ABSTRACT
OBJECTIVE: Comparative assessment of pediatric intensive care. DESIGN: Prospective multicenter study. SETTING: Four pediatric intensive care units in Moscow, the Russian Federation. PATIENTS: Consecutive unselected admissions (n = 583), < or = 14 yrs of age, in a 6-month period. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Effectiveness was defined as the ratio of observed to predicted mortality, based on prediction by Pediatric Risk of Mortality (PRISM) severity of illness scoring. Efficiency (on the day of admission only) depended on either a mortality risk of > 1% or the administration of intensive care unit-dependent therapies. In all four hospitals, observed mortality rates were higher than expected, with a range of standardized mortality ratios between 1.10 and 1.83 (mean 1.32). The excess mortality was found in the low- and medium-risk strata (risk of mortality of < 1% to 15%). Admission efficiency ratings did not fluctuate greatly between institutions (mean 60.4%, range 55.7 to 65.9). CONCLUSIONS: We provided a quantitative description and assessment of pediatric intensive care in Moscow. Moderate efficiency may reflect a low threshold for ICU admission due to poor nurse/patient ratios on the wards. Effectiveness in the low- and medium-risk strata is below standard, as compared with a Western reference population. Excess mortality was concentrated in the low- and medium-risk strata, and can only partially be explained by the inclusion of co-morbidity. Future analysis should focus on specific treatment protocols, protocol adherence, and the determination of infectious and therapeutic complications.
