Hospital experience predicts outcomes after high-risk geriatric surgery

Abstract

Background: Geriatric patients require specialized perioperative care, yet the impact of geriatric surgery proportion (a measure of experience) and geriatric surgery volume, on clinical outcomes is unknown. This study analyzes the association between proportion and volume and clinical outcomes after high-risk geriatric surgery.

Methods: Using the 2014 National Inpatient Sample, hospital encounters for older adults (≥65 years) undergoing high-risk geriatric surgery were identified. Geriatric surgery volume was defined as a hospital's annual volume of geriatric patients undergoing high-risk geriatric surgery. Geriatric surgery proportion was calculated as volume divided by the sum of high-risk surgeries in all ages. Hierarchical multivariable regression models identified predictors of inpatient mortality, postoperative length of stay, and discharge to nursing facility.

Results: There were an estimated 514,950 hospital encounters for older adults undergoing high-risk geriatric surgery from 3,115 hospitals. Mean proportion was 0.53 ± 0.19; median volume was 60 cases per year, ranging from 5 to 3,235. After adjustment, comparing the 90th to 10th percentiles, higher proportion was associated with decreased mortality (odds ratio [95% confidence interval] 0.81 [0.73-0.88]; P < .001) and shorter postoperative length of stay (-4.44% (-5.49 to -3.39%); P < .0001). Higher volume was not associated with mortality but was associated with longer length of stay (7.76% [6.75-8.77%]; P < .0001) and decreased discharge to nursing facility (0.87 [0.79-0.95]; P= .003).

Conclusion: Treatment of geriatric patients at hospitals with the highest proportion of high-risk geriatric surgery, or the most experience, is associated with improved outcomes. High-proportion hospitals should be examined to understand the mechanisms by which better quality geriatric surgical care is achieved, while lower-proportion hospitals may be targets for quality improvement efforts.

Figure 1.

Unadjusted inpatient mortality, median length of stay, and discharge to nursing facility by decile of geriatric surgery proportion and volume. Outcomes are calculated as the overall value within each decile, then plotted against the median of each decile. For example, the first decile of geriatric surgery proportion is 0.01 to 0.36 and the median is 0.31. The inpatient mortality within this decile is 5.44%. Thus, the corresponding point in panel (a) is (0.31, 5.44).