Characteristics and Practice Patterns of Non-certified Surgeons Treating Medicare Patients.

OBJECTIVE: The objective of this study is to explore the patient characteristics and practice patterns of non-certified surgeons who treat Medicare patients in the United States. SUMMARY BACKGROUND DATA: While most surgeons in the United States are board-certified, non-certified surgeons are permitted to practice in many locations. At the same time, surgical workforce shortages threaten access to surgical care for many patients. It is possible that non-certified surgeons may be able to help fill these access gaps. However, little is known about the practice patterns of non-certified surgeons. METHODS: A 100% sample of Medicare claims data from 2014-2019 were used to identify practicing general surgeons. Surgeons were categorized as certified or non-certified in general surgery​​ based on data from the American Board of Surgery. Surgeon practice patterns and patient characteristics were analyzed. RESULTS: A total of 2,097,206 patient cases were included in the study. These patients were treated by 16,076 surgeons, of which 6% were identified as non-certified surgeons. Compared to certified surgeons, non-certified surgeons were less frequently fellowship-trained (20.5% vs. 24.2%, P=0.008) and more likely to be a foreign medical graduate (14.5% vs. 9.2%, P<0.001). Non-certified surgeons were more frequently practicing in for-profit hospitals (21.2% vs. 14.2%, P<0.001) and critical access hospitals (2.2% vs. 1.3%, P<0.001), and were less likely to practice in a teaching hospital (63.2% vs. 72.4%, P<0.001). Compared to certified surgeons, non-certified surgeons treated more non-White patients (19.6% vs. 14%, P<0.001) as well as a higher percentage of patients in the two lowest socioeconomic status (SES) quintiles (36.2% vs. 29.2%, P<0.001). Operations related to emergency admissions were more common amongst non-certified surgeons (68.8% vs. 55.7%, P<0.001). There were no differences in gender or age of the patients treated by certified and non-certified surgeons. CONCLUSION: For Medicare patients, non-certified surgeons treated more patients who are non-White, of lower SES, and in more rural, critical-access hospitals.

Evaluating Educational Outcomes Using Patient Outcomes of New Surgeons Performing Partial Colectomy Compared to Cholecystectomy.

PURPOSE: Despite ongoing efforts to improve surgical education, surgical residents face gaps in their training. However, it is unknown if differences in the training of surgeons are reflected in the patient outcomes of those surgeons once they enter practice. This study aimed to compare the patient outcomes among new surgeons performing partial colectomy-a common procedure for which training is limited-and cholecystectomy-a common procedure for which training is robust. METHOD: The authors retrospectively analyzed all adult Medicare claims data for patients undergoing inpatient partial colectomy and inpatient cholecystectomy between 2007 and 2018. Generalized additive mixed models were used to investigate the associations between surgeon years in practice and risk-adjusted rates of 30-day serious complications and death for patients undergoing partial colectomy and cholecystectomy. RESULTS: A total of 14,449 surgeons at 4,011 hospitals performed 340,114 partial colectomy and 355,923 cholecystectomy inpatient operations during the study period. Patients undergoing a partial colectomy by a surgeon in their 1st vs 15th year of practice had higher rates of serious complications (5.22% [95% CI, 4.85%-5.60%] vs 4.37% [95% CI, 4.22%-4.52%]; P < .01) and death (3.05% [95% CI, 2.92%-3.17%] vs 2.83% [95% CI, 2.75%-2.91%]; P < .01). Patients undergoing a cholecystectomy by a surgeon in their 1st vs 15th year of practice had similar rates of 30-day serious complications (4.11% vs 3.89%; P = .11) and death (1.71% vs 1.70%; P = .93). CONCLUSIONS: Patients undergoing partial colectomy faced a higher risk of serious complications and death when the operation was performed by a new surgeon compared to an experienced surgeon. Conversely, patient outcomes following cholecystectomy were similar for new and experienced surgeons. More attention to partial colectomy during residency training may benefit patients.

Management of Massive Upper Gastrointestinal Hemorrhage Secondary to Gastrosplenic Fistula.

