Skip to main content

Explanatory Multivariate Modeling for Disability, Pain, and Costs in Patients with Spine Pain

 

By

spine-pain1

Explanatory Multivariate Modeling for Disability, Pain, and Costs in Patients with Spine Pain

Christopher Green, SPT

Background

Spine pain has been a clinical interest of mine for as long as I have considered physical therapy as a career. From my earliest internship with a spine-focused therapist to completing a formal clinical rotation at the UNC Spine Center, I have been absorbed with the evaluation and treatment of spine pain. Of particular interest to me was why some patients seemed to get better much quicker than others. This recognition led to curiosity, and if by fate, I became aware of a research proposal for a capstone project seeking to answer these same questions.  Our UNC DPT program director, Dr. Deborah Givens, PT, PhD, DPT along with Dr. Chad Cook, PhD, PT, FAAOMPT formed a collaborative research group with myself, my friend and fellow DPT student Anthony Pastore, and two very bright Duke students to study prognostic factors of spine pain.

Dr. Charles Thigpen, PhD, PT, ATC, who holds research positions with ATI Physical Therapy and the University of South Carolina, provided our research group with a dataset of direct access physical therapy patients who were seen for spine-related pain. For the purposes of my Capstone project, I teamed up with a Duke student to complete focused research on the low back literature, while Anthony and another Duke student turned their attention to the cervical spine. This endeavor has led to a variety of products relating to low back pain and spine pain, including a manuscript that we plan to publish in the next year, as well as an abstract for submission to the 2018 American Physical Therapy Association’s Combined Sections Meeting (CSM). Due to copyright issues, I am unable to post our manuscript on my Capstone page at this time. However, you will find links to the abstract and supporting products below.

Overview

Low back pain (LBP) is one of the most common musculoskeletal complaints that adults experience worldwide, with a global point occurrence estimated at 9.4%.2 In 2010, a United States survey estimated that 29% of adults had experienced low back pain (LBP) in the last months.3 While many LBP sufferers may have good outcomes regardless if they receive treatment, a significant percentage of the population eventually experience chronic or reoccurring pain and disability. In fact, 24% to 33% of individuals who experience activity-limiting LBP will continue to have reoccurring episodes throughout their lifetime.4,5 

LBP is also responsible for a significant economic encumbrance. A 2006 review revealed that LBP is accountable for healthcare costs upwards of $100 billion per year.6 Additionally, individuals experiencing spine-related pain were responsible for 58% more in healthcare spending when contrasted with those who do not have spine pain.7 Unfortunately, this increase in spending is not associated with significant improvements.7

 

Statement of Need

Although a variety of studies have examined prognostic factors in relation to low back pain, studies that have explored outcomes in patients who were seen via direct access for physical therapy are absent. The inclusion of this type of patient is significant because they represent a departure from traditional physician-referred patients in terms of levels of acuteness. Additionally, a novel aspect of this Capstone project is the inclusion of investigating how certain prognostic factors influence insurance claims. Identification of prognostic factors in terms of both disability and cost are important in improving patient outcomes and decreasing the economic burden of spine-related pain.

Products

As a result of delving into the low back pain (LBP) literature over the past year, I have developed a variety of products related to prognostic LBP studies. For my Evidence Based Practice II class, I produced a Critically Appraised Topic that assessed several studies that examined the relationship between prognostic factors and the Oswestry Disability Index (ODI), a commonly used outcome measure for LBP-related disability. As part of an independent research class, I generated a review of LBP literature that included a survey of commonly studied prognostic factors, a review the psychometric properties of the ODI, a discussion of multivariate logistic regression analysis, and a consideration of LBP clinical prediction rules. Finally, upon completion of our research group’s manuscript, I drafted an abstract of our work for submission to the 2018 CSM.

Find links to my products below:

Self-Assessment and Reflection

Both Dr. Givens and Dr. Cook have been instrumental in consulting and advising me on my Capstone products. To evaluate my abstract, I used the format and requirements necessary for abstract submission by CSM. Evaluation of the manuscript has come in the form of discussion and formal edits by our research group and our group advisors. It was drafted using the format required for submission to Spine and followed the Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis ((TRIPOD) recommendations, an evidence-based checklist used for reporting prediction models.8

This Capstone project has revealed to me the breadth and depth of effort required to produce quality original research materials. I have gained a better understanding of what previously felt like exotic statistical analysis, and have a much deeper appreciation for data analysis in general. One of the greatest aspects of my learning process through this project was the continual refinement of my academic writing skills.

