Socio-Economic and Clinical Factors Are Associated with In-Hospital Mortality in Patients Admitted for Diffuse Large B-Cell Lymphoma.

Abstract

Clinical, pathologic, and molecular predictors of prognosis and survival in diffuse large B-cell lymphoma (DLBCL) are well-described. Factors associated with differences in in-hospital mortality for patients admitted with a diagnosis of DLBCL, however, are less well understood. The objectives of this study were to assess the demographic and socio-economic factors associated with death in patients hospitalized with DLBCL. Data from the Nationwide Inpatient Sample of the Healthcare Cost and Utilization Project for patients admitted with a primary diagnosis of DLBCL during the years 2001 through 2002 were analyzed. Simple and multivariable logistic regression analyses were used to identify patient and admission characteristics associated with in-hospital mortality. 12,296 admissions and 1504 in-hospital deaths (12.2%) were included in our sample. In unadjusted analyses, patient’s age, insurance status, and co-morbid conditions in addition to the hospital type (rural versus urban versus teaching), urgency of admission, and source of admission were all significantly associated with in-hospital mortality (p<0.05). Conversely, race, gender, and the average income of the patient’s home zip code were not significantly associated with in-hospital mortality. When controlling for other factors, age, insurance type, patient co-morbidities, urgency of admission, and source of admission were independently associated with in-hospital death; race, gender, income status, and hospital type were not associated with inpatient mortality in the multivariable analysis. In adjusted analyses, compared to patients with private insurance as primary payer, patients with Medicare had significantly lower odds and patients whose payer was classified as “other” had significantly higher odds of inpatient death. Additionally, Native American patients were 4.2 (95% CI = 1.4–12.7) times as likely to die during hospitalization as white patients. No association between average income and inpatient death in DLBCL was observed. Our study suggests that inequalities in deaths exist for patients hospitalized for DLBCL. Advanced age, multiple medical co-morbidities, and emergent admission were expectedly associated with an increased probability of in-hospital death. Interestingly, the risk of death was also associated with insurance status. One reason for this finding may be better access to the healthcare system by patients covered by Medicare including, not only access to treatment, but also home health and other services that facilitate discharge. Further investigation to determine the etiology and impact of these differences is warranted. Selected Data and Association with In-Hospital Mortality Adjusted OR 95% C.I. Adjusted OR 95% C.I. Age (p=0.020) Insurance (p=0.001) 18-35 Ref. Private Ref. 36–45 1.05 0.65–1.70 Medicare 0.78 0.62–0.98 46–55 1.35 0.88–2.07 Medicaid 0.89 0.63–1.26 56–65 1.57 1.04–2.38 Self-pay 1.10 0.71–1.71 66–75 1.94 1.25–3.02 No charge 1.18 0.32–4.33 76–85 1.81 1.15–2.84 Other 2.03 1.32–3.12 >85 2.05 1.23–3.40 Race (p=0.067) Zip Code Median Income (p=0.334) White Ref. $1–24,999 Ref. Black 1.22 0.93–1.61 $25,000–34,999 1.09 0.76–1.63 Hispanic 0.89 0.64–1.23 $35,000–44,999 1.26 0.87–1.84 Asian/Pacific Islander 1.23 0.65–2.42 >$45,000 1.09 0.87–1.34 Native American 4.21 1.40–12.73 Other 0.83 0.48–1.44