Does “6-Clicks” Day 1 Postoperative Mobility Score Predict Discharge Disposition After Total Hip and Knee Arthroplasties?

Does “6-Clicks” Day 1 Postoperative Mobility Score Predict Discharge Disposition After Total Hip and Knee Arthroplasties?

Mariano E. Menendez, MD, Charles S. Schumacher, MD, David Ring, MD, PhD, Andrew A. Freiberg, MD, Harry E. Rubash, MD, Young-Min Kwon, MD, PhD
The Journal of Arthroplasty. Volume 31, Issue 9, September 2016, Pages 1916–1920

 

Abstract

Background

The use of inpatient rehabilitation services after total joint arthroplasty (TJA) is an important driver of episode-of-care costs. We determined the utility of a new standardized instrument collected during the immediate postoperative period, the Activity Measure for Post-Acute Care (AM-PAC) “6-Clicks” Mobility score, in predicting discharge disposition after TJA and its accuracy in estimating prolonged hospital stay, readmissions, and emergency department (ED) visits.

Methods

Using our institutional database, we retrospectively reviewed 744 patients undergoing primary total hip (40%) or knee (60%) arthroplasty for osteoarthritis during 2014. The AM-PAC Mobility score was prospectively collected by physical therapists within 24 hours of surgery. We constructed 2 multivariable logistic regression models for each study outcome: (1) a base model containing age, sex, Charlson Comorbidity Index, and procedure type and (2) the AM-PAC model including the aforementioned variables and this score. The predictive performance of these models was assessed and compared using the area under the receiver operating characteristic (AUC) curve.

Results

The AM-PAC model provided better prediction of discharge disposition (AUC = 0.777) than the base model (AUC = 0.716; 22% relative improvement). Although the AM-PAC model performed 32% and 27% better than the base model in estimating prolonged hospital stay and ED visits, the model’s predictive performance was poor (prolonged stay: AUC = 0.639; ED visit: AUC = 0.658). The AM-PAC model also showed poor discrimination of readmissions (AUC = 0.657), and there was no relative improvement in predictive performance compared to that of the base model.

Conclusion

The AM-PAC “6-Clicks” Mobility score is a valid, simple tool for predicting discharge disposition after TJA.

Keywords

  • total hip arthroplasty;
  • total knee arthroplasty;
  • discharge;
  • mobility;
  • AM-PAC;
  • resource utilization

With >1 million surgeries performed annually in the United States alone, total hip arthroplasty (THA) and total knee arthroplasty (TKA) are the most common inpatient procedures for Medicare beneficiaries [1]. Optimizing resource utilization in elective total joint arthroplasty (TJA) is increasingly relevant as the demand is projected to surpass 4 million procedures by 2030 [2] and payers shift toward bundled payments to share the financial risk of providing services with hospitals and providers 3, 4 and 5. Postdischarge care is an important target for cost containment as it accounts for nearly half of total episode-of-care payments for TJA [5], with inpatient rehabilitation services being the primary driver 5 and 6. Despite the benefits of direct patient care and sustained physical therapy, it is estimated that >$3 billion is spent annually on inpatient rehabilitation services after TJA 7 and 8. The considerable hospital-level variation in discharge to inpatient rehabilitation suggests that there is a potential opportunity for optimization from both the patient parameters including preoperative optimization and expectations, as well as provider parameters including surgical technique, anesthesia, perioperative care, and communication strategies [6].

Efforts to predict discharge to an inpatient rehabilitation facility after TJA have traditionally relied on preoperative surveys rather than postoperative functional assessments 9, 10 and 11. The Activity Measure for Post-Acute Care (AM-PAC) “6-Clicks” Mobility [12] is a new standardized instrument that allows physical therapists to assess basic mobility in hospitalized patients. The accuracy of the AM-PAC “6-Clicks” Mobility in predicting discharge disposition of medical and surgical inpatients has been recently reported [13]. However, it remains unclear whether it is useful in predicting discharge destination or healthcare resource utilization in patients after TJA. Unlike scores derived from preoperative surveys such as the Risk Assessment and Prediction Tool [11] that include items about disparate domains and susceptible to recall bias (eg, preoperative walking distance, use of gait aids, community supports), the AM-PAC “6-Clicks” Mobility score is collected postoperatively by physical therapists within 24 hours and all of its items are related to basic mobility tasks that every THA and TKA inpatient performs [12]. We determined whether the AM-PAC “6-Clicks” Mobility score is useful in predicting destination of TJA patients after discharge. In addition, we determined its accuracy in estimating prolonged hospital stay, readmissions, and emergency department (ED) visits.

