There has been much interest concerning residency training and how it relates to patient safety, clinical outcomes, and the increasing costs associated with teaching hospitals. Training residents requires a balance of patient safety while also providing educational opportunities for the doctors in training, and little research has been done to assess at which level orthopaedic residents should be involved in surgery. A study on resident participation in laparoscopic cholecystectomies demonstrated a longer operative time and a higher rate of operative complications in junior versus senior resident cases (1). A study of arthroscopic anterior cruciate ligament (ACL) reconstruction showed significantly longer surgical times, thus leading to higher cost for procedures performed by a teaching service (2). Furthermore, it was shown that teaching residents to perform total knee arthroplasty (TKA) resulted in increased operating room time and resource consumption when compared to a TKA performed on a private service (3). Although previous studies have shown an increased operative time (1-4), a comparison of total hip and knee arthroplasties performed on a teaching service versus a private practice service did not show a difference in any clinical outcome between the groups (4). Another study evaluated whether the training level of a resident was related to the rate of pediatric prescription errors in an academic emergency department and found that error rates were no higher for the lower-level residents. This study also found no difference in error rates between the beginning and end of the academic year (5).
In the present study, our goal was to measure patient safety in regards to junior vs senior resident involvement in the operating room. Ligament injuries were chosen as the measure of patient safety because they are directly related to the operative skill level of the surgeon, their avoidance is important, and they may be readily determined. Postoperative alignment was not used because all decisions regarding placement of cutting jigs were reviewed and possibly adjusted by the attending surgeon before bone cuts were made. Operative time was not used because it was consistently less for both junior and senior residents than the upper limit of 2 hours of tourniquet time. Estimated blood loss could be considered as another measure of safety, but a tourniquet was used in all cases and intraoperative blood loss was minimal.
In addition, previous authors have noted a higher level of ligament injury in patients with increased body mass index (BMI) (6,7). Therefore, the patients’ BMI was also recorded to determine if junior- or senior-level residents had a higher proportion of patients with a higher BMI.
Although one might expect the complication rate to be higher with junior residents, after training residents since 1996 the senior author felt that the incidence of complications was not higher for junior residents with appropriate supervision. We hypothesized that there would be no difference in intraoperative ligament injuries between TKA performed by junior or senior residents and no difference in ligament injuries between TKA performed in the months at the beginning or end of the rotation or in rotations at the beginning or the end of the postgraduate year.
MATERIALS & METHODS
After institutional review board approval, the operative reports of all TKAs performed by a single surgeon in a teaching hospital between January 1, 2008 to December 31, 2012, were reviewed to determine the incidence of ligament injury occurring during the operation. Only cases in which the same attending surgeon was directly supervising were included in order to eliminate any variability in supervising attending physicians’ styles or level of experience. Demographic and hospital data, including patient BMI, were collected prospectively for all patients. Using the operative reports, the postgraduate level of the primary surgeon, the occurrence of ligament injury, and the date of the operation were recorded. Patients with preoperative ligament injuries were excluded from the study.
Postgraduate year (PGY) 1, 2, and 3 were defined as junior residents and PGY 4 and 5 were defined as senior residents. The rotation duration of the arthroplasty service is 3 months. All residents rotated as both a junior and as a senior on the arthroplasty service for a total of 6 months. In all surgeries, the resident was the primary surgeon with the attending surgeon directly supervising and assisting. Typically, the junior resident had seen the senior resident or attending surgeon perform several TKAs and understood the technique, surgical approach, and preparation. The senior residents had previous experience during their junior years of training as the primary surgeon in performing a TKA, but the total number of knee replacements that the junior residents had observed prior to the start of the arthroplasty rotation was variable and depended on their rotations in medical school. The experience level of the senior residents prior to beginning the arthroplasty rotation was generally 30-50 TKAs that they had performed as a junior resident.
The attending surgeon was present for the critical portion of the operation in all patients, from incision until subcutaneous closure. The posterior cruciate ligament was retained, if it appeared functional, in all patients. An attempt was made to protect the ligaments with retractors and to avoid avulsing ligaments. A ligament was classified as injured during the operation if there was no evidence of preoperative ligament damage and the ligament was torn, stretched, cut by the saw, avulsed from its insertion, or in any way appeared compromised during surgery.
To compare the differences between age, gender, and BMI for the patients treated by junior and senior residents, an analysis of variance (ANOVA) with Bonferroni correction was used.
To determine if there was a difference between junior and senior residents performing the surgery with regards to intraoperative ligament injury a multinomial logistic regression model was applied. A Chi-square test was used to compare junior and senior residents and their relationship to the rate of PCL and MCL injuries since these were the two most common ligament injuries. All statistical analysis was performed using SPSS 20 and p<0.05 was considered a statistically significant difference.
The rate of ligament injury was not determined in cases without resident involvement. A previous study (4) noted a complication rate of 3% of TKAs performed by a private practice service, but the literature regarding the rate of ligament damage in TKAs performed without resident involvement is scarce.
This study identified 346 consecutive TKA operations. There was no statistically significant difference in age (p=0.10), gender (p=0.10), or BMI (p=0.46) for the group of patients treated by junior residents versus senior residents. The most common preoperative diagnosis was osteoarthritis. Among the patients, 289 were female, 57 were male, and the average age was 57. The average BMI was 35.4 and the average BMI for junior and senior residents was 32.7 and 32.0, respectively. 143 (41%) of the operations were performed by a junior resident and 203 (59%) were performed by a senior resident. With regard to the junior residents, 140 operations were performed by second-year residents and 3 operations were performed by third-year residents. As for the senior residents, 196 were in their fourth year of training and 7 were in their fifth year.
The overall incidence of ligament injury was 7.4% (26 of 346 surgeries). There was no significant difference (p=0.58) in the occurrence of injury in cases performed by junior residents, 9 of 143 (6.3%), when compared to cases performed by senior residents, 17 of 203 (8.4%). However, there were more PCL injuries in TKAs performed by junior residents when compared to senior
residents (p=0.76) and there were more MCL injuries in TKAs performed by senior residents than junior residents (p=0.05). These were the two most common ligaments damaged by the junior and senior residents, respectively. There was also no difference in ligament damage with regards to the specific month within a rotation (p=0.69) or when the rotation took place (p=0.65) (Table 1). The overall rate of specific ligament damage can be seen in Table 2.
There was a significant difference (p=0.03) in ligament injuries between all BMI categories greater than 30 when compared to patients with a BMI of less than 30. A patient with BMI of 30-40 was ten times more likely to be at risk for an injury, a BMI of 40-50 conferred eight times the risk, and a patients with a BMI of greater than 50 was nine times more likely to be at risk than someone with a BMI of less than 30.