Showing papers by "Alexander J. McNeil published in 1993"
TL;DR: Applications of Gibbs sampling in medicine, involving longitudinal, spatial, covariate measurement and survival models, are reviewed, including applications in immunology, pharmacology, transplantation, cancer screening, industrial epidemiology and genetic epidemiology.
Abstract: SUMMARY We review applications of Gibbs sampling in medicine, involving longitudinal, spatial, covariate measurement and survival models. Applications in immunology, pharmacology, transplantation, cancer screening, industrial epidemiology and genetic epidemiology are discussed.
191 citations
TL;DR: It is suggested that the stricter, biologically more plausible, case definition used in Scotland of two consecutive CD4 cell counts of < or = 200 x 10(6)/l [CD4(200) (x 2)] should be adopted--not as a new definition of AIDS, but as an additional important state of severe HIV-related immunodeficiency (SHRID).
Abstract: OBJECTIVE: To establish whether various accepted and proposed AIDS definitions have clinical and biological validity: because the Centers for Disease Control and Prevention (CDC) and World Health Organization (WHO) reclassifications of AIDS are important not only for describing the epidemiology of HIV disease but also to individual patients. SETTING: Regional Infectious Diseases Unit, City Hospital, Edinburgh, Scotland, UK. PATIENTS: We analysed the disease progression of 532 HIV-seropositive individuals seen at the City Hospital, Edinburgh, up to the end of July 1991. MAIN OUTCOME MEASURES: Annual numbers of potentially reportable cases from the Edinburgh City Hospital Cohort according to three proposed AIDS case definitions based on: (1) first lymphocyte count < or = 1000 x 10(6)/l; (2) first CD4 cell count < or = 200 x 10(6)/l; or (3) first of two consecutive CD4 cell counts < or = 200 x 10(6)/l. Lifetables to death (irrespective of cause) from month of satisfying the above case definitions, and proportion of patients who satisfied each definition in their calendar year of enrollment in the cohort are reported. RESULTS: There is a threefold increase in patients in the Edinburgh City Hospital Cohort defined as having AIDS under the 1987 and the proposed 1992 CDC definitions--a substantial change for patients and epidemiologists alike. That they are describing different immunodeficiency states is clear from lifetable analysis, which reveals median survivals of 20 and 50 months under the 1987 and the proposed 1992 AIDS definitions, respectively. For epidemiological purposes, redefinitions based on the WHO proposed classification of HIV disease using either a lymphocyte count < or = 1000 x 10(6)/l or a CD4 cell count < or = 200 x 10(6)/l are broadly interchangeable. They are not equally effective for monitoring individual progression (CD4 cell count is superior). Both, for different reasons, lack biological plausibility. CONCLUSIONS: We therefore suggest that the stricter, biologically more plausible, case definition used in Scotland of two consecutive CD4 cell counts of < or = 200 x 10(6)/l [CD4(200) (x 2)] should be adopted--not as a new definition of AIDS, but as an additional important state of severe HIV-related immunodeficiency (SHRID). Median survival under the CD4(200) (x 2) case definition was 40 months in the Edinburgh cohort. We have illustrated differences in CD4(200) case ascertainment between injecting drug users and other HIV-infected patients in the Edinburgh City Hospital Cohort. We recommend that surveillance centres should ascertain date of first immunological monitoring as well as date of SHRID diagnosis in order to identify differential case ascertainment.
28 citations