Economic Disparities in Acute Myeloid Leukemia Readmissions in the United States over a Decade

BACKGROUND

Hospital readmissions are an important measure of health care quality and hospitalization outcomes. Unplanned hospital readmissions are a significant financial burden on the US health care system, costing an estimated $15-20 billion annually. Compared to the general population, cancer patients experience higher rates of unplanned readmissions; this is a common complication of acute myeloid leukemia (AML) therapy, with estimated rates ranging from 30% to 60%. Patients with low socioeconomic status are at an even greater risk of readmission, resulting in poorer health outcomes. Consequently, identifying and preventing avoidable readmissions can significantly improve patient quality of life and reduce hospital costs. However, there is a dearth of comprehensive literature examining the relationship between socioeconomic status and readmissions in patients with AML. This study aims to assess the variation in readmission trends over the past decade among adult AML patients in the US, stratified by household income.

METHODS

This was a retrospective interrupted trends study involving adult (≥ 18 years) AML hospitalizations in the US, utilizing the Nationwide Readmissions Database (NRD) for the years 2010, 2012, 2014, 2016, and 2018. Using unique hospitalization identifiers, we obtained the index hospitalizations. A subsequent hospitalization involving the same patient within 30 days was classified as a readmission. We excluded patients with neutropenic fever (NF) during the index hospitalization, as well as elective and traumatic admissions and readmissions. The median household income (MHOI) quartiles variable stratifies patients as 1-low income (LIQ), 2-middle income, 3-upper middle income, or 4-high income (HIQ). We compared trends in the 30-day all-cause readmission rate (30ACR) and 30-day NF-specific readmission rate (30NFR) between LIQ and HIQ groups. Multivariate logistic trend analysis was employed to calculate odds of trend in 30ACR and 30NFR, adjusted for age and sex. The threshold for statistical significance was a p-value of <0.05.

RESULTS

The study included 181,432 index hospitalizations, with 24.1% belonging to the LIQ and 24.6% belonging to the HIQ. The mean age of patients during the index hospitalizations was 61.6 years (SD: 17.6), with a significantly increasing mean age correlating with higher MHOI. A total of 44,417 hospitalizations were readmissions within 30 days of the index hospitalization, with 25.0% in the LIQ compared to 23.2% in the HIQ. The mean age at readmission was 59.9 years (SD: 17.7), with a higher mean age correlating with higher MHOI. The majority of readmissions were male (53.5%), with no significant variation by MHOI.

The overall 30ACR overall was 29.7%, which varied per year, showing a trend towards increasing odds of 30-day readmission (p-trend <0.001). There was a trend towards increasing 30ACR from 2010 to 2018 (p-trend <0.001) with patients in the LIQ having higher odds of readmission compared to those in the HIQ. The 30NFR was 5.1% with patients in the HIQ having a higher odd of 30NFR compared to LIQ. Readmission mortality declined from 10.0% in 2010 to 7.8% in 2018. There was a trend towards decreased odds of inpatient mortality following readmission, adjusted for age, sex, and MHOI (p-trend <0.001).

CONCLUSION

To the best of our knowledge, this is the largest epidemiological study to date examining the impact of socioeconomic deprivation on AML readmissions. Our findings reveal a significantly increasing trend of 30-day all-cause readmission rates for AML over the past decade, with consistently higher rates in low-income quartile compared to high-income quartile. This underscores the urgent need for evidence-based transition of care programs and post-discharge interventions tailored to address the specific social needs of this vulnerable population.