Lower AM-PAC 6-Clicks Basic Mobility Score Predicts Discharge to a Postacute Care Facility Among Patients in Cardiac Intensive Care Units

Abstract

Objective

The objective of this study was to determine the ability of the Activity Measure for Post-Acute Care “6-Clicks” Basic Mobility Short Form to predict patient discharge destination (home vs postacute care [PAC] facility) from the cardiac intensive care unit (ICU), including patients from the cardiothoracic surgical ICU and coronary care unit.

Methods

This retrospective cohort study utilized electronic medical records of patients in cardiac ICU (n = 359) in an academic teaching hospital in the southeastern region of United States from September 1, 2017, through August 31, 2018.

Results

The median interquartile range age of the sample was 68 years (75–60), 55% were men, the median interquartile range 6-Clicks score was 16 (20–12) at the physical therapist evaluation, and 79% of the patients were discharged to home. Higher score on 6-Clicks indicates improved function. A prediction model was constructed based on a machine learning approach using a classification tree. The classification tree was constructed and evaluated by dividing the sample into a train-test split using the Leave-One-Out cross-validation approach. The classification tree split the data into 4 distinct groups along with their predicted outcomes. Patients with a 6-Clicks score >15.5 and a score between 11.5 and 15.5 with primary insurance other than Medicare were discharged to home. Patients with a 6-Clicks score between 11.5 and 15.5 with Medicare insurance and those with a score ≤11.5 were discharged to a PAC facility.

Conclusion

Patients with lower 6-Clicks scores were more likely to be discharged to a PAC facility. Patients without Medicare insurance had to be significantly lower functioning, as indicated by lower 6-Clicks scores for PAC facility placement than those with Medicare insurance.

Impact

The ability of 6-Clicks along with primary insurance to determine discharge destination allows for early discharge planning from cardiac ICUs.

Introduction

Cardiovascular disease is the leading cause of death and disability across the globe and in the United States.^1–3 Its pathophysiology can lead to poor aerobic capacity, decreased strength, and poor exercise tolerance resulting in decreased function.⁴ Cardiovascular disease can be treated conservatively or surgically.⁵ After cardiac surgery (such as coronary artery bypass grafts or valve replacements), critically ill cardiovascular patients are treated in a cardiothoracic surgery intensive care unit (CTS-ICU) and medically treated patients with cardiac conditions are usually managed in the coronary care unit (CCU).⁶ Partly due to advanced surgical, anesthesiologic, and critical care, cardiovascular patients with an increased number of comorbidities, advanced age, frailty, and psychosocial issues are now considered surgical candidates.⁷ Besides surgical or medical treatments, patients are stabilized in cardiac ICU with life-saving interventions such as initiation and titration of vasopressor support, antiarrhythmic agents, ventilatory support, or invasive support (e.g., intra-aortic balloon pump).^5,8,9 Patients who require cardiac ICU treatment due to critical illness from cardiovascular disease are at high risk for complications due to medical instability, prolonged ventilation, and sedation,¹⁰ leading to increased morbidity and mortality rates¹¹ and poor functional outcomes.⁵

Functional decline after CCU or CTS-ICU admission can be a result of muscle weakness, decreased aerobic capacity, and neuropsychological insult that occur following medical and surgical interventions.⁵ Prolonged bed rest following cardiac ICU admission increases the risk of these complications.⁶ Early mobility and rehabilitation in CCU and CTS-ICU can prevent decline in functional status¹² and cognitive deficits,¹³ decrease length of stay (LOS),⁶ and increase likelihood of discharge home.¹⁴ Although early mobility in cardiac ICU has been demonstrated to be safe and feasible,⁶ barriers to early mobility and rehabilitation include patient safety concerns due to critical illness, communication, and staffing.¹⁵ Depending on the severity of illness and resultant functional deficits, patients may require additional rehabilitation after discharge from the acute care hospital to improve function.¹⁴

