Association of AM-PAC “6-Clicks” Basic Mobility and Daily Activity Scores With Discharge Destination (2024)

Search for other works by this author on:

Article history

Abstract

Introduction

Implementation of health care cost containment measures, such as the Inpatient Prospective Payment System1 and Hospital Readmissions Reduction Program,2 has forced inpatient facilities to optimize care and develop strategies to reduce overall length of stay (LOS). Minimizing hospitalization days can lower the risk of developing health care–associated infections3 and functional decline as well as reduce incremental inpatient cost.4 In 2016, there were 35.7 million hospitalizations with a mean LOS of 4.6days, with an average cost of $11,700.5 Health care administrators face increasing pressure to lower costs and increase efficiency of care while maintaining quality and patient safety. LOS is often utilized as a barometer to measure a health care facility’s efficiency.6 It is essential for hospitals, both in regard to patient safety and the institutions’ long-term financial performance, to be adept at promptly identifying the most appropriate next level of care for a patient to prepare for timely discharge, reduce unnecessary post-acute care, and mitigate costly readmissions.

As experts in identifying functional impairments, participating in discharge planning is one of the most important contributions that physical therapists and occupational therapists (OT) can make in an acute care hospital.7 Discharge planning is a complex process that requires coordinated communication between the patient, their family, and all health care providers.7 Physical therapists are experts at discharge planning who rely on both clinical judgment and objective measures of functional status to inform the interdisciplinary team of their recommendations. Systematic use of a reliable and validated functional outcome measure facilitates efficient transition to the most appropriate next level of care8 and optimizes hospital LOS.

Of the several objective measures of functional status for use in the hospital, the Activity Measure for Post-Acute Care (AM-PAC)9 Short Form 6-Clicks measure10 offers several advantages. The construct aligns with post-acute measures to provide a seamless transition of care from acute to post-acute rehabilitation settings, the theoretical basis of 6-Clicks is strong, the applicability covers a wide range of diagnoses and conditions, and the tool can be easily and efficiently completed at every physical therapist and/or OT visit to track progress. 6-Clicks measures include 2 domains, basic mobility and daily activity, and focus on the difficulty patients have with 6 functional tasks in each domain.

Previous research has investigated how well the 6-Clicks scores predict discharge disposition.10–13 Very good to good prediction was reported,10 but methodological limitations caution adoption of the 6-Clicks measures for discharge prediction. In 1 study,10 approximately 20% of the patients had missing discharge disposition information. Missing discharge destination values were imputed from physical therapist or OT recommendations for discharge, allowing the potential for misclassification bias. A recent study assessing the predictive ability in 8 hospitals did not include the Daily Activity 6-Clicks and only assessed the Basic Mobility.12 Two studies were limited to specific diagnoses, including patients with total joint arthroplasties11 and patients with stroke,13 limiting the generalizability to the breadth of patients seen by physical therapists and OTs in the acute hospital.

The use of 6-Clicks scores to help rehabilitation professionals guide discharge planning is promising on the basis of previous studies; however, there remains insufficient knowledge on predictive properties across a setting that warrants replication to address methodological limitations. Therefore, the purpose of this study was to assess the predictive ability of AM-PAC 6-Clicks scores at physical therapist and/or OT initial evaluation for community versus institution discharge in patients who received inpatient physical therapy and/or occupational therapy. Because prediction focuses on a dichotomous outcome (community vs institution), the use of 6-Clicks scores for more clinically relevant discharge planning is also of benefit. Therefore, an additional purpose of this study was to examine the association of AM-PAC 6-Clicks scores at initial physical therapist and/or OT evaluation with discharge destination, categorized in 4 levels: home/self-care (HOME), home with home health (HHA), skilled nursing facility (SNF), and inpatient rehabilitation facility (IRF).

Methods

Patient Population and Data Source

This retrospective cohort study was conducted using electronic health records (EHR) from a tertiary care academic hospital in the southwest United States. The 268-bed hospital specializes in treating complicated and rare disorders, including difficult-to-treat cancers, organ transplants, cardiovascular diseases, and neurodegenerative disorders.

