HSR21-049: Fragmentation of Care in the Multidisciplinary Management of Pancreatic Cancer: Luxury or Liability?

Authors:
Lindsay A. GilThe Ohio State University Wexner Medical Center, Columbus, OH

Search for other papers by Lindsay A. Gil in
Current site
Google Scholar
PubMed
Close
 MD
,
Mariam F. EskanderThe Ohio State University Wexner Medical Center, Columbus, OH

Search for other papers by Mariam F. Eskander in
Current site
Google Scholar
PubMed
Close
 MD, MPH
,
Samilia Obeng-GyasiThe Ohio State University Wexner Medical Center, Columbus, OH

Search for other papers by Samilia Obeng-Gyasi in
Current site
Google Scholar
PubMed
Close
 MD, MPH.
,
Bridget A. OppongThe Ohio State University Wexner Medical Center, Columbus, OH

Search for other papers by Bridget A. Oppong in
Current site
Google Scholar
PubMed
Close
 MD
,
Yaming LiThe Ohio State University Wexner Medical Center, Columbus, OH

Search for other papers by Yaming Li in
Current site
Google Scholar
PubMed
Close
 MD, MS
, and
Allan TsungThe Ohio State University Wexner Medical Center, Columbus, OH

Search for other papers by Allan Tsung in
Current site
Google Scholar
PubMed
Close
 MD
View More View Less
Full access

Introduction: Despite advances in the multidisciplinary treatment of pancreatic cancer, survival rates remain low. Modifiable factors include patterns of care and receipt of guideline concordant care (GCC). In this study, we define populations receiving fragmented care (treatment at multiple facilities) and its implications on receipt of GCC. Methods: The National Cancer Database (NCDB) 2006-2016 was queried for patients with pancreatic adenocarcinoma who underwent resection and adjuvant chemotherapy. GCC defined as receipt of chemotherapy within 12 weeks of surgery, per NCCN guidelines. Patients divided into 2 groups: those who underwent resection and chemotherapy at the same facility (unified care) vs. different facilities (fragmented care). Intergroup characteristics compared on bivariable analysis. Multivariable logistic regression utilized to identify predictors of fragmentation and evaluate the relationship between fragmentation and GCC. Results: Of 28,688 patients, 16,354 (57%) received fragmented care. Age, race, insurance, education, income, facility type, rurality, and distance travelled were significantly different between groups (p<0.001). On multivariable analysis, Black race (OR 0.6, CI 0.55-0.65), Asian race (OR 0.73, CI 0.62-0.86; Ref White), Medicaid insurance (OR 0.78, CI 0.69 0.88) and uninsured status (OR 0.5, CI 0.42-0.59; Ref private) were associated with unified care. Moreover, care at a comprehensive community program (OR 0.57, CI 0.49-0.67), academic program (OR 0.23, CI 0.2-0.27), or integrated network cancer program (OR 0.52, CI 0.44-0.62) predicted unified care when compared to a community program. Treatment in the Southeast (OR 1.49, CI 1.32-1.58), Central east (OR 1.3, CI 1.2-1.39), Central west (OR 1.28, CI 1.17-1.39), and West (1.55, CI 1.42-1.69) vs the Northeast was associated with higher odds of fragmentation (Figure 1). Living in an urban (OR 1.48, CI 1.36-1.60) or rural (OR 1.38, CI 1.13-1.68) area vs a metropolitan area was associated with fragmented care. 20.7% of patients with fragmented care vs 13.7% with unified care did not receive GCC (p<0.001). On multivariable analysis, fragmented care was associated with not receiving GCC (OR 0.59, CI 0.55-0.63). Conclusion: Medicaid insurance, uninsured status, Black race and Asian race were associated with a higher probability of receiving unified care, which in turn was associated with receipt of GCC. Area of residence appeared to be the main driver of fragmentation.

Figure 1.
Figure 1.

Fragmentation of Care by Geographic Region in the United States

Citation: Journal of the National Comprehensive Cancer Network 19, 3.5; 10.6004/jnccn.2020.7773

Corresponding Author: Lindsay A. Gil, MD
  • Collapse
  • Expand
  • View in gallery
    Figure 1.

    Fragmentation of Care by Geographic Region in the United States

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 293 106 6
PDF Downloads 173 4 0
EPUB Downloads 0 0 0