Date of Award

January 2013

Document Type

Open Access Thesis

Degree Name

Medical Doctor (MD)



First Advisor

Paul G. Barash

Second Advisor

David G. Silverman

Subject Area(s)

Medicine, Surgery, Health care management


Within the ambulatory surgical setting, existing risk prediction models focus predominantly on postoperative factors of nausea, vomiting, and pain, but do not uniformly specify preoperative predictors of outcomes across multiple surgical specialties. Identification of preoperative markers, specifically those that are reversible, is key to improving risk stratification and designing patient-specific clinical interventions. Recent work shows that preoperative gait speed is a promising marker of postoperative morbidity and mortality within the inpatient surgical population. However, it remains to be explored whether gait speed is sensitive enough to delineate discharge and postoperative outcomes within the ambulatory surgical population.

We sought to determine which specific preoperative factors independently predict discharge readiness outcomes among ambulatory surgical patients. To address this aim and following Institutional Review Board (IRB) approval, we performed a cross-sectional analysis of data from a prospective observational study of 602 ambulatory surgical patients. The primary outcomes were: 1) Time to home discharge readiness from the ambulatory post-anesthesia care unit (PACU), and 2) 24-h postoperative occurrence of nausea, vomiting and bleeding. We evaluated the occurrence of unanticipated admissions from the ambulatory PACU to ancillary care units (inpatient wards and critical care) as a post hoc secondary outcome. Preoperative measures were gait speed (6.1 m divided by the average time to walk 6.10 meters), mean arterial pressure, heart rate, demographic factors and other clinical covariates. Statistical analysis was done with SAS, version 9.2® (Cary, NC), and p<0.05 was considered statistically significant.

Participants were 54.3% female, the mean gait speed was 0.90 ± 0.18 m/s, and the median home discharge readiness time was 89 minutes (interquartile range 61-126). Multivariable Cox regression analyses showed that gait speed (≥1 m/s vs. < 1 m/s) was an independent predictor of time to home discharge readiness after adjustment for covariates (adjusted hazard ratio = 1.25 (95% CI: 1.03-1.50), p = 0.02). For the primary outcomes, independent predictors of home discharge readiness ≤90 minutes were: preoperative heart rate, mean arterial pressure, and gait speed (adjusted odds ratio = 2.33 {95% CI: 1.52-3.54}, p<0.0001), when all other covariates are held constant. Monte-Carlo Cross validation (using 2x104 iterations) showed the mean percentage of correctly classified predictions by our model was 65.6 (95% CI: 61.8-69.4). However, gait speed was not independently associated with 24-h postoperative complications, p=0.35. Predictors of unanticipated admissions included the history of cardiac surgery and prior hospitalizations, and gait speed (adjusted odds ratio = 0.54 {95% CI: 0.38-0.82}, p=0.003).

We present the first cross-validated prediction model of outcomes in the ambulatory surgical setting and identify preoperative heart rate, mean arterial pressure and gait speed as three important modifiable factors, which independently associate with home discharge readiness time ≤90 minutes. Our findings underscore the importance of preoperative measures and elements of patients' history for potential risk stratification and resource allocation. We conclude that a focus on reversible clinical markers may help identify those patients at risk for delayed discharge in the ambulatory surgical setting.