Table 2b. Chloroquine or Hydroxychloroquine With or Without Azithromycin: Selected Clinical Data

Last Updated: October 09, 2020

The information in this table may include data from preprints or articles that have not been peer reviewed. This section will be updated as new information becomes available. Please see ClinicalTrials.gov for more information on clinical trials that are evaluating CQ, HCQ, and AZM.

The Panel has reviewed other clinical studies of HCQ with or without AZM and studies of CQ for the treatment of COVID-19.1-11 These studies have limitations that make them less definitive and informative than the studies discussed here. The Panel’s summaries and interpretations of some of those studies are available in the archived versions of the COVID-19 Treatment Guidelines.

Table 2b. Chloroquine or Hydroxychloroquine With or Without Azithromycin: Selected Clinical Data
Study Design Methods Results Limitations and Interpretation
Randomised Evaluation of COVID-19 Therapy (RECOVERY) Trial12

Open-label RCT with multiple arms, including a control arm; in 1 arm, hospitalized patients received HCQ (n = 11,197) 

 

This is a preliminary report that has not yet been peer reviewed. 

Key Inclusion Criteria: 

  • Clinically suspected or laboratory-confirmed SARS-CoV-2 infection

Key Exclusion Criteria: 

  • Patients with prolonged QTc intervals were excluded from HCQ arm. 

Interventions:

  • HCQ 800 mg at entry and at 6 hours, then HCQ 400 mg every 12 hours for 9 days or until discharge 
  • Usual SOC 

Primary Endpoint:

  • All-cause mortality at Day 28 after randomization 

Number of Participants:

  • HCQ (n = 1,561) and SOC (n = 3,155) 
  • Study enrollment ended early after investigators and trial-steering committee concluded that the data showed no benefit for HCQ. 

Participant Characteristics:

  • Mean age was 65 years in both arms; 41% of patients were aged ≥70 years. 
  • 90% of patients had laboratory-confirmed SARS-CoV-2 infection. 
  • 57% of patients had ≥1 major comorbidity: 27% had diabetes mellitus, 26% had heart disease, and 22% had chronic lung disease. 
  • At randomization, 17% of patients were receiving invasive mechanical ventilation or ECMO, 60% were receiving oxygen only (with or without noninvasive ventilation), and 24% were receiving neither. 
  • Use of AZM or another macrolide during the follow-up period was similar in both arms, as was use of dexamethasone. 

Outcomes: 

  • No significant difference in 28-day mortality between the 2 arms; 418 patients (26.8%) in HCQ arm and 788 patients (25.0%) in SOC arm had died by Day 28 (RR 1.09; 95% CI, 0.96–1.23; P = 0.18). 
  • A similar 28-day mortality for HCQ patients was reported during the post hoc exploratory analysis that was restricted to the 4,234 participants (90%) who had a positive SARS-CoV-2 test result. 
  • Patients in HCQ arm were less likely to survive hospitalization and had a longer median time to discharge than patients in SOC arm.  
  • Patients who received HCQ and who were not on invasive mechanical ventilation at baseline had an increased risk of requiring intubation and an increased risk of death. 
  • At the beginning of the study, the researchers did not record whether a patient developed a major cardiac arrhythmia after study enrollment; however, these data were later collected for 698 patients (44.7%) in HCQ arm and 1,357 patients (43.0%) in SOC arm.
  • No differences between the arms in the frequency of supraventricular tachycardia, ventricular tachycardia or fibrillation, or instances of AV block that required intervention. 

Limitations: 

  • Not blinded 
  • Information on occurrence of new major cardiac arrythmia was not collected throughout the trial. 

