The importance of identifying and modifying unemployment predictor variables in the evolution of a novel model of care for low back pain in the general population

Background ContextCare for low back pain (LBP) is costly, fragmented and, in non-compensation populations, rarely specifically addresses factors associated with maintaining employment status or return to work (RTW).PurposeThis study aimed to identify modifiable independent risk factors for (1) a negative work status at presentation and (2) a change in work status during treatment in a cohort of LBP patients. The results are intended to inform improvement in best-evidence care pathways to maximize societal outcomes and overall value of a new model of care.Study Design/SettingA prospective observational study was carried out. Inclusion criteria: Work-eligible, non-workers compensation patients with recurrent or persistent LBP ≥6 weeks and ≤12 months. Setting: The Inter-professional Spine Assessment and Education Clinics (ISAEC)—a novel Government-funded shared-care model of management for LBP.MethodsThis study used the following methods: (1) Cross-sectional analysis of baseline data from the initial ISAEC consultation (t0) from December 2012 to April 2014. Work status at t0 was dichotomized as employed (E) or underemployed (UE; unemployed, modified work duty, or disability). Multivariate logistic regression modeling was used to determine independent predictors of UE status at t0. (2) Bivariate analysis of longitudinal data from t0 to 6 months (t1) to identify risk factors for work status change. Employment journey categorized into four groups: Et0/Et1—employed at t0 and employed at t1; Et0/UEt1—employed at t0 and underemployed at t1; UEt0/Et1—underemployed at t0 and employed at t1; UEt0/UEt1—underemployed at t0 and underemployed at t1.ResultsThis study yielded the following results: (1) Initial consultation data on 462 consecutive patients (Et0=344, UEt0=118). Multivariate logistic regression identified legal claim, depression, smoking, and higher STarT Back (or Oswestry Disability Index [ODI]) score as independent risk factors for UEt0. (2) Overall UE rate did not significantly change during longitudinal analysis (n=178, UEt0=25.5%, UEt1=22.9%). However, 10.5% of Et0 became UEt1 (Et0/Et1=102, Et0/UEt1=12). Bivariate analysis identified elevated baseline ODI score as the only significant predictor variable for UEt1 in Et0 cohort (p=.0101). Conversely, ISAEC improved the employment status in 41% of UEt0 to Et1 (UEt0/Et1=16, UEt0/UEt1=23), and the absence of depression was significant for predicting RTW (p=.0001).ConclusionsFrom a societal perspective, employment status as an outcome measure is paramount in assessing the value of a new model of care for LBP. Mitigation strategies for the predictor variables identified will be included in ISAEC pathways to translate clinical improvement into societal added value.