Discharge planning is a process of transitioning patient care from one setting to the next based on medical and functional considerations. Discharge planning involves an interdisciplinary care team along with the patient and their family.¹² Anticipating discharge needs while a patient is in a cardiac ICU is challenging due to rapid changes in condition, high acuity, and uncertain recovery pathways.¹⁶

Postacute care (PAC) facility–based discharge is considered if the patient’s medical and/or rehabilitation needs require continued skilled care that cannot be safely received in the patient’s home environment.¹⁷ Patients with complex cardiac and co-morbid conditions, cardiac surgery, and older age are more likely to need PAC facility–based discharge after a cardiac ICU stay.^18,19 In the United States, types of facilities include skilled nursing facilities (SNF), inpatient rehabilitation facilities (IRF), or long-term acute care hospital (LTAC).¹⁷ LTACs offer inpatient medical care for patients with long-term complex medical needs.²⁰ Patients discharged to LTAC require prolonged recovery from acute illness (eg, patients with tracheostomy requiring extended ventilator weaning),²⁰ and rehabilitation services are provided if the patient can safely participate.²¹ IRF and SNF are the most common types of PAC facility placement.¹⁷ The IRF plan of care standard is 3 h/d of individual therapy, at least 5 d/wk of at least 2 rehabilitation disciplines.²² SNF placement allows for a longer LOS and less intensive therapy timelines with therapy minutes titrated to patient tolerance.¹⁷ Considerations for location of discharge include the patient’s physical abilities, severity of illness, insurance provider, caregiver support, the amount of therapy a patient can tolerate per day, and availability of beds in the facility.²³ Completion of rehabilitation and medical plan of care at the facility result in the patient being discharged to the next appropriate level of care such as return to home or, if needed, transition to another facility or nursing home.¹⁷ Early prediction of discharge location and involving physical therapists in discharge planning in the cardiac ICU may reduce ICU and hospital LOS, readmission rates, health care utilization costs, and facilitate home discharge.^7,24

The Activity Measure for Post-Acute Care (AM-PAC) Inpatient Basic Mobility Short Form (“6-Clicks”) was developed, validated, and shown to predict hospital discharge destination when performed on physical therapist evaluation in the acute care hospital.^25,26 Although outcome measures are often underutilized in acute care practice due to barriers such as lack of time or lack of knowledge of an appropriate tool, 6-Clicks provides a fast, easy to use, and easy to interpret measure.^27,28 6-Clicks assesses 6 basic mobility tasks scored out of 24 (higher scores indicate higher function)²⁶ and provides a wide range of scoring to allow for patients with a very low level of performance and very high level of performance to both be evaluated by the same tool, essential in ICU practice.²⁵ To our knowledge, the utility of 6-Clicks for discharge planning from cardiac ICUs has not yet been examined. Patients who have cardiovascular disease requiring ICU admission with conservative or surgical intervention are at high risk for medical and functional decline.^18,19 Patients recovering from cardiovascular events and cardiac surgery benefit from a structured rehabilitation program of physical therapy and physical activity, lifestyle modification advice, and psychological health interventions reducing cardiovascular mortality, preventing hospitalization, and improving quality of life.²⁹ Although previous studies have shown the utility of the 6-Clicks to predict discharge destination for cardiac patients as part of the cohort, these patients were not investigated individually or in the ICU.²⁶ Early discharge planning is crucial in cardiovascular patients due to their likelihood of needing additional in-facility rehabilitation after ICU admission.¹⁸ The aim of the study was to determine the ability of 6-Clicks to predict discharge destination (home vs PAC facility) when a physical therapist evaluation was performed at an academic hospital’s CTS-ICU and CCU. We hypothesized that a lower score on 6-Clicks (lower physical function) would predict discharge to a PAC facility.

