Increased complication and mortality among non-index hospital readmissions after brain tumor resection is associated with low-volume readmitting hospitals

Restricted access

OBJECTIVE

Fragmentation of care following craniotomy for tumor resection is increasingly common with the regionalization of neurosurgery. Hospital readmission to a hospital (non-index) other than the one from which patients received their original care (index) has been associated with increases in both morbidity and mortality for cancer patients. The impact of non-index readmission after surgical management of brain tumors has not previously been evaluated. The authors set out to determine rates of non-index readmission following craniotomy for tumor resection and evaluated outcomes following index and non-index readmissions.

METHODS

Retrospective analyses of data from cases involving resection of a primary brain tumor were conducted using data from the Nationwide Readmissions Database (NRD) for 2010–2014. Multivariate logistic regression was used to evaluate the independent association of patient and hospital factors with readmission to an index versus non-index hospital. Further analysis evaluated association of non-index versus index hospital readmission with mortality and major complications during readmission. Effects of readmission hospital procedure volume on mortality and morbidity were evaluated in post hoc analysis.

RESULTS

In a total of 17,459 unplanned readmissions, 84.4% patients were readmitted to index hospitals and 15.6% to non-index hospitals. Patient factors associated with increased likelihood of non-index readmission included older age (75+: OR 1.44, 95% CI 1.19–1.75), elective index admission (OR 1.19, 95% CI 1.08–1.30), increased Elixhauser comorbidity score ≥2 (OR 1.18, 95% CI 1.01–1.37), and malignant tumor diagnosis (OR 1.32, 95% CI 1.19–1.45) (all p < 0.04). Readmission to a non-index facility was associated with a 28% increase in major complications (OR 1.28, 95% CI 1.14–1.43, p < 0.001) and 21% increase in mortality (OR 1.21, 95% CI 1.02–1.44, p = 0.032) in initial analysis. Following a second multivariable logistic regression analysis including the readmitting hospital characteristics, low procedure volume of a readmitting facility was significantly associated with non-index readmission (p < 0.001). Readmission to a lower-procedure-volume facility was associated with a 46%–75% increase in mortality (OR 1.46–1.75, p < 0.005) and a 21%–35% increase in major complications (OR 1.21–1.34, p < 0.005). Following adjustment for volume at a readmitting facility, admission to a non-index facility was no longer associated with mortality (OR 0.90, 95% CI 0.71–1.14, p = 0.378) or major complications (OR 1.09, CI 0.94–1.26, p = 0.248).

CONCLUSIONS

Of patient readmissions following brain tumor resection, 15.6% occur at a non-index facility. Low procedure volume is a confounder for non-index analysis and is associated with an increased likelihood of major complications and mortality, as compared to readmission to high-procedure-volume hospitals. Further studies should evaluate interventions targeting factors associated with unplanned readmission.

ABBREVIATIONS APR-DRG = All Patient Refined DRG; DRG = Diagnosis Related Group; HCUP = Healthcare Cost and Utilization Project; ICD-9-CM = International Classification of Diseases, Ninth Revision, Clinical Modification; NRD = Nationwide Readmissions Database.

Supplementary Materials

  • Table S1 (PDF 395 KB)
Article Information

Contributor Notes

Correspondence Casey A. Jarvis: Keck School of Medicine, University of Southern California, Los Angeles, CA. cajarvis@usc.edu.INCLUDE WHEN CITING Published online October 4, 2019; DOI: 10.3171/2019.6.JNS183469.Disclosures Dr. Mack reports direct stock ownership in Rebound Therapeutics, Cerebrotech, Endostream, and Viseon and a consultant relationship with Medtronic, Penumbra, Stream Biomedical, Rebound Therapeutics, The Stroke Project, and Viseon.
Headings
TrendMD
Metrics

Metrics

All Time Past Year Past 30 Days
Abstract Views 280 280 78
Full Text Views 33 33 12
PDF Downloads 15 15 4
EPUB Downloads 0 0 0
PubMed
Google Scholar