Mortality derived from 5-year survival in patients with Alzheimer disease.

ABSTRACT

OBJECTIVE: The objective of the authors of the source article was to investigate survival and course of the disease in a registry of patients with Alzheimer disease diagnosed from 1987-1996. The objective of this article is to derive expected mortality, age/sex-matched as closely as possible to data available in the article, and to derive comparative mortality from the survival results at 5 years. METHODS: The cohort of 521 patients with newly recognized senile dementia (Alzheimer disease) was drawn from a health organization in western Washington with an enrollment of 23,000 members age 60 years and older. After initial selection, a careful evaluation was made to confirm the diagnosis. The cohort was followed to death or 2001, with follow-up (FU) ranging from 0.2 to 14 years (mean 5.2 years). The authors used elaborate statistical methods in their analysis of results. A detailed description is given in the text of this article on the derivation of both observed and expected mean annual mortality rates to obtain excess death rates (EDR) and mortality ratios (MR) as indices of excess mortality averaged over 5 years of FU. RESULTS: All patients were age 60 or older, mean age was 80.2 years, and females outnumbered males, 66% to 34%. The overall EDR, all patients combined, was 37 extra deaths per 1000 per year. For all males EDR was 52; and for all females, EDR was 33 per 1000 per year. The corresponding MR values were 142%, 149% and 141%. EDR and MR increased with test scores measuring severity of cognitive impairment, with physical features of the severity of the dementia, and especially with the presence of comorbid diseases such as stroke, coronary heart disease (CHD) and congestive heart failure (CHF). With a mean age of 80 years, MR values are relatively low in comparison with EDR, owing to the high mean expected mortality. CONCLUSION: An approximate pattern of increased mortality has been found in a cohort of senile dementia patients in the Group Health Cooperative, in the area of Seattle, Washington, despite some uncertainty attributable to absence of sex and race distribution data within each of the 4 individual age groups.