Methods

Data Source

This study was a retrospective secondary analysis of the Cerner PowerChart (Kansas City, MO, USA) electronic medical record data from a large academic teaching hospital. The study’s protocol was approved by Emory University’s Institutional Review Board (#00114595). The hospital’s Clinical Data Warehouse (CDW) provided data on age, sex, date, and time of admission to and discharge from hospital and ICU, discharge location, insurance, physical therapist evaluation date and time, dates of physical therapist treatments in ICU, and floor (where the patient is still being managed at the acute hospital but outside of the ICU). The CDW also provided the Acute Physiologic Assessment and Chronic Health Evaluation (APACHE) IV score, an estimate of ICU mortality utilizing physiologic variables, age, and underlying health.³⁰ The APACHE IV score is performed within the first 24 hours of ICU admission and is the only general ICU scoring system developed and validated for patients after cardiac surgery.^30,31 Additional data were extracted using manual chart review that included the 6-Clicks score from the physical therapist evaluation in the cardiac ICU, patient’s living situation, number of missed physical therapist treatments, and confirmation of discharge destination. Medical stability at the first physical therapist evaluation attempt was extracted through documentation made by the physical therapist attempting to evaluate the patient; if the patient was not stable, evaluation was deferred at least 24 hours until the patient was safely able to be evaluated. If the patient was unable to be evaluated due to poor medical status and therefore not given a 6-Clicks score, they were not included. All manually extracted data were quality checked by a team member for accuracy. The number of days from ICU admission to physical therapist consult, number of days from ICU admission to physical therapist evaluation, number of days from physical therapist consult to physical therapist evaluation, ICU LOS, hospital LOS, and total number of physical therapist treatments in ICU and hospital were calculated from the extracted data. Demographic data of patients admitted to CCU or CTS-ICU but not evaluated by a physical therapist during their ICU admission were provided by the CDW for comparison with those who were evaluated by a physical therapist. These data included age, sex, date and time of admission to and discharge from hospital and ICU, and APACHE IV score.

Setting and Procedures

This study’s population included patients admitted to the CTS-ICU and CCU in a large, metropolitan academic teaching hospital in the southeastern region of the United States.

Participants

Patients included were individuals 18 years and older admitted to the CTS-ICU and CCU during the 2018 fiscal year (September 1, 2017, to August 31, 2018) and discharged on or before August 31, 2018. Patients who were consulted in the CCU or CTS-ICU and evaluated by a physical therapist in the CCU or CTS-ICU were included. 6-Clicks was scored only by the physical therapist in the CCU and CTS-ICU. Individuals with a missing 6-Clicks score at the physical therapist ICU evaluation were excluded. We did not include a physical therapist evaluation and concurrent 6-Clicks score that occurred after the patient discharged to the floor from the cardiac ICU. A physical therapist consult may be termed an “order” in other hospital systems, but consult best describes the physical therapists’ autonomy to evaluate and treat patients as they deem appropriate once consulted. Although direct access is expanding, in most hospital systems across the United States as well as in our health system, physical therapists do not have the autonomy to see a patient without a consult or “order” being placed.³² However, as per American Physical Therapy Association’s guidelines, physical therapists do have the autonomy to defer or discharge a consultation if a patient is unsafe to or cannot participate in an evaluation and have the autonomy to select patients to continue with physical therapist treatment based on the patient’s response to intervention.³³

Patients were excluded when their discharge location was not based on patient functional status, including patients who left against medical advice, discharged to hospice at home or inpatient, transferred to another hospital, or died during their stay. Additionally, patients who discharged to a LTAC but could not participate in physical therapy due to inability to follow commands or maintain arousal were excluded because they were unable to engage in rehabilitation. Patients were also excluded if they were readmitted to the same or another ICU during their hospital stay or the physical therapist consult was entered prior to the patient’s admission to the CTS-ICU/CCU and consult was not re-requested during the ICU stay (Fig. 1).

Outcome: Discharge Destination

Discharge destination was classified as facility or home. A patient’s discharge destination was classified as “facility” if the patient was discharged to an inpatient institution for additional care and rehabilitation. Facilities included SNF, an IRF, or a LTAC. The discharge destination was classified as “home” if a patient was discharged back to their nursing home (for those admitted to the hospital from a nursing home), went home with or without a caregiver, or received home health or outpatient therapy.

