Ivermectin in combination with doxycycline for treating COVID-19 symptoms: a randomized trial

Objective: We evaluated whether ivermectin combined with doxycycline reduced the time to clinical recovery in adults with COVID-19 infection.

Methods: This was a randomized, blinded, placebo-controlled trial in patients with mild-to-moderate COVID-19 symptoms randomly assigned to treatment (n = 200) and placebo groups (n = 200). The primary outcome was duration from treatment to clinical recovery. Secondary outcomes included disease progression and persistent COVID-19 positivity by RT-PCR.

Results: Among 556 screened patients, 400 were enrolled and 363 completed follow-up. The mean patient age was 40 years, and 59% were men. The median recovery time was 7 (4–10) and 9 (5–12) days in the treatment and placebo groups, respectively (hazard ratio, 0.73; 95% confidence interval, 0.60–0.90). The number of patients who recovered within 7 days was 61% and 44% in the treatment and placebo groups, respectively (hazard ratio, 0.06; 95% confidence interval, 0.04–0.09). The proportion of patients who remained RT-PCR positive on day 14 and whose disease did not progress was significantly lower in the treatment group than in the placebo group.

Conclusions: Patients with mild-to-moderate COVID-19 infection treated with ivermectin plus doxycycline recovered earlier, were less likely to progress to more serious disease, and were more likely to be COVID-19 negative by RT-PCR on day 14.

Methods
Patients

Patients were enrolled between 1 June 2020 and 30 August 2020. The inclusion criteria for enrollment were as follows: age greater than 18 years, positive COVID-19 RT-PCR test within 3 days prior to enrollment, and mild to moderately severe COVID-19 infection. Patients who were unable to receive oral medications, were pregnant or breastfeeding, had severe COVID-19 symptoms (defined as tachypnea [>30 breaths/minute] and hypoxia [SpO2 < 90%] at room air and requiring supplemental oxygen), were admitted to intensive care or high-dependency units, or had known hypersensitivities to ivermectin or doxycycline were excluded from participation.

Trial design

This was an investigator-initiated, placebo-controlled trial. The independent data safety monitoring board of Dhaka Medical College monitored treatment safety and efficacy. Routine investigations were performed at the pathology laboratory of Dhaka Medical College. RT-PCR tests were performed at the virology laboratory of Dhaka Medical College and Bangabandhu Sheikh Mujib Medical University. Ivermectin, doxycycline, and placebo preparations were provided by a popular pharmaceutical company.

Eligible patients were randomly assigned to the treatment group or the placebo group at a 1:1 ratio on day 1 of the trial. Group assignment was not stratified according to disease severity.

The allocation schedule was created with a list of random numbers generated using a random number generator program by the head of the Department of Medicine of Dhaka Medical College. Group assignment was concealed in sequentially numbered, opaque, sealed envelopes. The randomization code was maintained by the pharmaceutical company. Both the investigators and the patients were blinded to the treatment allocation.

Decoding was performed at the end of the trial under the supervision of the principal of the institute.

The trial was conducted in accordance with the principles of the Declaration of Helsinki and Good Clinical Practice guidelines, and was approved by the ethical review committee of Dhaka Medical College Hospital (ERC-DMC/ECC/2020/117). Dhaka Medical College is the largest tertiary care teaching hospital and the largest COVID-19 treatment hospital in Bangladesh. The trial was registered retrospectively at ClinicalTrials.gov (identifier: NCT04523831). Written informed consent was obtained from all patients prior to participation. The study is reported according to the Equator Network Guidelines.

Study interventions

The treatment group received a single dose of ivermectin 12 mg and doxycycline 100 mg, twice daily for 5 days, in addition to standard of care. Standard of care included administration of paracetamol, antihistamines, cough suppressants, vitamins, oxygen therapy according to indication and need, low molecular weight heparin according to indication, appropriate other broad-spectrum antibiotics, remdesivir injection, other antiviral drugs, and other drugs for associated comorbid conditions.

The placebo group received placebo in addition to standard of care.