Interpretation: 

  • HCQ does not decrease 28-day all-cause mortality when compared to the usual SOC in hospitalized patients with clinically suspected or laboratory-confirmed SARS-CoV-2 infection.  
  • Patients who received HCQ had a longer median length of hospital stay, and those who were not on invasive mechanical ventilation at the time of randomization were more likely to require intubation or die during hospitalization if they received HCQ. 
Hydroxychloroquine and Hydroxychloroquine Plus Azithromycin for Mild or Moderate COVID-1913
Open-label, 3-arm RCT in hospitalized patients (n = 667) 

Key Inclusion Criteria: 

  • Aged ≥18 years 
  • Clinically suspected or laboratory-confirmed SARS-CoV-2 infection 
  • Mild or moderate COVID-19 
  • Duration of symptoms ≤14 days 

Key Exclusion Criteria: 

  • Need for >4 L of supplemental oxygen or ≥40% FiO2 by face mask 
  • History of ventricular tachycardia 
  • QT interval ≥480 ms 

Interventions: 

  • HCQ 400 mg twice daily for 7 days plus SOC  
  • HCQ 400 mg twice daily plus AZM 500 mg daily for 7 days plus SOC 
  • SOC alone 

Primary Endpoint: 

  • Clinical status at Day 15, as assessed by a 7-point ordinal scale among the patients with confirmed SARS-CoV-2 infection 

Ordinal Scale Definitions:

  1. Not hospitalized, no limitations 
  2. Not hospitalized, with limitations 
  3. Hospitalized, not on oxygen 
  4. Hospitalized, on oxygen 
  5. Hospitalized, oxygen administered by HFNC or noninvasive ventilation 
  6. Hospitalized, on mechanical ventilation 
  7. Death 

Number of Participants:

  • Modified ITT analysis included patients with laboratory-confirmed SARS-CoV-2 infection (n = 504). 

Participant Characteristics:

  • Mean age was 50 years. 
  • 58% of patients were men. 
  • At baseline, 58.2% of patients were ordinal level 3; 41.8% were ordinal level 4. 
  • Median time from symptom onset to randomization was 7 days. 
  • 23.3% to 23.9% of patients received oseltamivir. 

Outcomes: 

  • No significant difference between the odds of worse clinical status at Day 15 for patients in HCQ arm (OR 1.21; 95% CI, 0.69–2.11; P = 1.00) and patients in HCQ plus AZM arm (OR 0.99; 95% CI, 0.57–1.73; P = 1.00). 
  • No significant differences in secondary outcomes of the 3 arms, including progression to mechanical ventilation during the first 15 days and mean number of days “alive and free of respiratory support.” 
  • A greater proportion of patients in HCQ plus AZM arm (39.3%) and HCQ arm (33.7%) experienced AEs than those in SOC arm (22.6%). 
  • QT prolongation was more common in patients who received HCQ plus AZM or HCQ alone than in patients who received SOC alone, but fewer patients in SOC arm had serial electrocardiographic studies performed during the follow-up period. 

Limitations:  

  • Not blinded 
  • Follow-up period was restricted to 15 days. 

Interpretation:

  • Neither HCQ alone nor HCQ plus AZM improved clinical outcomes at Day 15 after randomization among hospitalized patients with mild or moderate COVID-19. 
Hydroxychloroquine Versus Standard of Care for Mild or Moderate COVID-1914
Multicenter, randomized, open-label trial (n = 150) 

Key Inclusion Criteria: 

  • Aged ≥18 years 
  • Laboratory-confirmed SARS-CoV-2 infection  

Key Exclusion Criteria: 

  • Severe conditions, including heart, liver, or kidney disease 
  • Inability to take oral medications 
  • Pregnancy or breastfeeding 

Interventions:  

  • HCQ 1,200 mg once daily for 3 days, then HCQ 800 mg once daily for 2 weeks (in patients with mild or moderate COVID-19) or 3 weeks (in patients with severe disease) 
  • SOC 

Primary Endpoint: 

  • Negative conversion of SARS-CoV-2 by Day 28 

Number of Participants:

  • HCQ (n = 75) and SOC (n = 75)  

Participant Characteristics: 

  • Patients were randomized at a mean of 16.6 days after symptom onset. 
  • 99% of patients had mild or moderate COVID-19. 

Outcomes: 

  • HCQ arm and SOC arm had similar negative PCR conversion rates within 28 days (85.4% of participants vs. 81.3% of participants) and similar times to negative PCR conversion (median of 8 days vs. 7 days). 
  • No difference in the probability of symptom alleviation between the arms in the ITT analysis. 

