Presentation
HE20 - The PA Effect: How Physician Assistants Drive Down Length of Stay in Orthopedic Surgery
SessionPoster Session 1
DescriptionBackground:
The rising incidence of osteoarthritis combined with an aging population has led to a substantial increase in hip and knee arthroplasty procedures (Xiao et al., 2022). Given joint arthroplasty procedures are among the most common elective surgeries in the United States, there have been consistent efforts to optimize these surgeries, reduce costs, and improve efficiency (Sloan et al., 2018).
Patient length of stay (LOS), a measure of healthcare efficiency, is associated with patient outcomes, hospital costs, and resource utilization (Rotter et al., 2008). A shorter LOS reduces the risk of hospital-acquired infections and allows for better recovery environments, which may improve patient outcomes (Hassan et al., 2010). Additionally, reducing LOS provides significant cost savings for healthcare systems, particularly for high-volume procedures, by increasing the availability of critical resources, such as beds, equipment, and staff (Vendittoli et al., 2019). Prolonged patient stays may increase the workload for surgical, post-operative care, and nursing teams, making LOS reduction essential for avoiding burnout and maintaining staff well-being. As healthcare systems seek to enhance efficiency for high volume procedures, optimizing patient LOS is essential for improving outcomes for both patients and staff.
One mechanism to assist in optimizing surgeries is to incorporate physician assists in arthroplasty procedures. Physician assistants (PA) have been integrated into various surgical specialties to “assist licensed physicians in patient surgery” and “function in all areas of the peri-operative environment,” such as in the outpatient clinic and post-anesthesia care unit (AASPA, n.d.). In 2015, approximately 9.4% of clinically active PAs in the United States were employed in orthopedics (Chalupa et al., 2016). Although limited research exists on the impact of orthopedic surgical PAs on patient and procedure outcomes, their inclusion on other surgical teams has been associated with greater early discharge rates, reduced resident workload, and improved collaboration of physicians and nurses (Dies et al., 2016; Halvachizadeh et al., 2022). Given the growing demand for joint arthroplasty procedures and efforts to enhance efficiency and reduce LOS, it is prudent to investigate the role of PAs in improving surgical outcomes. With this foundation in mind, the purpose of this talk is to elucidate the impact of leveraging physician assistants during arthroplasty procedures on length of stay.
Methods:
A retrospective chart review was conducted for all conventional and robot-assisted hip or knee arthroplasty procedures performed between December 2021 and May 2023 at a non-profit academic medical center in Southern California. Variables extracted included patient demographics (age, gender, race, ethnicity), health information (comorbidities, ASA classification, BMI, ambulatory status), procedure details (procedure type, time of day), and patient outcomes (LOS, estimated blood loss, complications, discharge disposition, skin-to-skin time), as well as surgical team composition.
Eligible cases included those performed within the study timeframe, with complete data available, involving the patient’s first orthopedic procedure (hip arthroplasty, knee arthroplasty, robot-assisted knee arthroplasty (note robot-assisted hip arthroplasty cases are not performed at the study site)). Cases were excluded if records were flagged as confidential, incomplete, or involved revision surgeries.
Descriptive statistics were calculated for all variables. An ANOVA was conducted to examine the differences in LOS across the three procedure types. Hierarchical regression was used to assess whether comorbidities (nDx), ASA, age at encounter, BMI, and PA presence influenced LOS for patients undergoing orthopedic surgery.
Results:
The study included 1848 procedures (872 Hip Arthroplasty; 522 Knee Arthroplasty; 454 Robot-Assisted Knee Arthroplasty). Mean LOS was 55.95 hrs (SD = 76.28).
A one-way ANOVA was conducted to examine differences in LOS across the three procedure types. LOS was significantly different between procedure groups, F (2, 1848) = 7.172, p=0.05. LOS was greatest for patients who underwent hip arthroplasty (61.78 hrs ± 99.47), followed by knee arthroplasty (55.65 hrs ± 52.89) and robot-assisted knee arthroplasty (45.12 hrs ± 36.06). Tukey post hoc analysis revealed a statistically significant difference (p < 0.05) between hip and robot-assisted knee arthroplasty (6.130, 95% CI (-2.12 to 14.38)), as well as between knee and robot-assisted knee arthroplasty (10.527, 95% CI (0.96-20.10)), p = 0.031.