Gastrosplenic fistula (GSF) is seen secondary to the development of a fistulous track between the stomach and spleen and/or splenic vessels. It is most commonly seen in patients with diffuse B-cell lymphoma, who usually present with symptoms of abdominal pain and weight loss. GSF has also been seen in patients with gastric adenocarcinoma, Hodgkin's lymphoma, peptic ulcer disease, splenic abscesses, and post gastric sleeve resection. Less than 25% of the patients with GSF may present with upper gastrointestinal bleed (UGIB). This presentation of GSF is common with benign causes including peptic ulcer disease. UGIB secondary to GSF, while rare, requires prompt identification and intervention, to avoid catastrophic outcomes. We discuss the case of a 64-year-old female with GSF, who presented with sentinel bleed followed by hemorrhagic shock, secondary to a B-cell lymphoma, who was managed with a partial gastrectomy, splenectomy, and distal pancreatectomy, with favorable outcomes.

How Do Resident Surgeons Identify Operative Case Complexity? An Analysis of Resident versus Attending Perceptions.

OBJECTIVE: Accurate recognition of patient-related complexity of an operation is critical for appropriate surgical decision making. It is not yet understood whether general surgery residents are able to accurately assess the relative complexity of a given operative case. This study investigates the agreement of case complexity ratings between residents and attending surgeons and explores whether resident-related factors correlate with any discordance in perception of patient-related operative complexity. DESIGN: Residents and attending surgeons rated the relative complexity of completed cases on a 3 point scale via the SIMPL (Society for Improving Medical Professional Learning) operative assessment smartphone app. Additional trainee demographic data, autonomy ratings, and performance ratings were also obtained from the SIMPL registry for each rated case. Complexity agreement was defined as an equal rating between the resident and attending and assigned a value of zero. Over-estimate ratings were assigned a positive value and under-estimate ratings were assigned a negative value. Trends in complexity agreement were analyzed using descriptive statistics and mixed-effects models. RESULTS: A total of 43,179 general surgery cases were rated by 1946 categorical general surgery residents and 1520 attending surgeons between 2015 and 2020. Residents and attendings agreed on case complexity in 63.23% of cases, while the residents overestimated complexity in 13.37% of cases and underestimated complexity in 23.40% of cases. Every level of resident except post-graduate year 2 had similar rates of agreement about the complexity of a procedure, while residents who received a higher autonomy rating were more likely to be in agreement with the faculty raters (OR 1.12, 95% CI 1.06-1.19). CONCLUSIONS: The results of this study suggest that general surgery residents inaccurately perceive the patient-related complexity of a given case approximately one third of the time. Greater experience and operative autonomy appear to be associated with higher complexity agreement. Future research into factors influencing perceived case complexity may provide insight into how to best implement new teaching for surgical residents regarding the concept of case complexity.

US general surgical trainee performance for representative global surgery procedures.

BACKGROUND: Many US general surgery residents are interested in global surgery, but their competence with key procedures is unknown. METHODS: Using a registry managed by the Society for Improving Medical Professional Learning (SIMPL), we extracted longitudinal operative performance ratings data for a national cohort of US general surgery residents. Operative performance at the time of graduation was estimated via a Bayesian generalized linear mixed model. RESULTS: Operative performance ratings for 12,976 procedures performed by 1584 residents in 52 general surgery programs were analyzed. These spanned 17 of 31 (55%) procedures deemed important for global surgical practice. For these procedures, the probability of a graduating resident being deemed competent to perform a procedure was 0.95 (95% confidence interval 0.86-1.00) but was less than 0.9 for 3 observed procedures. CONCLUSION: Our results highlight gaps in the preparedness of US general surgery trainees to perform procedures deemed most important for global surgery settings.

Feedback-Seeking Behavior and Practice Readiness for General Surgery.