Primary Independent Variable: AM-PAC 6-Clicks Basic Mobility

The standard of care for physical therapists at the academic teaching hospital was to administer 6-Clicks on all patients during evaluation but not on subsequent visits. The 6-Clicks tool is a 6-item scale, with each item representing a basic mobility task scored from 1 (total assist) to 4 (independent). The total scale is scored from 6 to 24, with lower numbers indicating lower functional ability.²⁵ 6-Clicks is designed to measure activity limitations in the domains of bed mobility, transfers, ambulation, and stairs and has been validated for use in assessing patients with a wide range of medical and surgical conditions in the acute care setting²⁵ with high inter-rater reliability.³⁴

Predictor Variables

Age, sex, living situation (alone or with a caregiver), and insurance (Medicare, Medicaid, private, other), APACHE IV score, and medical stability at the physical therapist’s evaluation (yes/no) were included in the predictive model.

Descriptive Variables

We descriptively examined the following clinical variables: number of days from ICU admission to physical therapist consult, number of days from ICU admission to physical therapist evaluation, number of days from physical therapist consult placement to physical therapist evaluation, ICU LOS, hospital LOS, number of missed physical therapist treatments, and total number of physical therapist treatments in ICU and hospital.

Data Analysis

Descriptive statistics and univariate analysis were computed using Statistical Package for Social Sciences (SPSS, IBM, Armonk, NY, USA), version 26.0. We checked the normality of the independent variables using the Shapiro–Wilk test. Univariate analyses were conducted to assess the association of independent variable and covariates to predict discharge destination (facility/home).

Prediction Model

We constructed a prediction model based on a machine learning approach using Classification tree using Python. This approach allows higher-order feature (predictor) interactions and picks the best features to split the data into the chosen classes. The model was based on features representing patient demographics: age (stratified by ≥65 and <65 years), gender, insurance type (Medicare, Medicaid, private, or none), if the patient lived alone (yes/no), medically stable at first physical therapist evaluation attempt (yes/no), and clinical scores of 6-Clicks and APACHE IV. Scikit-learn, a machine learning python package, was used to construct, evaluate, and optimize the classification tree.³⁵ The classification tree was constructed and evaluated by dividing our sample into a train–test split using Leave-One-Out cross-validation.³⁶ In Leave-One-Out cross-validation, each sample is used once as a test set and the remaining samples makeup the training set. Although this results in a large number of train-test sets (equal to the number of samples), it allows to maximize the utilization of the limited data in training the model.

There were 359 valid samples available to build the prediction model. However, the data were highly imbalanced, with n = 74 samples representing outcome “facility” and n = 285 representing outcome “home.” This imbalance in data could train a model that was heavily biased towards the majority outcome sample. To balance the data, we utilized Synthetic Minority Over-sampling Technique³⁷ to oversample minority class and create an equal number of cases representing both the classes (n = 285 × 2). The new dataset containing 570 samples was used to create the classification tree model using standard criteria for growing and pruning (threshold change in Gini index). To estimate the model’s performance, we calculated the cross-validated (K = 5 folds) accuracy scores, and precision and recall values. We performed a sensitivity analysis after excluding 3 patients coded as “home” discharge because they returned to their prior living situation, which was a nursing home.

Results

Descriptives

Because the data for continuous variables were not normally distributed as revealed by the Shapiro–Wilk test, we present the median count of the variables in the Table. Of the 359 patients, 197 (55%) were males, and the median age (interquartile range) was 68 (70–60) years, ranging from 28 to 93 years. A physical therapist evaluation from ICU admission took approximately 3 (4.9–1.7) days. The average ICU LOS and hospital LOS was 2.8 (5.1–1.7) days and 8.1 (13–5.3) days. When comparing means, patients evaluated by physical therapists in the ICU were older, had a longer ICU and hospital LOS, and were more medically complex based on higher APACHE IV score than patients not evaluated by physical therapists in the ICU (data not shown).

