No Evidence for Time-Dependent Treatment Effect of Oseltamivir for Adults Hospitalized With Influenza A
Year: 2025
Type: article
Abstract: To the editor—We read with interest the recent article by Lewis et al [1]. In their study, starting treatment with oseltamivir early (on the day of hospital admission) compared with delayed (after the first day) or no treatment was associated with the primary outcome, lower peak pulmonary disease severity during hospital admission. Additionally, in a sensitivity analysis that excluded those patients who did not receive oseltamivir, the authors found that early treatment was associated with lower peak pulmonary severity compared with delayed treatment, which the authors interpreted as evidence that early oseltamivir administration was beneficial. We commend the authors for this study but believe any assumption that the observed associations are explained by a causal relationship is premature, given the high risk for unmeasured confounding and selection bias inherent in the study design. Using the collated data provided in the article, we sequentially compared the primary outcome in the no-treatment cohort to both the early-treatment and delayed-treatment cohorts (Figure 1). This revealed no significant difference between the early- and no-treatment groups in the proportion of patients whose peak pulmonary severity worsened during hospital admission (14.2%, 59 of 415 versus 17.7%, 22 of 124; absolute risk difference, −3.5%; 95% confidence interval [CI], −11.0 to 4.0; Fisher exact P value = .390). However, more patients in the delayed-treatment cohort (31.6%, 95 of 301) experienced worsening peak pulmonary severity compared with patients in the no-treatment cohort (17.7%; 22 of 124; absolute risk difference, 13.8%; 95% CI, 5.3 to 22.4; Fisher exact P value = .004). Further, in analysis of the differences in the secondary binary outcomes, we found no differences in either the early- or delayed-treatment cohorts compared with the no-treatment cohort, aside from more patients requiring acute vasopressors or kidney replacement therapy in the delayed-treatment group compared with the no-treatment group (11.8%, 34 of 287 versus 5.1%, 6 of 118; absolute risk difference, 6.8%; 95% CI, 1.3 to 12.2; Fisher exact P value = .043). Percent of patients who had worsening peak pulmonary severity during hospital admission by early treatment (n = 415), delayed treatment (n = 301), and no treatment (n = 124) with oseltamivir. We observed that delayed treatment was associated with worse peak pulmonary disease compared with both early treatment and no treatment. If a time-dependent causal effect was the primary factor to explain observed outcome differences between the groups, one would expect the early-treatment group to have better outcomes than both the delayed- and no-treatment groups and the delayed-treatment group to have better outcomes, or at least no worse outcomes, than the no-treatment group. While we appreciate that the authors adjusted for several potential confounders, the additional analyses reported here led us to infer that the observed between-group differences more likely resulted from unmeasured confounders and selection bias as opposed to a treatment effect from early oseltamivir therapy. While the authors report that their results are consistent with those from other observational studies [1], the true clinical effect of oseltamivir for patients hospitalized for influenza remains unclear due to a paucity of randomized, controlled trials in this population [2, 3]. We suggest that this study provides additional equipoise regarding the effectiveness of oseltamivir in hospitalized patients and that randomized, controlled trials are needed. Potential conflicts of interest. C. W. J. has been an investigator on studies funded by AstraZeneca, Vapotherm, Abbott, and Ophirex. All other authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
Source: Clinical Infectious Diseases
DOI: 10.1093/cid/ciaf117
Authors: Joseph Naji, Christopher W Jones, Brian W Roberts
Language: English
Subfield: Epidemiology
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W4408324869