Patient Selection in Outpatient Total Joint Arthroplasty
A recently published study recommends a risk stratification scoring system to determine if patients are appropriate candidates for outpatient surgery. What are the pros and cons of this scoring system?
William J. Long, MD, FRCSC, and Robert Borzio, MD
Meneghini RM, Ziemba-Davis M, Ishmael MK, Kuzma AL, Caccavallo P. Safe Selection of Outpatient Joint Arthroplasty Patients with Medical Risk Stratification: The “Outpatient Arthroplasty Risk Assessment Score.” J Arthroplasty. 2017 Aug;32(8):2325-2331. doi: 10.1016/j.arth.2017.03.004. Epub 2017 Mar 14.
Summary and Clinical Relevance
Traditionally, total joint arthroplasty (TJA) patients were monitored in a hospital setting to assess for acute complications, deliver pain control, and supervise physical therapy. Recently, advances in multidisciplinary care coordination, perioperative protocols, and careful patient selection have introduced the option for outpatient TJA. [1-3]
Surgeons considering this model expect to benefit from the ability to control peri-surgical costs and acquire ownership in ambulatory surgery centers. Bertin  found that the average hospital bill for outpatient total hip arthroplasty (THA) was $4000 less than for an inpatient THA. If episode cost targets and quality metrics are based on regional performance, safely performing outpatient surgery may be the answer. This led Meneghini et al  to develop a risk stratifying system to determine appropriate patient selection for outpatient TJA.
Several studies have identified preoperative risk factors associated with outpatient TJA complications. Sher et al  examined 7474 primary TJAs among 120,847 TJA patients who were discharged within 24 hours after surgery. Using multivariate analysis, they identified the following independent predictors for adverse events or readmission:
- Age older than 80 years (odds ratio [OR] 4.16, P = .001)
- Smoking (OR 1.61, P = .03)
- Bleeding-causing disorders (OR 2.56, P = .01)
- American Society of Anesthesiologists (ASA) class 3 or 4 (OR 1.42, P < .05)
- Severe adverse events (SAE) pre-discharge (OR 13.13, P < .0001).
SAE was defined as death, myocardial infarction, cerebrovascular accident, renal failure, pulmonary embolism, venous thromboembolism, sepsis, septic shock, unplanned intubation, paraplegia, deep wound infection, organ/space infection, and return to the operating room.
Lee et al  retrospectively reviewed 1012 consecutive elective primary hip and knee arthroplasty patients to identify risk factors associated with postoperative complications and determine who should not undergo short-stay arthroplasty. They found that 6.9% of cases required additional physician interventions, 84% occurring more than 24 hours postoperatively. Independent multivariate risk factors for developing late (>24 hours) complications included:
- Chronic obstructive pulmonary disease (adjusted OR 4.16)
- Congestive heart failure (adjusted OR 9.71)
- Coronary artery disease (adjusted OR 2.80)
- Cirrhosis (adjusted OR 8.43)
They concluded that patients with these conditions should not undergo short-stay or outpatient TJA.  Unfortunately, knowledge of specific risk factors cannot quantify overall risk and guide the surgeon to precise patient selection.
Outpatient TJA complications rates continue to be debated in the literature. Lovecchio et al  also found higher complication rates (7.5% vs 5.6%) after matching for outpatients. Bleeding that required transfusion was the most common complication, occurring at similar rates after surgery but at higher rates post-discharge in outpatients. 
Springer et al  examined 30-day hospital readmissions, urgent care or ER visits, and other complications 30 days after outpatient TJA. Their study showed a higher rate of unplanned 30-day hospital readmissions following outpatient TJA (11.7% vs 6.6%), although this was not statistically significant. 
Otero et al  showed an increased complication rate in patients discharged on POD0 versus POD1. They attributed this to improper patient selection.  Berger et al examined a national database of 169,406 outpatient TJA patients and showed fewer complications in outpatient TJA. [1,2]
Despite differing views on complication rates, authors seem to agree on 1 factor: appropriate patient selection is paramount to success in outpatient TJA.
This provided the impetus for Meneghini et al  to develop a risk assessment score. They conducted a retrospective review of 1120 consecutive primary TJAs in an early discharge program. After excluding confounders, they were left with 980 TJAs performed between 2011 and 2016.
From 2011 to 2013, patients were told that the goal discharge date was POD2, reduced to POD1 from 2014 to 2015. From 2015-2016, all patients except those who were identified by OARA score to be same-day surgery candidates (<60 OARA score) continued to be told they would be discharged on POD1.
