Abstract
OBJECTIVE
To investigate whether deranged physiologic parameters at admission to an ICU in Tanzania are associated with in-hospital mortality and compare single deranged physiologic parameters to a more complex scoring system.
DESIGN
Prospective, observational cohort study of patient notes and admission records. Data were collected on vital signs at admission to the ICU, patient characteristics, and outcomes. Cutoffs for deranged physiologic parameters were defined a priori and their association with in-hospital mortality was analyzed using multivariable logistic regression.
SETTING
ICU at Muhimbili National Hospital, Dar es Salaam, Tanzania.
PATIENTS
All adults admitted to the ICU in a 15-month period.
MEASUREMENTS AND MAIN RESULTS
Two hundred sixty-nine patients were included: 54% female, median age 35 years. In-hospital mortality was 50%. At admission, 69% of patients had one or more deranged physiologic parameter. Sixty-four percent of the patients with a deranged physiologic parameter died in hospital compared with 18% without (p < 0.001). The presence of a deranged physiologic parameter was associated with mortality (adjusted odds ratio, 4.64; 95% CI, 1.95-11.09). Mortality increased with increasing number of deranged physiologic parameters (odds ratio per deranged physiologic parameter, 2.24 [1.53-3.26]). Every individual deranged physiologic parameter was associated with mortality with unadjusted odds ratios between 1.92 and 16.16. A National Early Warning Score of greater than or equal to 7 had an association with mortality (odds ratio, 2.51 [1.23-5.14]).
CONCLUSION
Single deranged physiologic parameters at admission are associated with mortality in a critically ill population in a low-income country. As a measure of illness severity, single deranged physiologic parameters are as useful as a compound scoring system in this setting and could be termed "danger signs." Danger signs may be suitable for the basis of routines to identify and treat critically ill patients.
Collapse