Regression 1: All Procedure Types
A hierarchical multiple regression was run to determine if the addition of patient factors (nDx, BMI, Age, ASA) and then procedural factors (PA presence) improved the prediction of LOS among patients undergoing hip arthroplasty, knee arthroplasty, and robot-assisted knee arthroplasty). The first model of patient factors to predict LOS was statistically significant, R2 =.152, F (4,1833) = 82.428, p < .05; adjusted R2 = .151. The addition of procedural factors (PA Presence) to the prediction of LOS (model 2) led to a statistically significant increase in R2 of .047, F(1,1832) =91.092, p < 0.05; adjusted R2 =.197, explaining 19.7% of the variance in LOS. Number of diagnoses (nDx) was the strongest predictor, significantly increasing LOS (B = 3.167hr) (𝛽 = .348, 𝑝 < . 05), followed by PA presence, which significantly reduced LOS (B = -35.960hr) (𝛽 = −.217, p<.05).
Regression 2: Hip and Knee Arthroplasty
A second hierarchical multiple regression was run to determine if the addition of patient factors (nDx, BMI, Age, ASA) and then procedural factors (PA presence) improved the prediction of LOS among patients undergoing just hip arthroplasty and knee arthroplasty. The first model of patient factors to predict LOS was statistically significant, R2 =.169, F (4,1379) = 70.213, p < .05; adjusted R2 = .167. The addition of procedural factors (PA Presence) to the prediction of LOS (model 2) led to a statistically significant increase in R2 of .047, F(1,1378) = 76.095, p < 0.05; adjusted R2 =.214 explaining 21.4% of the variance in LOS. Number of diagnoses (nDx) was the strongest predictor, significantly increasing LOS (B =3.490hr) (𝛽 = .355, 𝑝 < . 05), followed by PA presence, which significantly reduced LOS (B = -39.160hr) (𝛽 = −.220, p<.05).
Regression 3: Knee Arthroplasty and Robot-Assisted Knee Arthroplasty
A final hierarchical multiple regression was run to determine if the addition of patient factors (nDx, BMI, Age, ASA) and then procedural factors (PA presence) improved the prediction of LOS among patients undergoing knee arthroplasty and robot-assisted knee arthroplasty. The first model of patient factors to predict LOS was statistically significant, R2 =.130, F (4, 967) = 36.130, p < .05; adjusted R2 = .126. The addition of procedural factors (PA Presence) to the prediction of LOS (model 2) led to a statistically significant increase in R2 of .043, F (2,965) = 33.590, p < 0.05; adjusted R2 =.168 explaining 16.8% of the variance in LOS. Number of diagnoses (nDx) was the strongest predictor, significantly increasing LOS (B =1.890hr) (𝛽 = .305, 𝑝 < . 05), followed by PA presence, which significantly reduced LOS (B = -17.268hr) (𝛽 = −.171, p<.05).
Overall, the presence of a PA on the surgical team significantly reduced LOS across all procedure types, with the greatest impact observed in robot-assisted knee arthroplasty. Across all models, the diagnosis (nDx) variable was consistently the strongest contributor, further emphasizing the importance of case complexity in determining LOS.
Conclusion:
The presence of PAs significantly reduced LOS across all types of hip and knee arthroplasty. These findings highlight the value of incorporating PAs into orthopedic surgical teams. Given the increasing demand for joint arthroplasties, integrating PAs may offer a practical solution for enhancing surgical team efficiency while maintaining high standards of care. Additional research is needed to explore their long-term impact on team dynamics, patient safety, and staff retention, particularly in high-demand specialties.
