The Effect of Patient Financial Liability on Physical Therapy Utilization And
Total Page：16
File Type：pdf, Size：1020Kb
University of South Carolina Scholar Commons Theses and Dissertations Fall 2019 The Effect of Patient Financial Liability on Physical Therapy Utilization and Patient Reported Outcomes for Patients With Low Back Pain: An Instrumental Variable Analysis Adam D. Lutz Follow this and additional works at: https://scholarcommons.sc.edu/etd Part of the Exercise Science Commons Recommended Citation Lutz, A. D.(2019). The Effect of Patient Financial Liability on Physical Therapy Utilization and Patient Reported Outcomes for Patients With Low Back Pain: An Instrumental Variable Analysis. (Doctoral dissertation). Retrieved from https://scholarcommons.sc.edu/etd/5594 This Open Access Dissertation is brought to you by Scholar Commons. It has been accepted for inclusion in Theses and Dissertations by an authorized administrator of Scholar Commons. For more information, please contact [email protected]. THE EFFECT OF PATIENT FINANCIAL LIABILITY ON PHYSICAL THERAPY UTILIZATION AND PATIENT REPORTED OUTCOMES FOR PATIENTS WITH LOW BACK PAIN: AN INSTRUMENTAL VARIABLE ANALYSIS by Adam D. Lutz Bachelor of Science Louisiana State University, 2005 Doctor of Physical Therapy University of North Florida, 2010 ________________________________________________ Submitted in Partial Fulfillment of the Requirements For the Degree of Doctor of Philosophy in Exercise Science The Norman J. Arnold School of Public Health University of South Carolina 2019 Accepted by: Shana E. Harrington, Major Professor Charles A. Thigpen, Committee Member Ellen Shanley, Committee Member John M. Brooks, Committee Member Cheryl L. Addy, Vice Provost and Dean of the Graduate School © Copyright by Adam D. Lutz, 2019 All Rights Reserved. ii Dedication To my wife, Johna Lutz, who has sacrificed and supported me greatly to allow me to follow my dreams and is a proud and devoted partner beyond my worth. To my children, Harper, Amelia, & Anders Lutz, who inspire me every day to be a better person and worthy example. To my parents, David & Nina Lutz, who have never accepted anything less than excellence and have been immensely supportive of my academic pursuits. iii Acknowledgements To Chuck, who inspired a love of research in me long ago and has put me in a position to succeed from the very beginning. To Shana, who has been an advocate for and valuable critical eye of my work—always ensuring a higher quality, more impactful product. To Ellen, who has always taken time out of her busy schedule to advise me, buy me coffee, critique and better my work, brainstorm, or simply walk and talk. To John, who has been an excellent educator and mentor—and always taken the time to meet with me to explain complex methods until I understood. To Tom Denninger, Ryan Geary, Cole Chapman, and Brett Windsor, who have been outstanding mentors, contributors, and collaborators throughout my studies. iv Abstract Low back pain (LBP) is the most common orthopedic complaint managed by physical therapists in the United States. Insurance plans that require greater patient financial liability per physical therapy visit are thought to disincentivize participation in recommended physical therapy plans of care. The primary purpose of this investigation was to use instrumental variable regression to investigate the effect of insurance- mediated patient financial liability per visit on physical therapy visit count and functional patient reported outcomes. The secondary purpose of the study was to examine the generalizability of results to the broader excluded patient population—and beyond to patients that seek physical therapy for LBP. Insurance-mediated patient financial liability per visit was a poor instrument with an inconsistent effect that trended against the hypothesized direction. The instrument had no statistically or clinically important effect on visit count, and no subsequent clinically important change in patient reported outcome. A comparison of baseline patient information and “index” values for predicted visit count and change in patient reported outcome identified few potential clinically important differences, yielding the conclusion that results of the primary investigation are likely generalizable to a broader patient population with LBP. These results suggest patient financial liability per visit is a poor instrument with currently available data; however, any results from large observational analyses would likely generalize to a larger patient population with LBP. v Table of Contents Dedication .......................................................................................................................... iii Acknowledgements ............................................................................................................ iv Abstract ................................................................................................................................v List of Tables .................................................................................................................... vii List of Figures .................................................................................................................. viii Chapter 1: Introduction & Review of Literature..................................................................1 Chapter 2: Methodology ....................................................................................................15 Chapter 3: Instrumental Variable Regression Analysis .....................................................26 Chapter 4: Generalizability of Comparative Effectiveness Research with Missing Data ....................................................................................................42 Chapter 5: Discussion & Concluding Remarks: ................................................................55 Referenced Works ..............................................................................................................59 vi List of Tables Table 2.1: Variable Consideration by Instrumental Regression Stage for Research Question 1 ........................................................................................................25 Table 3.1: Description of Total Patients, Included Sample, and Quintiles of Patient Financial Liability Per Visit .............................................................................35 Table 3.2: First Stage Instrumental Variable Regression Coefficients Predicting Physical Therapy Visit Count ..........................................................................36 Table 3.3: Second Stage Instrumental Variable Regression Coefficients Predicting MDQ Change ...................................................................................................38 Table 4.1: Comparison of Included and Excluded Patients ...............................................49 Table 4.2: Regression Coefficients Predicting Physical Therapy Visit Count: Derived from Included Patient Cohort (Note: Coefficient for State of Physical Therapy Services Excluded from Table) .........................................................50 Table 4.3: Regression Coefficients Predicting MDQ Change: Derived from Included Patient Cohort (Note: Coefficient for State of Physical Therapy Services Excluded from Table) ......................................................................................52 vii List of Figures Figure 1.1: Factors Known or Theorized to Influence PRO Change .................................13 Figure 1.2: Factors Known or Theorized to Influence Physical Therapy Visit Count ......14 Figure 3.1: Representative Theoretical Model (Not Actual Data) .....................................40 Figure 3.2: Study Inclusion/Exclusion Flow Diagram ......................................................41 Figure 4.1: Histogram of Predicted Physical Therapy Visit Count by Inclusion Status....54 Figure 4.2: Histogram of Predicted MDQ Change by Inclusion Status ............................54 viii Chapter 1 Introduction & Review of Literature Background Low back pain (LBP) is the most common orthopedic complaint in the United States (U.S.) with an estimated lifetime prevalence of 80%.1 Evidence-based guideline care recommendations have been introduced to promote clinically and cost-effective management—and reduce costly, unwarranted variation in the evaluation and management of patients seeking medical care for LBP.2,3 Early care guidelines co- endorsed by the American College of Physicians and the American Pain Society in 2007 promoted screening for psychological risk-factors known to influence the development of chronic LBP and a thorough physical exam that categorized patients into defined classification groups to help guide appropriate interventions.2 These guidelines also strongly endorsed limited use of imaging (i.e., x-rays, computed tomography [CT], and magnetic resonance imaging [MRI]), limited use of over the counter medication (i.e., acetaminophen, non-steroidal anti-inflammatory drugs [NSAIDS]), and promoted patient education that encouraged appropriate self-care, maintaining an active lifestyle, and an understanding of the expected course of their condition.2 More recently, guideline recommendations co-endorsed by the American College of Physicians and American Academy of Family Physicians in 2017 recommended classifying patients as acute/subacute or chronic—then provided evidence-based, non-invasive treatment recommendations for each classification.3 For patients with acute or subacute LBP, 1 superficial heat, massage, acupuncture, spinal manipulation, NSAIDS, and skeletal muscle