Acknowledgements

First and foremost, I would like to thank Dr. Givens and Dr. Cook for deciding to unite two great physical therapy programs in a collaborative research effort, and for including Anthony and I in this great process. I owe Dr. Givens many thanks for her patience and guidance throughout the completion of my products. I appreciate Dr. Cook for assisting me in better understanding advanced data analysis, and logistic regressions in particular. This project would not have been possible without Dr. Thigpen, who provided our dataset and valuable feedback along the way. Lastly, I’m indebted to all the UNC DPT faculty and students with whom I’ve had the pleasure of spending the past 3 years with. I could not imagine a better group of people.

References

  1. Spine Pain. University of Utah Health. https://healthcare.utah.edu/spine/spine-pain.php. Published 2017. Accessed April 20, 2017.
  2. Hoy D, March L, Brooks P, et al. The global burden of low back pain: estimates from the Global Burden of Disease 2010 study. Ann Rheum Dis. 2014;73:968-974. doi:10.1136/annrheumdis-2013-204428.
  3. National Center for Health Statistics. Summary Health Statistics for U.S. Adults: National Health Interview Survey, 2010. 2010;10(252). https://www.cdc.gov/nchs/data/series/sr_10/sr10_252.pdf. Accessed April 6, 2017.
  4. Wasiak R, Pransky G, Verma S, Webster B. Recurrence of low back pain: definition-sensitivity analysis using administrative data. Spine (Phila Pa 1976). 2003;28(19):2283-2291. doi:10.1097/01.BRS.0000085032.00663.83.
  5. Stanton TR, Henschke N, Maher CG, Refshauge KM, Latimer J, McAuley JH. After an episode of acute low back pain, recurrence is unpredictable and not as common as previously thought. Spine (Phila Pa 1976). 2008;33(26):2923-2928. doi:10.1097/BRS.0b013e31818a3167.
  6. Katz JN. Lumbar disc disorders and low-back pain: socioeconomic factors and consequences. J Bone Joint Surg Am. 2006;88 Suppl 2(suppl_2):21-24. doi:10.2106/JBJS.E.01273.
  7. Martin BI, Deyo RA, Mirza SK, et al. Expenditures and Health Status Among Adults With Back and Neck Problems. JAMA. 2008;299(6):656. doi:10.1001/jama.299.6.656.
  8. TRIPOD Checklist : Prediction Model Development. Julius Center. https://www.tripod-statement.org/TRIPOD/TRIPOD-Checklists/TRIPOD-Checklist-Prediction-Model-Development. Published 2017. Accessed April 14, 2017.

6 Responses to “Explanatory Multivariate Modeling for Disability, Pain, and Costs in Patients with Spine Pain”

  1. Chris Green

    Carla,
    I think spine pain is so interesting to me because it’s such a big problem in health care, so much has been studied/written about it, and yet, it still poses such a conundrum as far as finding an agreement on the best way to approach the problem. It seems like even the Clinical Practice Guidelines have more suggestions for what is not worth doing rather than definitive positions on what to do. I think a lot of reasons contribute to this, most especially the sheer number of patients who experience LBP which creates a heterogeneous population that makes generalized recommendations difficult. To make matters more difficult, as you well know, a clear pathoanatomical etiology for back pain is at best elusive, creating a complexity for both clinicians and researchers. Findings from my research that show imaging is often unnecessary, expensive, and even sometimes producing a pernicious effect on our patients’ perceptions of their pain impacts how I would like to treat back pain. I say indirectly because although we cannot order imaging, we can educate patients on the benefits and risks of asking their doctors for imaging. I am also inspired to form a closer relationship with physicians and surgeons to open the lines of communication to keep an ongoing discussion of when imaging will best serve our patients. The finding of “active” treatment approaches leading to lower costs might influence the way I treat, but without going into too much detail, the “active” treatment was determined by the ratio of active CPT codes to passive CPT codes. This doesn’t necessarily provide much insight into the *type* of active intervention performed. However, throughout my own research of prognostic LBP studies and LBP studies in general, it seems like you can find evidence to support a variety of treatment interventions for LBP, but also a number of studies proposing one approach is better than another. My main takeaway from that research is that any active treatment is better than only a passive one (only patient education, traction, some other passive modality, etc). My intuition based off this information tells me that having a variety of interventional approaches may be most beneficial to serving such a heterogenous population of LBP sufferers. I think clinicians may gravitate more towards one approach or another, whether it be MDT vs manual therapy vs exercise-intensive vs lumbar stabilization focus, for example. At this point in my nascent career, I wish I could say I knew whether it was better to specialize in one interventional style or approach and have others as adjunct/secondary approaches, or if keeping a generalized initial approach leads to better patient outcomes. Regardless, I do believe that the prognostic literature to date suggests that chronic LBP sufferers often require more resources and time than those with acute durations. That may make me more inclined to schedule CLBP patients with more frequency as opposed to someone with acute symptoms, for example. All LBP patients seem to benefit from treatment as early as possible, which emphasizes the importance of direct access and PT as an initial point of contact for LBP sufferers of any chronicity. This may circle back to the importance of forming quality relationships with physicians and surgeons, and to advocate for our profession in terms of our value in treating persons with LBP, especially as the initial point of contact for a patient experiencing a LBP episode.

