Abstract ID: 2227

Primary Topic: Health Services Research/Cost Effectiveness
Secondary Topic: Manual medicine/body work (including chiropractic and massage)
Tertiary Topic: Mind-body (including meditation and yoga)

Comparing the effectiveness and cost-effectiveness of pharmacologic and nonpharmacologic options for chronic low back pain
Patricia M. Herman, ND, PhD, RAND Corporation, Santa Monica, CA, United States; Tara A Lavelle, PhD, , Boston, MA, United States; John Luke Irwin, MPH, RAND Corporation, Santa Monica, CA, United States

Late Breaker: No


Chronic low back pain (CLBP) is one of the most prevalent types of chronic pain and the subject of many randomized trials, each comparing two or three complementary and/or conventional medicine interventions. Economic simulation models offer one method by which the effectiveness of all (studied) interventions can be directly compared, and their cost-effectiveness determined even if costs were not included in the original study.

Methods/Session Format

As part of our NCCIH-funded Center of Excellence in Research in CIH we built a Markov model for CLBP based on datasets from 12 large studies plus published results from 12 others. Costs were estimated from Medical Expenditure Panel Survey data. Tests of internal and external consistency were conducted and to ensure face validity the models were built using input from a panel of policy-maker representatives.

Patricia Herman is the only speaker for this oral presentation.


The intervention arms included in the models were limited to those with sufficient evidence: spinal manipulation, mobilization, acupuncture, massage, exercise, yoga, cognitive behavioral therapy (CBT), interdisciplinary rehabilitation, opioids, NSAIDs, injections, prolotherapy and usual MD care. Based on available data for a typical CLBP cohort the most effective interventions were yoga, flexion-distraction (a type of mobilization), active trunk exercise, individualized acupuncture and relaxation massage. However, the most cost-effective therapies were yoga, CBT, exercise, flexion-distraction and multidisciplinary rehabilitation. For a cohort with severe CLBP, yoga was still highly effective and cost-effective, but injections of different types (including prolotherapy) and spinal manipulation join the mix. Two studies of tramadol and one of prescription NSAIDs were included for the severe CLBP cohort (the only population in which we found useable studies of these) and compared poorly to the nonpharmacologic interventions.


CLBP demands good strategies that are both clinically effective and financially responsible. These models provide useful flexible tools by which to examine what we know about the relative effectiveness and cost-effectiveness of the large number of interventions commonly used for CLBP under different assumptions and scenarios.