[Effect of misclassification of the place of residence on incidence estimates: the example of child cancer in a local health authority in Rome, Italy]. / Misclassificazione della residenza nelle schede di Dimissione Ospedaliera ed effetti sulle stime di incidenza: l'esempio dei tumori in età pediatrica nella ASL RMD.

A study on cancer of the lymphatic and hematopoietic tissue among residents aged 0-14 years was conducted by the Local Health Unit RMD (Rome, Italy; period 2003-09; codes of the International Classification of Diseases, Ninth Revision, Clinical Modification: 200-208). Age and gender Standardized Mortality and Hospitalization Ratios were computed in order to compare observed and expected cases, using municipal rates as reference. Place of residence at the time of admission, as recorded in the Hospital Registry, was compared with the information recorded in the Municipal Registers and the correlation between the two sources was calculated by Cohen's Kappa. No mortality nor morbidity excesses were observed in the study area. Although 14% of children were not confirmed as being resident at the time of admission, the Cohen's Kappa indicates a strong correlation between the Municipal Registry and the Hospital Registry (84%). The analyses restricted to children with ascertained residence did not yield different results. For those whose residence was not confirmed, the mismatch of information between the Municipality Registry and the Hospital Registry needs to be clarified.