Work-Related Chronic Low Back Pain

Return-to-Work Outcomes After Referral to Interventional Pain and Spine Clinics

Rathin N. Vora, MD, MPH; Bruce A. Barron, MD, MS; Anthony Almudevar, PHD; Mark J. Utell, MD


Spine. 2012;37(20):E1282-E1289. 

In This Article


Study Population

A total of 230 patients who met the inclusion criteria were entered into the study. Of the 230 patients, 122 were referred to a PTC/SC (Yes PTC/SC) and 108 were not enrolled in PTC/SC (No PTC/SC) (Figure 1).

Figure 1.

A schematic representation of patient selection and outcome measurements. The fi gure shows that every patient with work-related chronic low back pain seen in OEM and FLOHS within the study period was part of the study. Of total 211 patients, 54% were referred to a pain treatment center/spine clinic and 46% were not. Their pain and functional outcomes were measured at 6 months and 1 year after being seen at pain treatment center/spine clinic in the referred group and at 6 months and 1 year after being seen at OEM/FLOHS in the nonreferred group

Primary Outcome Measures

There were no significant differences between the referral and nonreferral groups at baseline with regard to age, sex, race, smoking history, body mass index, history of back injury and LBP type, or other comorbid conditions (Table 1). The prevalence of muscle relaxants, narcotics, and physical therapy was significantly higher in the referral group (Yes PTC/SC). However, the baseline pain, function levels, and strength demand of the job were similar in both groups. We also refit the logistic model using narcotics as a covariate and found that it was not significantly associated with outcome in any of the models. We think that the reason for the higher percentage of magnetic resonance images in the referral group was due to the fact that a majority of the PTCs/SCs need magnetic resonance images of the lumbar spine to determine whether a particular type of intervention would be warranted rather than because of a difference in severity of disease between the 2 groups at baseline.

To assess change in function and pain levels over time, paired differences comparing 6 months to baseline and 12 months to 6 months were performed separately for each referral type (Table 2). For example, a patient who at baseline was "not working" (functional level = 3) and improved at 6 months to the "return to modified work" (functional level = 2) would have a functional improvement of −1 from 6 months to baseline. A significant decrease in pain and significant improvement in work status was observed at 6 and 12 months for the nonreferral group (No PTC/SC) only.

Evaluating Outcomes Based on Clinical Site of Initial Evaluation

To further understand our findings for the overall outcome measures, we explored the possibility of potential differences in patient groups between the 2 clinics, OEM versus FLOHS (Table 3 and Table 5). At baseline, we observed statistically significant differences in several covariates (e.g., sex, smoking, narcotics) by clinic site. In addition, baseline functional scale level was significantly higher (indicating poorer functional status) for the OEM group (Table 3). There was no significant difference in the level of decline among the referral group, but across the nonreferral group, the decline in pain and improvement in function were much greater in FLOHS than in OEM. This difference reached statistical significance in almost all pain and functional change categories (Table 4).

Because sex distribution and rates of smoking and narcotic use differed significantly between sites, differences in functional/pain levels attributable to these covariates were investigated within each clinic/site pair; however, no differences were observed.

Interaction of TSI and Clinical Referral Status

We also evaluated the effect of TSI on various outcomes (Table 5). There was a significantly large difference in TSI between sites (OEM = 173 d, FLOHS = 6 d, P < 0.001). This difference far exceeded any differences attributable to the other covariates or to clinic referral status. A clear trend was observed in functional improvement with TSI as can be seen in Figure 2. There was a strong decreasing association within the No PTC/SC group between TSI and FI. There were no significant associations within the PTC/SC group (Figure 2). Furthermore, among the nonreferral group, the rate of FI among the subjects with the higher levels of TSI did not differ noticeably from the referral group. This did not seem to be the case within FLOHS site model and may have been attributed to the relative lack of high TSI subjects. Thus, the rate of FI differed significantly between clinic referral groups but only among patients with low values of TSI.

Figure 2.

Predicted function improvement rate as a function of TSI. The fi gure shows the changes in functional improvement (return-to-work status) as a function of time since initial injury. The upper 2 graphs include all patients in the study comparing the nonreferral group with the referral group. The lower graphs separate OEM and FLOHS and compare the nonreferral group with the referral group in each clinic separately. The graphs show that the higher functional improvement was seen in the nonreferral group overall, with the largest contribution coming from FLOHS. Within the referred group, there was no change in functional improvement overall and in each clinic group separately. TSI indicates time since injury.

Relationship Between Function and Pain Scales

To examine the association between pain and functional scales, the means for the pain scale were stratified by functional scale level ( Table 6 ) at baseline, 6 months, and 12 months. There was a positive association between pain and function levels, which became larger with time in the study. The findings indicate that as the pain worsens, functional/return-to-work outcomes worsen