Fusaro R E, Bacchetti P, Jewell N P
Division of Biostatistics and School of Optometry, University of California, Berkeley 94720-7380, USA.
Biometrics. 1996 Mar;52(1):211-25.
The proportions of gay men presenting with various AIDS diagnoses display temporal trends. In particular, the proportion of initial diagnoses reported as Kaposi's sarcoma (KS) has declined over time. Epidemiologists have hypothesized that (a) KS may require a cofactor, whose prevalence has declined over time, or (b) KS may have a shorter incubation period than other presenting diagnoses. We examine whether this latter hypothesis, considered in a competing risks framework, could account for the observed decline in KS. We nonparametrically estimate the relevant cause-specific hazard functions from the doubly-censored data of the San Francisco City Clinic Cohort by maximizing a roughness penalized likelihood using an EM algorithm. These estimates suggest that differences in the underlying cause-specific hazard functions account for a substantial portion of the observed diagnoses trends.
呈现各种艾滋病诊断的男同性恋者比例呈现出时间趋势。特别是，最初诊断为卡波西肉瘤（KS）的比例随时间下降。流行病学家推测，（a）KS可能需要一种辅助因素，其流行率随时间下降；或者（b）KS的潜伏期可能比其他呈现的诊断更短。我们研究在竞争风险框架下考虑的后一种假设是否可以解释观察到的KS下降情况。我们通过使用期望最大化（EM）算法最大化粗糙度惩罚似然，从旧金山城市诊所队列的双重删失数据中进行非参数估计相关的特定病因风险函数。这些估计表明，潜在特定病因风险函数的差异占观察到的诊断趋势的很大一部分。
