Analysis of MEPS data to discern the chronicity of back pain in young people

Intro: Medical Expenditures Panels Survey (MEPS) data largely have reported only cross-sectional or time-trend analyses of prevalence or health service use for back pain. Pooled MEPS longitudinal files have added utility for studying small population subgroups (e.g. children), and for examining patterns of utilization or pain over time. Data/Methods: Medical Expenditures Panel Survey (MEPS) data pooled across MEPS 2-year Longitudinal Panels 5-through-10 (MEPS Survey Years 2000-2006). Of the total pooled cohort of MEPS respondents with backpain (N=10,753), 6.7% were youth age 7-20 (n=718). This session will present a detailed report of a methodological protocol for discerning the chronicity of back pain in longitudinal MEPS All-5-Rounds data, using Round-by-Round pattern mapping to generically operationalize “episodes-of-pain” and “episodes-of-care”. Results: Of 718 youth (age 7-20) with back pain, 75% reported backpain in only a single MEPS round, 19% in 2 or more consecutive rounds (chronic continuous backpain) and 6% during intermittent rounds (recurrent episodic backpain). Of the 718 youth, 549 used ambulatory health services for backpain, and for 89% of those 549 (n=486) their back pain “episodes-of-pain” patterns were consistent with their “episodes-of-care” patterns. Discussion: Large-scale population-based survey data such as MEPS can be pooled across years and panels, in order to increase sample sizes and cell sizes for conducting focused analyses of small population groups such as children with back pain, subgrouped by chronicity. A companion presentation in this session (Ward&Smith) will further discuss the challenges of subgrouping (e.g. by race, gender) for robust multivariate modeling of such data.