Acute Myeloid Leukemia in the Elderly: Trends in Mortality Rates, Hospital Cost, Length of Stay, and Complication Rates in Patients Aged 60 and Older

Background: Acute myeloid leukemia (AML) is associated with a high mortality rate. Advancing age is a risk factor associated with poor prognosis and an increased rate of chemotherapy-related complications in patients with AML. We aimed to evaluate trends in cost of hospitalizations, length of stay, mortality rates, and complication rates in patients aged 60 years and older who were admitted for active AML. We also sought to elucidate differences in these outcomes in teaching and non-teaching institutions.

Methods:We queried the Nationwide Inpatient Sample (NIS) between 1999 and 2013 using the ICD-9 codes 205.00, 205.01, 206.00, and 206.01 for acute myeloid and acute monocytic leukemias in the primary diagnosis field. Admission data regarding total cost, length of stay (LOS), and in-hospital mortality was extracted. This data was trended over the 15-year interval and comparisons were made between teaching and nonteaching institutions. Incidence of in-hospital complications including clostridium difficile infection (CDI), bacteremia, sepsis, pneumonia, venous thromboembolism (VTE), neutropenic fever, candidiasis, urinary tract infection (UTI), and acute respiratory failure were determined and compared in subsets of teaching and nonteaching hospitals. Frequency of bone marrow transplant was also determined in both hospital settings.

Results: A total of 51,684 (weighted n=247,747) admissions for AML occurred from 1999-2013. Of these 31,004 admissions (weighted n=148,683) were in patients aged 60 and older. Most of these elderly admissions occurred at teaching institutions (n=17,593, weighted n=84,829). In-hospital mortality was higher in patients aged 60 and greater (23.68%) compared to those less than 60 (13.7% (p<.0001)). For patients 60 and older, mortality has decreased by approximately 40% during the 15-year interval (p<.0001). Specifically, in-hospital mortality was 30.21% in 1999 and 18.05% in 2013. In comparing teaching and non-teaching hospitals, mortality rate was not found to have a statistically significant difference (p=.4473). Complication rates due to VTE, bacteremia, febrile neutropenia, pneumonia, and UTI increased during this time period. Rates of CDI and candidiasis did not have a statistically significant difference over time. Rates of acute respiratory failure, neutropenic fever, bacteremia, VTE, sepsis, and CDI were higher at teaching than at non-teaching institutions (p<.0001). Rates of UTI were higher at non-teaching (9.62%) than at teaching institutions (8.43% (p=.004)). Differences in the rate of pneumonia and candidiasis were not statistically significant between the two hospital settings. Rates of bone marrow transplant have roughly doubled from .23% in 1999 to .51% in 2013 (p=.0079) and occurred more frequently in teaching (0.54%) than in non-teaching (0.24%) hospitals (p=.0017). Mean LOS (days) is relatively unchanged over the 15- year interval (p=.2277), however, cost has increased dramatically (p=.0001). Total cost in 1999 was $46,833(±1,508), whereas in 2013 it was $146,965(±4,296). Mean LOS and cost were higher at teaching (17.16, $122,257±1,221) compared with nonteaching (10.57, $65,448±993) institutions (p=.0001).

Conclusions: For patients admitted with a primary diagnosis of active AML, in-hospital mortality was markedly higher in patients aged 60 and older compared with those less than 60. In the elderly, in-hospital mortality decreased dramatically between 1999 and 2013. Many factors may contribute to the decrease in mortality in this population including the use of less-aggressive cytotoxic chemotherapy, such as low-dose cytarabine or hypomethylating agents, improved adherence to preventative practices including the use of high-efficiency particulate air filtration, and prophylactic antibiotics. In patients older than 60, LOS and total cost were higher in teaching institutions, although in-hospital mortality was similar. In general, complication rates were higher at teaching hospitals, which may be a consequence of increased medical complexity and more aggressive therapy offered at these hospitals. For instance, bone marrow transplant rates were much higher in teaching than in non-teaching hospitals. Further research is required to determine the exact factors and practice differences contributing to the discrepancies between teaching and non-teaching institutions.