Experimental procedures

The baseline demographic and clinical characteristics of patients were recorded. The date of random assignment was considered as day 1, and all patients received their initial study intervention on day 1. Outpatients were followed daily until at least 3 days of clinical recovery were observed. Clinical recovery was defined as a normal body temperature of 36.1C to 37.2C maintained for at least 3 days, significantly improved respiratory symptoms (respiratory rate < 25/minute, no dyspnea), and oxygen saturation greater than 93% without assisted oxygen inhalation, as recommended by the World Health Organization and the national guidelines of Bangladesh.

Hospitalized patients were followed from day 1 through day 14 or until discharge or clinical improvement, whichever occurred later. Clinical status and vital signs (including respiratory status) were recorded daily. Adverse events as defined by the Medical Dictionary for Regulatory Activities (MedDRA) were documented.

Laboratory tests were performed on day 1, and included the following: complete blood count, random blood glucose, creatinine, alanine transaminase, C-reactive protein, ferritin, and D-dimer. Chest radiography or chest CT scanning were performed as needed. For outpatients, test results were documented at the next visit. Hospitalized patients were tested on days 1 and 7 and when ordered by the treating physicians. COVID-19 RT-PCR testing was performed 14 days after the initial positive test in all patients.

Outcome measures

The primary outcome was the number of days required for clinical recovery from day 1. Clinical recovery was divided into three categories: early recovery within 7 days, intermediate recovery within 7 to 11 days, and late improvement requiring 12 or more days for recovery. Secondary outcomes were disease progression through mild, moderate, severe, or death, and the proportion of patients who continued to test positive for COVID-19 on day 14. Adverse drug reactions (adverse events assumed to be caused by the study intervention) were also recorded. Mild disease was defined as symptoms of an upper respiratory tract viral infection, including mild fever, dry cough, sore throat, nasal congestion, headache, muscle pain, anosmia, and malaise. Moderate respiratory symptoms such as cough and shortness of breath were observed without signs of severe pneumonia. Severe disease included severe dyspnea, tachypnea (> 30 breaths/minute), and hypoxia (SpO2 < 90% at room air). These classifications were according to the World Health Organization and national guidelines of Bangladesh. The co-investigators assessed the outcome, graded the disease, and documented adverse reactions.

Statistical methods

We sought to enroll as many patients as possible in a short time. As a result, a priori power calculations were not performed. Instead, 200 patients were enrolled in each group and it was assumed that 20% of patients in each group would be lost to follow-up. A post-hoc power calculation [2SD2 (Zα/2+ Zβ) 2/d2]14 using a two-tailed test, an alpha level of 0.05, and SD 6 indicated that the enrollment of 141 patients in each group (170 before 20% attrition) would result in an 80% chance of detecting a 2-day difference in median recovery time.

Intention-to-treat analysis was performed. The difference in median recovery time was determined using Kaplan–Meier analysis and a log-rank test. Details of patients who were lost to follow-up, had died, or had withdrawn from the trial due to adverse effects were censored on the final study day. Differences in the clinical improvement in the three disease categories were analyzed using logistic regression. Hazard ratios (HRs) were calculated using Cox regression analysis. Categorical variables are presented as n (%), normally distributed continuous variables as mean (standard deviation [SD]), and skewed continuous variables as median (interquartile range [IQR]).

Subgroup analysis was performed using the logistic regression analysis. HRs were calculated using Cox regression analysis. Age group, sex, fever, cough, respiratory distress, hypertension, diabetes, and presence of any comorbidity were adjusted. Subgroups consisted of age (less than vs. more than 60 years), sex, presence or absence of any comorbidities, mild disease, moderate disease, hospital admission, and interval between disease onset and drug application interval (less than vs. more than 4 days from the onset of the disease). A forest plot was constructed to evaluate differences in the primary outcome among the subgroups. No corrections were made for multiple comparisons. All tests were two-tailed and the alpha was set at 0.05. Data were analyzed using SPSS software, version 20 (Armonk, NY: IBM Corp, USA).