After institutional review board approval by Mayo Clinic and Northern Arizona University, data were extracted from the EHR for all hospital admissions between October 1, 2015, and August 31, 2018, who received inpatient physical therapy and/or OT (n=18,624; Fig. 1). These 18,624 admissions represented 13,368 unique patients. Admissions of patients who were discharged alive (n=183 died during hospital stay) from the acute care hospital were included. Patients who were discharged against medical advice, to hospice, or to long-term care were excluded (n=895), leaving 17,456 admissions representing 12,816 patients in the final cohort (Fig. 1). Of these patients, 98.7% had 6-Clicks Basic Mobility scores recorded for the initial physical therapist or OT evaluation, and 86.4% had 6-Clicks Daily Activity scores on the initial physical therapist or OT evaluation.

Outcome

The primary outcome was the discharge destination obtained from the EHR. For the first purpose of assessing the ability of the 6-Clicks scores to predict discharge disposition, patient destinations were dichotomized to community (with and without HHA) and institution (SNF and IRF). The outcome was also restricted to those discharged home dichotomized to HOME and HHA. For the second purpose of examining the association of 6-Clicks scores with discharge destination, a 4-level nominal variable was used for the outcome: HOME, HHA, SNF, and IRF.

6-Clicks Basic Mobility and Daily Activity Measures

The independent variables for this study were the 6-Clicks Basic Mobility and Daily Activity scores completed by the physical therapist or OT at the initial evaluation. The 6-Clicks measures are reliable14 and validated15 short forms based on the AM-PAC instruments.9 The 6-Clicks measures are standardized functional assessment instruments for use in acute care hospitals that measure 3 domains: basic mobility, daily activity, and applied cognition.10 For this study, only basic mobility and daily activity scores were used. The Basic Mobility domain (version 1) includes 6 items scored on a 1-to-4 scale, with 3 measures based on level of difficulty completing the task, and 3 items based on level of assistance required to perform the task. The Daily Activity domain includes 6 items scored on a 1-to-4 scale based on the level of difficulty for the patient to complete the task. The scores for the 6 items are summed from 6 to 24, with lower scores corresponding to lower functional levels. Because the 6-Clicks measures were derived from the AM-PAC, which includes an item bank, standardized t scores (mean=50 and SD=10) can be generated to compare the patient’s activity limitation.

Statistical Analysis

Descriptive statistics were calculated as means and standard errors for continuous variables and percentages for categorical variables. Significance of demographic and clinical variables by discharge destination was assessed with repeated-measures ANOVA for continuous variables and generalized estimating equations for categorical variables.

Statistical analyses for both purposes of this study analyzed the Basic Mobility and Daily Activity 6-Clicks scores separately. On July 31, 2017, the hospital updated the tool’s standardized intake script from version 1 to version 2 of the 6-Clicks Basic Mobility form. Although the changes in the measure were minor, to maintain consistency and allow comparison with previously published literature,10 only version 1 6-Clicks Basic Mobility scores were included (October 1, 2015–July 30, 2017). This decreased the sample size to 10,380 for Basic Mobility (Fig. 1).

To assess the ability of 6-Clicks scores at initial the physical therapist or OT evaluation to predict discharge, receiver operator characteristic (ROC) curves assessed several cut-points to determine the best Basic Mobility and Daily Activity scores for discharge destination dichotomized as community (with or without HHA) versus institution (SNF and IRF) as well as HOME versus HHA as an exploratory analysis. Cut-points included the t scores corresponding to raw scores of 6 to 24 as well as the cut-point previously published (42.9 for Basic Mobility and 39.4 for Daily Activity). ROC curves plot the sensitivity and 1-specificity for each cut-point to identify the point on the curve that maximizes true positives and negatives. From the ROC curve, the area under the curve (AUC) is calculated to represent the overall accuracy of the test,16 with values closer to 1.0 representing high sensitivity and specificity and values closer to 0.5 representing low sensitivity and specificity.

For the second purpose of this paper to examine the association between 6-Clicks Basic Mobility and Daily Activity scores and the discharge destination, a 4-level outcome (HOME, HHA, SNF, and IRF) was used. The independent variable was the dichotomous cut-point score from the first objective. Two random coefficients multinomial logistic regression models were used to assess the association between discharge destination and AMPAC 6-Clicks Basic Mobility and Daily Activity scores. The random coefficient included an intercept of patient to allow for cluster correlated data (ie, more than 1 admission per patient). Adjusted odds ratio (OR) and 95% CIs were calculated. Marital status was assessed as an effect modifier. Potential confounders were age, sex, race, marital status, insurance, Elixhauser co-morbidity score,17 and hospital LOS. Confounders were included in the final model if there was a significant association with the exposure and outcome and the effect size changed more than 10%.18 These models were adjusted for confounders of age, sex, marital status, insurance, Elixhauser co-morbidity score, and hospital LOS.

