Abstract
AIMS: Critical appraisal of secondary data made available by the OECD for the time frame 2000-2011. METHODS: Comparison of trends and variation of amputations in people with diabetes across OECD countries. Generalized estimating equations to test the statistical significance of the annual change adjusting for major potential confounders. RESULTS: A total of 26 OECD countries contributed to the OECD data collection for at least 1 year in the reference time frame, showing a decline in rates of over 40 %, from a mean of 13.2 (median 9.4, range 5.1-28.1) to 7.8 amputations per 100,000 in the general population (9.9, 1.0-18.4). The multivariate model showed an average decrease equal to -0.27 per 100,000 per year (p = 0.015), adjusted by structural characteristics of health systems, showing lower amputation rates for health systems financed by public taxation (-4.55 per 100,000 compared to insurance based, p = 0.002) and non-ICD coding mechanisms (-7.04 per 100,000 compared to ICD-derived, p = 0.001). Twelve-year decrease was stronger among insurance-based financing systems (tax based: -0.16 per 100,000, p = 0.064; insurance based: -0.36 per 100,000; p = 0.046). CONCLUSIONS: In OECD countries, amputation rates in diabetes continuously decreased over 12 years. Still, in 2011, one amputation every 7 min could be directly attributed to diabetes. Although interesting, these results should be taken with extreme caution, until common definitions are improved and data quality issues, e.g., a different ability in capturing diabetes diagnoses, are fully resolved.