Physician Work Efforts Reflected by Millions of Primary Care Transactions In An Electronic Health Record
Physician time is a key resource in health services delivery. Understanding how physicians spend their clinical time is essential, given the need to understand practical capacity, guide staffing and support models, as well as to improve the accuracy of payment for primary care services. Fee-for-service (FFS) payments are based on physicians’ report of their resource use linked to face-to-face clinical encounters with their patients. The Resource-Based Relative Values Scale (RBRVS) was built, for some specialties, however, on survey responses to vignettes from about 20 physicians. Many analysts have questioned the accuracy of such reports in reflecting the true efforts in providing medical services. Recent delivery system innovations such as patient-centered medical homes have been changing how physicians spend their time. More objective data on physician work efforts in more recent time are needed to guide delivery systems and public policy.
Methods
This novel study examines physician work effort by leveraging the access-time stamp functionality of an electronic health record (EHR) of a large multispecialty group practice. The unobtrusive approach assigns physician time spent to particular work efforts. Over 24 million EHR activity records were categorized into face-to-face visits and “desktop medicine” (ordering medications, checking laboratory results, and communicating with patients) and then analyzed for patterns of care time by 336 to 390 physicians (117~130 family physicians, 119~146 internists, and 95~114 pediatricians) serving 656,745 patients in 2011~2014. Some clinics in the system had been recognized by NCQA as Level 3 (N=5), 5 clinics as Level 2 (N=5) patient-centered practices while 6 clinics didn’t apply for NCQA recognition. We fitted linear mixed effect models for total EHR-logged time spent face-to-face with patients in exam rooms and on desktop medicine with covariates including physicians’ years of experience using EHR, clinical FTE, average patient Charlson score, average patient age, and NCQA recognition level of their clinic.
Results
Access log suggests that 55% of physicians’ time-in-clinic can be attributed to patient care, of which 25% involved face-to-face visits and 30% on desktop medicine. From 2011~2014, for each in-person patient visit, an average of 11.1~9.9 minute was logged inside the exam room and 6.3~7.6 minutes outside the room on the day of the visit. The average daily number of face-to-face visits declined from 15 to 13 patients whereas daily number of patients touched by desktop medicine activities increased from 42 to 44. Physicians in Level 2 clinics and clinics that didn’t apply for NCQA recognition spent relatively less time on face-to-face visits and less time on desktop medicine, compared to those in Level 3 clinics.
Conclusions
The findings suggest physicians spend less than half of their time-in-clinic seeing patients face-to-face. While time on face-to-face visits has been declining in recent years, more time is being spent on desktop medicine. Many desktop medicine activities–e.g., care coordination—are of high value to delivery systems and to patients. The staffing, scheduling and support as well as provider payment models of primary care need to encourage and support patient-centered allocation of physician work efforts.