Limitations: 

  • Unclear how the overall rate of symptom alleviation was calculated 
  • Study did not reach target sample size. 

Interpretation:  

  • This study demonstrated no difference in the rate of viral clearance between HCQ and SOC. 
High-Dose Chloroquine Versus Low-Dose Chloroquine15
Randomized, double-blind, Phase 2b study in hospitalized adults (n = 81) 

Key Inclusion Criteria: 

  • Aged ≥18 years 
  • Clinically suspected COVID-19 
  • At least 1 of the following conditions: 
    • Respiratory rate >24 rpm 
    • Heart rate >125 bpm 
    • SpO2 <90% on room air 
    • Shock 

Interventions: 

  • CQ 600 mg twice daily for 10 days (high dose) 
  • CQ 450 mg twice daily for 1 day, then CQ 450 mg for 4 days (low dose) 

Primary Endpoint:

  • Mortality by Day 28 

Number of Participants: 

  • High-dose CQ (n = 41) and low-dose CQ (n = 40) 
  • Planned study sample size was 440 participants, but study was stopped by the study’s DSMB. 

Participant Characteristics: 

  • All patients also received ceftriaxone plus AZM. 
  • 89.6% of patients received oseltamivir. 

Outcomes: 

  • Overall fatality rate was 27.2%. 
  • Mortality by Day 13 was higher in high-dose arm than in low-dose arm (death occurred in 16 of 41 patients [39%] vs. in 6 of 40 patients [15%]; P = 0.03). This difference was no longer significant after controlling for age (OR 2.8; 95% CI, 0.9–8.5). 
  • Overall, QTcF >500 ms occurred more frequently in high-dose arm (18.9% of patients) than in low-dose arm (11.1%). 
  • In the high-dose arm, 2 patients experienced ventricular tachycardia before death. 

Limitations:  

  • More older patients and more patients with a history of heart disease were randomized into the high-dose arm than into the low-dose arm. 

Interpretation: 

  • Despite the small number of patients enrolled, this study raises concerns about an increased risk of mortality when high-dose CQ is administered in combination with AZM and oseltamivir. 
Hydroxychloroquine in Nonhospitalized Adults with Early COVID-1916
Randomized, placebo-controlled trial in the United States and Canada (n = 491) 

Key Inclusion Criteria:

  • ≤4 days of symptoms that were compatible with COVID-19 
  • Either laboratory-confirmed SARS-CoV-2 infection or high-risk exposure within the previous 14 days 

Key Exclusion Criteria:

  • Aged <18 years 
  • Hospitalized 
  • Receipt of certain medications 

Interventions:  

  • HCQ 800 mg once, then HCQ 600 mg in 6 to 8 hours, then HCQ 600 mg once daily for 4 days 
  • Placebo 

Primary Endpoints: 

  • Planned primary endpoint was ordinal outcome by Day 14 in 4 categories: not hospitalized, hospitalized, ICU stay, or death.
  • Because event rates were lower than expected, a new primary endpoint was defined: change in overall symptom severity over 14 days, assessed on a 10-point, self-reported, visual analog scale 

Number of Participants: 

  • Contributed to primary endpoint data: HCQ (n = 212) and placebo (n = 211) 

Participant Characteristics: 

  • 241 patients were exposed to people with COVID-19 through their position as health care workers (57%), 106 were exposed through household contacts (25%), and 76 had other types of exposure (18%). 
  • Median age was 40 years. 
  • 56% of patients were women.  
  • Only 3% of patients were Black.  
  • Very few patients had comorbidities: 11% had hypertension, 4% had diabetes, and 68% had no chronic medical conditions. 
  • 56% of patients were enrolled on Day 1 of symptom onset. 
  • 341 participants (81%) had either a positive PCR result or a high-risk exposure to a PCR-positive contact. 