All analyses were completed using SAS (version 9.4, SAS Institute, Inc., Cary, NC). Alpha was set to .05 for statistical significance testing.

Results

The final cohort included 17,546 admissions, representing 12,816 unique patients. Most of the admissions were discharged to community (73.5%), either without (54.0%) or with (19.5%) HHA (Tab. 1). The remaining admissions (26.5%) were discharged to institutions, with most (19.3%) going to SNF and a smaller percentage (7.2%) going to IRF. Patients who were discharged to institutions were older, especially those going to a SNF. A higher proportion of males were discharged to community, either with or without HHA, and a higher proportion of females were discharged to an SNF. Racial minorities had a lower proportion being discharged to institutions and a higher proportion with HOME discharge compared with white patients. A higher proportion of patients who were married or who had a partner were discharged to community. Marital status was not a significant effect modifier for Basic Mobility (P=.13) or Daily Activity (P=.16). Admission diagnosis was varied, with only the top 15 most frequent in Table 1, including total joint arthroplasty (14.1%) and heart transplant (1.1%). Most of the admissions with total joint arthroplasties (79.3%) did not go to an extended care facility (ECF), with the majority (76.8%) having HOME discharge. Admissions with other orthopedic DRG codes had a higher proportion going to ECF, especially SNF discharge. Admissions for transplants have HOME discharge, but those with cardiac, neurological, and medical DRGs had a higher proportion with institutional discharge. Physical and/or occupational therapy was initiated early in the hospital stay. The median hospital LOS was on average 4days (interquartile range [IQR]=2–7), and initial physical therapist and/or OT evaluation was completed on average 1.3days (IQR=0.80–2.3days) after hospital admission.

Prediction of community versus institution discharge was assessed with ROC curves and AUC for version 1 of 6-Clicks. The ROC curve showed 40.78 as the t score that maximized sensitivity (0.71) and specificity (0.74) for the 6-Clicks Basic Mobility, with an AUC of 0.80 (95% CI = 0.80–0.81; Fig. 2). The t score of 40.78 corresponds to a raw score of 16. The data from version 2 (n=4649) were also analyzed, and the same raw score (16) was found to maximize sensitivity and specificity (0.74 and 0.71, respectively) with good prediction (AUC=0.80 [95% CI = 0.79–0.81]). For the outcome that compared HOME versus HHA, the same cut-point had similar sensitivity (0.76) but very low specificity (0.38) and AUC (AUC=0.59 [95% CI = 0.58–0.60]). For 6-Clicks Daily Activity, the ROC curve showed the point that maximized sensitivity (0.68) and specificity (0.81), with an AUC of 0.81 (95% CI = 0.80–0.82; Fig. 3). The t score of 40.22 corresponds to a raw score of 19. When comparing HOME with HHA, a different cut-point had the best sensitivity (0.65) and specificity (0.54) but still low and low prediction overall (AUC=0.63 [95% CI = 0.62–0.65]).

The t scores cut-points from the ROC curves (40.78 for Basic Mobility and 40.22 for Daily Activity) were used to assess the association of 6-Clicks scores with discharge destination using a 4-level nominal variable for discharge. After adjusting for age, sex, marital status, insurance, Elixhauser co-morbidity score (categorical), and hospital LOS (median split), there was a significant association between 6-Clicks Basic Mobility and Daily Activity scores with discharge destination (Tab. 2). Patients with 6-Clicks Basic Mobility and Daily Activity scores below the cut-point had higher odds of discharge to HHA, SNF, and IRF compared with patients with scores above the cut-point. There was a monotonic increase in ORs as IRF>SNF>HHA>HOME with 6-Clicks Daily Activity. With the Basic Mobility 6-Clicks measure, there were also significant differences between HOME and HHA, and both of these with SNF and IRF, but SNF and IRF were not significantly different. With Daily Activity 6-Clicks measure, all destinations were significantly different.

Discussion

The purpose of this study was to better understand the relationship between 6-Clicks scores completed at the physical therapist and/or OT initial evaluation and discharge destination from the hospital. Good prediction was found with community versus institution discharge for both Basic Mobility and Daily Activity 6-Clicks scores and cut-points with moderately good sensitivity and specificity. These cut-points from the physical therapist and/or OT initial evaluation were significantly associated with clinically relevant discharge destinations, which remained after adjusting for demographic and clinical confounders.

