Table 1.
Characteristics of the included studies
| Author [Ref.] | ID | Country | Design | Participants (n) | Treatment | Comparison | Outcomesa | Resultsb | Risk of Bias (tool)c |
|---|---|---|---|---|---|---|---|---|---|
| Studies with a “no CQ/HCQ” control group | |||||||||
| Chen et al. [11] | NCT04261517 | China | Single-centre RCT | Patients with COVID-19 (n = 30) | HCQ sulphate 400 mg, oral formulation, daily for 5 days (n = 15) | Standard care (n = 15) | Negative conversion rate of SARS-CoV-2 nucleic acid in respiratory pharyngeal swab at day 7 or death at day 14; adverse events | Negative nucleic acid negative in 86.7% of the HCQ group and 93.3% of the control group (P > 0,05). No deaths at last follow-up. Averse events in 26.7% of the HCQ group and 20% of the control group (p > .05) | High risk of bias (Rob2) |
| Geleris et al. [12] | NA | USA | Single-centre prospective observational study | Hospitalized patients with COVID-19 (n = 1376) | HCQ 600 mg twice on day 1, then 400 mg daily for 4 days; with or without AZ 500 mg on day 1 and then 250 mg daily for 4 more days (n = 811) | Treatment regimens not including HCQ (n = 565) | Time to intubation or death | HCQ group: higher risk of primary outcome (hazard ratio - HR 2.37; 95% CI,1.84 to 3.02). Findings not confirmed by neither multivariate analysis (HR 1.00; 95% CI, 0.76–1.32) nor propensity score analyses (HR 1.04; 95% CI, 0.82 to 1.32). No association between AZ use and primary outcome (HR 1.03; 95% CI, 0.81 to 1.31). | Moderate risk of bias (ROBINS-I tool) |
| Huang et al. [13] | NA | China | Single-centre RCT | Hospitalized patients with confirmed COVID-19 (n = 22) | CQ 500 mg orally twice daily for 10 days (n = 10) | Lopinavir/Ritonavir 400 mg/100 mg twice daily for ten days (n = 12) | Viral clearance; improvements on lung CT; length of hospitalization; adverse events | Discharge at day 14 was registered for 100% of the patients in the CQ group vs the 50% of the Lopinavir/ Ritonavir group. Five patients experienced a total of 9 adverse events in the CQ group. | High risk of bias (Rob2) |
| Mahevas et al. [14] | NA | France | Multicentre nonrandomized prospective study | Hospitalized patients with COVID-19 (n = 173) | HCQ 600 mg daily, started in the first 48 h after hospitalization (n = 84) | No HCQ in the first 48 h (n = 89) | Survival without transfer to the ICU at day 21; QT | Survival rate without transfer to the ICU at day 21 was 76% in the treatment group and 75% in the control group (weighted hazard ratio 0.9, 95% confidence interval 0.4 to 2.1; Eight (10%) of the HCQ group experienced electrocardiographic modifications requiring discontinuation. | Moderate risk of bias (ROBINS-I) |
| Rosenberg et al. [15] | NA | USA | Multicentre retrospective observational study | Hospitalized patients with COVID-19 (n = 1438) | HCQ alone (most common dosage 400 mg first dose then 200 mg daily); HCQ in combination with AZ; AZ alone (most common dosage 500 mg daily) (n = 1217) | Treatment regimens not including HCQ or AZ (n = 221) | In-hospital mortality; adverse events | The HCQ + AZ group had the highest in-hospital mortality (25.7%). No significant differences in mortality between HCQ + AZ (aHR, 1.35 [95% CI, 0.76–2.40]), HCQ alone (aHR, 1.08 [95% CI, 0.63–1.85]), or AZ alone (aHR, 0.56 [95% CI, 0.26–1.21]), compared with neither drug in the adjusted analysis. Cardiac arrest more likely in HCQ + AZ than neither drug (aOR, 2.13 95% CI, 1.12–4.05), and in HCQ alone vs AZ alone (aOR, 2.97 95% CI, 1.56–5.64) | Moderate risk of bias (ROBINS-I) |
