1.1 Background/rationale

With an increasing number of total joint arthroplasty (TJA) surgeries performed annually, optimizing time spent in the operating room (OR) has never been more paramount.¹ Many factors influence OR efficiency preoperatively, intraoperatively, and postoperatively. For example, overlapping arthroplasty cases in two different rooms has been shown to significantly reduce the total work hours, increase the number of cases performed per day, yet does not compromise patient safety or outcomes ^2–7. Staff and vendor turnover between cases, as well as having a consistent staff day-to-day, can also heavily influence OR efficiency by impacting workflow and collaboration of the multidisciplinary OR team ^8–10.

Higher surgeon volume has also been strongly reported in the literature as improving patient outcomes, as it is associated with lower rates of complication, infection, readmission, reoperation, shorter length of hospital stay, and higher likelihood of being discharged home.^11,12 Although this has been conjectured to be due to greater surgeon experience and confidence performing these procedures, the investigations generally report total annual cases, rather than the number of cases performed each day or the duration of each case.

Prior studies have shown that it is possible to optimize OR efficiency to better suit the staff, surgeon, and patient in both the hospital and surgical center settings while reducing associated costs ^10,13–15. Lonner et al. demonstrated that by simply eliminating redundant or underutilized instruments, set-up time was decreased by 6.5 min in total knee arthroplasty (TKA) and 9.3 min in total hip arthroplasty (THA).¹³ In an investigation by Attarian et al., the authors found that optimizing all elements of the OR process, such as patient, nurse, and surgeon factors as well as team variations and OR setup and breakdown, resulted in a decrease in room turnover time by 25 min and increased their daily average number of TJA cases per OR by 29%.¹⁵

1.2 Objective

This study investigates the effect on 90-day patient outcomes following the optimization of OR efficiency in a high-volume ASC setting. We compared OR efficiency and early patient outcomes prior to, and directly following, the implementation of processes intended to improve the OR efficiency. This study builds on the existing literature by incorporating efficiency changes at every element of the OR procedure in the setting of the ambulatory surgical center. We hypothesized that no differences in complication or revision rates would be found between groups.

2.1 Study design

After an institutional review board determined that this study met the requirements for exemption, our institutional research database was queried for all adult patients who underwent primary hip and knee arthroplasty performed by one fellowship-trained joint arthroplasty surgeon at a high-volume institution from July 2023 to December 2024.

The control group was comprised of 272 primary hip and knee arthroplasties performed between July 2023 and January 2024 prior to any optimization measures being implemented (traditional cohort). From February 2024 to April 2024 optimization measures were being sequentially implemented, and cases done during this transitional window were excluded from the outcome and efficiency analyses. The exclusion of the transition period was necessary to evaluate the steady-state performance of the optimize protocol, rather than the implementation process itself. There were 227 cases performed between May 2024 and December 2024, after optimization measures were fully implemented (optimization cohort).

Optimization was carried out with an artificial intelligence OR data and insights platform (DEO, Ann Arbor, MI). The platform uses a HIPAA compliant, anonymized software to record OR process flows and parallelization, taking over 50 datapoints per surgery. Following several days of data collection, insights are given regarding opportunities to increase operational efficiency and OR throughput, including material preparation, patient preparation, operative stages (exposure, bone preparation, implant trialing, wound closure), breakdown, OR turnover, and empty OR time. Feedback on areas for improved efficiency are given that are unique to each facility. At our facility, feedback focused strongly on non-operative portions of the operative day. Simulated models using a digital twin, showing areas for improvement in parallelization, are then provided for potential implementation.

The software identified multiple opportunities for efficiency optimization in the areas of preparation efficiency, technique adjustments, communication and coordination, and parallelization. For preparation efficiency initiatives, minimizing patient preparation idle time, optimizing staff allocation during preparation, ensuring timely gowning, eliminating pre-drape idle time, and reducing material idle time after trays were opened were identified as targets. For technique adjustments, attention was paid towards optimizing closing times and reducing patient positioning time. Communication and coordination improvements included increased communication between the OR team and the patient preoperative preparation team as well as implementing “trigger points” to better facilitate the movement of the fellow and lead surgeon between parallel ORs. Parallelization efficiency targets included decreasing patient lag time from when materials are first opened to when the patients first enters the OR, and reducing empty OR time in parallel rooms between cases. These changes were tested, implemented, and refined from February 2024 to April 2024.

