Prevalence, Determinants, and Outcomes of Prolonged Mechanical Ventilation in Patients Admitted into the Intensive Care Unit at Tenwek Hospital, Kenya

Abstract

Abstract In intensive care units (ICUs) worldwide, prolonged mechanical ventilation (PMV) affects approximately 30% of mechanically ventilated patients, exacting a heavy toll through heightened morbidity, mortality, and escalating healthcare costs. Yet, despite advances in critical care, the predictors and outcomes of PMV remain poorly understood in low-resource settings, where resource constraints amplify these challenges. This study sought to ascertain the prevalence, determinants, and consequences of PMV among mechanically ventilated adult ICU patients at Tenwek Hospital, Kenya. A retrospective cohort design was employed to review the medical records of adult ICU patients who required mechanical ventilation for more than 7 consecutive days from January to December 2024. Data collection encompassed three key domains: ventilatory parameters, patient–ventilator interactions, and clinical status, supplemented by pertinent laboratory and imaging findings. Descriptive statistics characterised the cohort, while inferential analyses—comprising multivariable logistic regression for PMV predictors and Cox proportional hazards models for time to extubation—identified associations at the p < 0.05 significance level. Variable selection was guided by a conceptual framework, with adjusted odds ratios (ORs) and 95% confidence intervals (CIs) reported; model robustness was maintained by ensuring at least 10 events per predictor. Of 173 mechanically ventilated adults, 72.3% experienced PMV (>7 days). Multivariable logistic regression pinpointed acute respiratory distress syndrome (ARDS; adjusted OR = 5.25, 95% CI: 2.35–11.75) and chronic obstructive pulmonary disease (COPD; adjusted OR = 5.28, 95% CI: 2.38–11.73) as the foremost predictors, trailed by pneumonia (adjusted OR = 1.82, 95% CI: 0.80–4.14) and sepsis (adjusted OR = 1.56, 95% CI: 0.69– 3.52). Daily sedation vacations curtailed PMV odds by 81% (adjusted OR = 0.19, 95% CI: 0.08–0.46), while early mobility protocols diminished them by 37% (adjusted OR = 0.63, 95% CI: 0.28–1.42). Cox analysis revealed an ICU mortality rate of 69.9%, with moderate disease severity associated with delayed extubation (hazard ratio [HR] = 0.53, 95% CI: 0.30–0.93, p = 0.025). In conclusion, PMV imposes a formidable burden in this Kenyan ICU context, driven by respiratory pathologies and modifiable management gaps. We recommend embedding structured daily sedation interruptions and early mobility protocols into routine ICU practice to shorten ventilation duration, reduce resource strain, and improve patient outcomes in similar low-resource environments.