Comparisons of cause of death verification methods and costs in the lipid research clinics program mortality follow-up study

Classification of causes of death by a nosologist is standardized and relatively inexpensive, although the quality of the data recorded on death certificates has been subjected to criticism and the level of detail may not allow examination of specific circumstances of the death. A costly alternative is classification by a panel of physicians reviewing additional clinical information. The costs associated with both methods of classification as utilized by the Lipid Research Clinics Program Mortality Follow-up Study are presented along with their advantages and disadvantages. For deaths among the elderly (65 years or older), determination of the underlying cause of death may be difficult, especially as regards cardiovascular disease causes of death using only the death certificate, and impossible for classifying sudden death. Both classification methods are compared for 268 deaths among the elderly and 155 deaths among the nonelderly males and females participating in the study. The kappa statistics for observer agreement indicate moderate agreement (k = 0.58) and were not significantly different between the two age-groups (p greater than 0.10). Thus, the agreement between the methods was not affected significantly by the age of the deaths being classified. However, examination of the agreement chart showed a preference by the nosological classification for cardiovascular classification in the elderly population. The effect of the different classifications upon the analysis of the predictive value of high-density lipoprotein cholesterol on coronary heart disease mortality showed little difference between both methods of classification.