Type and frequency of healthcare encounters can predict poor surgical outcomes in anterior cruciate ligament reconstruction patients

Background: Several challenges are associated with collecting clinically meaningful post-operative outcomes. The widespread implementation of electronic medical records (EMR) offers a new opportunity to evaluate surgical outcomes using routinely collected data in these systems. This study evaluated whether surgical outcomes can be ascertained from EMR's hospital and outpatient encounters. Specifically, we evaluated anterior cruciate ligament reconstructions (ACLR) outcomes. Methods: A retrospective cohort study of 6985 ACLRs performed between 2/2005-9/2012 was conducted. Patient encounters during days 1-90 and days 91-180 after ACLR surgery were the exposures of interest. Nine hospital and eight outpatient encounter types were evaluated. The main endpoint of the study was revision surgery six months after ACLR. Results: The cohort was 66.7% male, the mean age was 28 (standard deviation = 11) years-old, and the incidence of revision was 1.5% (n = 105). After adjustments, in days 1-90 post-ACLR, compared to patients with 0-4 orthopedic office visits, patients with 5-9 (hazard ratio (HR) = 9.9, 95% confidence interval(CI), 4.3-23.2) and those with 10 or more (HR = 13.8, 95%CI, 5.6-33.8) visits had a higher risk of revision. In days 91-180, patients with any outpatient hospital encounters (HR = 2.5, 95%CI 1.4-4.5) had a higher risk of revision than patients without visits. Additionally, patients with 4-5 regular office visits (HR = 3.8 times, 95%CI, 2.0-7.0) had a higher risk of revision surgery than those with 0-1 visits. Discussion: The number of post-operative outpatient visits was associated with ACLR revision surgery. Using EMR encounters to assess surgical outcomes is a viable option for monitoring ACLR patients. The simple assessment of visit types and number of encounters alone can provide valuable information regarding the normal course of rehabilitation of a surgical patient and possible deviation from this normal course. In large cohorts of patients, this type of patient surveillance can assist surgeons with monitoring their patients.