Abstract
Background: The healthy worker survival effect is a complex and poorly controlled bias affecting occupational studies. Little is known about its behaviour in samples of computer based workers. This descriptive study examined the healthy worker effect in relation to muscle performance and neck pain in call-centre operators.
Methods: 28 call-centre workers were examined for postural muscle performance. Neck pain was measured using a validated questionnaire. Analysis was by multivariate regression of the cohort, stratified by years of employment.
Results: Analysis using the entire sample failed to display any relationship between neck pain and muscle performance. Further examination revealed survival/secondary selection within the sample, with pain and poor muscle performance reduced after 6 years employment. Stratified analysis demonstrated that deep flexor muscle performance and posture were statistically significantly associated with neck pain. Each stratum explained a greater percentage of the variance in questionnaire scores, increasing from 42.3% for the entire sample to 95.1% for the 4 years employed subgroup.
Conclusion: The healthy worker survival effect is demonstrable in occupations with lower physical demands. It should be considered in analysis since it may obscure causal associations. There is no single method for controlling this bias and greater understanding of its components is required.