Materials and Methods

Patient Population and Measured Outcomes

This retrospective cohort study was conducted at a tertiary academic hospital in the United States. After institutional review board approval, we initially identified 1100 consecutive adult (aged ≥18 years) patients undergoing either elective primary THA or TKA arthroplasty in 2014, as identified by Current Procedural Terminology codes 27130 and 27447. To achieve a more homogeneous sample, we excluded patients without a primary International Classification of Diseases, 9th Revision, Clinical Modification diagnosis code for hip (715.15, 715.25, 715.35, 715.95) or knee (715.16, 715.26, 715.36, 715.96) osteoarthritis (n = 356) [14]. The final study cohort included 744 patients, of whom 60% (n = 446) underwent TKA and 40% (n = 298) THA. The primary outcome of interest was discharge disposition, which was divided into 2 groups: home (routine discharge) and rehabilitation (nonroutine discharge), with the latter being defined as any inpatient rehabilitation facility or skilled nursing facility [11]. Secondary study outcomes included prolonged hospital stay (>75th percentile or 3 days [15]), 30-day readmission, and 30-day ED visit. Consistent with the Centers for Medicare and Medicaid Services risk adjustment model, we considered data on covariate patient characteristics that might be associated with the study outcomes such as age, sex, and preexisting medical conditions—quantified with the Charlson Comorbidity Index 16 and 17.

Activity Measure for Post-Acute Care “6-Clicks” Basic Mobility Scores

The AM-PAC “6-Clicks” Basic Mobility is a validated instrument to assess basic mobility in patients receiving post–acute care [12]. It is derived from the original AM-PAC instrument, which measures 3 functional domains: basic mobility, daily activities, and applied cognition [18]. The AMPAC “6-Clicks” Mobility includes 6 items related to basic mobility tasks: (1) difficulty turning over in bed, (2) sitting down and standing up from a chair, (3) moving from lying on back to sitting on the side of the bed, (4) moving to and from a bed to a chair, (5) walking in hospital room, and (6) climbing 3-5 steps with a railing (Supplementary Fig. 1). Each item is scored on a 4-point Likert scale from 1 to 4 (overall score range: 6-24), with lower scores indicating a greater degree of limitation [12]. The AM-PAC “6-Clicks” Mobility score was prospectively collected by physical therapists within 24 hours of surgery and entered into the electronic health record as part of the documentation of therapist visits. Physical therapists rated each item by observing patients’ performance [12]. Although we did not directly assess interrater reliability of AM-PAC “6-Clicks” Mobility in the present study, a previous study demonstrated that this instrument has very high interrater reliability (intraclass correlation coefficient = 0.85) among physical therapists [19].

Statistical Analysis

We used independent-samples t tests to examine the association of the AM-PAC “6-Clicks” Mobility score with nonroutine discharge, prolonged hospital stay, readmissions, and ED visits. Multivariable logistic regression modeling was performed to determine whether the AM-PAC “6-Clicks” Mobility score was a useful predictor of our study outcomes. We constructed 2 main regression models: a base model that included only patient characteristics and procedure type and the AM-PAC “6-Clicks” Mobility model that contained the aforementioned variables and this score. The predictive performance of the base model and the AM-PAC “6-Clicks” Mobility model was assessed and compared using the area under the receiver operating characteristic curve [20]. The area under the receiver operating characteristic curve (AUC), also referred to as the C-statistic, is a measure of discrimination that quantifies the ability of the model to accurately predict the value of an observation’s response 21 and 22. Value <0.7 is indicative of poor discrimination, between 0.7 and 0.8 acceptable discrimination, between 0.8 and 0.9 excellent discrimination, and >0.9 outstanding discrimination [23]. In addition to the absolute improvement in predictive performance, we computed the relative improvement in predictive ability of the AM-PAC “6-Clicks” Mobility model to the base model by calculating the difference between the 2 AUCs in percent beyond the level of chance performance (AUC = 0.50) [24].