Univariate Analyses

Mann–Whitney U test revealed that patients were more likely to be discharged to a PAC facility if they had lower function on 6-Clicks, were older, have Medicare as their primary insurance, were not medically stable, had increased mortality risk on APACHE IV, increased time from ICU admit to physical therapist consult and evaluation, longer ICU and hospital LOS, higher ICU and total physical therapist sessions, and missed physical therapist treatments (Table).

Prediction Model

The classification tree split the data into 4 distinct groups along with their predicted outcomes: (1) individuals with a 6-Clicks score >15.5 were discharged to home; (2) individuals with a 6-Clicks score between 11.5 and 15.5 with primary insurance other than Medicare were discharged to home; (3) individuals with a 6-Clicks score between 11.5 and 15.5 with Medicare insurance were discharged to a PAC facility; and (4) individuals with a 6-Clicks score ≤11.5 were discharged to a PAC facility (Fig. 2). The model had a predictive ability to classify individuals discharged to home with a precision of 0.93 and recall of 0.73. The model had a predictive ability to classify individuals discharged to a PAC facility with a precision of 0.43 and recall of 0.80. Overall model accuracy calculated using cross-validation with 5 folds was 0.80 ± 0.01. Figure 3 shows a confusion matrix representing the model’s performance on the dataset. After excluding 3 patients from the home discharge group, the sensitivity analysis revealed a similar trend; however, the 6-Clicks score increased by 2 points across all domains.

Discussion

To our knowledge, this is the first study to examine the ability of the 6-Clicks to predict discharge from physical therapist evaluation in cardiac ICU. Our findings demonstrated that 6-Clicks scores at first physical therapist visit in cardiac ICU could be utilized to provide objective data to facilitate early discharge planning. Using a machine learning approach, we were able to accurately predict discharge destination for 80% of cardiac ICU patients only by using 6-Clicks and utilization of Medicare insurance. The results of this study revealed that a 6-Clicks score ≤11 on physical therapist evaluation in the cardiac ICU accurately predicts discharge to a facility and patients with a 6-Clicks score ≥16 were discharged to home. In a previous study of patients of all diagnoses during an acute hospital stay, patients with a 6-Clicks score <20 upon physical therapy evaluation were predicted to discharge to a facility at 86% accuracy, which was a higher cutoff score (16 vs 20) than our current study.²⁵ Two studies that evaluated 6-Clicks within 48 hours of hospital admission revealed a cut-off score of ≥17 or 18 to home discharge and ≤12 to PAC facility discharge in a diverse patient population.^38,39 The lower cut-off score in our study for PAC facility discharge may be because cardiac ICU patients are more sick and debilitated than patients in acute hospitalization requiring institutionalization. Three patients were included in the home group because they were residing in a nursing home without rehabilitation services and returned to the same prior living situation. When these 3 patients were removed from the study, the 6-Clicks score trend increased by 2 points (≥18 discharged to home, ≤13 discharged to a facility), in line with prior studies. This could be because we excluded the low-functioning individuals from the analysis, overall increasing the 6-Clicks score.

Insurance type was also a predictor in the model, where patients with Medicare and a 6-Clicks score between 11 and 15 were discharged to a facility; otherwise, they were discharged to home. This indicates that patients without Medicare had to be functionally much more debilitated at physical therapist evaluation (median 6-Clicks score 9) compared with patients with Medicare to discharge to facility (median 6-Clicks score 11). Medicaid insurance has been shown to be a barrier to access to in-facility rehabilitation services due to low reimbursement.⁴⁰ Additionally, because age is a determinant of access to Medicare coverage, Medicare and age may be interlinked, although age was not a predictor of discharge to facility vs home in this model. In our sample, 90% of patients ≥65 years had Medicare and 36% of patients <65 years had Medicare. When comparing between those who went home and those who went to a facility, age was significantly higher in patients who were discharged to a facility. This is consistent with prior literature indicating that factors associated with home discharge include younger age, being married, and having better functional and cognitive status.⁴¹