2.2 Setting

Only cases performed at 2 ambulatory surgical centers were included, as these were the two locations where the efficiency optimization measures were implemented. The ASC sites included in this study opened in 2013 and 2021 respectively, and the operating surgeon first began operating at these sites in 2021. Both sites are structured very similarly in terms of patient flow and available staff support. The OR staff including the circulating nurses, anesthesia care team, and surgical technologists have undergone turnover, as many intuitions frequently experience. These personnel changes have remained limited but consistent throughout the optimization efforts. The fellowship-trained surgeon operates with the support of an adult total joint reconstruction fellow and a physician assistant.

2.3 Participants

Demographic data including sex, age at time of surgery, BMI, and time to follow-up were analyzed (Table 1). The age at the time of surgery was higher in the optimized cohort (66.5 vs 64.8 years, P = 0.03), and there were no significant differences between the two groups concerning sex, height, weight and BMI. Minimum follow-up was 90-days to capture early failures reasonably attributable to intraoperative variations. The average follow-up in the traditional cohort was statistically longer than that of the optimized cohort (0.7 vs 0.3 years, P < 0.01).

Patients under the age of 18, or who underwent revision TJA were excluded. Patients with persistent end-organ failure with inadequate medical optimization or a body mass index (BMI) greater than 50 kg/m² were not ASC candidates and were treated in an inpatient setting.

The breakdown of procedure type was similar in both cohorts, where 35% of the cases in the traditional cohort were THA, 38% were TKA, and 27% were unicompartmental knee arthroplasty (UKA) as shown in Table 1. In the optimized cohort, 38% of the cases were THA cases, 41% were TKA, and 20% were UKA cases (Table 1). The two groups were also similar in terms of primary diagnosis (Table 2). In both cohorts the most common primary diagnosis was osteoarthritis (94.1%, 93%), followed by post-traumatic arthroplasty (2.9%, 2.2%), and osteonecrosis (1.1%, 1.3%).

2.4 Variables

The primary outcomes of this investigation were 90-day complications and revisions for each cohort. Secondary outcomes included the average number of cases on full-schedule OR days, observed total open OR time (opening of the sterile tray to patient exiting the OR), time spent operating, case volume and case distribution, as well as patient-reported outcome measures (PROMs).

The PROMs were analyzed retrospectively and included the Patient-Reported Outcomes Measurement Information System (PROMIS) Global Health Physical (GHP) and Mental (GHM) scores, Knee dysfunction and Osteoarthritis Outcome Score, Joint Replacement (KOOS, JR), and Hip dysfunction and Osteoarthritis Outcome Score, Joint Replacement (HOOS, JR) collected preoperatively and at 4-months postoperatively. Each PROM was compared using previously reported minimal clinically important difference (MCID) values to determine clinical importance.^16,17

2.5 Data sources/management

All available postoperative revision and complication data, as well as patient-reported outcome measures (PROMs) were collected for both groups. With patient identifiers removed, the electronic medical records were also queried to retrospectively obtain the operating surgeon's case volume. Two full-scheduled OR days were observed prior to and following the implementation of process changes to compare case makeup and volume, total OR time, and OR parallelization. The surgeon's daily case volume was compared using the same timeline as described above for the traditional and optimized cohorts. The efficiency analysis only included full days when the surgeon was scheduled from 7:00am to 5:00pm, as the intent was to compare surgeon efficiency and case volume prior to and following optimization implementation, and partially- or un-scheduled OR days would influence this analysis.

2.6 Bias

To minimize contamination between study groups, cases performed during the transitional period from February 2024 to April 2024, when optimization measures were being sequentially implemented, were excluded from the outcome and efficiency analyses. Both cohorts were treated at the same ambulatory surgical centers by the same fellowship-trained surgeon, limiting variability related to surgical technique, institutional protocols, and care setting.

2.7 Study size

The study included 499 primary hip and knee arthroplasty cases performed at two ambulatory surgical centers, consisting of 272 cases in the traditional cohort and 227 cases in the optimization cohort. All eligible cases meeting inclusion criteria during the defined study periods were included in the analysis.