| Tang et al. [16] | ChiCTR2000029868 | China | Multicentre RCT | Hospitalized patients with confirmed COVID–19 (n = 150) | HCQ 1200 mg daily for 3 days, then 800 mg daily for the remaining days (total treatment duration: 2–3 weeks) (n = 75) | Standard care (n = 75) | Negative conversion of SARS–CoV–2 at day 28; adverse events | Similar median time to negative conversion in HCQ group (8 (95% confidence interval 5 to 10) days) vs standard care group (7 (5 to 8) days) (hazard ratio 0.85, 95% C·I 0.58–1.23; P = .34 by log rank test). Adverse events in 9% of HCQ non–recipients vs 30% of HCQ recipients. | Some concerns (Rob2) |
| Yu et al. [17] | NA | China | Single-centre retrospective observational study | Critically ill patients with COVID-19 (n = 550) | HCQ 200 mg twice daily for 7–10 days (n = 48) | Standard care (n = 502) | Mortality; hospital length of stay; level of IL-6 | HCQ associated with a significantly decreased fatality risk (HR: 0.31; 95% CI: 0.16–0.61; P = .001; adjusted HR: 0.36; 95% CI: 0.18–0.75; P = .006). The time of hospital stay before patient death was lower for the HCQ group, compared to NHCQ group (P < .05). The levels of IL-6 were significantly reduced in the HCQ group. | Serious risk of bias (ROBINS-I) |
| Studies without a “no CQ/HCQ” comparison group | |||||||||
| Borba et al. [18] | NCT04323527 | Brazil | Single-centre parallel,double-masked,randomized,phaseIIb clinical trial | Hospitalized patients with severe acute respiratory syndrome and suspected diagnosis of COVID-19 (n = 81) | CQ 600 mg (four tablets containing 150 mg CQ base) twice daily for 10 days (n = 41) | Day 0: CQ 450 mg (three tablets containing 150 mg CQ base) and 1 placebo tablet twice daily; Day 1 to day 4: 450 mg (three tablets containing 150 mg) plus 1 placebo tablet once a day followed by 4 placebo tablets once daily; Day 5 to day 9: four placebo tablets twice daily (n = 40) | Reduction in lethality at day 28 by at least 50% in the high-dosage group compared with the low-dosage group; adverse events | Lethality until day 13 in 39% of high-dosage group vs. 15% of low-dosage group. QTc > 500 milliseconds in 18.9% of the high-dosage group vs. 11.1% of low-dosage group. Ventricular tachycardia in 2.7% of the high-dosage group experienced ventricular tachycardia. CK increase in 50.0% of high-dosage group vs 31.6% of low-dosage group. The high-dosage group was associated with lethality (odds ratio, 3.6; 95% CI, 1.2–10.6). | Low risk of bias (Rob2) |
| Gautret et al. [20] | NA | France | Single-centre single arm retrospective observational study | Patients with positive PCR test for SARS-CoV-2 (n = 80) | HCQ 200 mg, three times daily for 10 days and AZ 500 mg on day and 250 mg daily for the next 4 days | No comparison | Clinical outcome; ICU admission at day 3; results of PCR and cultures; length of stay | The 81.3% of the patients had favourable outcome and were discharged, 15% required oxygen therapy, 93% had negative viral load at day 8 | Poor quality score (NOS) |
| Million et al. [19] | NA | France | Single-centre single arm retrospective observational study | Patients with positive PCR test for SARS-CoV-2, both admitted in day-care or conventional unit. (n = 1061) | HCQ 200 mg 3 times daily for 10 days and AZ 500 mg on day 1 and 250 daily for 4 days | No comparison | Clinical outcome; ICU admission or death; hospitalization ≥10 days; results of PCR and cultures; adverse events | Good clinical outcome and virological cure were achieved in 91.7% of patients at day 10. Adverse events occurred in 2.4% of the patients. | Poor quality score (NOS) |
