Higher Frequency of Acute Occupational Therapy Services Is Associated With Reduced Hospital Readmissions

Method

Participants

The study sample consisted of 17,618 Medicare patients with a HRRP-qualifying diagnosis. We identified HRRP diagnoses in Epic using International Statistical Classification of Diseases and Related Health Problems 10th Revision (World Health Organization, 2019) codes. Only acute care patients who were enrolled in Medicare were included. Patients who left the hospital against medical advice, transferred to another acute care hospital, died during the index hospitalization, were discharged to hospice, or died within 30 days of discharge were excluded from the sample (Figure 1). These exclusion criteria minimized the competing risk of readmission related to death because this indicator does not reflect quality of care (Kumar et al., 2019).

OPEN IN VIEWER

Figure 1.

Cohort creation.

Results

The 30-day readmission rate for the total sample was 13% (Table 1). Patients with COPD had the highest readmission rate (14.9%), followed by patients with HF (14.8%) and PN (13.8%). Patients who were readmitted were younger (M ± SD = 69.8 yr ± 12.8 yr vs. 71.9 yr ± 12.7 yr), were more likely to be Black or African-American (31.1% vs. 25.1%), were more likely to discharge to a nursing facility or other location (i.e., assisted living facility, intermediate care facility, rehab and psychiatric, or other; 24.6% vs. 21.0%), had a higher comorbidity count (5.9 ± 3.2 vs. 4.4 ± 3.3), and were more likely to have an ICU stay (37.7% vs. 27.2%). Patients who were readmitted also had a lower duration of occupational therapy services (27.1 ± 8.3 min per day vs. 29.6 ± 10.8 min per day) and a lower frequency of services (74% ± 26% treatment days vs. 82% ± 24% treatment days).

Adjusted results from the receipt of the occupational therapy services model, occupational therapy duration model, and occupational therapy frequency model with the primary outcome of readmission are presented in Table 2. Patients who received occupational therapy services had significantly higher odds of readmission (odds ratio [OR] = 1.18, 95% confidence interval [CI] [1.07, 1.31]). Patients who received higher frequencies of services had significantly lower odds of readmission (OR = 0.99, 95% CI [0.99, 1.00]). Patients who received higher durations of services did not have significantly higher or lower odds of readmission.

Younger age and higher comorbidity count were both significantly associated with increased odds of readmission across all three models (see Table 2). Conversely, patients who had the diagnoses of THA and TKA had significantly lower odds of readmissions across all three models. Patients who were discharged to a nursing facility; had an ICU stay; or had a primary diagnosis of COPD, HF, or PN had significantly higher odds of readmission in the occupational therapy received model. ICU stay was also significantly associated with increased odds of readmission in the occupational therapy duration model. Race, sex, and social support were not statistically significant in any of the three models.

Results of the χ² analyses and Fisher’s exact tests comparing the occupational therapy CPT codes for readmitted versus not-readmitted patients with HRRP-qualifying diagnoses are presented in Table 3. A higher proportion of not-readmitted patients received ADL/self-care training (80.6% vs. 74.9%; p = .005). Conversely, a higher proportion of readmitted patients, in comparison with patients who were not readmitted, received therapeutic exercise (4.3% vs. 1.7%; p < .001).

Discussion

Reducing readmissions and the associated poor patient outcomes and high costs is a top priority in the U.S. health care system and, accordingly, is the target of HRRP (McIlvennan et al., 2015). In this study of more than 17,000 Medicare patients with HRRP-qualifying diagnoses, we found that receipt of occupational therapy services in the hospital was associated with higher odds of readmission; however, patients who received acute occupational therapy services had reduced odds of readmission if they received higher frequencies of those services while hospitalized. In addition, patients who were not readmitted to the hospital received more ADL/self-care training than patients who were readmitted to the hospital. These results provide important guidance on how occupational therapy services can be leveraged to reduce readmissions.

It is not surprising that the isolated examination of receipt of occupational therapy services was associated with higher odds of readmission. Patients who receive occupational therapy services while hospitalized appear to be older, have a higher number of comorbidities and chronic conditions, and be discharged to a location other than home (Andrews et al., 2015). All of these demographics are known to be associated with higher readmission risk (Horney et al., 2017; Pedersen et al., 2017; Silverstein et al., 2008). Including functional status in the analyses would have helped to clarify the relationship between receipt of occupational therapy services and readmissions. Impaired functional status is known to be associated with higher risk of readmission and consequently is the reason that occupational therapy services are used in the acute care setting (Middleton, Downer, et al., 2018; Middleton, Graham, & Ottenbacher, 2018). However, large amounts of data on functional status were missing at the study site, making it impossible to include functional status as a variable in the analyses.

This barrier is common when investigating the relationship between acute occupational therapy services and outcomes. In the acute care setting, no universal or required standardized measure of functional status exists. As a result, documentation on functional status is at the discretion of individual occupational therapy practitioners and facilities. Future studies should consider implementation of a standardized measure of functional status at the study site to ensure that all potentially confounding variables can be included in the study.

