Risk COVID-19 Patients: Experience from Case Series of 54 Patients in Long-Term Care Facilities

54 patients in long term care facilities. 6% death with HCQ+AZ compared to 22% using a naive indirect comparison.

Macias et al., medRxiv, 10.1101/2020.05.16.20104141 (Preprint)

Similar incidence of Coronavirus Disease 2019 (COVID-19) in patients with rheumatic diseases with and without hydroxychloroquine therapy

Very small retrospective study of rheumatic disease patients, sample size is too small for statistical significance (HCQ 0.5-4.0%, no-HCQ 0.4-2.7%). Confirmed cases were 1 HCQ and 2 no-HCQ, confirmed+likely cases were 1 HCQ and 3 no-HCQ. 1 HCQ and 2 no-HCQ patients were admitted to hospital. We do not think a conclusion can be drawn based on these sample sizes.

There are very significant differences between the groups, for example 30% of the HCQ group have SLE vs. 2.5% of the no-HCQ group. SLE patients have a 5.7 times relative risk of pneumonia according to [1], whereas the relative risk with glucocorticoids and TNF-α inhibitors is significantly lower [2]. Two more recent studies with rheumatic disease/autoimmune condition patients provide higher confidence.

[1] ncbi.nlm.nih.gov/pmc/articles/PMC4516647/

[2] academic.oup.com/rheumatology/article/52/1/53/1830871

hospitalization, ↓25.5%, p=1.00
COVID-19 case, ↑49.0%, p=0.53

Yu et al., Science China Life Sciences, 2020 May 15, 1-7, doi:10.1007/s11427-020-1732-2 (Peer Reviewed)

Low Dose of Hydroxychloroquine Reduces Fatality of Critically Ill Patients With COVID-19

Retrospective, 550 critically ill patients. 19% fatality for HCQ versus 47% for non-HCQ, RR 0.395, p=0.002.

The levels of inflammatory cytokine IL-6 were significantly reduced from 22.2 pg/mL to 5.2 pg/mL (p<0.05) at the end of the treatment in the HCQ group but there was no change in the control group.

death, ↓60.5%, p=0.002

Mahévas et al., BMJ 2020, 369, doi: https://doi.org/10.1136/bmj.m1844 (Peer Reviewed)

Clinical efficacy of hydroxychloroquine in patients with covid-19 pneumonia who require oxygen: observational comparative study using routine care data

Observational study of 181 patients with advanced disease requiring oxygen showing no benefit for HCQ. Power of study appears too low to support conclusions [1].

None of the 15 patients receiving HCQ+AZ were transferred to intensive care or died compared to 23% overall.

[1] bmj.com/content/369/bmj.m1844/rapid-responses

death, ↑20.0%, p=0.75

Okour et al., Journal of Pharmacokinetics and Pharmacodynamics, May 13, 2020, doi:10.1007/s10928-020-09689-x (Peer Reviewed)

Hydroxychloroquine and azithromycin as potential treatments for COVID-19; clinical status impacts the outcome

Odds of PCR-positive decrease by 53% for each unit increase in HCQ log-concentration. Similarly, the odds decrease by 61%, and by 12% for each day increase, and for azithromycin co-treatment, respectively. Computes the minimum HCQ concentration needed based on severity, and corresponding dosage regimens. A loading dose is found to be important. For LRTI and URTI patients the addition of AZ is needed. Extended analysis of Gautret et al. using the observed HCQ concentrations and pharmacokinetic analysis to compute concentrations for all days.

Cassione et al., Annals of the Rheumatic Diseases, doi:10.1136/annrheumdis-2020-217717 (Letter)

COVID-19 infection in a northern-Italian cohort of systemic lupus erythematosus assessed by telemedicine

Survey of 165 SLE patients, 127 on HCQ. 8 patients with suspected COVID-19 and 4 confirmed cases. No mortality, one ICU case. 7 patients had no symptoms despite contact with a COVID-19 patient.

No adjustment for concomitant medications or severity of SLE. Confounding by indication.

COVID-19 case, ↑49.6%, p=0.59

Shabrawishi et al., medRxix, doi:10.1101/2020.05.08.20095679 (Preprint)

Negative nasopharyngeal SARS-CoV-2 PCR conversion in response to different therapeutic interventions

Retrospective 93 hospitalized patients in Saudi Arabia showing a non-statistically significant 15% reduction in PCR positive results at day 5, RR 0.85, p = 0.65. The treatment group had significantly more severe illness and significantly more male patients.

risk of no virological cure at day 5, ↓14.7%, p=0.66

Rosenberg et al., JAMA, May 11, 2020, doi:10.1001/jama.2020.8630 (Peer Reviewed)

Association of Treatment With Hydroxychloroquine or Azithromycin With In-Hospital Mortality in Patients With COVID-19 in New York State

Restrospective observational late stage study showing no significant differences but calling for clinical trials.

Zervos et al. [1] point out serious limitations that they say should be corrected on the record: patients receiving HCQ with or without AZ were overall sicker on presentation and had multiple other risk factors including much higher risk based on ethnicity; patients receiving HCQ were more likely to be obese, diabetic, have chronic lung disease, and cardiovascular conditions; yet these sicker patients had approximately the same mortality rates compared to patients with a milder course of the disease and less risk factors. However, the authors conclude that "there are no significant benefits." It is noteworthy that HCQ was associated with a significant survival benefit in a larger cohort of patients from New York City as reported by Mikami et al [2].