Outcomes: 

  • Compared to the placebo recipients, HCQ recipients had a nonsignificant 12% difference in improvement in symptoms between baseline and Day 14 (-2.60 vs. -2.33 points; P = 0.117). 
  • Ongoing symptoms were reported by 24% of those in HCQ arm and 30% of those in the placebo arm at Day 14 (P = 0.21). 
  • No difference in the incidence of hospitalization (4 patients in the HCQ arm vs. 10 patients in placebo arm); 2 of 10 placebo participants were hospitalized for reasons that were unrelated to COVID-19. 
  • A higher percentage of patients in HCQ arm experienced AEs than patients in placebo arm (43% vs. 22%; P < 0.001). 

Limitations:  

  • This study enrolled a highly heterogenous population. 
  • Only 227 of 423 participants (53.7%) were confirmed PCR-positive for SARS-CoV-2. 
  • Changing the primary endpoint without a new power calculation makes it difficult to assess whether the study is powered to detect differences in outcomes between the study arms. 
  • This study used surveys for screening, symptom assessment, and adherence reporting. 
  • Visual analog scales are not commonly used, and their ability to assess acute viral respiratory infections in clinical trials has not been validated. 

Interpretation:  

  • The study has some limitations, and it did not find evidence that early administration of HCQ reduced symptom severity in patients with mild COVID-19. 
Hydroxychloroquine in Nonhospitalized Adults with Mild COVID-1917
Open-label RCT in Spain (n = 353) 

Key Inclusion Criteria: 

  • Laboratory-confirmed SARS-CoV-2 infection 
  • <5 days of mild COVID-19 symptoms 

Key Exclusion Criteria: 

  • Moderate to severe COVID-19 
  • Severe liver or renal disease 
  • History of cardiac arrhythmia 
  • QT prolongation 

Interventions: 

  • HCQ 800 mg on Day 1, then HCQ 400 mg once daily for 6 days 
  • No antiviral treatment 

Primary Endpoint: 

  • Reduction in SARS-CoV-2 viral load, assessed using nasopharyngeal swabs on Days 3 and 7 

Secondary Endpoints: 

  • Disease progression up to Day 28 
  • Time to complete resolution of symptoms 

Number of Participants: 

  • ITT analysis: HCQ (n = 136) and control (n = 157)
  • 60 patients were excluded from the ITT analysis due to negative baseline RT-PCR, missing RT-PCR at follow-up visits, or consent withdrawal.

Participant Characteristics: 

  • Mean age was 41.6 years. 
  • 67% of patients were woman. 
  • Majority of patients were health care workers (87%). 
  • 53% of patients reported chronic health conditions. 
  • Median time from symptom onset to enrollment was 3 days (IQR 2–4 days).  
  • Most common COVID-19 symptoms were fever, cough, and sudden olfactory loss. 

Outcomes: 

  • No significant difference in viral load reduction between control arm and HCQ arm at Day 3 
    (-1.41 vs. -1.41 log10 copies/mL; difference of 0.01; 95% CI, -0.28 to 0.29), or at Day 7 (-3.37 vs. -3.44 log10 copies/mL; difference of -0.07; 95% CI, -0.44 to 0.29). 
  • No difference in the risk of hospitalization between control arm and HCQ arm (7.1% vs. 5.9%; risk ratio 0.75; 95% CI, 0.32–1.77). 
  • No difference in the median time from randomization to the resolution of COVID-19 symptoms between the 2 arms (12.0 days in control arm vs. 10.0 days in HCQ arm; P = 0.38). 
  • A higher percentage of participants in the HCQ arm than in the control arm experienced AEs during the 28-day follow-up period (72% vs. 9%). Most common AEs were GI disorders and “nervous system disorders.” 
  • SAEs were reported in 12 patients in control arm and 8 patients in HCQ arm. SAEs that occurred among patients in HCQ arm were not deemed to be related to the drug. 

Limitations: 

  • Open-label, non-placebo-controlled trial  
  • Study design allowed for the possibility of drop-outs in control arm and over-reporting of AEs in HCQ arm. 
  • The intervention changed during the study; the authors initially planned to include HCQ plus DRV/COBI. 
  • The majority of the participants were relatively young health care workers. 