The rising incidence of osteoarthritis combined with an aging population has led to a substantial increase in hip and knee arthroplasty procedures (Xiao et al., 2022). Given joint arthroplasty procedures are among the most common elective surgeries in the United States, there have been consistent efforts to optimize these surgeries, reduce costs, and improve efficiency (Sloan et al., 2018).
Patient length of stay (LOS), a measure of healthcare efficiency, is associated with patient outcomes, hospital costs, and resource utilization (Rotter et al., 2008). A shorter LOS reduces the risk of hospital-acquired infections and allows for better recovery environments, which may improve patient outcomes (Hassan et al., 2010). Additionally, reducing LOS provides significant cost savings for healthcare systems, particularly for high-volume procedures, by increasing the availability of critical resources, such as beds, equipment, and staff (Vendittoli et al., 2019). Prolonged patient stays may increase the workload for surgical, post-operative care, and nursing teams, making LOS reduction essential for avoiding burnout and maintaining staff well-being. As healthcare systems seek to enhance efficiency for high volume procedures, optimizing patient LOS is essential for improving outcomes for both patients and staff.
One mechanism to assist in optimizing surgeries is to incorporate physician assists in arthroplasty procedures. Physician assistants (PA) have been integrated into various surgical specialties to “assist licensed physicians in patient surgery” and “function in all areas of the peri-operative environment,” such as in the outpatient clinic and post-anesthesia care unit (AASPA, n.d.). In 2015, approximately 9.4% of clinically active PAs in the United States were employed in orthopedics (Chalupa et al., 2016). Although limited research exists on the impact of orthopedic surgical PAs on patient and procedure outcomes, their inclusion on other surgical teams has been associated with greater early discharge rates, reduced resident workload, and improved collaboration of physicians and nurses (Dies et al., 2016; Halvachizadeh et al., 2022). Given the growing demand for joint arthroplasty procedures and efforts to enhance efficiency and reduce LOS, it is prudent to investigate the role of PAs in improving surgical outcomes. With this foundation in mind, the purpose of this talk is to elucidate the impact of leveraging physician assistants during arthroplasty procedures on length of stay.
Methods:
A retrospective chart review was conducted for all conventional and robot-assisted hip or knee arthroplasty procedures performed between December 2021 and May 2023 at a non-profit academic medical center in Southern California. Variables extracted included patient demographics (age, gender, race, ethnicity), health information (comorbidities, ASA classification, BMI, ambulatory status), procedure details (procedure type, time of day), and patient outcomes (LOS, estimated blood loss, complications, discharge disposition, skin-to-skin time), as well as surgical team composition.
Eligible cases included those performed within the study timeframe, with complete data available, involving the patient’s first orthopedic procedure (hip arthroplasty, knee arthroplasty, robot-assisted knee arthroplasty (note robot-assisted hip arthroplasty cases are not performed at the study site)). Cases were excluded if records were flagged as confidential, incomplete, or involved revision surgeries.
Descriptive statistics were calculated for all variables. An ANOVA was conducted to examine the differences in LOS across the three procedure types. Hierarchical regression was used to assess whether comorbidities (nDx), ASA, age at encounter, BMI, and PA presence influenced LOS for patients undergoing orthopedic surgery.
Results:
The study included 1848 procedures (872 Hip Arthroplasty; 522 Knee Arthroplasty; 454 Robot-Assisted Knee Arthroplasty). Mean LOS was 55.95 hrs (SD = 76.28).
A one-way ANOVA was conducted to examine differences in LOS across the three procedure types. LOS was significantly different between procedure groups, F (2, 1848) = 7.172, p=0.05. LOS was greatest for patients who underwent hip arthroplasty (61.78 hrs ± 99.47), followed by knee arthroplasty (55.65 hrs ± 52.89) and robot-assisted knee arthroplasty (45.12 hrs ± 36.06). Tukey post hoc analysis revealed a statistically significant difference (p < 0.05) between hip and robot-assisted knee arthroplasty (6.130, 95% CI (-2.12 to 14.38)), as well as between knee and robot-assisted knee arthroplasty (10.527, 95% CI (0.96-20.10)), p = 0.031.