| Molina et al. [21] | NA | France | Single-centre rospective observational study | Hospitalized patients for COVID-19 (n = 11) | HCQ 600 mg daily for 10 days and AZ 500 mg on day 1 and 250 mg daily from day 2 to day 5 | No comparison | Virological clearance; clinical outcome; prolonged QTc; adverse events | In one patient, HCQ and AZ were discontinued after 4 days because of a prolongation of the QT interval. | Poor quality score (NOS) |
| Studies addressing cardiac adverse events only | |||||||||
| Bessiere et al. [26] | NA | France | Single-centre, retrospective, observational study | Patients admitted in ICU for COVID-19 (n = 40) | HCQ 200 mg, twice a day, for 10 days, with or without AZ 250 mg, daily, for 5 days | No comparison | Prolonged QTc; ΔQTc; adverse events | Prolonged QTc was observed in 14 patients (36%). Median ΔQTc was 35 (10–66). The treatment was ceased in 17.5% of the patients following ECG abnormalities and in the 25% of the patients for acute renal failure. | Poor quality score (NOS) |
| Chorin et al. [24] | NA | USA | Multicentre retrospective observational study | Hospitalized patients with COVID-19 (n = 251) | HCQ 400 mg twice daily on day 1, then 200 mg twice daily for 4 days and AZ 500 mg once daily for 5 days | No comparison | QTc prolongation | QTc > 500 ms in 28/222 (13%) patients with QRS < 120. JTc > 410 in 4/29 (14%) patients with QRS >120. ΔQTc>60 in 51/251 (20%) patients with any QRS | Poor quality score (NOS) |
| Maraj et al. [25] | NA | USA | Single-centre retrospective observational study | Hospitalized patients with COVID-19 (n = 91) | Hydroxychloroquine and azithromycin | No comparison | Significant QTc prolongation; ventricular tachyarrhythmias | Excessive QTc prolongation occurred in 23%. Significant ventricular arrhythmias occurred in 2 patients (1 classic TdP and 1 polymorphic VT degenerating into VF). Additional QT prolonging agents (especially propofol) independently associated with QTc prolongation [any drug: adjusted OR 3.69, CI (1.22, 11.20), p = .02; propofol: adjusted OR 3.28, CI (1.06, 10.17), p = .04]. | Poor quality score (NOS) |
| Mazzanti et al. [22] | NA | Italy | Single-centre prospective observational study | Hospitalized patients with COVID-19 (n = 150) | HCQ for a median of 9 days (IQR 5–11), at a daily dosage of 400 mg (97%), or 600 mg (3%). In 67% of cases, HCQ with LopinavirRitonavir (35%), AZT (26%), or LR + AZT (6%) | No comparison | QTc prolongation | Median ΔQTc 18 ms (IQR 2–34 ms; p < .001; baseline QTc available only for 79 patients); 9% QTc mild prolongation; 4% intermediate and 2% severe (≥500 ms) | Poor quality score (NOS) |
| Mercuro et al. [23] | NA | USA | Single-centre, retrospective, observational study | Hospitalized patients with COVID-19 (n = 90) | HCQ 400 mg twice on day 1, then 400 mg once daily for 4 days (n = 37) | HCQ and azithromycin (n = 53) | ΔQTc; prolonged QTc; adverse events | Patients receiving HCQ alone had a median ΔQTc of 5.5 IQR −15.5–34.25) vs the 23 ms median (IQR 10–40) of HCQ and AZ group; p = .03. The 19% of HCQ alone group vs 21% of combination group developed QTc ≥ 500 ms. The 3% of HCQ alone group vs 13% of combination group had ΔQTc >60 ms. One patient developed torsades de pointes in the combination group. | Serious risk of bias (ROBINS-I) |
| Peng et al. [29] | NA | Malaysia | Single-centre nonrandomized observational study | Hospitalized patients with COVID-19 (n = 13) | HCQ with or without azithromycin | No comparison | Tisdale score and QTc prolongation | Mean Tisdale risk score 7.5 ± 1.45, with 69.2% at intermediate risk of QT prolongation. QT prolongation in 38.5% of the patients | Poor quality score (NOS) |
