Variability in Gastrostomy Tube Placement for Intracerebral Hemorrhage Patients at US Hospitals.

OBJECTIVE: We sought to characterize the variability among US hospitals with regard to gastrostomy tube placement for inpatients with intracerebral hemorrhage (ICH).

METHODS: Using the Nationwide Inpatient Sample, we examined variations in the annual rate of gastrostomy tube placement from 2002 to 2011 for ICH patients admitted to hospitals with 30 or more annual ICH admissions. We then directly compared, among these hospitals, their individual frequencies of gastrostomy tube placement for ICH patients over the same time period. To quantify variability among hospitals, we used multilevel multivariable regression models accounting for a hospital random effect, adjusted for patient-level and hospital-level factors predictors of placement.

RESULTS: Gastrostomy tube placement rates did not significantly change from 2002 to 2011 (9.8 to 8.7 per 100 admissions; P trend = .57). Among 690 hospitals with 38,080 ICH hospitalizations during this period, 10.4% of patients had a gastrostomy tube placed (n = 3976). Variation in the rate of placement among individual hospitals was large, from 0% to 34.4% (interquartile range 5.7%-13.6%). For a regression model controlling for patient and hospital covariates, the median odds ratio was 1.36 (95% confidence interval 1.28-1.44), indicating that if a patient moved from one hospital to another with a higher intrinsic propensity of placement, there was a 1.36-fold median increase in the odds of receiving a gastrostomy tube, independent of patient and hospital factors.

CONCLUSIONS: Variation in gastrostomy tube placement rates across hospitals is large and may in part reflect differences in local practice patterns or patient and surrogate preferences.