    Two other aspects of my research may have an effect on how I treat patients with LBP. The literature seems to be mixed on the contributions of psychosocial factors to the prognosis of LBP in terms of disability and outcomes. However, I think there is enough evidence to include something like the FAB-Q as an outcome measure for my LBP patients. It takes a nominal amount of effort to enter it into something like FOTO or an EMR, and could provide extra evidence for opportunities where I could positively contribute to a patient’s care. Emerging evidence in chronic pain strongly suggests that psychosocial factors play an important role in the experience of chronic pain, and the fact that many LBP sufferers reach a chronic stage begs the inclusion of the possibility that psychosocial factors may indeed by playing a role in my patient’s LBP. Secondly, although we did not account for this in our formal research project, the importance of work status is also a strong prognostic indicator in much of the literature. Including a patient’s work status on their entrance forms and including investigation of occupational information into every patient subjective evaluation may reveal important aspects of a patient’s prognosis.
    My Capstone project and my accompanying research may have produced more questions than answers when it comes to patient care, but I think they are important questions that still have a direct and indirect impact on how I could best treat patients with LBP. I am as curious as ever and hope to play an important role in improving approach to treating LBP.

    Reply

  2. Chris Green

    Debby,
    Indeed there is a massive amount of literature, and I hope to continue being more acquainted with the breadth of studies available! I also hope that our efforts with this project will provide some influence to further research in using statistical models and a checklist (like TRIPOD) to report on prognostic studies. Although potentially difficult, I also believe including costs is something future researchers should really strive to employ, especially because of reasons that you touch on with your question. I think I have demonstrated in my literature review that although a variety of MCIDs have been reported for the ODI, the fact the each exists in the literature is basis for requesting more visits from an insurance company. If I can demonstrate a patient is improving in a clinically meaningful way, or that they are very close to meeting that improvement, I can find studies that support a value appropriate for my patient. Perhaps more pertinent to an insurance companies interests would be the fact that “active” physical therapy has proven to incur less overall insurance claims. Documenting the types of active interventions and conveying the fact that our PT treatments have been shown to be less expensive to them in the longterm may persuade insurance providers to extend PT visits.

    Reply

  3. Chris Green

    Monica,
    Thank you for taking some time to review my Capstone materials. I hope some of the resources within my products will help you in your spine rotation! We did seem to have some findings consonant with what we’ve been learning- namely that imaging is expensive, and perhaps not always necessary. Patients with a higher baseline ODI and female gender were actually more likely to meet a lower and higher threshold for meeting a minimal clinically important difference as a result of physical therapy treatment. In other words, those two groups were likely to see greater meaningful changes in disability as measured by the ODI. The high ODI group makes sense because they have more room to improve. The female gender findings actually run counter to a lot of what has been reported elsewhere in the literature. Many studies found that the female gender was actually a negative prognostic indicator of disability outcomes. It is hard to say why this difference exists in our study versus what others have found. Perhaps direct access physical therapy allowed them to be seen sooner, leading to better outcomes? At this stage, we can only speculate. It’s our hope that other studies will use the direct access model of treatment and follow TRIPOD guidelines to confirm or refute what we have found thus far.

    Reply

  4. Carla Hill

    Chris,
    Very interesting Capstone project. It’s obvious you learned a lot about spine pain and statistics through this effort. How will the findings of your research impact your patient care, either directly or indirectly?

    Reply

  5. Debby Givens

    Chris – it’s great to see the work pulled together. As I’ve said many times, you picked a topic with a massive amount of literature! Let’s consider the time when you are practicing in an outpatient PT setting and you have a big caseload of people with low back pain. Most have policies that cover 6 visits of PT but some will give more visits if the PT requests and shows the patient is making progress. What parameters will you use to make the case that a patient should have more PT visits, based on what you’ve learned from your reading and research?

    Reply

  6. Monica Anderson

    Hi Chris,

    I was eager to review your capstone products because I will be spending a considerable amount of time at an outpatient spine clinic this summer. Your findings on the cost benefits of active interventions for low back pain further confirm my preference for this form of treatment approach. I will keep this in mind this summer! When reading the results of the study, it makes sense that participants with higher initial ODI were more likely to have less-than desirable outcomes with physical therapy treatment. However, I’m intrigued that gender had such a profound role on outcomes. In your research was there there any reasoning for why this may have been in your patient population? And it seems your research resonates with what we have learned throughout our curriculum, in that ordered imaging may incur unnecessary costs, that may not correlate with better outcomes in patients with low back pain. I am excited to see the results of your abstract and manuscript submission!

    Best,

    Monica

    Reply

Leave a Reply