Characterizing Outpatient Problem List Completeness and Duplications in the Electronic Health Record

Abstract

Purpose: To characterize rates of problem list completeness and duplications in outpatient problem lists and to identify any relationships between problem list completeness and duplications and disease type, demographics, and disease severity. Methods: We performed a retrospective analysis of electronic health record data from Partners HealthCare. Using a combination of vital sign measurements, lab results, and ICD-10 codes, we identified patients with at least one of eleven chronic diseases: Crohn’s disease, ulcerative colitis, depression, schizophrenia, opioid use disorder, hepatitis B, hepatitis C, hypertension, diabetes, asthma, and epilepsy. We then collected additional information regarding the number of disease-related entries on each patient’s problem list. We also collected additional information, including vital sign measurements, lab results, medication orders, and hospital admissions, to assess the severity of disease in each patient. Finally, we performed two-proportion z-tests to compare the rates of completeness and duplications by disease type, demographics, and disease severity. Results: Rates of problem list completeness varied from 35.1% in hepatitis B infection to 92.3% in asthma, whereas rates of problem list duplication varied from 3.4% in hepatitis B to 26.4% in diabetes. Except in certain diseases like hypertension and asthma, there was no consistent relationship between demographic factors and rates of completeness and duplication. There was also no consistent relationship between rates of completeness and disease severity, except in the cases of asthma and epilepsy. However, for eight of the eleven diseases, there was a clear relationship in which increasing disease severity is associated with increased rates of duplication. Conclusions: Problem list completeness and duplication demonstrated substantial variation across the eleven studied diseases. Rates of completeness and duplication are not consistently affected by demographic factors unless a certain disease is especially prevalent in a certain demographic group. While rates of completeness were also not consistently affected by disease severity, rates of duplication showed a relatively consistent positive correlation with disease severity.