Factors Predict 30-Day Hospital Readmissions for Patients With Hematologic Malignancies

Adult patients with hematologic malignancies (i.e., acute leukemias, myelodysplastic syndromes, and aggressive lymphomas [diffuse large B-cell lymphoma and Burkitt’s lymphoma]) treated at the Cleveland Clinic between January 2011 and February 2016 were included in the study. The researchers defined 30-day unplanned readmission as hospitalization within 30 days of index admission for any reason other than planned chemotherapy. Patients undergoing bone marrow transplantation were excluded from the study.

The researchers collected data on demographics (i.e., age, race, sex, highest education attained, median income, and payer status), clinical characteristics (i.e., disease type, symptoms at index admission and readmission, disease status, body mass index at discharge, absolute neutrophil count [ANC] at discharge and readmission, and discharge on intravenous [IV] antibiotics or narcotics), hospitalization (i.e., reason for readmission and length of stay at index admission and readmission), and discharge characteristics (i.e., discharged to and readmitted from).

The researchers observed 259 30-day unplanned readmissions in 157 patients (107 had a single 30-day unplanned readmissions, while 50 [32%] had more than two). More than half of patients were male (59%), and the median patient age was 66 years (interquartile range [IQR] = 48–75). Acute myeloid leukemia was the most common diagnosis (44%), and half of patients had relapsed or refractory disease. Median patient income was $51,700, and most (86%; IQR = 80–90) had greater than a high school education. Almost half (49%) had Medicare coverage, 12% had Medicaid, and 36% had private insurance.

Following unplanned readmissions, patients were discharged to home (50%), home with home health (32%), a nursing facility (15%), or hospice (3%). Ten percent were discharged on IV antibiotics, 44% on opioids, and 48% on psychotropic drugs.

The primary source for 30-day unplanned readmissions was neutropenic fever with or without a confirmed source (61%), and 59% had symptoms at presentation. Median ANC at the time of readmission was 940 cells/ml (IQR = 100–3700) and 870 cells/ml (IQR = 190–3570) at the time of unplanned readmission. The median length of stay during a 30-day unplanned readmission was five days (IQR = 3–10), and the median time from prior discharge to readmission was 11 days (IQR = 5–19).

Most 30-day unplanned readmissions originated from the outpatient clinic, emergency department, or patient’s home (46%), followed by a non-healthcare facility (27%) and outside hospital transfers (22%).

In multivariate analysis, factors that independently predicted subsequent 30-day unplanned readmission included:

• ANC less than 2,000 cells/ml at last discharge (odds ratio [OR] = 7.5, p = 0.0002)
• Constitutional symptoms (i.e., fevers, chills, sweats, and severe fatigue at index admission presentation; OR = 5.84, p = 0.002)
• Gastrointestinal (GI) symptoms (OR = 5.11, p = 0.009)
• If patient was transferred from an outside facility (OR = 3.26, p = 0.005)
• Febrile neutropenia as the reason for index admission (OR = 3.18, p = 0.0001)
• Relapsed or refractory disease (OR= 2.94, p = 0.009)
• Higher education (OR = 2.08, p = 0.01)

“Our findings suggest directing unplanned readmission reduction strategies to optimize symptom control (constitutional and GI) prior to discharge and design specific measures for the cohort with relapsed/refractory disease who might further benefit with individualized care from a multidisciplinary care team, and closer ambulatory follow-up,” the researchers concluded. “Increased risk of additional 30-day unplanned readmission with education may reflect better understanding of care complexities, resulting in a lower threshold for seeking care with a change in medical condition, or possibly a surrogate for better healthcare access.”