Trends in Hospital Cost, Length of Stay, Complication Rates, and Mortality in Patients with Acute Lymphoblastic Leukemia in Teaching Versus Nonteaching Institutions: Data from the Nationwide Inpatient Sample from 1999-2014

Background: With advancements in therapies as well as prevention of treatment-related complications, the five-year survival rate for acute lymphoblastic leukemia (ALL) has more than quadrupled since 1960. Nevertheless, adult ALL has continued to have a high mortality rate. In this study, we evaluated trends in hospital cost, length of stay, complication rates and inpatient mortality in adult patients admitted with active ALL from 1999 through 2014. We also aimed to clarify differences in outcomes between teaching and nonteaching institutions.

Methods: We analyzed the Nationwide Inpatient Sample between 1999 and 2014 using the ICD-9 codes 204 and 204.02 for acute lymphoblastic and acute lymphocytic leukemia in the primary diagnosis domain. The sample was weighted to approximate the full inpatient population of the U.S. over the 16- year interval. Total cost was adjusted for inflation using data from the U.S. Bureau of Labor Statistics. Total hospital cost, length of stay (LOS), complication rates, and in-hospital mortality data were extracted and comparisons were made between teaching and nonteaching institutions. Common in-hospital complications such as clostridium difficile infection (CDI), sepsis, pneumonia, venous thromboembolism (VTE), neutropenic fever, and urinary tract infection (UTI) were extracted by ICD-9 codes and analyzed. Chi square analysis was performed to examine differences in the categorical variables. Frequency of bone marrow transplant were also examined in both settings.

Results: A total of 13,704 (weighted N=67,744) admissions with a primary ALL diagnosis were identified from 1999-2014. The majority of admissions occurred at teaching institutions for all ages under 60, while the majority of patients aged 60 and above occurred at non-teaching institutions (p< 0.0001). Total hospital charges increased dramatically from $58,293 (± $2,872) in 1999 to $224,450 (± $10,015) in 2014, or $157,651 when adjusted for inflation (p<0.0001). Total hospital charges were greater at teaching, $153,763, as compared to nonteaching institutions, $83,785 (p<0.0001). Overall mean LOS increased over the 16-year interval from 15 days in 1999 to 18 days in 2014 (p<0.04). LOS was also greater at teaching institutions with 19 days compared to 12 days at nonteaching sites (p<0.004). Throughout the 16 year time interval, complication rates due to CDI, VTE, pneumonia, and UTI increased, while rates of sepsis, neutropenia, and fever were not statistically different. No significant statistical difference was noted in the rates of UTI, pneumonia or sepsis between the institutional settings. In teaching versus nonteaching institutions, rates of neutropenia, 16.3% v. 7.9%, and fever, 13.7% v. 7.5%, were both significantly higher (p<0.001). Rates of CDI, 5.4% v. 2.6% (p<0.006), and VTE, 2.6% v. 1.3% (p<0.01) were also higher in teaching institutions. Rates of bone marrow transplant decreased from 4.67% in 1999 to 3.14% in 2014 with no statistical significance. However, more transplants were performed in teaching institutions 5.27%, as compared to nonteaching institutions 1.39% (p<0.001). Overall in-patient mortality showed a significant decrease from 16.7% in 1999 to 7.6% in 2014 (p<0.0001).

Conclusions: The majority of admissions for ALL between 1999-2014 occurred at teaching institutions. Concomitantly, total cost and mean LOS were greater at teaching institutions when compared to nonteaching. Though multifactorial, we hypothesize that one cause of this trend can be attributed to the greater frequency of bone marrow transplants at teaching institutions, increasing medical complexity of cases and resource requirement. Although certain complication rates increased, overall in-patient mortality demonstrated a striking improvement during this 16-year time interval. This can be credited to better recognition of, surveillance for and treatment of known complications, improved education of ancillary staff, and the advent of tyrosine kinase inhibitors within well established guidelines. Nevertheless, further research is required to determine why complication rates overall have not shown significant rectification.