No date for the PROM: the association between patient-reported health events and clinical coding in primary care

Objective: It is unclear whether data from patient-reported outcome measures (PROMs) are captured and used by clinicians despite policy initiatives. We examined the extent to which fall risk and urinary incontinence (UI) reported on PROMS and provided to clinicians prior to a patient visit are subsequently captured in the electronic medical record (EMR). Additionally, we aimed to determine whether the use of PROMs and EMR documentation is higher for visits where PROM data was provided to clinicians. Design: We conducted a cross-sectional patient-reported risk assessment survey and semi-structured interviews with clinicians to identify themes related to the use of PROMs. Setting: Fourteen primary care clinics in the US (eight intervention and six control clinics), between October 2013 and May 2015. Participants: Primary care clinicians and older adult (>= 66 years) patients completing a 46-item health risk assessment, including PROMs for fall risk and UI. Intervention: Risk assessment results provided to the clinician or nurse practitioners prior to the clinic visit in intervention clinics; data was not provided in control clinics. Main outcome: 1) Agreement between ICD-9 codes of fall risk or UI in the EMR and patient-reports, and 2) clinician experience of PROMs use and impact on coding. Results: A total of 505 older adult patients were included in the study, 176 at control clinics and 329 at intervention clinics. While patient reports of fall risk and UI were readily captured by PROMs, this information was only coded in the EMR between 3% - 14% of the time (poor Kappa agreement). Intervention clinics performed slightly better than control clinics. Clinician interviews (n = 16) revealed low use of PROMs data with multiple barriers cited including poor access to data, high quantity of data, interruption to workflow, and a lack of training on PROMs. Conclusions: Current strategies of providing PROMs data prior to clinic visits may not be an effective way of communicating important health information to busy clinicians; ultimately resulting in underuse. Better systems of presenting PROMs data, and clinician training on the importance of PROMs and their use, is needed.