Results

Measured Outcomes

Our study population included 414 (56%) women and 330 men with a mean age of 66 (range: 23-92) years and Charlson comorbidity score of 1.6 (range: 0-8) who underwent elective joint arthroplasty with standardized postoperative clinical care pathways. The overall rate of nonroutine discharge was 39% (THA: 33%; TKA: 43%). The average length of hospital stay was 2.9 (range: 1-25) days (for THA, 2.8 [1-9] days; for TKA, 3.0 [1-25] days). Nearly one-fifth of patients experienced a prolonged hospital stay (overall: 18%; THA: 17%; TKA: 18%). The overall 30-day hospital readmission (including readmission from outpatient office) and ED visit rates were 7.4% (THA: 4.0%; TKA: 9.6%) and 3.6% (THA: 2.0%; TKA: 4.7%), respectively.

Activity Measure for Post–Acute Care “6-Clicks” Basic Mobility Scores

Lower AM-PAC “6-Clicks” Mobility scores were associated with nonroutine discharge (10.6 vs 12.8 for home discharge; P < .001) and prolonged hospital stay (10.8 vs 12.2 for normal stay; P < .001), however not with readmissions and ED visits ( Table 1). Overall and procedure-specific nonroutine discharge disposition rates increased steadily with decreasing AM-PAC “6-Clicks” Mobility scores (Fig. 1). The AM-PAC “6-Clicks” Mobility model provided better prediction of nonroutine discharge (AUC: 0.777, 95% CI: 0.744-0.811) than the base model (AUC: 0.716, 95% CI: 0.679-0.754; Fig. 2), albeit with slightly overlapping CIs. However, as the CIs overlap by less than half the length of one CI arm, this P value is <.05 [25]. The results were consistent in procedure-specific subgroup analyses, although the score was better in predicting discharge disposition after TKA (AUC AM-PAC model vs base model: 0.771 [95% CI: 0.728-0.814] vs 0.691 [95% CI: 0.642-0.741]) than after THA (AUC AM-PAC model vs base model: 0.776 [95% CI: 0.718-0.833] vs 0.749 [95% CI: 0.689-0.809]). In terms of relative improvement in predictive performance, the AM-PAC model performed 22% better than the base model (Table 2). Although the AM-PAC model was more accurate than the base model in estimating hospital stay and ED visits (relative improvement of 32% and 27%, respectively), the model’s predictive performance was poor (prolonged hospital stay: AUC: 0.639; ED visit: 0.658). The AM-PAC “6-Clicks” Mobility model also showed poor discrimination of readmissions (AUC: 0.657), and there was no relative improvement in predictive performance compared to that of the base model (Table 2).

Discussion

With the health policy focus on shifting financial risk to providers, optimizing resource allocation in elective TJA has become a national priority. The use of post–acute inpatient rehabilitation services is an important driver of episode-of-care payments 5 and 6. We aimed to determine the utility of a new standardized instrument collected within 24 hours of the postoperative period, the AM-PAC ““6-Clicks” Mobility score, in predicting nonroutine discharge after TJA and its accuracy in predicting prolonged hospital stay, readmissions, and ED visits.

