Time to Resolution of Iron Deficiency: A 10-Year Single State Retrospective Review

Abstract

Introduction: Iron deficiency can have clinical consequences even without anemia, including fatigue, exercise intolerance and mood disorders. The time from laboratory diagnosis to resolution of iron deficiency has not been previously evaluated. Methods: In a retrospective review of electronic medical record (EMR) data from a single state health system, we identified patients meeting laboratory criteria for iron deficiency (defined as a ferritin level < 25 ng/mL). Patients with at least one follow-up ferritin level of any value within 3 years were included in the final cohort. Patients with a follow-up ferritin > 50 ng/mL at any time point within 3 years were classified as having resolved iron deficiency. Diagnoses of interest included iron deficiency anemia (IDA), as well as other conditions where iron deficiency is routinely assessed and treated irrespective of anemia, namely heart failure (HF) and restless leg syndrome (RLS). Descriptive statistics and time to event analyses (via Cox proportional hazards regression) were performed to determine time to resolution of iron deficiency in the cohort and characteristics predictive of achieving resolution. Results: We identified 13,084 patients with iron deficiency detected between 2010 to 2020. The majority of patients were female (86.6%) and White (79.3%); 57.3% had an EMR diagnosis of IDA (Table 1). We found that 5,485 (41.9%) had resolution of iron deficiency within 3 years of diagnosis, while 7,599 (58.1%) had unresolved iron deficiency. Among those with resolved iron deficiency, the median time to resolution was 1.9 years (IQR 0.8, 3.9 years). Factors associated with resolution of iron deficiency included increasing age (aHR 1.42 [95% CI 1.31 - 1.54], p < 0.001 for age > 60), male gender (1.49 [1.39 - 1.59], p < 0.001) and a diagnosis of HF (1.23 [1.11 - 1.36], p < 0.001) (Table 2). Black race (0.75 [0.68 - 0.83], p < 0.001) and absence of an EMR diagnosis of IDA, HF or RLS (0.75 [0.71 - 0.81], p < 0.001) were associated with a lower likelihood of iron deficiency resolution. Conclusion: Prolonged and unresolved iron deficiency were prevalent in this cohort, with young age, female gender and Black race increasing risk of non-resolution. Interventions are needed to raise awareness of the clinical consequences of non-anemic iron deficiency and of diagnostic ferritin thresholds.