[Paper Review] Efficacy of Hydroxychloroquine as Prophylaxis for Covid-19

Maria Angélica Ehara Watanabe|arXiv (Cornell University)|Jul 18, 2020

COVID-19 Clinical Research Studies7 references4 citations

TL;DR

This study re-analyzes the Boulware et al. randomized, double-blind, placebo-controlled trial using time-dependent symptom onset data and simple linear regression. It finds that hydroxychloroquine significantly reduces symptomatic COVID-19 incidence when administered within 0–2 days of viral exposure, with relative risk reductions of 72% (day 0), 48.9% (day 1), and 29.3% (day 2), but no significant effect beyond 3 days post-exposure.

ABSTRACT

Limitations in the design of the experiment of Boulware et al[1] are considered in Cohen[2]. They are not subject to correction but they are reported for readers' consideration. However, they made an analysis for the incidence based on Fisher's hypothesis test for means while they published detailed time dependent data which were not analyzed, disregarding an important information. Here we make the analyses with this time dependent data adopting a simple regression analysis. We conclude their randomized, double-blind, placebo-controlled trial presents statistical evidence, at 99% confidence level, that the treatment of Covid-19 patients with hydroxychloroquine is effective in reducing the appearance of symptoms if used before or right after exposure to the virus. For 0 to 2 days after exposure to virus, the estimated relative reduction in symptomatic outcomes is 72% after 0 days, 48.9% after 1 day and 29.3% after 2 days. For 3 days after exposure, the estimated relative reduction is 15.7% but results are not statistically conclusive and for 4 or more days after exposure there is no statistical evidence that hydroxychloroquine is effective in reducing the appearance of symptoms. Our results show that the time elapsed between infection and the beginning of treatment is crucial for the efficacy of hydroxychloroquine as a treatment to Covid-19.

Motivation & Objective

• To re-analyze the Boulware et al. clinical trial data using time-dependent symptom onset data not fully utilized in the original study.
• To assess whether hydroxychloroquine reduces symptomatic COVID-19 incidence when administered shortly after viral exposure.
• To determine the time window during which hydroxychloroquine is most effective as a prophylactic intervention.
• To provide statistical evidence for hydroxychloroquine's efficacy based on regression analysis of symptom incidence over time.

Proposed method

• The study re-analyzes time-dependent symptom onset data from the Boulware et al. trial, which recorded symptom development at multiple time points post-exposure.
• A simple linear regression model is applied to assess the relationship between time since exposure and symptom incidence across treatment and placebo groups.
• The analysis focuses on relative risk reduction in symptomatic outcomes at different intervals post-exposure (0–4+ days).
• Statistical significance is evaluated at the 99% confidence level to determine efficacy.
• The study contrasts its findings with the original analysis, which used Fisher’s exact test on aggregated data without time resolution.
• Results are reported as relative risk reductions with 99% confidence intervals to quantify treatment effect over time.

Experimental results

Research questions

• RQ1 Does hydroxychloroquine reduce the incidence of symptomatic COVID-19 when administered within 0–2 days after viral exposure?
• RQ2 Is there a statistically significant reduction in symptom onset when hydroxychloroquine is initiated 3 days after exposure?
• RQ3 Does the efficacy of hydroxychloroquine diminish when treatment begins 4 or more days after exposure?
• RQ4 How does the time interval between viral exposure and treatment initiation affect the relative risk of developing symptoms?
• RQ5 Can a time-dependent regression model better capture the efficacy of prophylactic hydroxychloroquine than the original aggregate analysis?

Key findings

• Hydroxychloroquine use within 0 days of exposure is associated with a 72% relative reduction in symptomatic outcomes, with 99% confidence.
• For treatment initiated 1 day after exposure, the relative reduction in symptomatic cases is 48.9%, statistically significant at the 99% confidence level.
• When treatment begins 2 days after exposure, the estimated relative reduction is 29.3%, also significant at 99% confidence.
• For treatment initiated 3 days after exposure, the relative reduction is 15.7%, but the result is not statistically conclusive.
• No statistical evidence supports hydroxychloroquine efficacy when treatment begins 4 or more days after exposure.
• The time elapsed between viral exposure and treatment initiation is a critical determinant of hydroxychloroquine’s prophylactic effectiveness.