This cohort study was designed as undertaken to evaluate the protective effect of Arsenicum album 30C against COVID-19.
Participants were enrolled in a homeopathy intervention (HI) cohort (who received Arsenicum album) or in a non-intervention (NI) cohort (who received no systematic intervention) from COVID-19 containment areas of Delhi. Individuals of age 5 years or above were given four medicated pills of Arsenicum album 30C, while those from 1 to 5 years old were given two medicated pills in each dose.
The analysis included 10,180 individuals residing in 11 COVID-19 containment areas in Delhi, out of which 6,590 individuals were in the HI cohort and 3,590 individuals were in the NI cohort. The overall protective effect of Arsenicum album 30C was 83% (95% confidence interval [CI], 76.77 to 88.17): 45 cases per 6,590 (8.34 per 10,000 person-weeks) in the Arsenicum album 30C group versus 143 cases per 3,590 (45.01 per 10,000 person-weeks) in the NI cohort. The protective effect of Arsenicum album 30C against laboratory-confirmed COVID-19 was 74% (95% CI, 55.08 to 85.41): 18 cases per 6,590 (3.32 per 10,000 person-weeks) in the Arsenicum album 30C group versus 38 cases per 3,590 (11.85 per 10,000 person-weeks) in the NI cohort.
The authors concluded that the use of Arsenicum album 30C was associated with some protection against probable and laboratory-confirmed COVID-19 in a containment-zone setting. Randomized controlled trials are needed to confirm or refute these results.
It is remarkable, I feel, that the authors conclude Arsenicum album 30C was associated with some protection. All too often enthusiasts of homeopathy claim a causal link where there is none – but not this time, and I wonder why.
Unfortunately, I was unable to read the full text of the paper (it’s behind a paywall and I would be grateful for anyone to make it available). Thus, I cannot comment on one of the most crucial questions related to the study: how were the patients divided into the two groups?
It is clear that it was not by randomization. Yet only randomization would have created two fully comparable groups. The most likely explanation for the findings of this trial is therefore that the two groups differed in terms of one or more prognostic factors. This would explain why a group of patients receiving a placebo (Arsenicum album C30 is a dilution of Arsenic at a ratio of 1: 1000000000000000000000000000000000000000000000000000000000000 and therefore is a pure placebo [unless, of course, one believes in homeopathic magic) experience different outcomes from those receiving nothing.
As I said, the answer can only be found by studying the precise selection criteria used in this study. Until this is cleared up, I can only say three things for sure:
- A causal link between the treatment and the result is highly unlikely.
- It is regrettable that researchers do not use randomization for potentially important trials.
- It seems unethical to encourage placebo use for the prevention of a serious illness.
I just received the full text from one of the authors. This is what they say about the allocation of the participants:
“Participants were enrolled in two cohorts: the homeopathy intervention (HI) cohort and the non-intervention (NI) cohort. Recruitment to the cohorts was at cluster level (containment area): the clusters were allotted to the HI cohort or the NI cohort as per convenience.”
I am afraid, this tells me very little, and my concerns noted above still apply.
A few other points are of relevance:
- The study was conducted between April and August 2020. This begs the question of why it took 2 years to publish the findings.
- The outcomes were verified via telephone. This means that social desirability might have influenced the results.
- The paper also confirms that there were many important differences between the groups that might have prognostic significance.
- The conclusion at the end of the paper does imply causality in stronger terms than the abstract: “The use of Arsenicum album 30C may help protect against COVID-19 infection. Randomized controlled trials are needed
to confirm or refute our findings.