| Saleh et al. [28] | NA | USA | Multicentre prospective observational study | Hospitalized patients with COVID-19 (n = 201) | CQ: 500 mg oral twice daily at day1, followed by 500 mg once daily for 4 days; or HCQ: 400 mg oral twice daily at day 1 followed by 200 mg twice daily for 4 days; In combination with AZ 500 mg oral or i.v. daily for 5 days (n = 119) | CQ alone: 500 mg oral twice daily at day1, followed by 500 mg once daily for 4 days; HCQ alone: 400 mg oral twice daily at day 1 followed by 200 mg twice daily for 4 days (n = 82) | QT prolongation with torsades de pointes | No cases of torsades de pointes or arrhythmogenic death. Baseline QTc did not differ between monotherapy vs. combination group (440.6 ± 24.9 ms vs. 439.9 ± 24.7 ms, p = .834). QTc was significantly longer in the combination vs monotherapy group (470.4 ± 45.0 ms vs. 453.3 ± 37.0 ms, p = .004). Seven patients (3.5%) required discontinuation due to QTc prolongation. | Serious risk of bias (ROBINS-I) |
| Van den Broek et al. [27] | NA | The Netherlands | Single-centre retrospective observational study | Hospitalized patients suspected for COVID-19 (n = 95) | CQ loading dose 600 mg, followed by 300 mg twice daily (starting 12 h after the loading dose), for a total of 5 days | No comparison | ECG alterations, PR, QRS and QTc intervals | Mean QTc prolongation of 35 ms (95% CI 28–43 ms) using computerized interpretation and 34 ms (95% CI 25–43 ms) on manual interpretation; 23% of the patients had a QTc > 500 ms. Heart rate mean difference was −10 bpm, PR interval mean difference was 8 ms and QRS interval mean difference was 6 ms. | Poor quality score (NOS) |
| Studies on prophylactic use | |||||||||
| Boulware et al. [30] | NA | USA and Canada | Multicentre RCT | Asymptomatic non-hospitalized adults with high or moderate risk exposure to SARS-CoV-2 (n = 821) | HCQ 800 mg (4 tablets) once, then 600 mg (3 tablets) 6 to 8 h later, then 600 mg (3 tablets) daily for 4 days (5 days,19 tablets total) (n = 414) | Placebo (n = 407) | COVID-19 related symptoms | 13.0% of the participants turned symptomatic by 14 days. Symptoms appearance did not differ significantly between those receiving HCQ (11.8%) and those receiving placebo (14.3%), p = .35 | Low risk of bias (Rob-2) |
| Gendelman et al. [33] | NA | Israel | Multicentre retrospective observational study | Individuals tested for COVID-19 (n = 14,520) | HCQ regimens (in use prior to COVID-19 for other clinical reasons) (n = 36) | No HCQ regimens (n = 14,484) | SARS-CoV-2 infection rate | Hydroxychloroquine was prescribed to subjects with SARS-CoV-2 infection com- pared to those without in rate of 0.23% vs 0.25% (p = .877) of the population | Serious risk of bias (ROBINS-I) |
| Konig et al. [34] | NA | USA | Multicentre retrospective observational study | Patients with SLE and COVID-19 (n = 80) | HCQ or chloroquine (in use prior to COVID-19 for other clinical reasons) (n = 51) | No HCQ or chloroquine regimens (n = 29) | SARS-CoV-2 infection rate; hospitalization rate | Fifty-one (64%) patients were already using hydroxychloroquine or chloroquine before SARS-CoV-2 infection. Frequency of hospitalization did not significantly differ between antimalarial users versus non-users | Serious risk of bias (ROBINS-I) |
| Nagaraya et al. [32] | NA | India | Multicentre retrospective observational study | Healthcare workers in “direct contact”(66.9%), “indirect contact”(27.9%) or “no contact” (3.4%) with patients (n = 166) | HCQ | No comparison | Adverse events | At least one adverse event was experienced by 37.9% participants | Poor quality score (NOS) |
