Outcomes from coronavirus disease 2019 among hospitalized patients depend on demographic factors, clinical presentation, and the level of hospital care provided. The investigators conducted a prospective observational analytical study to describe clinico-epidemiological characteristics, hospital-care interventions, and clinical outcomes among adults admitted with SARS-CoV-2 infection in Nepal, and to assess whether age statistically mediates relationships between selected hospital interventions and survival.
Consecutive adult patients with laboratory-confirmed SARS‑CoV‑2 infection were recruited at a tertiary care teaching hospital in Nepal between March 30, 2022 and June 30, 2023. The cohort comprised 348 adults who were followed prospectively until hospital discharge or in‑hospital death.
Clinical and epidemiological data, comorbidities, symptoms, and in-hospital interventions were recorded. Associations with survival were examined using Chi-square or Fisher’s exact tests where appropriate, and multivariable binary logistic regression for adjusted analyses. Mediation analyses employed Hayes’ PROCESS Macro version 4.2, Model 4, with 5,000 bias-corrected bootstrap samples to test whether age group statistically mediated relationships between specific interventions (steroid therapy, oxygen therapy, respiratory support) and survival.
Of 348 participants, 277 (79.6%) survived and 71 (20.4%) died during the index hospitalization. Older adults represented 41.1% of the study population. The most frequent comorbidities were hypertension (29.3%) and diabetes mellitus (23.0%). The leading presenting symptoms reported were fever (81.0%), cough (75.9%), and dyspnea (48.0%).
In unadjusted analyses, receipt of steroid therapy, oxygen therapy, and respiratory support were each significantly associated with survival. The report provides these associations as unadjusted findings; more detailed effect estimates and stratified counts are presented in the manuscript.
When potential confounders were accounted for in multivariable binary logistic regression, age group and the presence of dyspnea at presentation remained independently associated with hospital survival. The manuscript reports these covariates as the key independent predictors after adjustment for other measured factors.
Mediation testing targeted three intervention–survival relationships: steroid therapy–survival, oxygen therapy–survival, and respiratory support–survival. Using Hayes’ PROCESS Model 4 with 5,000 bias-corrected bootstrap samples, the study found that age group partially statistically mediated the association between steroid therapy and survival. The reported indirect effect for age mediating the steroid–survival relationship was −0.1371 with a 95% bootstrap confidence interval of −0.3053 to −0.0154. The direct effect of steroid therapy on survival remained significant (c′ = −2.8670; p = 0.0049), indicating partial rather than complete mediation.
No significant age-mediated indirect effects were observed for oxygen therapy or respiratory support in the analyses reported.
The authors explicitly note that this mediation finding is statistical in nature and should not be interpreted as proving a causal pathway, since age is a pre‑existing characteristic that temporally precedes treatment exposure.
Ethical approvals were obtained from the Research Ethics Committee of Lincoln University College, Malaysia; the Nepal Health Research Council, Kathmandu; and the Institutional Review Committee of Chitwan Medical College Teaching Hospital, Nepal. An amendment to include the mediation objective and use of Hayes’ PROCESS Macro was reviewed and approved by the Lincoln University College committee. Written informed consent was obtained from participants or legally authorized representatives when needed. Participant information was handled confidentially and only de‑identified data were used for analysis. The authors state that all data are available in the manuscript and/or supporting information files.
This prospective hospital series from Nepal identifies age group and dyspnea as independent predictors of in‑hospital survival among adults with SARS‑CoV‑2 infection. Unadjusted associations linked steroid therapy, oxygen therapy, and respiratory support with survival, but mediation analysis suggested age partially accounts for the relationship between steroid use and survival while the direct effect of steroids remained significant. No age-mediated effects were detected for oxygen or respiratory support.
The report cautions that the mediation result is statistical rather than causal because age temporally precedes treatment exposure; the observational design limits causal inference. The manuscript provides the detailed statistical methods, effect estimates, and bootstrapped confidence intervals for the mediation models.
Overall, the study highlights the role of age and respiratory presentation in hospital outcomes for SARS‑CoV‑2 in a resource‑constrained setting and documents ethical oversight and data availability for further scrutiny.