Abstract Rationale Respiratory High Dependency Units (HDUs) provide critical support for patients with severe respiratory compromise who do not yet require invasive ventilation or full intensive care. However, limited evidence exists regarding factors predicting ICU transfer or mortality in such settings, especially in resource-limited contexts. This study aimed to identify clinical predictors associated with ICU transfer and in-hospital mortality among patients admitted to a respiratory HDU. Methods A retrospective analysis was conducted among 200 patients admitted to a tertiary hospital respiratory HDU. Clinical and laboratory variables including heart rate, respiratory rate, oxygen requirement, serum creatinine, pH, and vasopressor use were evaluated. Univariable and multivariable logistic regressions were performed to determine predictors of ICU transfer and in-hospital mortality, expressed as adjusted odds ratios (AOR) with 95% confidence intervals (CI). Results A total of 200 patients admitted to the High Dependency Unit (HDU) were included in the study. The mean age of patients was 67.89 ± 13.49 years, and 66.5% were male. 37% required ICU transfer and 12.5% died during hospitalization. Tachycardia (AOR = 2.30, 95% CI 1.17–4.53; p = 0.016) and tachypnea (AOR = 3.86, 95% CI 1.94–7.65; p 0.001) independently predicted transfer to the ICU. For in-hospital mortality, tachycardia (AOR = 3.30, 95% CI 1.19–9.16; p = 0.022), tachypnea (AOR = 5.23, 95% CI 1.89–14.51; p = 0.0015), vasopressor use (AOR = 4.21, 95% CI 1.41–12.52; p = 0.010), and elevated serum creatinine (AOR = 6.15, 95% CI 1.13–33.45; p = 0.036) were significantly associated with mortality. Conclusions Tachycardia and tachypnea consistently predicted both ICU transfer and in-hospital mortality among respiratory HDU patients, highlighting their role as early warning signs of deterioration. Vasopressor use and elevated serum creatinine further identified patients at high risk of death. In settings where continuous monitoring and advanced investigations are limited, these simple bedside indicators can guide timely escalation of care and optimize resource allocation. This abstract is funded by: None
Timshina et al. (Fri,) studied this question.