| Not peer reviewed, pre-print studies | |||||||||
| Ahmad et al. [43] | NA | USA | Multicentre retrospective observational study | Residents in long-term facilities with suspected or confirmed COVID-19 (n = 54) | Doxycycline 100 mg p.o. twice daily for 7 days and HCQ (200 mg p.o. three times daily for 7 days or 400 mg p.o. twice daily at day 1, then 400 mg daily for 6 days) | No comparison | Clinical recovery | All patients improved within 6 days of DOXY-HCQ initiation. 93% (n = 50) did not display any side-effects of DOXY-HCQ. 2% (n = 1) had a seizure and HCQ was immediately terminated. 9 patients did not complete the 7-day course of DOXY-HCQ due to hospital transfers, death, or side-effects. | Poor quality score (NOS) |
| Carlucci et al. [36] | NA | USA | Single-centre retrospective observational study | Hospitalized patients with confirmed SARS-CoV-2 infection (n = 932) | HCQ 400 mg load followed by 200 mg twice daily for five days, and AZ 500 mg once daily and zinc sulfate 220 mg capsule containing 50 mg elemental zinc twice daily for five days (n = 411) | HCQ 400 mg load followed by 200 mg twice daily for five days and azithromycin 500 mg once daily (n = 521) | Length of hospital stay; mechanical ventilation; oxygen flow rate; FiO2; ICU admission; length of ICU stay; death/hospice; intubation; discharge destination | In the unadjusted analysis, no outcomes were associated with zinc use. In bivariate logistic regression analysis, the addition of zinc sulfate was associated with decreased mortality or transition to hospice (OR 0.511, 95% CI 0.359–0.726), need for ICU (OR 0.545, 95% CI 0.362–0.821) and need for invasive ventilation (OR 0.562, 95% CI 0.354–0.891) | Serious risk of bias (ROBINS-I) |
| Chen et al. [35] | ChiCTR2000029559 | China | Single-centre randomized controlled trial | Hospitalized not severe or critical patients with confirmed COVID-19 (n = 62) | HCQ 200 mg twice daily for 5 days (n = 31) | Standard care (n = 31) | Adverse events and time to clinical recovery | Compared with the control group, the body temperature and cough remission recovery times were significantly shortened in the HCQ treatment group. Two patients had mild adverse reactions in the HCQ treatment group. | Some concerns (Rob2) |
| Ip et al. [42] | NA | USA | Multicentre retrospective observational study | Hospitalized patients with COVID-19 (n = 2512) | Hydroxychloroquine, or hydroxychloroquine with azithromycin, or azithromycin alone (n = 1914) | No HCQ or AZ (n = 598) | In-hospital mortality; adverse events | No significant association between survival and any use of 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 in combination with azithromycin (HR, 0.98 [95% CI, 0.75–1.28]). | Moderate risk of bias (ROBINS-I) |
| Kim et al. [37] | NA | Republic of Korea | Single-centre retrospective observational study | Hospitalized patients with COVID-19 (n = 270) | HCQ 200 mg tablets twice daily with or without antibiotics (n = 22) | Lopinavir 200 mg/ritonavir 50 mg tablets twice daily with or without antibiotics or standard of care (n = 248) | Time to complete or probable viral clearance; time to discharge; time to symptom resolution; adverse events | Time to viral clearance was significantly shorter with HQ plus antibiotics compared to Lopinavir/Ritonavir plus antibiotics (HR 0.49; 95% CI, 0.28–0.87) or conservative treatments (HR, 0.44; 95% CI, 0.25–0.78). Hospital length of stay was also shortest for patients treated with HQ plus antibiotics compared to other treatment groups. Both HQ and Lopinavir/Ritonavir showed side effects, none serious. | Serious risk of bias (ROBINS-I tool) |
