Table 1 Crude incidence of events and adjusted risk following primary shoulder arthroplasty with and without preoperative CT scans. CI, confidence interval; VTE, venous thromboembolism
Source: Navarro RA, et al: “New Preoperative Planning Technologies in 8,117 Elective Shoulder Arthroplasty Procedures: Trends and Outcomes”


Published 12/20/2023
Rebecca Araujo

Rising Use of Preoperative Planning Technologies in TSA Not Linked to Reduced Risk of Aseptic Revision and Complications

A study presented at the AAOS 2023 Annual Meeting examined recent trends in use of preoperative planning technologies prior to total shoulder arthroplasty (TSA) and found that deployment of these technologies did not necessarily improve outcomes. The study was presented as a poster by Ronald A. Navarro, MD, FAAOS, FAOA, director of clinical affairs and professor of clinical science in the Department of Orthopaedic Surgery at the Kaiser Permanente Bernard J. Tyson School of Medicine. Dr. Navarro is also a member of the AAOS Now Editorial Board.

In particular, the investigators assessed the use of preoperative CT imaging and patient-specific instrumentation (PSI) in order to assess the widespread use of these technologies and any associated changes in surgical outcomes. The researchers evaluated the use of planning technologies in shoulder arthroplasties across the Kaiser Permanente system. In their study, the authors explained, “Post-market surveillance on the adoption of new technologies and their outcomes, outside of controlled trials in community settings, is often limited. More recently, many of these involve preoperative planning technologies directed at improving the durability of the glenoid component.”

“We wanted to understand how new technologies can benefit preoperative planning in advance of shoulder arthroplasty. We also wanted to see how these technologies are adopted by surgeons in our system over time, especially as we have to be free of conflict,” Dr. Navarro told AAOS Now. “We wanted to evaluate how these technologies might affect outcomes in patients undergoing shoulder arthroplasty.”

Overall, he reported, “We observed a trend of increasing use of preoperative CT scans for shoulder arthroplasty that peaked in 2017, while utilization of PSI has continued to increase.”

This cohort study was conducted using data from Kaiser Permanente’s shoulder arthroplasty registry, “a surveillance tool with 100 percent coverage that records information on patient-, procedure-, implant-, surgeon-, and hospital-related variables,” according to the study authors. Inclusion criteria were patients aged ≥18 years who underwent anatomical TSA or reverse TSA (RTSA) between 2015 and 2020. The investigators assessed whether preoperative planning technologies were used, and they defined technology use as either:

  • CT scan as a proxy for preoperative planning software utilization
  • PSI computer software system supported by implant manufacturer

Data on preoperative planning were assessed starting in 2015 for TSA and 2017 for RTSA. The authors noted that, during this time period, “There was no separate cost for use of PSI within the organization.”

Hazard ratios (HRs) were calculated for the risk of aseptic revision, and odds ratios (ORs) were used to calculate the likelihood of 90-day adverse events. Adverse events captured included an ED visit, readmission, deep infection, and venous thromboembolism (VTE). Incidences of revision and adverse events were compared between procedures that used preoperative planning technologies and procedures that used conventional radiography planning. Regression model variates included age, gender, BMI, American Society of Anesthesiologists classification, Elixhauser Comorbidity Burden, procedure type, Walch glenoid classification, and the use of upper-extremity walking aids. In addition, a cluster term was included in the model to adjust for surgeon differences.

In total, the analysis included 8,117 arthroplasty procedures performed by 130 surgeons at 40 hospitals. The mean patient age was 70.6 years, and 44.7 percent of patients were male.

Regarding the use of preoperative CT scans, in the total cohort, 31 percent of procedures utilized CT (n = 2,527). There was an upward trend in use in TSA between 2015 and 2017, peaking at 36 percent for TSA; in 2017, CT use was 42 percent for RTSA. By 2020, these rates declined to 20 percent and 28 percent, respectively. Compared with patients who did not receive CT scans, the CT group had a higher proportion of Walch type B or C native glenoids and an average of 13-minute longer operative time.

PSI software was utilized in 11 percent of procedures overall (n = 400). One percent of TSAs in 2015 used PSI; this rate rose to 25 percent in 2020. For RTSA, PSI use also rose between 2017 and 2020, from 5 percent to 16 percent, respectively. This technology was more commonly used in male patients. Compared with procedures without PSI software, the PSI group had more Walch type B/C native glenoids and a mean 10-minute longer operative time. Procedures using PSI were also more likely to be performed by a higher-volume surgeon. Notably, 28 percent of patients without PSI had a preoperative CT scan (n = 875).

Table 1 lists the incidence of aseptic revision and 90-day events associated with the use of preoperative CT scans. The investigators reported no significant differences in aseptic revision associated with the use of either preoperative planning technology (CT group, HR = 1.22, P = 0.257; PSI group, HR = 1.44, P = 0.311). Regarding other adverse events, the CT group had a lower likelihood of 90-day ED admission (OR = 0.84, P = 0.021) and a higher likelihood of 90-day VTE (OR = 1.79, P = 0.007) compared with patients who did not receive CT. There were two cases of 90-day deep infection among patients who received PSI, which was significantly greater likelihood than in patients without PSI (OR = 6.25, P = 0.038).

Given these findings, Dr. Navarro commented, “We cannot suggest the need for advanced imaging and use of PSI on cases to improve outcomes, as we were unable show that the use leads to these better outcomes. While we believe preoperative preparation for complex cases is important and helps to enhance the depth of anatomical understanding in advance of the case, we also feel an expectation of utilization of this technology in all case complexities by all surgeons may not be warranted.”

The authors noted that higher likelihood of VTE and deep infection associated with CT and PSI use, respectively, identified in this study were grounds for further study. Additionally, Dr. Navarro added, “Cost-effectiveness studies may be needed to discern whether these technologies add value to patient care.”

Regarding the limitations of this study, Dr. Navarro said, “Only associations, not causality, are reported from this observational study. The healthcare system did not reliably collect patient-reported outcomes during the study period.” In addition, some cases of PSI use may not have been captured in the registry, and CT imaging may not have been specifically used with preoperative planning software. “Some surgeons may have ordered a CT scan to study the glenoid morphology without enhanced proprietary preoperative planning or PSI,” he said.

Dr. Navarro’s coauthors of “New Preoperative Planning Technologies in 8,117 Elective Shoulder Arthroplasty Procedures: Trends and Outcomes” are Priscilla Hannah Chan, MS; Michael L. Pearl, MD, FAAOS; Heather Ann Prentice, MPH, PhD; and Matthew D. McElvany, MD.

Rebecca Araujo is the managing editor of AAOS Now. She can be reached at