This study was approved both by the Teikyo University Institutional Review Board (IRB No. 12-030-3) and by the Teikyo University Committee on Conflict of Interest (TU-COI 12–201). We conducted the present research in accordance with the guidelines set by the Ethical Principles for Medical Research Involving Human Subjects in DECLARATION OF HELSINKI. The consent to participate was waived because our present study was a retrospective observational study. The need for consent was waived by the Teikyo University Institutional Review Board (IRB No. 12-030-3) as above. Since the present study is based on our three previous ones, we intentionally used the similar methods in data collection and analysis framework to those described in our previous ones [1–3].
Data collection
We collected data from all the surgical procedures performed in the main operating rooms of Teikyo University Hospital from April 1 through September 30 in 2013-18. Teikyo University Hospital is located in metropolitan Tokyo, Japan, serving a population of ~ 1,000,000. It has 1,152 beds and has a surgical volume of approximately 9,000 cases annually. The necessary information for the present study was extracted from surgical records in its electronic medical record system [1–3].
Exclusion criteria for surgery were similar to our previous studies [1–3]. First, we excluded surgical procedures performed under local anesthesia by surgeons to equalize resource utilization. Second, oral, dermatologic and ophthalmic surgical procedures were excluded because most of their cases were minor surgeries. Third, the surgical procedures were excluded if the patients die within one month after surgery to maintain a constant quality outcome of surgery. Fourth, the surgical procedures which were not reimbursed under the surgical fee schedule were excluded. Fifth, the surgical procedures were excluded if their records were incomplete for any reason [1–3].
Analysis framework
Similar to our previous studies [1–3], output-oriented Charnes-Cooper-Rhodes model of DEA was used. This model was particularly relevant because of its ability to employ multiple inputs and outputs [11]. It can be applied for evaluating outputs while controlling multiple inputs (i.e. resources). In this analysis, we focused on the surgeons’ activity and their clinical decision as we did in our previous studies [1–3]. A decision making unit (DMU) is defined as the entity that is regarded as responsible for converting inputs into outputs in DEA [12]. We defined in this study the DMU as a surgeon with the highest academic rank that scrubbed in the surgery. All the inputs and output are under the control of a DMU. Inputs were defined as (1) the number of medical doctors who assisted surgery (assistants), and (2) the time of surgical operation from skin incision to skin closure (surgical time). We defined the output as the surgical fee for each surgery [1–3]. It is classified as K000- K915 in the Japanese surgical fee schedule and is called “K codes.” Each surgical procedure is assigned to one of the K codes which correspond with surgical fees [13–16].
Japan has maintained universal health insurance system for more than half a century. Most health care providers are reimbursed on a fee-for-service basis according to the fee schedule that set prices uniformly at the national level [4]. This fee schedule is revised every two years at the Central Social Insurance Medical Council [13–16]. Our study periods represented four surgical fee schedules because the fee schedule was revised in Japan on April 1st of 2014, 2016 and 2018. We also compiled the data in four periods; 2013; 2014/2015; 2016/2017; 2018.
We added all the inputs and outputs of the surgical procedures for each DMU during the study period in each year and in each surgical fee schedule, and calculated his/her efficiency scores using DEA-Solver-Pro Software (Saitech, Inc., Tokyo, Japan) [17]. The efficiency scores must take a value greater than zero and less than or equal to one, and the most efficient surgeons are given the score of 1 [6].
All the surgeons analyzed were members of one of the following ten surgical specialties; cardiovascular surgery, emergency surgery, general surgery, neurosurgery, obstetrics & gynecology, orthopedics, otolaryngology, plastic surgery, thoracic surgery and urology [1–3]. We compiled their efficiency scores in their surgical specialties in each year and in each surgical fee schedule. By comparing the median efficiency scores of surgical specialties, we inferred Gini coefficients. We inferred Gini coefficients and their standard errors in each year of the median of efficiency scores in each surgical specialty using Bootstrap methods [10].The Gini coefficients all lie between 0 and 1, and the most equal distribution of efficiency scores are given the Gini coefficient of 0.
Statistical analysis
We used Stata Data Analysis and Statistical Software (Stata 14, StataCorp LP, College Station, Texas, U.S.A.) for our statistical analysis. We compared the Gini coefficients between the years and between the surgical fee schedules using the methods described by Davidson [10]. Briefly, if the independent samples are drawn from two populations by Bootstrap methods, they are assumed to distribute normally. Therefore, the Gini coefficients were compared by one way analysis of variance. A p-value < 0.05 was considered statistically significant.