OBJECTIVE: To analyze the relationship between feedback-seeking behavior, operationalized as the number of trainee-requested evaluations, with ratings of surgical trainees' operative autonomy and performance. DESIGN: We analyzed operative assessment data using the System for Improving and Measuring Procedural Learning's smartphone-based assessment app called Society for Improving Medical Professional Learning (SIMPL) OR. Using cross-classified mixed effects models, we analyzed the association between trainee-requested SIMPL OR app evaluations and both trainee performance and autonomy ratings. Models included covariates for requested evaluations, PGY-year, month of the academic year, and patient-related case complexity. Random effects for program, procedure, rater, and trainee were also included to account for correlations among evaluations. Only ratings for procedures deemed Core to general surgery were included. SETTING: Operative assessment data using the SIMPL OR app requested by categorical U.S. general surgery residents between September 2015 to April 2021. PARTICIPANTS: A total of 61 general surgery residency programs, encompassing 2190 categorical general surgery residents. RESULTS: A total of 58,104 SIMPL app operative assessments were analyzed. Autonomy scores were weakly but positively associated with number of trainee-requested evaluations (B = 0.002, p < 0.001). Trainee-requested evaluations were also statistically associated with operative performance scores ( B = 0.002, p < 0.001). CONCLUSIONS: The propensity of a resident to seek feedback using the SIMPL app was weakly associated with higher operative autonomy ratings and higher operative performance ratings. While regular feedback is important for monitoring performance over time, more direct approaches related to the quality of feedback that trainees receive may be needed to better assess the relationships between feedback-seeking behavior and operative autonomy as well as performance.

How Do Programs Measure Resident Performance? A Multi-Institutional Inventory of General Surgery Assessments.

OBJECTIVE: To perform an inventory of assessment tools in use at surgical residency programs and their alignment with the Milestone Competencies. DESIGN: We conducted an inventory of all assessment tools from a sample of general surgery training programs participating in a multi-center study of resident operative development in the United States. Each instrument was categorized using a data extraction tool designed to identify criteria for effective assessment in competency based education and according to which Milestone Competency was being evaluated. Tabulations of each category were then analyzed using descriptive statistics. Interviews with program directors and assessment coordinators were conducted to understand each instrument's intended use within each program. SETTING: Multi-institutional review of general surgery assessment programs. PARTICIPANTS: We identified assessment tools used by 10 general surgery programs during the 2019 to 2020 academic year. Programs were selected from a cohort already participating in a separate research study of resident operative development in the United States. RESULTS: We identified 42 unique assessment tools used. Each program used an average of 7.2 (range 4-13) unique assessment instruments to measure performance, of which only 5 (11.9%) were used by at least 1 other program in our sample. Of all assessments, 59.5% were used monthly or less frequently. The majority (66.7%) of instruments were retrospective global assessments, rather than discrete observed performances. There were 4 (9.5%) instruments with established reliability or validity evidence. Across programs there was also significant variation in the volume of assessment used to evaluate residents, with the median total number of evaluations/trainee across all Milestone Competencies being 217 (IQR 78) per year. Patient care was the most frequently evaluated Milestone Competency. CONCLUSIONS: General surgical assessment systems predominantly employ non-standardized global assessment tools that lack reliability or validity evidence. This variability makes it challenging to interpret and compare competency standards across programs. A standardized assessment toolkit with established reliability and validity evidence would allow training programs to measure the competence of their trainees more uniformly and understand where improvements in our training system can be made.

Novel method to link surgical trainee performance data to patient outcomes.

BACKGROUND: A significant roadblock in surgical education research has been the inability to compare trainee performance to the outcomes of those surgeons after they enter independent practice. We describe the feasibility of an innovative method to link trainee performance data with patient outcomes. METHODS: We extracted surgeon NPI numbers from Medicare claims data for common general surgery procedures between 2007 and 2017. Next, American Board of Surgery (ABS) trainee performance data was cross-referenced with additional resources to supplement NPI data. The patient and trainee datasets were linked using NPI number and a linkage rate was calculated. RESULTS: We identified 12,952 unique surgeons in the Medicare file. Medicare surgeons were matched with ABS records by NPI number, with 96.2% (n = 12,460) of surgeons linked successfully. CONCLUSIONS: We demonstrated a novel process to link patient outcomes to trainee performance. This innovation can enable future research investigating the relationship between surgical trainee performance and patient outcomes in independent practice.