Completeness and underestimation of cancer mortality rate in Iran: a report from Fars Province in southern Iran.

BACKGROUND The incidence and mortality rates of cancer are increasing worldwide, particularly in the developing countries. Valid data are needed for measuring the cancer burden and making appropriate decisions toward cancer control. We evaluated the completeness of death registry with regard to cancer death in Fars Province, I. R. of Iran. METHODS We used data from three sources in Fars Province, including the national death registry (source 1), the follow-up data from the pathology-based cancer registry (source 2) and hospital based records (source 3) during 2004 - 2006. We used the capture-recapture method and estimated underestimation and the true age standardized mortality rate (ASMR) for cancer. We used log-linear (LL) modeling for statistical analysis. RESULTS We observed 1941, 480, and 355 cancer deaths in sources 1, 2 and 3, respectively. After data linkage, we estimated that mortality registry had about 40% underestimation for cancer death. After adjustment for this underestimation rate, the ASMR of cancer in the Fars Province for all cancer types increased from 44.8 per 100,000 (95% CI: 42.8 - 46.7) to 76.3 per 100,000 (95% CI: 73.3 - 78.9), accounting for 3309 (95% CI: 3151 - 3293) cancer deaths annually. CONCLUSION The mortality rate of cancer is considerably higher than the rates reported by the routine registry in Iran. Improvement in the validity and completeness of the mortality registry is needed to estimate the true mortality rate caused by cancer in Iran.