Higher APACHE IV score, longer ICU and hospital LOS and increased number of physical therapist sessions, and number of missed physical therapist sessions were associated with facility discharge. APACHE IV score indicates a higher likelihood of mortality in ICU patients and is a marker of severity of illness. Additionally, longer ICU LOS and more physical therapist sessions during hospitalization indicate the significant rehabilitation needs among cardiac ICU survivors.⁵ Patients waiting on facility placement while continuing to be managed in the acute hospital may concurrently have both more physical therapist sessions and more missed physical therapist sessions due to prolonged physical therapist plan of care. These medically complex patients with longer LOS may require more physical therapist sessions and benefit from a PAC facility discharge for continued rehabilitation to improve long-term functional outcomes.⁶ Frailty and decreased function are associated with increased hospitalization burden and longer hospital LOS among cardiac patients.⁴² In this study, the most frequent, documented, missed physical therapist sessions were due to medical instability (eg, low blood pressure or arrhythmia) and scheduling conflicts (eg, patient procedure, surgery, or hemodialysis). Rare occurrences of missed physical therapy were due to patient refusal. Patients who miss therapy due to medical instability may have fewer opportunities to mobilize and exercise safely, resulting in decreased strength and mobility.⁴³ These patients may therefore benefit from a PAC facility discharge to optimize their recovery and prevent long-term deficits and mortality.⁴⁴

Patients discharged to facility received a physical therapist consultation significantly later than those who discharged to home, which delayed the physical therapist evaluation of patients discharged to a facility. This is likely because patients discharged to a facility may have increased medical complexity, as also demonstrated by higher APACHE IV scores. The speed of patient stabilization varies widely and although the goal is to extubate and stabilize patients as quickly as possible, this may not be feasible depending on the patient and their care needs.⁴⁵ Administrative logistics may have delayed the physical therapist consultation²⁴ because rounding and interdisciplinary discussions occur generally in the mornings at our institution, and patients who stabilize or whose cardiac surgery occurs later that day may not be consulted until the following morning. Assessment of rounding and consulting practices may be considered at the institutional level to prevent delays in physical therapist consults and improve patient outcomes.

Additionally, a significant number of patients did not receive a physical therapist consult in both CTS-ICU and CCU; further research may examine the outcomes of this patient population compared with the patients who received physical therapy in the cardiac ICU. Patients not evaluated by a physical therapist in the ICU had a shorter ICU LOS, so they were likely not in the ICU long enough for a physical therapist consult to be requested and a physical therapist evaluation to be completed. Our study reveals that the lower 6-Clicks scores is predictive in more critically ill, older, and more debilitated patients when evaluated in ICU compared with those who were evaluated on the floor. Future studies must compare our findings with patients evaluated by physical therapists on the floor after an ICU stay to determine the ability of 6-Clicks to predict discharge destination in those who had a shorter LOS in the ICU. The findings of this study will improve the efficiency of early discharge planning by learning our health system in the cardiac ICUs to improve care delivery. Implementation and adoption of such a learning health systems approach will improve outcomes of both patients and the health systems.⁴⁶ Our next step will be towards identifying and implementing strategies to enable physical therapist consultations among patients who would benefit from rehabilitation.

