The identifier concept: clinical parameters to stratify hospital patient costs for patients with hypertension.

Hospital administrators, nurses, and physicians must strive to improve efficiency as new payment systems constrain health care spending. Our purpose in this project was to confirm the hypothesis that four clinical variables (emergency or intensive care unit [ICU] admission, and blood or plasma product use) could predict hospital costs and outcome for adult patients with hypertension. We analyzed hospital costs (exclusive of physician fees) for all patients with either a primary or secondary diagnosis of hypertension (N = 4289) receiving treatment during a 2-year period at an academic medical center. In addition to predicting outcome (i.e., whether the patient lived or died), these four clinical variables predicted differences in hospital cost per patient, and financial risk per Diagnosis-Related Group (DRG) as measured by percentage of "outliers" (patients whose length of stay exceeds the cutoff point) or profit/loss per patient under DRG prospective hospital payment. The cost and predictive effect on cost and mortality of the variables were also cumulative; patients with hypertension with one variable had two times the cost of patients with no variables; patients with hypertension with two variables, 2.2 times the cost; patients with three variables, 4.3 times; and patients with four variables, 6.8 times. The mortality of patients with no variables was 0.8%; one variable, 7.7%; two variables, 12.7%; three variables, 15.7%; and four variables, 25.8%. Patients with hypertension can thus be stratified by cost and outcome by clinical events that occur during the patient's hospital stay.(ABSTRACT TRUNCATED AT 250 WORDS)