Interpretation: 

  • Early administration of HCQ to patients with mild COVID-19 did not result in improvement in virologic clearance, a lower risk of disease progression, or a reduced time to symptom improvement. 
Observational Study on Hydroxychloroquine With or Without Azithromycin18
Retrospective, multicenter, observational study in a random sample of inpatients with COVID-19 from the New York Department of Health (n = 1,438) 

Key Inclusion Criteria: 

  • Laboratory-confirmed SARS-CoV-2 infection 

Interventions: 

  • HCQ plus AZM 
  • HCQ alone 
  • AZM alone 
  • Neither drug 

Primary Endpoint: 

  • In-hospital mortality 

Secondary Endpoint: 

  • Cardiac arrest and arrhythmia or QT prolongation on an ECG 

Number of Participants: 

  • HCQ plus AZM (n = 735), HCQ alone (n = 271), AZM alone (n = 211), and neither drug (n = 221) 

Participant Characteristics: 

  • Patients in the treatment arms had more severe disease at baseline than those who received neither drug. 

Outcomes: 

  • In adjusted analyses, patients who received 1 of the 3 treatment regimens did not show a decreased in-hospital mortality rate when compared with those who received neither drug. 
  • Patients who received HCQ plus AZM had a greater risk of cardiac arrest than patients who received neither drug (OR 2.13; 95% CI, 1.12–4.05). 

Limitations: 

  • This study has the inherent limitations of an observational study, including residual confounding from confounding variables that were unrecognized and/or unavailable for analysis. 

Interpretation: 

  • Despite the limitations discussed above, these findings suggest that although HCQ and AZM are not associated with an increased risk of in-hospital death, the combination of HCQ and AZM may be associated with an increased risk of cardiac arrest. 
Observational Study of Hydroxychloroquine Versus No Hydroxychloroquine in New York City19
Observational study in hospitalized adults with COVID-19 at a large medical center (n = 1,376) 

Key Inclusion Criteria: 

  • Laboratory-confirmed SARS-CoV-2 infection 

Key Exclusion Criteria: 

  • Intubation, death, or transfer to another facility within 24 hours of arriving at the emergency department 

Interventions: 

  • HCQ 600 mg twice daily on Day 1, then HCQ 400 mg once daily for 4 days 
  • No HCQ 

Primary Endpoint: 

  • Time from study baseline (24 hours after patients arrived at the emergency department) to intubation or death 

Number of Participants: 

  • Received HCQ (n = 811) and did not receive HCQ (n = 565) 

Participant Characteristics: 

  • HCQ recipients were more severely ill at baseline than those who did not receive HCQ.

Outcomes: 

  • Using propensity scores to adjust for major predictors of respiratory failure and inverse probability weighting, the study demonstrated that HCQ use was not associated with intubation or death (HR 1.04; 95% CI, 0.82–1.32). 
  • No association between concomitant use of AZM and the composite endpoint of intubation or death (HR 1.03; 95% CI, 0.81–1.31). 

Limitations: 

  • This study has the inherent limitations of an observational study, including residual confounding from confounding variables that were unrecognized and/or unavailable for analysis. 

Interpretation: 

  • The use of HCQ for treatment of COVID-19 was not associated with harm or benefit in a large observational study.
Observational Cohort Study of Hydroxychloroquine Versus No Hydroxychloroquine in France20
Retrospective, observational cohort study in hospitalized adults with severe COVID-19 pneumonia at 4 tertiary care centers (n = 181) 

Key Inclusion Criteria: 

  • Aged 18 to 80 years 
  • Laboratory-confirmed SARS-CoV-2 infection 
  • Required supplemental oxygen 

Key Exclusion Criteria: 

  • Started HCQ before hospital admission 
  • Received tocilizumab, LPV/RTV, or RDV within 48 hours of admission 
  • Organ failure requiring immediate ICU admission 
  • ARDS 

Interventions: 

  • HCQ 600 mg once daily 
  • No HCQ 

Primary Endpoint: 

  • Survival without transfer to the ICU at Day 21 

Secondary Endpoints: 

  • Overall survival rate at Day 21 
  • Survival rate without ARDS at Day 21 
  • Weaning from oxygen by Day 21 
  • Discharge from hospital to home or rehabilitation by Day 21 

Number of Participants: 

  • Received HCQ within 48 hours (n = 84), received HCQ beyond 48 hours (n = 8), and did not receive HCQ (n = 89) 

Participant Characteristics: 

  • In the HCQ arm, 18% of patients received concomitant AZM. 