Discharge destination from the hospital is a complex decision-making process affected by many variables, including demographic characteristics and clinical factors. Early identification of discharge destination, especially for institutions, can improve efficiencies as delays from hospital discharge can be due to lack of available beds in SNFs and IRFs.8 Functional status is an important predictor of discharge destination,8 but the lack of a universal and consistent measure historically made it difficult to identify key functional status items to be used in the hospital to help plan the most appropriate discharge destination.19

Good prediction was found for the 6-Clicks Basic Mobility and Daily Activity measures at physical therapist and/or OT initial evaluation and community versus institution discharge, but the same good prediction when assessing a subsample of just those patients who were discharged to the community (HOME vs HHA). The good prediction of community versus institution discharge is in agreement with others who reported prediction of 6-Clicks measures for discharge destination.10^,11 In a study of 62,540 patients in 1 health care system, the AUC for 6-Clicks scores from the first physical therapist or OT visit for Basic Mobility was 0.82 (95% CI = 0.82–0.83) and for Daily Activity was 0.73 (95% CI = 0.72–0.73).10 More recently, the 6-Clicks Basic Mobility score at 48hours after hospital admissions was better at prediction discharge destination (AUC=0.78) compared with demographic (e.g., age, zip code, insurance) and clinical (eg, vital signs, laboratory data) measures (AUC=0.66) in 19,963 admissions to a medical service in 8 different hospitals.12 The lack of prediction when comparing HOME with HHA requires future research, but it may be that the 6-Clicks cannot differentiate between those 2 discharge destinations. When specific patient groups were studied, the 6-Clicks Basic Mobility administered within 24hours of surgery was able to accurately classify community versus institution discharge (AUC=0.78 [95% CI = 0.74–0.81]) in 744 patients after total joint arthroplasty, and the inclusion of 6-Clicks Basic Mobility score improved the prediction over a model with patient preoperative characteristics and type of arthroplasty (hip and knee).11 Finally, the 6-Clicks Basic Mobility score predicted discharge destination after stroke, and adding demographics and the National Institutes of Health Stroke Score did not appreciably improve prediction over 6-Clicks Basic Mobility alone.13 The current study adds to the literature that both the 6-Clicks Basic Mobility and Daily Activity scores calculated early (1.3days [IQR=0.80–2.3days]) on a hospital stay can predict community versus institution discharge in a diverse group of patients.

The cut-points identified in this study differ from previously published work from Jette etal.10 Although similar, the t score as a cut-point for community versus institution discharge was 40.78 (sensitivity=0.71 and specificity=0.74) compared with 42.9 (sensitivity=0.82 and specificity=0.72) in Jette etal10 for Basic Mobility and 40.22 (sensitivity=0.68 and specificity=0.81) compared with 39.4 (sensitivity=0.73 and specificity=0.80) for Daily Activity. Like the current study, Jette etal10 assessed the initial physical therapist or OT 6 Clicks scores and categorized discharge to community versus institution. The study by Jette etal10 was limited by the absence of discharge destinations in approximately 20% of the records. In the EHR, the discharge destination data field was left blank if the patient was discharged home with self-care. However, a null value could also signify a missing discharge destination. In the case of absent values, discharge destination was imputed from the physical therapist or OT recommendations, as this was followed for other patients 83% of the time. That imputation may have led to misclassification for those patients who were not discharged to a destination where the physical therapist or OT recommended or if there were differences between physical therapist and OT recommendation. The differences found between the current study and Jette etal10 may also be due to variability in the percentage of admissions discharged to institutions (26.5% vs 56.8%, respectively) and hospital LOS (median [IQR]=4.0 [5] days vs 4.7 [9.6] days, respectively). Pfoh etal12 used a cut-point of 12 and reported 0.94 sensitivity but only 0.27 specificity for community versus institution discharge. The authors of this study did not specify which 6-Clicks version was used, so the t score is not known. Patient population variability and care delivery models may also account for the lack of a consistent cut-point.

