Repeatability of health economic data in COPD

Background: Little is known concerning the precision of self-reported health economic data. Aim: To investigate the repeatability of self-reported health economic cost and utilization data in subjects with chronic obstructive pulmonary disease (COPD). Methods: Twelve weeks after inclusion, a physician and a nurse in random order interviewed 29 subjects with post-bronchodilator COPD included from a research registry of COPD patients. The interview recorded healthcare utilization and costs, sick leave, exacerbations and quality of life (QoL). Variation of individual agreement of observations was described estimating kappa statistics and 95% limits of agreement. Results: Mean age was 63 years (standard deviation (SD) 10) and 17 participants were men. Average FEV(1)% predicted was 56% (SD 15). For sick leave, exacerbation, healthcare provider visit, change of medication, assistance and leisure time the kappa values were 1.00, 0.73, 0.73, 0.66, 0.63 and 0.54, respectively. The physician recorded fewer days of exacerbation and fewer contacts with healthcare providers than the nurse (p = 0.01 and p = 0.05, respectively). The 95% limits of agreement for costs of drugs were -690 to +710 Norwegian Kroner (NOK), -1200 to +899 NOK for costs of healthcare providers and -20 to +26 for QoL as measured by a visual analogue scale. Conclusion: Repeatability of economic data from COPD patients showed considerable variation. This issue should be addressed when analyzing cost data from interviews and when designing studies on health economy. (C) 2008 Elsevier Ltd. All rights reserved.