Utility of clinical pathway and prospective case management to achieve cost and hospital stay reduction for aortic aneurysm surgery at a tertiary care hospital.

Abstract

We reviewed our experience with a clinical pathway instituted in December 1993 for all nonurgent abdominal aortic aneurysm (AAA) surgery. We analyzed a reference group of 49 consecutive pre-pathway AAA patients (group I) and the 44 patients enrolled in the first year of the pathway (group II). On the basis of the interim review of data collected during the first year, pathway modifications were made, and 34 patients enrolled after these modifications (group III) were also analyzed. Comparison of groups I and II showed that institution of the pathway resulted in a marginally significant reduction in mean charges of 14.7% (p = 0.09), and a slight fall in mean length of stay (LOS) (13.8 vs 13.1 days, NS) and mortality rate (4.1% vs 2.3%, NS). For group II, a significant correlate (p < 0.05) of increased charges was fluid overload as diagnosed by chest radiograph. This recognition led to active efforts to reduce perioperative fluid administration. Comparison of groups II and III revealed that the practice modifications led to marked reduction in the incidence of fluid overload (73% vs 24%; p < 0.01), mean charges (30.4% reduction; p < 0.05), mean LOS (13.1 vs 10.2 days; p < 0.05), and median LOS (11 vs 8 days). Multiple regression analysis of all pathway patients showed that preoperative renal insufficiency is a significant predictor of both increased LOS (p < 0.01) and charges (p < 0.01), but that age, sex, and coronary disease were not predictive. Of the postoperative parameters analyzed, important correlates of increased charges were acute renal failure (p < 0.01) and fluid overload (p < 0.01). Institution of a clinical pathway for AAA repair resulted in significant charge reduction and a slight reduction in stay. Practice modifications based on interim data analysis yielded further significant reductions in charges and LOS, with overall per-patient charge savings (group I vs III) of 40.6% (p < 0.05) and overall LOS reduction of 3.5 days (p < 0.05). The reduction in actual charges was seen despite an overall increase in the hospital rate structure. Comparing groups I, II, and III, we found no indication of increasing mortality rate. Ongoing analysis has identified correlates of increased charges, potentially permitting identification of high-cost subgroups and more focused cost-control efforts. Rather than restricting management, clinical pathways with periodic data analysis may improve quality of care.