We found that patients’ mobility within the first 24 hours of TJA, assessed with the AM-PAC “6-Clicks” score, predicted discharge disposition better than a base model containing age, sex, medical comorbidity, and procedure type. Our findings suggest that preoperative variables alone do not accurately predict discharge disposition and support the notion that patients who do not meet physical therapy milestones during the immediate postoperative period are at greater likelihood of being discharged to an inpatient rehabilitation facility. Prompt identification of these patients might lead to faster discharge to rehabilitation and ultimately result in cost reductions through decreased hospital length of stay [26]. In this context, the Medicare rule that mandates a 3-night hospital stay as a precondition for inpatient rehabilitation coverage may need to be revisited [27]. Although the difference in the absolute improvement in predictive power between the AM-PAC “6-Clicks” and base models (AUC = 0.061) may not appear that large, it has been demonstrated that even slight improvements in the AUC can translate into important reductions in confounding bias [28]. The discriminative ability of the AM-PAC “6-Clicks” score in our study (AUC = 0.777) was lower than that reported by Jette et al [13] (AUC = 0.855), which might be partly explained by differences in patient population. Our analysis was limited to TJA inpatients, whereas the previous study considered all hospitalizations, irrespective of diagnosis [13]. The present study results suggest that early postoperative functional assessments such as AM-PAC “6-Clicks” Mobility score lead to enhanced predictive accuracy of discharge destination after elective joint arthroplasty, with the score being better in predicting discharge disposition after TKA than after THA.

Although the AM-PAC “6-Clicks” Mobility score outperformed the base model in predicting hospital stay and 30-day ED visits, its predictive performance was poor (AUC values <0.70). An AUC value approximating 0.70 is considered acceptable for discrimination and validation of methods for ongoing use [29]; we therefore could not validate the AM-PAC “6-Clicks” Mobility score for predicting either of these outcomes in the TJA setting. Similarly, the observation that the AM-PAC “6-Clicks” Mobility score was unable to predict 30-day readmissions after TJA any better than chance reflects the complexity and multifactoriality of readmission and the difficulty of reducing the risk to a few specific factors 30 and 31. Factors shown to influence readmission risk include patient characteristics (eg, age, race/ethnicity, comorbidity burden, marital status, education, income), index surgical admission complications, and hospital characteristics (eg, teaching status, surgical volume) 31, 32, 33, 34, 35 and 36.

Our analysis should be interpreted in light of its limitations. First, this study was performed at a single large academic center, which may limit generalizability to other practice settings. However, the score is related to basic mobility tasks that every THA and TKA inpatient performs postoperatively. Second, unmeasured patient-level factors such as social and emotional support may confound the association between the AM-PAC “6-Clicks” Mobility score and postoperative resource use 37, 38 and 39, suggesting that although this instrument is a useful research instrument in planning and analyzing care of patients, it alone may not consistently predict patient care decisions. Third, given that we were unable to capture return to care provided in an ED or hospital outside of our system, the incidence of ED visits and readmissions reported herein may be an underestimate. However, the review of the medical records of these patients at the latest follow-up showed that there was no documented additional ED visits or readmissions to outside our hospital system. Fourth, although combining the preoperative survey and the postoperative functional assessment has a potential to further optimize discharge prediction, combined preoperative and postoperative surveys were not performed in our study as preoperative surveys such as the Risk Assessment and Prediction Tool include baseline demographics such as age and sex which are already included in our base models. Finally, owing to the retrospective nature of this study, no a priori power analysis was performed. However, a post hoc power analysis was performed which demonstrated that a sample size of 744 patients provided 98% power to detect a small effect size difference (0.30) in the AM-PAC “6-Clicks” Mobility scores between routinely (n = 455) and nonroutinely (n = 289) discharged patients.

In conclusion, the AM-PAC “6-Clicks” Mobility score is a valid, simple tool collected during the immediate postoperative period for predicting nonroutine discharge in patients after THA and TKA. Additional research is needed to determine whether the AM-PAC “6-Clicks” Mobility score leads to optimizing postdischarge resource allocation.

