. 2023 Nov 16;10(1):e22299. doi: 10.1016/j.heliyon.2023.e22299

Table 4.

Bivariate (n = 3869) and multivariate (n = 3823) analysis of factors associated with self-medication on the recommendation of a family member.

Medicine	Bivariate analysis	Multivariate analysis according to
Medicine	Bivariate analysis	Self-medication	Positive COVID-19	Gender ♂	Age
Chloroquine	0.234	0.251	0.976	0.377	0.316
Hydroxychloroquine	0.968	0.899	0.887	0.414	0.265
Azithromycin	0.037 +	0.051	0.956	0.338	0.325
Penicillin	0.365	0.323	0.974	0.385	0.261
Other ABx	<0.001 +	<0.001 +	0.942	0.513	0.806
Warfarin	0.472	0.321	0.789	0.460	0.249
Ivermectin	0.664	0.348	0.934	0.381	0.270
Paracetamol	<0.001 +	<0.001 +	0.880	0.609	0.381
Ibuprofen	<0.001 +	<0.001 +	0.738	0.961	0.878
Other NSAIDs	<0.001 +	<0.001 +	0.990	0.559	0.248

COVID-19: coronavirus disease 2019. ABx: antibiotics. NSAIDs: non-steroidal anti-inflammatory drugs. P-values were obtained using generalized linear models, with the Poisson distribution and a logarithmic link function, and models for robust variance adjusted. In the multivariate model, we adjusted for educational level and country of residence. The sign indicates whether the prevalence of drug use was found to be higher (+) or lower (−).