A high-volume outpatient TJA surgeon with an internal medicine specialist on staff developed the OARA score based on 9 comorbidity areas:
- General medical (180 total possible points)
- Hematologic (325)
- Cardiac (385)
- Endocrine (165)
- Gastrointestinal (185)
- Neurologic/psychological (185)
- Renal/urology (220)
- Pulmonary (250)
- Infectious disease (65)
Meneghini et al  proposed that from a medical risk perspective, an OARA score between 0 and 59 would reflect appropriate safety for early discharge.
Their subjects were also separated by ASA-PS score and Charleston Comorbidity Index (CCI). The ASA-PS score is defined by 6 classes, of which only 4 were included in the sample population:
- Normal healthy patients
- Patients with mild systemic diseases
- Patients with severe systemic diseases
- Patients with severe systemic diseases that are a constant threat to life
ASA-PS classifications were grouped as lower risk (2 or less) and higher risk (3 or more). Romano CCI scores were also grouped as lower risk (0) and higher risk (1 or more). Group proportions were analyzed with the chi-square test and means were compared with student t test.
The study used binary logistic regression to examine the relationship between the OARA score (59 or less vs 60 or more), ASA-PS class (2 or less vs 3 or more), and Romano CCI (0 or less vs 1 or more) as predictors of the same-day or POD1 discharge (yes vs no). They also performed 2 additional multivariate analyses to control for variation in discharge expectations described above by year. Positive predictive values (PPV) and negative predictive values (NPV) were calculated as indicators of each measure’s diagnostic value.
Low OARA scores (59 or less) had a higher percentage of early discharges than low ASA-PS or Romano CCI (81.6%, 56.4%, 70.3%). In binary logistic regression of all cases, patients with low OARA were 2.0 times more likely to be discharged on POD0 or POD1 (95% confidence interval [CI], 1.4-2.8) versus 1.7 times with a low ASA-PS score. The Romano CCI did not demonstrate association (P = 0.301).
When data were stratified to the period 2014 to 2015, with anticipated discharge on POD1, the likelihood ratio increased to 2.5 for low OARA score and stayed the same at 1.7 for ASA-PS (Romano CCI was again not significant). When patients with low OARA score were given the option for same-day discharge in the period 2015 to 2016, the OR increased to 2.7 for OARA and 1.9 for ASA-PS (CCI remained insignificant).
Male patients were 2 times more likely to undergo early discharge (95% CI, 1.5-2.7; P < .001). Meneghini et al  also found that the PPV of lower OARA scores (81.6%) was significantly higher than the PPV of lower ASA-PS (56.4%; P < .001) and Romano CCI (70.3%; P = .002) scores. The NPV of the OARA score was significantly lower than for ASA-PA (P < .001) and Romano CCI (P = .003) scores. All-cause readmission rates were 1.8% for same-day discharge, compared with 2.4% for POD1 discharge and 3.5% for discharge on POD2 or later. Mean OARA scores for POD0, POD1, POD2, POD3, and greater than POD4 were 22.2, 30.7, 45.3, 62.6, and 78.3, respectively. Likewise, the ASA scores were 2.1, 2.5, 2.6, 2.7, and 3.0, and the CCI scores were 0.125, 0.471, 0.608, 0.754, and 0.846.
Meneghini et al  do a very good job of discrediting the ASA and CCI scores as tools for preoperative risk stratification for outpatient TJA. The ASA score lacks the specificity to stratify these patients with clinical relevance, as demonstrated by the fact that the mean ASA score between POD0 and POD3 is only 0.5. The CCI is a sum of weighted scores involving 19 comorbidities with limited application to arthroplasty; it did not show significance when evaluating early discharge. The ASA and CCI have been included in the risk factor assessment throughout arthroplasty literature and, therefore, it is worthwhile to appreciate their shortcomings.
There are several limitations to this study that the authors do not address. First, they describe the likelihood of “early discharge,” which includes POD0 and POD1. This limits the utility of their cutoff value of 59. If they used POD0 alone as their endpoint, a cutoff value for outpatient surgery would be more applicable. When TJA surgeons are considering the utility of this score, the most important aspect is a proper cutoff value with clinical significance. The authors fail to give this.
They describe a low OARA score as less than 60 and then show that patients with this low score are more likely to go home POD0 or POD1; however, they disregard the mean OARA scores. The mean OARA scores for POD2 and POD3 were 45.3 and 62.6, respectively, while the mean OARA score for POD0 was only 22.2. This implies the cutoff value for same-day discharge should be closer to 30. Using receiver operating curves to determine the optimal cutoff value and an endpoint of only POD0 could have given a clinically relevant cutoff value.