Outcomes: 

  • In the inverse probability of treatment-weighted analysis, there was no difference in survival rates without ICU transfer at Day 21 between the HCQ arm (76% of participants) and the non-HCQ arm (75%). 
  • No difference between the arms in the secondary outcomes of overall survival rate and survival rate without ARDS at Day 21. 

Limitations: 

  • This was a retrospective, nonrandomized study. 

Interpretation: 

  • There was no difference in the rates of clinically important outcomes between patients who received HCQ within 48 hours of hospital admission and those who did not. 
Retrospective Cohort Study of Hydroxychloroquine Versus No Hydroxychloroquine in Detroit, Michigan21
Comparative, retrospective cohort study in hospitalized patients with COVID-19 in the Henry Ford Health System in Michigan (n = 2,541) 

Key Inclusion Criteria: 

  • Laboratory-confirmed SARS-CoV-2 infection 

Interventions: 

  • HCQ 400 mg twice daily for 1 day, then 200 mg twice daily for 4 days 
  • AZM 500 mg for 1 day, then 250 mg once daily for 4 days 
  • HCQ plus AZM, at the above doses 
  • Neither drug 

Primary Endpoint: 

  • In-hospital mortality 

Number of Participants: 

  • HCQ alone (n = 1,202), AZM alone (n = 147), HCQ plus AZM (n = 783), and neither drug (n = 409) 

Participant Characteristics: 

  • HCQ plus AZM was reserved for patients with severe COVID-19 and minimal cardiac risks. 
  • Median patient age was 64 years (IQR 53–76 years); 51% of patients were men, 56% were African American, and 52% had a BMI ≥30. 
  • Median time to follow-up was 28.5 days (IQR 3–53 days). 
  • The mSOFA score was not available for 25% of patients. 
  • Corticosteroids were given to 79% of patients in the HCQ alone arm, 74% of patients in the HCQ plus AZM arm, and 35.7% of those on neither drug. 

Outcomes: 

  • Overall, crude mortality was 18.1%. When broken down by the different arms, mortality was 13.5% in HCQ alone arm, 20.1% in HCQ plus AZM arm, 22.4% in AZM alone arm, and 26.4% in the arm that received neither drug (P < 0.001). 
  • Mortality HRs were analyzed using a multivariable Cox regression model; the arm that received neither drug was used as the reference. HCQ alone decreased the mortality HR by 66% (P < 0.001). HCQ plus AZM decreased the mortality HR by 71% (P < 0.001).
  • Other predictors of mortality were age ≥65 years (HR 2.6; 95% CI, 1.9–3.3); White race (HR 1.7; 95% CI, 1.4–2.1); chronic kidney disease (HR 1.7; 95% CI, 1.4–2.1); reduced O2 saturation level on admission (HR 1.6; 95% CI, 1.1–2.2); and ventilator use at admission (HR 2.2; 95% CI, 1.4–3.0). 
  • A propensity-matched Cox regression result suggested a mortality HR of 0.487 for patients who received HCQ (95% CI, 0.285–0.832, P = 0.009). 

Limitations: 

  • This study evaluated 1 health care system with an institutional protocol for HCQ and AZM use. 
  • Because the study was not randomized and not blinded, there is a possibility of residual confounding. 
  • There was a lower rate of ICU admission among patients who did not receive HCQ, which suggests that this group may have received less aggressive care. 
  • Given that the RECOVERY trial showed that dexamethasone use conferred a survival benefit, it is possible that the findings were confounded by the imbalance in corticosteroid use among the arms. 

Interpretation: 

  • This study reported a mortality benefit in hospitalized patients with COVID-19 who received either HCQ alone or HCQ plus AZM compared to patients who received neither drug. However, there were substantial imbalances in corticosteroid use among the arms, which may have affected mortality.  
  • Because the study was retrospective and observational, it cannot control for confounders.