Appendix A. Supplementary Data

References

    • 2
    • S. Kurtz, K. Ong, E. Lau, et al.
    • Projections of primary and revision hip and knee arthroplasty in the United States from 2005 to 2030
    • J Bone Joint Surg Am, 89 (4) (2007), p. 780
    • 3
    • M. Chernew
    • Bundled payment systems: can they be more successful this time
    • Health Serv Res, 45 (5 Pt 1) (2010), p. 1141
    • 4
    • D.C. Miller, C. Gust, J.B. Dimick, et al.
    • Large variations in Medicare payments for surgery highlight savings potential from bundled payment programs
    • Health Aff (Millwood), 30 (11) (2011), p. 2107
    • 5
    • K.J. Bozic, L. Ward, T.P. Vail, et al.
    • Bundled payments in total joint arthroplasty: targeting opportunities for quality improvement and cost reduction
    • Clin Orthop Relat Res, 472 (1) (2014), p. 188
    • 6
    • P. Cram, B. Ravi, M.S. Vaughan-Sarrazin, et al.
    • What drives variation in episode-of-care payments for primary TKA? An analysis of Medicare administrative data
    • Clin Orthop Relat Res, 473 (11) (2015), p. 3337
    • 7
    • T.R. Mallinson, J. Bateman, H.Y. Tseng, et al.
    • A comparison of discharge functional status after rehabilitation in skilled nursing, home health, and medical rehabilitation settings for patients after lower-extremity joint replacement surgery
    • Arch Phys Med Rehabil, 92 (5) (2011), p. 712
    • 8
    • B. Sharareh, N.B. Le, M.T. Hoang, et al.
    • Factors determining discharge destination for patients undergoing total joint arthroplasty
    • J Arthroplasty, 29 (7) (2014), p. 1355
    • 9
    • W.K. Barsoum, T.G. Murray, A.K. Klika, et al.
    • Predicting patient discharge disposition after total joint arthroplasty in the United States
    • J Arthroplasty, 25 (6) (2010), p. 885
    • 10
    • M. Dauty, X. Schmitt, P. Menu, et al.
    • Using the Risk Assessment and Predictor Tool (RAPT) for patients after total knee replacement surgery
    • Ann Phys Rehabil Med, 55 (1) (2012), p. 4
    • 11
    • V.J. Hansen, K. Gromov, L.M. Lebrun, et al.
    • Does the Risk Assessment and Prediction Tool predict discharge disposition after joint replacement?
    • Clin Orthop Relat Res, 473 (2) (2015), p. 597
    • 12
    • D.U. Jette, M. Stilphen, V.K. Ranganathan, et al.
    • Validity of the AM-PAC “6-Clicks” inpatient daily activity and basic mobility short forms
    • Phys Ther, 94 (3) (2014), p. 379
    • 13
    • D.U. Jette, M. Stilphen, V.K. Ranganathan, et al.
    • AM-PAC “6-Clicks” functional assessment scores predict acute care hospital discharge destination
    • Phys Ther, 94 (9) (2014), p. 1252
    • 14
    • A. Mehrotra, E.M. Sloss, P.S. Hussey, et al.
    • Evaluation of centers of excellence program for knee and hip replacement
    • Med Care, 51 (1) (2013), p. 28
    • 15
    • J.B. Dimick, J.A. Cowan Jr., L.M. Colletti, et al.
    • Hospital teaching status and outcomes of complex surgical procedures in the United States
    • Arch Surg, 139 (2) (2004), p. 137
    • 16
    • R.A. Deyo, D.C. Cherkin, M.A. Ciol
    • Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases
    • J Clin Epidemiol, 45 (6) (1992), p. 613
    • 17
    • H. Quan, V. Sundararajan, P. Halfon, et al.
    • Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data
    • Med Care, 43 (11) (2005), p. 1130
    • 18
    • S.M. Haley, W.J. Coster, P.L. Andres, et al.
    • Activity outcome measurement for postacute care
    • Med Care, 42 (1 Suppl) (2004), p. I49
    • 19
    • D.U. Jette, M. Stilphen, V.K. Ranganathan, et al.
    • Interrater reliability of AM-PAC “6-Clicks” Basic Mobility and Daily Activity Short Forms
    • Phys Ther, 95 (5) (2015), p. 758
    • 20
    • M.E. Menendez, D. Ring
    • A comparison of the Charlson and Elixhauser comorbidity measures to predict inpatient mortality after proximal humerus fracture
    • J Orthop Trauma, 29 (11) (2015), p. 488

 