The derived cut-points for Basic Mobility and Daily Activity measures were significantly associated with more clinically relevant discharge destination than community versus institutions. This is in agreement with other studies that have reported function strongly related to discharge.8^,13^,20 Higher 6-Clicks Basic Mobility scores had significantly higher odds of discharge to SNF and IRF versus home.13 Louis Simonet etal8 reported dependency with showering and transferring at hospital day 3 increased the odds of discharge to institutions by 2.8 (95% CI = 1.3–6.0) and 2.8 (95% CI = 1.3–6.2), respectively. The studies by Covert13 and Louis Simonet,8 however, did not differentiate between home with and without services. Additionally, Covert etal13 did not report when the 6-Clicks Basic Mobility score was calculated. Using an activities of daily living (ADL) and instrumental ADL scoring system from the self-reported Medicare Current Beneficiary Survey, with stage 0 representing no activity limitations and stage IV representing complete activity limitations,21 Na etal20 reported a monotonic dose–response relationship between increasing activity limitations in ADL and instrumental ADL and odds of discharge to institutions and home with services compared with home discharge. Discharge to all institutions was considered together rather than considering discharge to SNF and IRF separately. The current study adds to these previous publications by considering destinations with post-acute rehabilitation individually, providing additional data for clinical decision-making and post-acute transition planning.

Limitations

This study is not without limitations. First, discharge planning can be affected by many factors, including cognition,22^,23 social determinants of health,24 and market-level factors.25 These factors were not part of the EHR in the current study, and planning discharge may not allow the reliance on a single instrument. Second, on July 31, 2017, the hospital changed from version 1 to version 2 of the 6-Clicks Basic Mobility; this version update resulted in a slight modification to the questions and t scores, although the raw scores remained the same. To maintain consistency of questions and t score as well as allow comparison with previously published literature, only version 1 data were analyzed, resulting in a decreased sample size from 12,624 to 7975. The results of version 2 analyses were substantially similar to that of version 1. An additional limitation is the fact that the reliability of the 6-Clicks short forms was not established in the physical therapist and OT who completed the forms for this study; however, good interrater reliability has been reported for the 6-Clicks Basic Mobility and Daily Activity instruments.14 Completion of 6-Clicks does require clinical judgment, which may be more variable in different samples of rehabilitation practitioners. The inclusion of Elixhauser co-morbidity score17 allows for adjustment by case mix to minimize confounding bias. However, as a tertiary care hospital, all ICD codes related to a comprehensive list of each of the patient’s comorbidities may not be captured, especially those conditions not directly related to the current hospitalization. Finally, this study was completed in 1 academic medical center that focuses on medically complex patient populations, and it may lack generalizability to other hospitals. Differences were reported compared with other studies performed at other academic medical centers; however, the overall conclusions pertaining to the utility of 6-Clicks Basic Mobility and Daily Activity short forms for discharge were consistent with prior studies.

In conclusion, this study builds the evidence base for the clinical usefulness of the 6-Clicks Basic Mobility and Daily Activity measures in the acute care hospital and works to address therapists’ perceptions of the 6-Clicks measures identified by Dewhirst etal.26 They reported a lack of clarity for the use of 6-Clicks for clinical decision-making, including discharge planning. AM-PAC 6-Clicks Basic Mobility and Daily Activity scores early in the hospital stay can be used to assist clinicians to predict a community versus institution discharge as well as specific discharge destinations. Identifying the most accurate cut-points may require consideration of individual hospital care delivery models or processes within subdivisions within a hospital as well as hospital or patient characteristics.

Author Contributions

Concept/idea/research design: M. Warren, J. Knecht, J. Verheijde, J. Tompkins

Writing: M. Warren, J. Knecht, J. Verheijde, J. Tompkins

Data collection: J. Knecht, J. Tompkins

Data analysis: M. Warren, J. Tompkins

Project management: M. Warren, J. Tompkins

Providing facilities/equipment: J. Knecht

Providing institutional liaisons: J. Tompkins

Clerical/secretarial support: J. Tompkins

Consultation (including review of manuscript before submitting): J. Knecht, J. Tompkins

Acknowledgments

At the time this manuscript was written, M.W. was affiliated with the Department of Physical Therapy and Athletic Training, Northern Arizona University, Flagstaff, Arizona, USA.

Funding

There are no funders to report for this study.

Ethics Approval

This study received institutional review board approval by Mayo Clinic and Northern Arizona University.

Disclosures

The authors completed the ICMJE Form for Disclosure of Potential Conflicts of Interest. They reported no conflicts of interest.

References

© The Author(s) 2021. Published by Oxford University Press on behalf of the American Physical Therapy Association. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com

Comments

0 Comments

I agree to the terms and conditions. You must accept the terms and conditions.

Email alerts

Citing articles via

More from Oxford Academic