Anemia at Discharge and Risk of Readmission in Elderly Patients

Title: Anemia at Discharge and Risk of Readmission in Elderly Patients Authors: Yeshanew Teklie MD, Niraj Patel MD; Victoria Campdesuner DO; Kayla Marini DO; Yorlenis Rodriguez DO; Karen M. Hamad MD, FAAP, FACP; Mary E. Geary PhD, RN; and Wilhelmine Wiese-Rometsch MD, FACPAffiliation: Florida State University Internal Medicine Residency Program at Sarasota Memorial Hospital, 1700 South Tamiami Trail, Sarasota FL 34239, 941-917-7799Type of Project: Practice-based Quality ImprovementIntroduction: Anemia is defined by the World Health Organization (WHO) as a hemoglobin (Hb) concentration less than 12.0 g/dL for females and less than 13.0 g/dL for males. Objective of this quality improvement initiative was to test if anemia independently is associated with readmission in elderly patients discharged from the medicine service of a community teaching hospital. Methods: We conducted a retrospective cohort study declared exempt by our Institutional Review Board involving electronic medical record data from patients at least 65 years old hospitalized between September 2009 and June 2019 discharged not to hospice care. Severity of anemia classified by the WHO was applied for Hb concentrations obtained at hospital discharge for females as mild (11.0 – 11.9 g/dL), moderate (8.0 – 10.9 g/dL), severe (<8.0 g/dL); and males as mild (11.0 - 12.9 g/dL), moderate (8.0 - 10.9 g/dL), severe (<8.0 g/dL). Time to readmission after index hospitalization within 30d, 90d, and 180d with mild, moderate, severe, or no anemia was compared using Kaplan Meier survival curves with covariates (age; sex; multiple chronic conditions; length of stay; APR-DRG Severity of Illness and Risk of Mortality; and discharge destination) controlled using Cox Proportional Hazard modeling with relative impact assessed using Random Forest modeling. Readmission after 180d was considered a new index hospitalization. Continuous variables summarized as mean (SD) or median (interquartile range) were contrasted using Kruskal-Wallis ANOVA. Categorical variables summarized as proportions were compared using chi square or Fisher exact test. Statistical tests were two-tailed with p<.05 considered significant.Findings: Among 13,526 inpatients with 18,793 discharges median age of 78 (14) years was statistically similar by sex (49% females, 51% males) exhibiting an anemia distribution of no (31.0%), mild (20.5%), moderate (47.1%) and severe (1.4%). Race distributed as 89.9% white; 5.9% African American or Black; and 4.2% Other. Females differentiated distribution across anemia spectrum at no (95.7%) (p<0.001), mild (26.4%) (p<0.001), moderate (51.0%) and severe (43.5%). No, mild, moderate and severe anemia corrected for putative confounders impacted (p<0.001) respectively cumulative risk of readmission at 30 d (7.0%, 7.9%, 17.2%, 21.3%), 90 d (12.0%, 15.3%, 28.1%, 34.8%) and 180 d (16.4%, 20.7%, 34.4%, 42.6%).Conclusion: An essential patient-centric question is whether anemia in elderly inpatients affects poor outcomes and/or whether anemia is a surrogate marker for underlying overt and/or subclinical disease(s). Although present quality improvement initiative was not designed to unravel mechanisms of anemia, we controlled for putative severity of illness confounders while demonstrating readmission risk escalating with severity of anemia. Implications for Practice: Findings herald readmission risk associated with “no anemia” as defined by the World Health Organization. Tailored anemia care could offer clinical advantages to mitigate risk for readmission.Reference: WHO. Hemoglobin concentrations for the diagnosis of anemia. Accessed May 16, 2020 at https://www.who.int/vmnis/indicators/haemoglobin.pdf