Alternative discharge destination following lobectomy: Analysis of a national quality improvement database

Abstract

Objective

To determine factors significantly associated with alternative discharge destination (ADCD) following lobectomy, including the modified 5-item Frailty Index (mFI-5).

Methods

Patients in the 2017-2020 NSQIP who underwent elective lobectomy and were admitted from home were included, with ADCD defined as a patient who was discharged to any nonhome location. Four multivariable logistic regression models for ADCD were evaluated for predictive power. Model A was created from backward selection of variables significantly associated with ADCD in bivariate analyses, model B was the mFI-5, model C was mFI-5 and a minimally invasive approach, and model D was mFI-5 and age group.

Results

Among the 15,868 patients, 687 (4.3%) experienced ADCD. Model A identified older age, hypertension, dyspnea, history of chronic obstructive pulmonary disease, and increased length of stay as significantly associated with ADCD. A minimally invasive approach was significantly protective of ADCD. Model A had the best predictive power of the models tested (C-statistic = 0.785). Model B, which assessed mFI-5 alone, had fair predictive power (C-statistic = 0.637). Adding surgical approach (C-statistic = 0.673; model C) or age group (C-statistic = 0.682; model D) as independent variables with mFI-5 improved model fit.

Conclusions

Patients who were frail or age >75 years were more likely to have postlobectomy ADCD. Although the variables identified in model A better predict ADCD, consideration of surgical approach or age with mFI-5 can help surgeons anticipate discharge destination following lobectomy.