    • 21
    • M. Greiner, D. Pfeiffer, R.D. Smith
    • Principles and practical application of the receiver-operating characteristic analysis for diagnostic tests
    • Prev Vet Med, 45 (1-2) (2000), p. 23
    • 22
    • A.N. Tosteson, M.C. Weinstein, J. Wittenberg, et al.
    • ROC curve regression analysis: the use of ordinal regression models for diagnostic test assessment
    • Environ Health Perspect, 102 (Suppl 8) (1994), p. 73
    • 23
    • J.J. Gagne, R.J. Glynn, J. Avorn, et al.
    • A combined comorbidity score predicted mortality in elderly patients better than existing scores
    • J Clin Epidemiol, 64 (7) (2011), p. 749
    • 24
    • S. Schneeweiss, P.S. Wang, J. Avorn, et al.
    • Improved comorbidity adjustment for predicting mortality in Medicare populations
    • Health Serv Res, 38 (4) (2003), p. 1103
    • 25
    • G. Cumming
    • Inference by eye: reading the overlap of independent confidence intervals
    • Stat Med, 28 (2) (2009), p. 205
    • 26
    • V.H. Hernandez, A. Ong, Z. Post, et al.
    • Does the Medicare 3-day rule increase length of stay?
    • J Arthroplasty, 30 (9 Suppl) (2015), p. 34
    • 27
    • L.A. Lipsitz
    • The 3-night hospital stay and Medicare coverage for skilled nursing care
    • JAMA, 310 (14) (2013), p. 1441
    • 28
    • S. Schneeweiss, J.D. Seeger, M. Maclure, et al.
    • Performance of comorbidity scores to control for confounding in epidemiologic studies using claims data
    • Am J Epidemiol, 154 (9) (2001), p. 854
    • 29
    • J. Grendar, A.A. Shaheen, R.P. Myers, et al.
    • Predicting in-hospital mortality in patients undergoing complex gastrointestinal surgery: determining the optimal risk adjustment method
    • Arch Surg, 147 (2) (2012), p. 126
    • 30
    • M.G. Hageman, J.K. Bossen, R.M. Smith, et al.
    • Predictors of readmission in orthopaedic trauma surgery
    • J Orthop Trauma, 28 (10) (2014), p. e247
    • 31
    • M.L. Barnett, J. Hsu, J.M. McWilliams
    • Patient characteristics and differences in hospital readmission rates
    • JAMA Intern Med, 175 (11) (2015), p. 1803
    • 32
    • R.E. Mednick, H.M. Alvi, V. Krishnan, et al.
    • Factors affecting readmission rates following primary total hip arthroplasty
    • J Bone Joint Surg Am, 96 (14) (2014), p. 1201
    • 33
    • R.P. Merkow, M.H. Ju, J.W. Chung, et al.
    • Underlying reasons associated with hospital readmission following surgery in the United States
    • JAMA, 313 (5) (2015), p. 483
    • 34
    • N.W. Mesko, K.R. Bachmann, D. Kovacevic, et al.
    • Thirty-day readmission following total hip and knee arthroplasty—a preliminary single institution predictive model
    • J Arthroplasty, 29 (8) (2014), p. 1532
    • 35
    • J.M. Saucedo, G.S. Marecek, T.R. Wanke, et al.
    • Understanding readmission after primary total hip and knee arthroplasty: who’s at risk?
    • J Arthroplasty, 29 (2) (2014), p. 256
    • 36
    • B. Zmistowski, C. Restrepo, J. Hess, et al.
    • Unplanned readmission after total joint arthroplasty: rates, reasons, and risk factors
    • J Bone Joint Surg Am, 95 (20) (2013), p. 1869
    • 37
    • M.D. Neuman, R.M. Werner
    • Marital status and postoperative functional recovery
    • JAMA Surg, 1 (2015)
    • 38
    • K.A. Schwarz
    • Predictors of early hospital readmissions of older adults who are functionally impaired
    • J Gerontol Nurs, 26 (6) (2000), p. 29
    • 39
    • O. Stundner, M. Kirksey, Y.L. Chiu, et al.
    • Demographics and perioperative outcome in patients with depression and anxiety undergoing total joint arthroplasty: a population-based study
    • Psychosomatics, 54 (2) (2013), p. 149

Check Also

Epidemiology and management of atlas fractures

Última actualización 29/09/20 Published: 30 January 2020 Epidemiology and management of atlas fractures Fiedler, N., Spiegl, …