Online Program


Physical Therapy in Home Health Care: Modeling Eligibility & Estimating Effectiveness
*Carlin Brickner, The Visiting Nurse Service of New York 
Robert J Rosati, The Visiting Nurse Service of New York 
Iordan Slavov, The Visiting Nurse Service of New York 

Keywords: propensity scores, home health care, physical therapy

Home health care is an important component in the rehabilitation of patients and prevention of hospitalizations after being referred from a hospital, physician or nursing home. It is believed that physical therapy plays an important role in a patient’s rehabilitation in home health care. However, there is limited research that tests the effectiveness of physical therapy in a home health care setting. This study is comprised of two phases: 1) Model eligibility for physical therapy 2) Create a comparable sample of eligible patients who received physical therapy and those who did not. A model was developed to predict the probability of receiving physical therapy from a large sample (n= 61573) of home health care admissions in the New York City metro area during 2007. This model used a wide variety of baseline data sources such as the Center for Medicare and Medicaid Services mandated Outcome and Assessment Information Set, medication counts, payer eligibility, and ICD9 groupings. A cut-off for the propensity score was defined by assessing the receiver operating characteristic curve and the results of an expert's assessment on a random sample of 100 patients. This cut off was used to determine patient eligibility for physical therapy. The second phase of this study uses this criterion to identify all patients admitted into care during the first half of 2008 who are eligible for physical therapy. Patients who did not receive physical therapy but had a physical therapy order were eliminated in order to rule out patient behavior as a factor for not receiving services. A second model was developed to estimate the probability that a patient, eligible for physical therapy, would not be recommended for physical therapy by their physician or nurse. Patients who were eligible, but did not receive physical therapy (n = 5639), were matched on their propensity scores from this second model to patients who were eligible and received physical therapy (n = 21155). From this matched sample, we provide estimates of the effects of physical therapy on hospital and emergency room utilization and on activities of daily living during home health care.