Musculoskeletal spine pain, including neck pain (NP) and low back pain (LBP), are complex, chronic conditions that can result in significant disability, and they are among the top ten most prevalent conditions affecting Veterans. However, little VA research focuses on patients with these disorders.
The objectives of this study were to:
1. describe the characteristics of veterans who seek care for NP and LBP, including socio-demographics and clinical characteristics;
2. describe the treatment provided to veterans with NP or LBP and the cost of that treatment;
3. describe the outcomes of care for veterans who seek treatment for NP and LBP, where outcomes are defined as incident episode duration and probability of first recurrence; and
4. compare episode duration for veterans with LBP who receive VHA/DoD guideline concordant care to those who do not.
This was a retrospective observational study to describe the natural history of back and neck pain in VA and to compare current practice to the VHA/DoD Clinical Practice Guideline for the Management of LBP and Sciatica in Primary Care. From VA administrative datasets (Patient Treatment File, National Patient Care Database, and Fee Basis), we identified all individuals with a diagnosis consistent with neck pain or back pain pain between FY2002 and FY2009, and compared this population with the VA population at large. We defined the incident event study population as individuals who had an initial spine pain event in FY2005-FY2007 (i.e., no previous care for spine pain in previous years' data) and followed them for 365 days after the first event using CPT and diagnosis codes. Using the FY2007 cohort alone, we identified individuals who had received more than a single visit of care for spine pain during the 90 days after the incident event (first episode). We categorized these patients as "recovered" if they did not receive any medical care for spine pain 91 to 365 days after their incident spine pain encounter, while continuing to receive VA care for other conditions during this period.
Objective 1: We identified 2,163,292 Veterans (24% of the total VA population) who had a mechanical or non-specific spine pain diagnosis between FY2002 and FY2009. Approximately 75% of the cases had back pain diagnoses and 15% had neck pain diagnoses at the incident event. The remaining 10% had combinations of back, neck, and non-specific spine pain diagnoses. The rate of arthritis in the 2009 cohort was 41% greater than the general VHA population, and the rate of headache was 65% greater than the general VHA population. The rates of mental health and substance use disorders was 40 - 65% greater in this population than the general VHA population. 93% of the population was male, with an average age of 56 years.
Selecting an incident cohort of veterans who had a first episode of neck or back pain in fiscal FY2005-FY2007 (n=685,544), the population was slightly older (57 years) than the overall cohort, 38% were estimated as overweight and 39% as obese. The cohort was predominantly white (64%), and 3% were homeless. Mean pain scores for patients with neck pain was 3.4 (SD 3.3) out of 10 at the incident event, and for patients with low back pain 3.8 (SD 3.4). Pain scores were missing at the incident event for 1.03% of the population. Severity was classified by ICD codes: 87% were least severe; 6.5% were moderately severe; and 6.2% were classified as most severe.
Objective 2: In the incident cohort (FY2005-2007) only 17% of the population had a VA return visit for spine pain within 30 days of the incident visit. Of these visits, 30.5% were to Primary Care, 23.6% were to Physical Therapy, 8% were to Prosthetics Services, 6.8% to admit/screening/emergency services, 6.1% to X-ray, MRI or CT scan, and 1.9% to Chiropractic Care in VA. In the subpopulation of patients with incident events in FY2007, the marginal increase in total costs in the year of the incident event compared to the previous year was $3500 per person for patients with neck or low back pain.
Objective 3. Using the FY2007 subpopulation and the above definition of recovered, 41% of the population of patients with neck pain recovered and 35% of the patients with low back pain recovered.
Objective 4: In the FY2007 cohort, controlling for patient demographics and clinical characteristics, medical and mental health comorbidities and substance use disorders, patients with low back pain who received any element of guideline-concordant care in the first 30 days following the incident event were 12% more likely to recover, compared to those who did not receive guideline-concordant care (OR 1.12, p<0.001; 95%CI 1.08-1.16). Adjusting for mean values of all covariates, the probability of recovery for patients who received guideline-concordant care for low back pain was improved 3% over those who did not receive guideline-concordant care (probability with guideline 36%, vs. 33% without).
Classifying guideline-concordant care by service type (Primary Care E&M, Physical Therapy evaluation, Chiropractic Care manipulation, Occupational Therapy evaluation, and Physical Medicine services), patients who received physical medicine services were 29% more likely to recover (OR 1.29; Robust SE 0.05, p-value<0.0001; 95%CI 1.20 - 1.39) than those who did not, and those who received chiropractic care were 26% less likely to recover in 90 days (OR 0.74; Robust SE 0.08; p-value=0.006; 95%CI 0.60 - 0.92) than those who did not receive chiropractic care.
Adjusting for mean values of all covariates, patients who received physical medicine services in the first 30 days of treatment were 45.4% more likely to recover than those who did not, and those who received physical medicine services and an Occupational Therapy or Physical Therapy evaluation were 40.7% more likely to recover than those who did not.
This study provides insight into the burden of back and neck pain in VA and the extent of variation in treatment and outcomes for veterans. Our findings suggest that when viewed in the aggregate, the low back pain guidelines have little effect on patient outcomes. However, when the components of the guidelines are disaggregated into service type, we see that patients who receive physical medicine services within 30 days of their first visit for low back pain are more likely to recover than those who do not.
External Links for this Project
Grant Number: I01HX000156-01
- Sinnott PL, Siroka AM, Shane AC, Trafton JA, Wagner TH. Identifying neck and back pain in administrative data: defining the right cohort. Spine (Philadelphia, Pa. : 1986). 2012 May 1; 37(10):860-74. [view]
- Sinnott PL. Guideline Concordant Care for LBP. Paper presented at: North American Spine Society Annual Evidence and Technology Spine Summit; 2013 Feb 28; Park City, UT. [view]