No abstract

The intake was adjusted for within-person variability by the method proposed by Iowa State University, using the software PC-SIDE, and energy-adjusted for the residual method. The validity for 23 nutrients was assessed by Spearman correlation coefficient and weighted k. Intra-class correlation coefficient and weighted k was used to reproducibility assessment. Results In the validation, the Spearman correlation coefficients ranged from 0.25 (riboflavin) to 0.57 (phosphorus) of which eight nutrients showed acceptable correlation (>0.4). In the reproducibility analysis, the intra-class correlation coefficients ranged from 0.18 (vitamin C) to 0.57 (niacin), of which five nutrients showed acceptable correlation. The weighted k ranged from 0.18 (sodium) to 0.67 (riboflavin) to reproducibility and from 0.19 (polyunsatured fat) to 0.56 (calcium) to validation. Conclusion The results support the use of this instrument to assess food intake in epidemiological studies conducted in São Paulo, Brazil. Background Several low-and middle-income countries, lack wellfunctioning population-based cancer registry. We evaluated the completeness of a pathology based cancer registry in Iran. We further studies evolutionary progress of the cancer registries worldwide. Methods We evaluated consistency of the incidence rates reported by national pathology-based cancer registry in Iran from 2004 to 2007. We further compared the incidence rates of the pathology-and population-based registries in a few regions, where both data were available. In addition, we studied the increasing trend in the number of population-based cancer registries worldwide, using the reports published in the Volumes IeIX of the monograph "Cancer in Five Continents." Results The Iranian pathology-based cancer registry, reports only about 60e70% of cancers. The underestimates were greater in cancers with poor-prognosis including lung, stomach, and oesophageal cancers. Almost four regional cancer registries were established every 10 years since 1960. However, the USA was an exception, where the number of cancer registries increased from 14 in 1998 to 44 regional registries in 2002, due to the advance infrastructure in the health informatics and ambitious initiatives by the Centers for Disease Control in the USA. Conclusions Pathology based cancer registry cannot provide reliable estimate for the cancer incidence rates, particularly in cancers with a poor prognosis. Developing countries should establish and support regional registries and expand their coverage gradually. Given the recent advances in the health informatics, small efforts will enhance the coverage of cancer registries worldwide, particularly in the less than middle income countries. SP3-79