2.8 Quantitative variables

Minimum follow-up was 90-days to capture early failures reasonably attributable to intraoperative variations. PROMs were collected preoperatively and at 4-months postoperatively and compared using previously reported minimal clinically important difference (MCID) values.

2.9 Statistical method

Differences among categorical variables were assessed using Chi-square tests for large groups and Fisher's exact tests when any of the expected cell counts were less than five. Continuous variables were compared with independent samples t-tests. A P-value of less than 0.05 was used as the threshold for statistical significance. Statistical analyses were performed using IBM SPSS (Statistical Package for the Social Sciences) for Windows, v27.0 (IBM Corp., Armonk, New York, United States).

3.1 Outcome data

Neither cohort had any revisions in the first 90 days postoperatively for aseptic causes (P > 0.05). Each cohort had 1 revision in the first 90 days postoperatively due to periprosthetic joint infection (PJI), as shown in Table 3. In the traditional cohort, a diabetic patient with a history of triple bypass presented 82 days postoperatively with an acute, hematogenous PJI and was revised 83 days postoperatively. In the optimized group, an obese, diabetic patient presented 15 days postoperatively with wound drainage and was revised 16 days postoperatively for a PJI. Neither cohort experienced any intraoperative complications. The traditional cohort went on to have 10 total complications (3.67%), 5 of which were surgical complications, and the optimized cohort had 7 total complications (3.08%), 3 of which were surgical complications within the first 90 days following surgery (P > 0.05) (Table 3).

3.2 Main results

Prior to the implementation of process changes to increase OR efficiency, a mean of 8.7 cases were performed each full-scheduled OR day for 2 parallel ORs at the ASCs included in this study (Fig. 1). Following process changes, a mean of 9.9 cases were performed on full-scheduled OR days with 2 parallel ORs at the corresponding sites (Fig. 1).

Close

Fig. 1 Operating Room Case Volume Prior to and Following Optimization Implementation. Data are presented as daily case load corresponding to the operating room schedule date. Trend lines are provided for the group averages for Traditional (8.7 cases/day) and Optimized (9.9 cases/day) cohorts.

Export to PPT

The optimized cohort saw an 11-min decrease in mean case duration (94 vs 83 min) when compared to the traditional cohort (Fig. 2). A similar trend was seen for each respective procedure type. For TKA, UKA, and THA there was a decrease of 8 min (92 vs 84 min), 4 min (85 vs 81 min) and 12 min (89 vs 77 min) respectively (Fig. 3).

Close

Fig. 2 Total Operating Room Time per Case. Data are presented in minutes. The mean case duration for the traditional (top) cohort and the optimized (bottom) cohort.

Export to PPT

Close

Fig. 3 Total Operating Room Time per Case for Each Procedure Type. Data are presented in minutes. The mean duration of total knee arthroplasty (TKA), unicompartmental knee arthroplasty (UKA), and total hip arthroplasty (THA) are compared for Traditional cohort (orange) and the Optimized cohort (blue).

Export to PPT

Table 4 summarizes the observed days in the OR where case distribution, case volume, and total OR time were recorded. After OR flow was optimized, the mean time needed for the operative and material portions of each case was reduced, as well as the delay between the beginning of a case and the start of patient preparation, the procedure duration, the time needed to close the incision, and the room and material breakdown at the end of each case (Fig. 4). Additionally, the material preparation and the material breakdown were more effectively parallelized to allow each to occur during patient preparation and incision closure respectively, reducing the mean case duration.

Close

Fig. 4 Material and Patient Parallelization and Associated Timing. Data are presented in minutes. The mean duration of each stage of the operating room process is outlined for both patient care and surgical materials, including preparation, procedure duration, incision closure, and breakdown.

Export to PPT

3.3 Other analyses

The PROMIS GHM score at 4-month follow-up was significantly higher in the optimized cohort (54.7 ± 8.0 vs 52.9 ± 7.8, P = 0.03), but it did not reach a clinically important difference (Table 5). The MCID values used to determine a clinically important difference from preoperative values are listed in Table 5. None of the other PROMs analyzed exhibited a statistically significant difference between the traditional and the optimized cohorts (P > 0.05).

