Abstract
Background/Aims Little is known about how physician work affects the categorization of a visit for billing purposes. Wide variation in physician billing practice is well recognized, but most physicians bill legitimately and provide appropriate documentation. We investigated factors that can account for variations in documentation, and thus the coded visit level, in a primarily fee-for-service practice setting.
Methods We examined adult established patient primary care visits (CPT = 99211–99215 or level 1–5) in a large multispecialty ambulatory care organization where physician compensation is based on work RVUs associated with visit coding. In 2008–2011, there were 1,506,603 visits by 352,938 unique patients to 382 unique providers. Over 90% of visits were level 3(57%) or level 4(34%). Data on all encounters were extracted from the electronic health records to measure scheduled visit length, physician work during and between visits (patient calls and messages, lab/imaging/medication/referral orders, and chart updates), patient case-mix (age, sex, comorbidities) and relationship with the physician (whether own primary care provider (PCP), years with the PCP, visits to the PCP in the past year). Physician specialty and years of practice, clinic site, and year were included in a multi-level random effects model.
Results Visits were more likely to be a level 4 (vs. 3) when scheduled for 30min or 40min (vs. 20min) (OR = 2.8 for both). The number of orders (OR = 1.3) and procedures conducted during the visit (OR = 1.2) were positive predictors of level 4 visit. Phone/email contacts with patients (OR = 1.2) and orders before the visit (OR = 1.3) were also positive predictors of level 4 coding. Patient age (OR = 1.01), number of comorbidities (OR = 1.05), own PCP visit (OR = 1.2), number of previous visits to the PCP (OR = 1.02) were positive predictors while longer years with the PCP (OR = 0.98) and female patients (OR = 0.94) were negative predictors of level 4 visits (all P <0.01). None of PCP characteristics was significant.
Conclusions Objective measures of PCP work effort, including scheduled appointment length, orders and procedures done during the visit, out-of-office messaging and pre-visit orders, and patient complexity, are associated with higher billing codes. The non-office visit effort may reflect more efficient use of time in the office to accomplish more things.
