Background and Goal of Study
 
Previous reports suggest that the addition of functional dependency to ASA classification is a strong predictor for mortality. Using data from Predictors, Risk factors, and Outcomes Following major Surgery study (PROFS) our aim is to assess the value of adding preoperative functional dependency to ASA 3 class to predict postoperative adverse events.

Materials and methods

An observational cohort study between 2015 Nov 1th and 2016 Feb 15th. The inclusion criteria were: age ≥18 years, ASA≥3, major or xmajor/complex abdominal, urogenital and orthopedic surgery. The functional dependency was classified as independent (class a) or dependent (class b), if there was any need of assistance for activities of daily living (ADL). The outcome measures were postoperative complications screened by Postoperative Morbidity Survey (POMS) (days 3, 7 and 10) and 30-day mortality. The POMS includes 10 domains, here we collapsed these into one composite outcome. 
The predictive value of adding functional dependency to ASA 3 was tested by uni and multivariate logistic regression analyses.

Results and discussion

Number of included patients was 1089, 13 were excluded due to violation of inclusion criteria, 3 were lost to follow up and. 1073 were analyzed. 
ASA 3b had higher risk for adverse postoperative events compared with ASA 3a. When age and urgency was added to the regression models, the significance were lost for all outcomes. ASA 4 had higher risk compared with ASA 3, which remains after adjustment for age and urgency.
The loss of significance might be attributed to the limited numbers of observations and the greater proportion of acute surgery in the ASA 3b group. Nevertheless, the effect size (odds ratio) is comparable with previous reports.

Conclusion

We could not confirm previous findings, and we suggest that the utility of reclassification of the ASA 3 group by functional dependency should be addressed by further research in a larger population.

Combining ADL status and ASA III classification to predict postoperative outcomes after major surgery (PROFS study NCT02626546)