Regression 1: All Procedure Types
A hierarchical multiple regression was run to determine if the addition of patient factors (nDx, BMI, Age, ASA) and then procedural factors (PA presence) improved the prediction of LOS among patients undergoing hip arthroplasty, knee arthroplasty, and robot-assisted knee arthroplasty). The first model of patient factors to predict LOS was statistically significant, R2 =.152, F (4,1833) = 82.428, p < .05; adjusted R2 = .151. The addition of procedural factors (PA Presence) to the prediction of LOS (model 2) led to a statistically significant increase in R2 of .047, F(1,1832) =91.092, p < 0.05; adjusted R2 =.197, explaining 19.7% of the variance in LOS. Number of diagnoses (nDx) was the strongest predictor, significantly increasing LOS (B = 3.167hr) (𝛽 = .348, 𝑝 < . 05), followed by PA presence, which significantly reduced LOS (B = -35.960hr) (𝛽 = −.217, p<.05).
Regression 2: Hip and Knee Arthroplasty
A second hierarchical multiple regression was run to determine if the addition of patient factors (nDx, BMI, Age, ASA) and then procedural factors (PA presence) improved the prediction of LOS among patients undergoing just hip arthroplasty and knee arthroplasty. The first model of patient factors to predict LOS was statistically significant, R2 =.169, F (4,1379) = 70.213, p < .05; adjusted R2 = .167. The addition of procedural factors (PA Presence) to the prediction of LOS (model 2) led to a statistically significant increase in R2 of .047, F(1,1378) = 76.095, p < 0.05; adjusted R2 =.214 explaining 21.4% of the variance in LOS. Number of diagnoses (nDx) was the strongest predictor, significantly increasing LOS (B =3.490hr) (𝛽 = .355, 𝑝 < . 05), followed by PA presence, which significantly reduced LOS (B = -39.160hr) (𝛽 = −.220, p<.05).
Regression 3: Knee Arthroplasty and Robot-Assisted Knee Arthroplasty
A final hierarchical multiple regression was run to determine if the addition of patient factors (nDx, BMI, Age, ASA) and then procedural factors (PA presence) improved the prediction of LOS among patients undergoing knee arthroplasty and robot-assisted knee arthroplasty. The first model of patient factors to predict LOS was statistically significant, R2 =.130, F (4, 967) = 36.130, p < .05; adjusted R2 = .126. The addition of procedural factors (PA Presence) to the prediction of LOS (model 2) led to a statistically significant increase in R2 of .043, F (2,965) = 33.590, p < 0.05; adjusted R2 =.168 explaining 16.8% of the variance in LOS. Number of diagnoses (nDx) was the strongest predictor, significantly increasing LOS (B =1.890hr) (𝛽 = .305, 𝑝 < . 05), followed by PA presence, which significantly reduced LOS (B = -17.268hr) (𝛽 = −.171, p<.05).
Overall, the presence of a PA on the surgical team significantly reduced LOS across all procedure types, with the greatest impact observed in robot-assisted knee arthroplasty. Across all models, the diagnosis (nDx) variable was consistently the strongest contributor, further emphasizing the importance of case complexity in determining LOS.
Conclusion:
The presence of PAs significantly reduced LOS across all types of hip and knee arthroplasty. These findings highlight the value of incorporating PAs into orthopedic surgical teams. Given the increasing demand for joint arthroplasties, integrating PAs may offer a practical solution for enhancing surgical team efficiency while maintaining high standards of care. Additional research is needed to explore their long-term impact on team dynamics, patient safety, and staff retention, particularly in high-demand specialties.
Authors
Event Type
Poster Presentation
TimeMonday, March 314:45pm - 6:15pm EDT
LocationFrontenac Foyer
Digital Health (DH)
Simulation and Education (SE)
Hospital Environments (HE)
Medical and Drug Delivery Devices (MDD)
Patient Safety and Research Initiatives (PS)