Predictors of extended length of stay related to craniotomy for tumor resection

Abstract

Background: Hospital length of stay (LOS) related to craniotomy for tumor resection (CTR) is a marker of neurosurgical quality of care. Limiting LOS benefits both patients and hospitals. This study examined which factors contribute to extended LOS (eLOS) at our academic center.

Methods: Retrospective medical record review of 139 consecutive CTRs performed between July 2020 and July 2021. Univariate and multivariable analyses determined which factors were associated with an eLOS (≥8 days).

Results: Median LOS was 6 days (IQR 3-9 days). Fifty-one subjects (36.7%) experienced an eLOS. Upon univariate analysis, potentially modifiable factors associated with eLOS included days to occupational therapy (OT), physical therapy (PT), and case management clearance (p < .001); and discharge disposition (p < .001). Multivariable analysis revealed that pre-operative anti-coagulant use (OR 10.74, 95% CI 2.64-43.63, p = .001), Medicare (OR 4.80, 95% CI 1.07-21.52, p = .04), ED admission (OR 26.21, 95% CI 5.17-132.99, p < .001), transfer to another service post-surgery (OR 30.00, 95% CI 1.56-577.35, p = .02), and time to post-operative imaging (OR 2.91, 95% CI 1.27-6.65, p = .01) were associated with eLOS. Extended LOS was not significantly associated with ED visits (p = .45) or unplanned readmissions within 30 days of surgery (p = .35), and both (p = .04; p = .04) were less likely following a short LOS (<5 days).

Conclusion: While some factors driving LOS related to CTR are uncontrollable, expedient pre- and post-operative management may reduce LOS without compromising care.

Keywords: Case management; Craniotomy for tumor resection; Disposition; Length of stay; Readmissions; Therapies.

Fig. 1

Flowchart demonstrating the patient selection strategy for the study. 162 billing records for supratentorial or infratentorial CTR (CPT codes 61,510, 61,512, and 61,518) with any attending neurosurgeon between July 1, 2020 and July 31, 2021 at our medical center were identified. 150 entries reflected unique surgeries, as 12 craniotomies were performed by two attending surgeons and therefore had a billing entry for each surgeon. Eight patients who received more than one operation during the same hospital admission as their CTR were analyzed separate from the primary cohort, as any additional trip to the operating room could impact length of stay in ways unrelated to the CTR. Ultimately, 139 CTRs were included in this retrospective review. There were 135 unique patients, as 4 patients underwent two CTRs during the period covered by this review; however, because each craniotomy resulted in a separate hospital admission, these craniotomies were still included in analysis.

Fig. 2

Flowchart of neurosurgical workflow for (a) all patients with any route of admission and (b) patients with extended length of stay and any route of admission. The flowchart demonstrates the typical neurosurgical workflow and the median (IQR) number of days between various steps for (a) all patients and (b) just patients with extended LOS who were admitted to our medical center and neurosurgical service through any route. Because not all data was available for every patient and because OT, PT, and case management were not involved in all cases, the number of subjects used to calculate each median is also indicated.