ASPO Abstracts
Concordance of HIV self-report and EHR data capture in cancer patients
Category: Behavioral Science & Health Communication
Conference Year: 2020
Abstract Body:
Purpose: The population living with both HIV and cancer is growing. Accurate information about oncology
patient HIV status is important because those with underlying HIV infections experience poorer cancer-specific
outcomes than the general cancer population. We assessed the sensitivity of three different metrics for
identifying HIV status at a large, NCI-designated comprehensive cancer center.
Method: Using the institution’s Health Research Information System, we extracted HIV data from the electronic
health record (EHR) at a comprehensive cancer center for patients admitted to the institution between January
2009 and September 2019. HIV status was determined via the following sources: 1) electronic patient
questionnaire self-report of an HIV/AIDS diagnosis or use of daily HIV prophylaxis pills (PrEP) or antiretroviral
therapy (ART); 2) ICD 9/10 diagnosis code affirming HIV status; and 3) evidence of administration of ≥1 HIV
medication (ART) regimen in EHR (of 40 regimen search terms). We report frequencies of patients identified
through each metric and concordance of data.
Results: 949 persons living with HIV were identified through at least one of the three data sources. 284 (30%)
self-reported an HIV diagnosis. The remaining 70% (N=665) failed to self-disclose HIV status and were
identified either through HIV medications or ICD diagnosis codes. Only 257 HIV diagnoses (27%) were
consistently identified through all three HIV identification metrics, with many patients only identified through a
singular metric (e.g., 154 (16%) via diagnosis code only; 192 (20%) via HIV medications only).
Conclusions: In our sample, most (70%) of those with HIV did not disclose this information on the standard
electronic new patient intake questionnaire. Thus, initial consultations in the oncology setting may fail to
consider HIV status in treatment planning and oncology care. Notably, only 27% of oncology patients living with
HIV were identified through all three metrics, indicating lack of concordance regarding HIV status in the EHR.
These findings highlight the potential for oncology care team members to be uninformed of clinically relevant
information, potentially leading to inadequate care for a vulnerable population.
Keywords: HIV disclosure, electronic health record