Health measurement using an audio computer-assisted self interview system to support the patient-centered medical home
Health measurement using computer-assisted self interviews
Introduction. Patient-reported outcomes are a logical component of the patient-centered medical home. We evaluated the feasibility of integrating patient-reported assessments through audio computer-assisted self interviews into the intake process of a general medicine clinic. We joined interview responses (health-related quality of life (QOL) and symptom burden) to electronic medical record data.
Methods. We custom-developed ACASI software, which was integrated into our clinical data warehouse. Standard instruments (e.g., NIH-PROMIS [QOL], PHQ-2 or 9 [depression screen]) were administered as patient interviews through touch-screen monitors installed in the waiting room of a general medicine clinic in the Cook County Health & Hospitals System. Patient responses were electronically stored in the data warehouse, summarized for clinicians, and joined to clinical data. We evaluated the mean per-question response duration and factors associated with a prolonged response time.
Results. We successfully developed and installed an ACASI system and joined the responses to EMR data. Symptom burden was strongly inversely correlated with both physical and mental QOL components (high symptom burden was associated with poor QoL). Three symptoms explained 57% of the variation in QoL (parameter estimate [95% CI]): pain (-2.9 [-2.6 to -3.2]), lack of energy (-2.6 [-2.2 to -3.0]), and shortness of breath (-0.8 [-0.7 to -1.2]). From 1,670 patient interviews, the mean per-question response time was 18 seconds. By multivariable analysis, factors associated with prolonged response time (seconds per question; 95% CI) were age, 80-89 (5.5; 4.1 to 7.0 seconds), 70-79 (5.1; 4.0 to 6.1), 60-69 years (3.4; 2.6 to 4.1), 50-59 years (1.4; 0.7 to 2.1); Spanish language (3.7; 2.8 to 4.7); no home computer use (2.6; 1.8 to 3.4); and mental quality-of-life below sample mean (0.8; 0.2 to 1.3).
Conclusions. Electronic medical record systems can be developed to integrate self-administered patient assessments. Per-question response times were not excessive for most patients, but were significantly longer for older patients, those who did not use a home computer, and for Spanish-language interviews.
EHR/EMR System or Vendor
Cerner, custom-developed software
Primary care or ambulatory clinics, Academic medical centers
Electronic Health/Medical Records (EHR or EMR), Patient-Reported Outcomes, Patient Registries
Geographic scope type
Locations of Focus
Chicago, Cook County, Illinois
Population Network Size
We custom developed an audio computer-assisted self interview software system and integrated the data into a clinical data warehouse hosted on a SQL Server platform.
CER/PCOR Study Priority Populations
Low-income groups, Minority groups, Individuals who need chronic care, Individuals who live in inner-city areas
We developed an infrastructure to collect patient reported outcomes and join the results to data from the EMR.
Outcome(s) of Interest
Patient quality of life and symptom burden and its association with health services utilization and comorbidity scores.
Acknowledgement of Funders
This study was funded by the Agency for Healthcare Research and Quality (AHRQ), grant number: R24 HS19481-01.