high prevalence of potential drug-drug interactions among patients treated with off-label therapies for covid-19

Background: during the first wave of the covid-19 pandemic, severe patients were treated with the off-label drugs hydroxychloroquine and lopinavir/ritonavir. the aim of the study was to determine the prevalence of potential drug-drug interactions (ddis) between hydroxychloroquine, lopinavir/ritonavir and concomitant medications used by hospitalized patients treated for covid-19 in costa rica. methods: we included all patients that received lopinavir/ritonavir or hydroxychloroquine as treatment for covid-19. clinical pharmacists reviewed the prescription profile of each patient and determined the probability and severity of any ddi through two databases (the lexi-interact program) and the micromedex online interaction checker. a logistic regression model was used to identify variables associated with the occurrence of potential ddis. results: we identified a total of 108 potential ddis in 34 inpatients (n=34). at least one of these ddis occurred in 27 patients (79.4%; 95% ci: 65.8-92.9%). a total of 70 ddis (64.8%) were classified as clinically relevant (grade d or x) by the clinical pharmacists. only the number of concomitant drugs was associated with the occurrence of a probable ddi. the most common drugs associated with any ddi were fentanyl (n=12, 11.1%), midazolam (n=11, 10.2%), and insulin (n=10, 10.2%). conclusion: a large proportion of patients treated with hydroxychloroquine and lopinavir/ritonavir for severe covid-19 were at risk for clinical meaningful ddis.