Clear communication with patients and families by providing rationale for discharge recommendations early on facilitates improved rehabilitation outcomes and patient satisfaction.²³ The consequences of incorrect discharge recommendation can be detrimental because patients who are discharged to a postacute care facility instead of home could be at higher risk for infection⁴⁷ or complications such as pressure ulcer.^48,49 On the other hand, patients who are recommended to discharge to home instead of a PAC facility may be at increased risk of falls or readmission.⁴⁹ Although discharge recommendations include multiple factors, evaluating patient’s functional status during physical therapy assessment is a major component.^50,51 6-Clicks is a valid physical function measure that provides the ICU interdisciplinary team the ability to assess patient mobility across a wide range of patients and is quick and easy to use.⁵¹

Limitations

This study was a single-year and single-site study; future studies must analyze longitudinal data across multiple sites. Limitations also included those inherent to retrospective research using electronic medical records such as reliability of documentation and biases in hospital practice. Although data involving physical therapist evaluation timestamps may be imprecise (because the timestamp is entered when the therapist submits the note rather than the time of assessment of the patient), these notes were submitted on the same day of service. Only physical therapist treatment dates and not timestamps were utilized for this study. Limitations in analysis were due to sample size imbalance in the outcome variable. To correct for the sample size imbalance, we over-sampled our data using the synthetic minority over-sampling technique and randomly under-sampled and ran 2 different classification tree models. The over-sampling model had a higher overall accuracy in the cross-validation scoring and was utilized for the study analysis. Both models provided strong support for the use of AM-PAC scores as a determinant for discharge destination.

Future Directions

Further research into the utilization of 6-Clicks in predicting discharge destination from ICU may include medical and additional surgical ICUs to understand the applicability of this tool in a heterogeneous sample. Future studies must also assess the functional outcomes of patients who received physical therapy only in the ICU, those who received physical therapy in an acute care hospital, and those who did not receive physical therapy during their acute care hospitalization; this may allow for improved utilization of physical therapy to address the patients who most need early intervention. Considering that patients who had more missed therapy sessions due to medical instability were more likely to discharge to a PAC facility, further research should investigate the ability to optimize therapeutic interventions and prevent patients from missing therapy. Early physical therapy may combat the effects of post-intensive care syndrome, including ICU-acquired weakness, but longer hospitalization and rehabilitation needs may still occur.⁵² Further research should continue to investigate combatting post-intensive care syndrome and ICU-acquired weakness through physical therapy as well as discharge planning to decrease LOS and optimize discharge destination.

The results of this study indicate that 6-Clicks and primary insurance provide information to facilitate early discharge planning for ICU physical therapists and aid in interdisciplinary communication with other rehabilitation therapists, social workers, physicians, and patients and families. Early patient-centered discharge planning may improve patients’ outcomes by decreasing LOS and readmission rates.^16,23,53

Author Contributions

Concept/idea/research design: K.C. Whitlock, M. Mandala, K.L. Bishop, V. Moll, J.J. Sharp, S. Krishnan

Writing: K.C. Whitlock, M. Mandala, K.L. Bishop, V. Moll, J.J. Sharp, S. Krishnan

Data collection: K.C. Whitlock, J.J. Sharp

Data analysis: K.C. Whitlock, M. Mandala, J.J. Sharp, S. Krishnan

Project management: K.C. Whitlock, K.L. Bishop, S. Krishnan

Providing facilities/equipment: K.L. Bishop, S. Krishnan

Providing institutional liaisons: K.L. Bishop, S. Krishnan

Clerical/secretarial support: S. Krishnan

Consultation (including review of manuscript before submitting): K.L. Bishop, S. Krishnan

Acknowledgments

The authors thank Donna Lipsius Gordon (director of rehabilitation therapy, Emory University Hospital Midtown) for providing administrative support and assisting with project development. They also thank Kejun Xu (Emory University CDW) for assistance with data extraction; and Natalie Edwards, Elizabeth Niedermair, and Gabrielle Ringenberg (Emory University, doctor of physical therapy students research course 915) for performing manual data extraction and quality checking the data.

Ethics Approval

The study protocol was approved by Emory University’s Institutional Review Board (#00114595).

Funding

There are no funders to report for this study.

Disclosures and Presentations

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

A portion of this study was presented as a platform presentation at the American Physical Therapy Association’s Combined Sections Meeting; February 11–15, 2020; Denver, Colorado, USA.

References