Accuracy of death certificates for motor neuron disease and multiple sclerosis in the province of Palermo in southern Italy

Mortality studies based on death certificates (DCs) are relatively inexpensive and easy to conduct. Therefore, they are frequently used to evaluate variations of geographical and temporal patterns, particularly in uncommon diseases. Recent surveys of motor neuron disease (MND) and multiple sclerosis (MS) based on official mortality statistics in Italy showed a decreasing trend of mortality from northern to southern Italy. To evaluate if DCs are homogeneously recorded in Italy and whether or not they can be considered a good instrument for mortality studies, we assessed the accuracy of DCs for MND and MS in the province of Palermo, Italy, and compared our results with those reported in other studies. We searched the archives of the neurological clinic of the University of Palermo for patients affected by one of the two diseases who were residents and were diagnosed in Palermo province. We found 157 patients with definite MND and 360 with clinically definite MS. Seventy-eight out of the 157 MND patients and 43 out of the 360 MS patients had died. Considering the underlying cause of death, the true-positive rate was 48% [95% confidence interval (CI) 36-60%] for MND and 46% (95% Cl 30-62%) for MS. The rates we found for MND were lower than those reported either in Italy or worldwide. This result suggests caution in formulating gradient hypotheses for MND. Concerning MS, the rate we found was low, but similar to that found by others. This indicates that a mortality survey of MS based on DCs is unlikely to be contributory.