Different Alternatives to Assess the Burden of Disease Using Attributable Fraction on a Disability Variable: The Case of Pain and Chronic Musculoskeletal Disorders in Chile

To estimate the burden of disease through 4 complementary procedures to years lived with disability (YLDs) using the concept of attributable fraction and including analysis of subdomains of disability. We explored the burden on disability for 7 common musculoskeletal disorders (CMD) using the 2009 to 2010 Chilean National Health Survey, which included the Community Oriented Programme for the Control of Rheumatic Disease Core Questionnaire to identify cases with CMD, and an 8-domain questionnaire for health state descriptions. We calculated the proportion of disability attributable to pain in the general population and people with CMD. We also estimated the burden of CMD expressed as YLD and as the proportion of the disability in the general population attributable to people with CMD, with a particular focus in the pain domain of disability. Second order of uncertainty around point estimations was also characterized. Pain domain of disability accounted for 23.4% of the total disability in the general population, and between 20% (fibromyalgia) to 27.1% (osteoarthritis of the hip) in people with some of the selected CMD. People with chronic musculoskeletal pain accounted for 21.2% of total disability from general population, which generated 1.2 million of YLD (6679 YLD/100 000 inhabitants). Chronic low back pain and osteoarthritis of the knee were in the top position of specific CMDs, explaining the highest national burden. Pain is an essential component of disability in people with CMD and also in the general population. The approach used can be easily applied to other health conditions and other domains of disability. •In addition to years lived with disability, attributable fraction methods have been proposed to assess the burden of disease on disability. These methods traditionally use dichotomous outcomes of disability (disabled/not disabled), which assume a uniform construct, without considering its multidomain composition.•This study proposes 4 approaches applied to chronic musculoskeletal disorders (CMDs) and the subdisability domain of pain in Chile, to estimate the burden of disease as a complement of years lived with disability: (1) proportion of disability attributable to pain in the general population, (2) in people with a CMD; (3) the proportion of the disability, (4) pain domain disability in general population attributable to people with a CMD.•Results showed the methods can be easily implemented using cross-sectional data and could be applied