Background:

The electronic health record (EHR) is recognized as a key component of the modern practice of medicine. Despite numerous advantages, EHR have enabled frequent copy–paste events (CPE). While previous studies show that CPE occur in up to 54% of progress notes, the significance of these CPE is unclear. We hypothesized that the frequency and location of CPE by physicians on the same service would differ from CPE created by physicians.

Methods:

All physician–generated notes for patients admitted to a general medicine service in a single year were included. Notes associated with 12 random patients were compared in a pair–wise fashion on a line–by–line level for CPE using a Microsoft Access database. False CPE (e.g. generic signatures and templates) were identified and removed. All CPE were then categorized and quantified into components of notes (e.g. physical exam, labs, impression, plan). Chi–squared and Fisher’s exact were used for statistical analysis comparing location of CPE based upon provider and service, with different service, different provider CPE as our baseline.

Results:

We observed a total of 57,125 CPE in 299 documents from our 12 patients. After we removed 46815 false CPE, 229 (76.6%) documents remained, containing 10,310 (18.0%) CPE. 1754 note components were copied, 1062 (60.5%) by the same provider on the same service, 562 (32.0%) by different providers on the same service (e.g. transition from one hospitalist to the next) and 127 (7.2%) by different providers on different services (e.g. between a consult and primary service). The location of CPE differed depending upon service. The comparative distributions of intra– and inter–provider CPE components generated from the same service were as follows: significant (16.7% vs. 17.3%) and insignificant (21.0% vs. 24.6%) portions of the plan; physical exam (24.9% vs. 19.4%); assessment (15.5% vs. 22.8%); and labs/studies (11.8% vs. 6.9%). The most common, inter–provider CPE components from different services were: lab/study (39.4%); insignificant portions of the plan (28.3%); past medical history (8.7%); and medications (6.3%). Inter–service, inter–provider CPE were significantly less frequent than intra–service CPE in physical exam, significant portions of the plan and assessment (P < 0.001 for all).

Conclusions:

CPE between physicians on different services are the most infrequent and differ substantially in the location of such events compared to CPE on the same service. CPE are likely to occur for different reasons between services versus within a service, which should be considered in future approaches in the evaluation of CPE.