Higher frequencies of occupational therapy services, but not higher durations of services, were associated with reduced odds of readmission for the patients in our study. Our findings are novel because the unique impact of frequency of services on outcomes has yet to be evaluated. Only the combined impact of frequency of occupational therapy and physical therapy services has been shown to reduce readmissions and death (Freburger et al., 2020). Our findings indicate that for each 10% increase in the frequency of treatment days, the odds of readmission were 1% lower. The insignificant finding on duration of services is consistent with the limited literature on the topic. Kumar et al. (2019) found that only higher durations of acute physical therapy services, not acute occupational therapy services, were significantly associated with reduced readmission risk. Our findings suggest that frequency of occupational therapy services may play a more important role in improving patient outcomes than duration of services.

The ORs for comorbidity count (OR = 1.12), as well as for THA and TKA (OR = 0.42), demonstrated a stronger association with readmission than the occupational therapy frequency variable (OR = 0.99). However, the importance of our findings should not be diminished. Comorbidity count and the diagnosis of THA or TKA are not modifiable factors, but occupational therapy frequency is a modifiable service delivery factor. To reduce readmissions, actionable items must be identified and leveraged. Our findings demonstrate that when patients are seen by occupational therapy practitioners more frequently while hospitalized, their odds of readmission are lowered. Future studies may want to explore exact dosage for frequency of services to identify the optimal frequency associated with reduced readmission risk for patients. Evidence on exact dosage would lead to best practice guidelines that optimize patient outcomes in the acute care setting.

In our study, we found that patients who received occupational therapy services and who were not readmitted received a significantly higher amount of ADL/self-care training while hospitalized compared with patients who were readmitted, according to CPT codes. Identifying the importance of ADL and instrumental activities of daily living (IADL) interventions, not only impairments with ADLs and IADLs, in preventing readmissions is a notable finding for the occupational therapy profession because ADLs and IADLs are uniquely within the scope of occupational therapy practice. The occupational therapy CPT code for ADL/self-care training includes interventions focused on ADLs, IADLs, and fall prevention education. In the postacute setting, impairments with ADLs/self-care, increased dependency in IADLs, and experiencing a fall after hospitalization have been shown to be associated with increased likelihood of readmission (DePalma et al., 2013; Galet et al., 2018; Greysen et al., 2015; Middleton, Downer, et al., 2018; Middleton, Graham, & Ottenbacher, 2018; Pisani et al., 2018).

Future studies should seek further clarification of the exact occupational therapy activities and interventions provided by occupational therapy practitioners according to CPT codes because the American Medical Association’s definitions for CPT codes allow for wide interpretation by clinicians. Last, more complex intervention profiles for patients organized by CPT codes are needed to accurately reflect acute care practice. In future studies, researchers should consider examining the interaction effects of different occupational therapy services on readmission risk or focus on investigating the specific impact of ADL/self-care training on readmission risk.

Study Limitations

Our study had several limitations. First, the occupational therapy services provided variable only included patients who had one service provided (i.e., one type of CPT code billed) for the entirety of the patient’s hospital stay. This approach reduced the sample size (n = 3,804) for the analyses by 54%. In addition, it may not be truly reflective of practice because multiple occupational therapy services can be provided to a single patient during their hospital stay. However, this approach allowed us to isolate the impact of a single occupational therapy service on readmission risk and minimize the confounding effects of receiving multiple services.

Second, because of the broad nature and lack of standardized guidelines available for occupational therapy practitioners to use when submitting CPT codes, it is unclear which specific occupational therapy interventions occurred when the ADL/self-care training CPT codes were submitted.

Third, each treatment CPT code was assigned the value of 15 min for the analysis of duration of occupational therapy services, producing only an estimate of minutes spent by practitioners in delivering services.

Fourth, generalizability of our study is limited by the data source and sample characteristics. The data originated from a single hospital and included only Medicare patients. Therefore, findings may not be generalizable to all settings where acute occupational therapy services are provided, non-Medicare patients, or Medicare patients who do not have a HRRP diagnosis.

Fifth, only within-system hospital readmissions were captured. Out-of-system readmissions were not available in the data source.

Finally, several covariates that could have confounded results were not included in the analyses because of data limitations, including patients’ functional status, dual eligibility for Medicare and Medicaid, additional therapy received in postacute care settings, cognitive status, and housing instability. These covariates may be options for future studies.

Implications for Occupational Therapy Practice

The findings of this study have the following implications for occupational therapy practice:

Occupational therapy practitioners in the acute care setting may consider delivering higher frequencies of occupational therapy sessions for patients while hospitalized to improve their outcomes, including reduced risk of readmission.

Conclusion

The findings of our study indicate that patients with HRRP-qualifying diagnoses who received occupational therapy services at higher frequencies had lower odds of readmission. In addition, patients who received occupational therapy services and who were not readmitted received more ADL/self-care training during sessions than patients who were readmitted. These findings help to further define occupational therapy’s unique role in reducing readmissions and may contribute to future work on evidence-based practice guidelines for improving patient outcomes.

Supplemental Material