Date of Award

6-2026

Degree Name

Doctor of Philosophy

Department

Interdisciplinary Health Sciences

First Advisor

Diane Dirette Ph.D., OT, FAOTA

Second Advisor

Robert Lyerla, Ph.D., MGIS

Third Advisor

Kieran Fogarty, Ph.D.

Fourth Advisor

Alessander Danna dos Santos PT, Ph.D.

Keywords

BMI, clinical outcomes, low back pain, medicaid, nonspecific low back pain, race

Abstract

The purpose of this research is to investigate demographic factors, including age, gender, body mass index (BMI), race, marital status, insurance type, severity of disease (initial pain level), and clinical factors such as number of visits (no. of visits) at physical therapy (PT), length of time at PT and the type of clinician either physical therapist versus physical therapy assistant (PTA), that most predict PT clinical outcomes in patients with nonspecific low back pain (NSLBP) and see if there is an interaction between demographic and clinical variables.

Methods

The current study is a retrospective, cross-sectional, observational study using secondary electronic medical records (EMRs) of patients receiving PT to examine associations among various factors influencing PT clinical outcomes. This exploratory study answered four research questions. The first research question analyzed whether demographic variables such as age, gender, BMI, race, marital status, insurance type, and initial pain level correlate with functional improvement, which is the Patient Specific Functional Scale (PSFS) change score, for patients receiving PT for NSLBP.

In addition to demographic variables, clinical variables such as the no. of PT visits, the length of treatment time, and visits provided by a physical therapist versus the PTA were also included to observe for correlations with functional improvement. A total of four research questions were generated based on their clinical relevance to the field of physical therapy (PT). These four research questions examined the association of demographic factors on clinical outcomes, the association between the number of visits and outcomes, the association between initial pain and outcomes, and finally, the interaction between demographic and clinical factors.

Results

Demographic factors such as BMI, marital status, and insurance type were statistically significant. BMI was statistically significant but had no clinical relevance. The subgroup in the Marital Status, which is the Divorced group, was statistically significant with a small effect size. The subgroup in the insurance type, which is patients on Medicaid, was also statistically significant with a small effect size. The study found no interaction effects between the demographic and clinical variables. The no. of visits and initial pain levels did not show statistical significance.

Conclusion

The best fitting variables that were associated with the PSFS change score were BMI, being divorced, and Medicaid insurance. These variables, however, contribute to only 8% of the variance in the clinical outcomes, which reveals that the other 92% of the variance could be attributed to other factors.

Access Setting

Dissertation-Open Access

Download

Share

COinS