4.1 Key results

This study aimed to evaluate the impact of optimization efforts on OR efficiency and patient outcomes in a high-volume ASC setting. Implementing optimization strategies led to an increase of 1.2 cases per full-scheduled OR day following a 3-month implementation period. There were no appreciable compromises in patient safety or postoperative outcomes. These results underscore the potential of OR workflow improvements, such as increasing efficiency of perioperative processes, avoiding delays in room turnover, and improving team dynamics to enhance the overall OR efficiency.

Implementing optimization strategies increased the mean case load from 8.7 to 9.9 cases per full OR day with 2 parallel ORs, as well as reduced the mean OR time by 11 min per case. The improvements in efficiency are further supported by a modest increase in total open OR time per day after optimization (increased from 7 h and 53 min to 8 h and 2 min). Due to the efficiency implementations, the 8-h OR scheduling template changed from allowing a maximum of 11 cases per day to 13 cases per day. Thus, improvements found in the parallelization of patient care and OR material management also shortened the duration of cases and allowed more cases to be performed in the same amount of time.

The patient outcomes were largely unchanged between the two cohorts, with no statistically significant differences observed in the revision rate (0.4% vs 0.4%, P > 0.05) or the complication rate (3.67% vs 3.08%, P > 0.05). No patient was revised for aseptic reasons, and the rate of periprosthetic joint infection was similar between the two cohorts (0.4%). These values align with the previously reported periprosthetic infection rates of 1.6% in the first two years following THA reported by Ong et al. and 1.55% in the first two years following TKA reported by Kurtz et al..^18,19 Additionally, of the PROMs analyzed, none showed a clinically important difference between the two groups, supporting similar patient outcomes in the group prior to optimization and following optimization.

The results in this study suggest that increasing OR throughput and daily case volume can result in similar patient outcomes. The workflow described in this study includes the use of overlapping ORs, with the same attending surgeon present for all critical surgical portions of each case. Prior studies have shown that using this method to increase OR throughput and surgeon case volume does not adversely affect patient outcomes, with no impact on intraoperative complications, complication rate, infection rate, or revision rate ^2,4–7. Our investigation builds on these results but does so in an intentional manner in the stand-alone ASC setting where dedicated orthopaedic ORs and operative staff are familiar with the procedures, team dynamic, and surgeon expectations. The use of dedicated orthopaedic operating rooms, has been shown to increase OR throughput, minimize perioperative processes, and reduce the operating time needed per procedure, without increasing adverse events.²⁰ Our results mirror those findings, with increased throughput while maintaining similar total open OR times and surgeon operating times. Team dynamics in the ASC play an essential role in improving efficiency as well. Previous studies having highlighted that turnover among OR staff, including circulating nurses and anaesthesiologists, can significantly increase operative time and decrease efficiency. We strongly agree the presence of surgeon-preferred staff, anaesthesiologist, or surgical technician can decrease surgical time and improve efficiency.^8,9 The use of an orthopaedic-specialized ASC likely has an impact on reducing variability in team dynamics, contributing to smoother OR workflows and consistent case throughput. Thus, the current study demonstrates overlapping ASC ORs combined with perioperative efficiency efforts can significantly increase case throughput without compromising patient safety or postoperative outcomes.

4.2 Limitations & strength

Despite these promising results, our study has potential limitations. The goal of the investigation was to determine if the willful implementation of more efficiency processes in the ASC contributed to increased early complications, revision, or worse PROMs. Therefore, the relatively short follow-up period of 90 days may not adequately capture the long-term effects of optimized OR processes; however, it is unlikely that long-term outcomes were influenced by OR speed and flow. Second, this is a retrospective review of prospectively collected data. Our institution has routinely collected standardized complication forms for over 30 years. Although it is possible complications were missed, it is unlikely given the institutional protocols in place for collecting adverse events. While we observed increased case volume, we did not account for patient complexity, which could have influenced the results. Additionally, while retrospective pre-post designs cannot fully eliminate temporal confounding or natural team maturation, the operating surgeon had been established at these centers for two years prior to data collection. The abrupt improvement in efficiency metrics following the specific intervention period argues against gradual secular trends as the primary cause. Other factors such as the aid of a physician's associate or an orthopaedic fellow during the operation, the use of cemented versus cementless implants and the addition of patellar resurfacing were also not accounted for and could have influenced our results. The BMI cutoff utilized at the included ASCs is up to 50 kg/m2, which is higher than most hospital settings.