| Magagnoli et al. [38] | NA | USA | Multicentre retrospective observational study | Male hospitalized patients with confirmed SARS-CoV-2 infection (n = 368) | HCQ or with azithromycin (n = 210) | No HCQ regimens (n = 158) | Discharge or death; need for mechanical ventilation | Death occurred in the 27.8% of the HC group vs 22.1% of the HC + AZ group, and 11.4% of no HC group. Mechanical ventilation in the 13.3% of the HC group, 6.9% of the HC + AZ group, and 14.1% of the no HC group. Risk of death from any cause was higher in the HC group (aHR, 2.61; 95% CI, 1.10 to 6.17; P = .03). No significant differences in the risk of ventilation among the groups. |
Serious risk of bias (ROBINS-I tool) |
| Mallat et al. [39] | NA | UAE | Single-centre retrospective observational study | Hospitalized patients with confirmed COVID-19 (n = 34) | HCQ 400 mg twice daily for 1 day, then 400 mg daily for 10 days (n = 23) | Non HCQ regimens (n = 11) | Time to SARS-CoV-2 negativity | The time to negativity was significantly longer in patients who received HCQ compared to those who did not receive the treatment (17 [13−21] vs. 10 [4–13] days, p = .023. This effect was confirmed in the adjusted analyses. | Serious risk of bias (ROBINS-I) |
| Ramireddy et al. [40] | NA | USA | Single-centre prospective observational study | Hospitalized patients with COVID-19 or under investigation (n = 98) | HCQ, azithromycin or their combination | No comparison | QTc prolongation | The 12% of the patients reached a critical level of QTc prolongation. Changes in QTc were highest with the combination treatment compared to either drug, especially vs. azithromycin alone (17 ± 39 vs. 0.5 ± 40 ms, p = .07). No patients manifested torsades de pointes. | Poor quality score (NOS) |
| Singh et al. [41] | NA | USA | Multicentre retrospective observational study | Hospitalized patients with COVID-19 (n = 3372) | HCQ treatment, with or without azithromycin (n = 1125) | Non HCQ treatment (n = 2247) | Mortality and need for mechanical ventilation at day 7 and day 14; adverse events | The estimated risk of mortality was similar in the treated vs control group at both day 7 (RR 1; 0.73–1.37) and day 14 (RR 1.04; 0.80–1.36). New ventricular arrythmias or cardiac arrest occurred in the 1.09% of the treated patients, with no significant differences compared to not treated. | Moderate risk of bias (ROBINS-I) |
The table shows the main characteristics of the included studies, as reported by the authors.
Data are reported as percentages, mean ± standard deviation, median (interquartile range) or pointed estimates, as reported by the authors.
AZ: azithromycin; CI: confidence interval; COVID-19: coronavirus disease 2019; CQ: chloroquine; CT: computer tomography; ECG: electrocardiogram; HCQ: hydroxychloroquine; HR: Hazard Ratio; ICU: Intensive care unit; NA: Not available; PCR: Polymerase Chain Reaction; Rob2: Revised tool for Risk of Bias in randomized trials; ROBINS-I: Risk Of Bias in Non-randomized Studies of Interventions; NOS: Newcastle Ottawa Scale; USA: United States of America; UAE: United Arab Emirates.
a
Under the column ‘Outcomes’ the main outcomes of the study and the evaluated safety outcomes are reported.
b
Under the column ‘Results’ the most relevant findings are reported. For synthesis purpose, full results are not reported and can be retrieved from correspondent full-text papers.