Unplanned hospitalizations among gastrointestinal cancer patients in the state of Texas.

Abstract

6634 Background: Hospitalization rates and cost of care for patients with Gastrointestinal (GI) malignancies are among the highest in the spectrum of cancers. It is poorly understood if these are driven by planned or unplanned hospitalizations. Methods: We conducted a retrospective cohort study using linked Texas Cancer Registry and Medicare claims data to describe the patterns of unplanned hospitalization among GI cancer patients in Texas. Our outcome, unplanned hospitalization, was defined as any emergent or urgent admission. Prior comorbidities were assessed using a modified Charlson score. Data were collected for 2 years from date of cancer diagnosis. Modified Poisson regression model was used to identify factors associated with the outcome. Results: 30,199 patients aged >66 years were included in the study. 25,096 patients had 70,256 in-patient claims. 53.6% of these claims were unplanned hospitalizations. The top 3 reasons for unplanned hospitalization were pneumonia, congestive heart failure and volume depletion. Unplanned hospitalization rates ranged from 57.9% (colon) to 67.4% (esophageal) among the different cancer types. After multivariate analysis, unplanned hospitalization was more likely among those with esophageal cancer (RR=1.15, CI 1.1-1.2), gastric cancer (RR=1.07, CI 1.04-1.11), pancreatic cancer (RR=1.04, CI 1.01-1.07) and rectal cancer (RR= 1.03, CI 1.00-1.06). Patients with regional and distant disease were at higher risk for unplanned hospitalization (RR=1.14, CI 1.11-1.16 and RR=1.13 CI 1.1-1.16 respectively), as well as those aged >80 (RR=1.04 CI 1.01-1.06), of black race (RR=1.06, CI 1.02-1.09), and living in census tract with >21% poverty level (RR=1.06 CI 1.03-1.09). Risk of unplanned hospitalization also increased with Charlson scores (p<.0001). Conclusions: The majority of hospitalizations were unplanned. Top reasons for unplanned hospitalization were diagnoses that are considered potentially preventable, and should be a focus for intervention. The disparity in unplanned hospitalizations among blacks and those living in poverty-stricken areas warrants further investigation as to whether this finding correlates with timely access to outpatient care in this population.