Second, the study was not performed in an ambulatory setting. The endpoint of POD0 or POD1 is less useful for surgeons considering this score for procedures done in ambulatory surgery centers. By performing the surgery in a hospital setting, patients with early postoperative complications such as hypotension and uncontrolled pain could easily be converted to an inpatient stay. Although performing a prospective randomized control trial in an ambulatory-only setting would give a clearer picture, the ethical ramifications limit such a trial.
Finally, the complexity of the scoring system was not addressed in the study. The authors describe the 9 comorbidity categories and maximum value for each category, but they do not describe how to calculate the score. The subjectivity involved in each individual category could limit the usefulness of the score. The authors imply that a specialized internal medicine physician is needed to obtain an accurate calculation of this score. Inter-observer reproducibility is important with such low cutoff values (less than 60 points), as the maximum score by adding the possible points in all 9 categories is a staggering 1960 points.
In conclusion, the authors provide ample evidence that ASA-PS and CCI scores are not useful preoperative risk stratification tools in outpatient TJA. They show that the OARA score has the potential to predict early discharge, but they need to give a proper cutoff value for same-day discharge, perform future studies in an ambulatory setting, and explain (and possibly simplify) the calculation of the OARA score.
William J. Long, MD, FRCSC, is clinical associate professor of orthopaedic surgery in the Division of Adult Reconstruction, Department of Orthopaedic Surgery, at NYU Langone Health – Hospital for Joint Diseases, New York, New York. Robert Borzio, MD, is a fellow in the Division of Adult Reconstruction, Department of Orthopaedic Surgery, at NYU Langone Health – Hospital for Joint Diseases, New York, New York.
- Berger RA. A comprehensive approach to outpatient total hip arthroplasty. Am J Orthop (Belle Mead NJ). 2007 Sep; 36(9)(Suppl):4-5.
- Berger RA. Outpatient total knee arthroplasty: pathways and protocols. Tech Knee Surg. 2009;8(2):115-8.
- Aynardi M, Post Z, Ong A, Orozco F, Sukin DC. Outpatient surgery as a means of cost reduction in total hip arthroplasty: a casecontrol study. HSS J. 2014 Oct;10(3):252-5. Epub 2014 Jul 12.
- Bertin KC. Minimally invasive outpatient total hip arthroplasty: a financialanalysis. Clin orthopaedics Relat Res 2005;435:154e63.
- Meneghini RM, Ziemba-Davis M, Ishmael MK, Kuzma AL, Caccavallo P. Safe Selection of Outpatient Joint Arthroplasty Patients with Medical Risk Stratification: the “Outpatient Arthroplasty Risk Assessment Score”. J Arthroplasty. 2017 Aug;32(8):2325-2331. doi: 10.1016/j.arth.2017.03.004. Epub 2017 Mar 14.
- Sher A, Keswani A, Yao DH, Anderson M, Koenig K, Moucha CS. Predictors of Same-Day Discharge in Primary Total Joint Arthroplasty Patients and Risk Factors for Post-Discharge Complications. J Arthroplasty. 2017 Sep;32(9S):S150-S156.e1. doi: 10.1016/j.arth.2016.12.017. Epub 2016 Dec 22.
- Courtney PM, Rozell JC, Melnic CM, Lee GC. Who Should Not Undergo Short Stay Hip and Knee Arthroplasty? Risk Factors Associated With Major Medical Complications Following Primary Total Joint Arthroplasty. J Arthroplasty. 2015 Sep;30(9 Suppl):1-4. doi: 10.1016/j.arth.2015.01.056. Epub 2015 May 27.
- Lovecchio F, Alvi H, Sahota S, Beal M, Manning D. Is Outpatient Arthroplasty as Safe as Fast-Track Inpatient Arthroplasty? A Propensity Score Matched Analysis. J Arthroplasty. 2016 Sep;31(9 Suppl):197-201. doi: 10.1016/j.arth.2016.05.037. Epub 2016 May 27.
- Springer BD, Odum SM, Vegari DN, Mokris JG, Beaver WB Jr. Impact of Inpatient Versus Outpatient Total Joint Arthroplasty on 30-Day Hospital Readmission Rates and Unplanned Episodes of Care. Orthop Clin North Am. 2017 Jan;48(1):15-23. doi: 10.1016/j.ocl.2016.08.002. Epub 2016 Oct 28.
- Otero JE, Gholson JJ, Pugely AJ, Gao Y, Bedard NA, Callaghan JJ. Length of Hospitalization After Joint Arthroplasty: Does Early Discharge Affect Complications and Readmission Rates? J Arthroplasty. 2016 Dec;31(12):2714-2725. doi: 10.1016/j.arth.2016.07.026. Epub 2016 Aug 9.