Premium
An Estimate of the Prevalence of Human Immunodeficiency Virus Infection in England and Wales by Using a Direct Method
Author(s) -
Giesecke Johan,
Johnson Anne,
Hawkins Anne,
Noone Ahilya,
Nicoll Angus,
Wadsworth Jane,
Wellings Kaye,
Field Julia
Publication year - 1994
Publication title -
journal of the royal statistical society: series a (statistics in society)
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.103
H-Index - 84
eISSN - 1467-985X
pISSN - 0964-1998
DOI - 10.2307/2983507
Subject(s) - human immunodeficiency virus (hiv) , virology , medicine , biology
SUMMARY Knowledge of the number of human immunodeficiency virus (HIV) seropositive people in a country and of their distribution in different transmission categories is important for health care planning, for risk estimations and for targeting public health interventions. We used data on seroprevalence of antibody to HIV in defined subgroups of the population attained from unlinked anonymous HIV testing surveys and combined these with estimates of the respective sizes of these subgroups from an interview study on sexual attitudes and lifestyles. Seroprevalence in those subgroups not covered by the HIV testing surveys was derived from a series of assumptions and approximations. The estimated numbers of HIV seropositive subjects in each subgroup were added to give an overall total for England and Wales at the end of 1991. The overall estimate was 27180 HIV seropositive people, with lower and upper bounds of 20200 and 34300 respectively. Homosexual and bisexual men were estimated to constitute about half of the total number who were seropositive, and heterosexually infected people constituted approximately a quarter. The greatest numerical uncertainties applied to the number of HIV seropositive men who have had a male partner, but not in the last 5 years, and to the number of HIV seropositive heterosexual men and women who have no recognized risk behaviour. Overlap between the various subgroups defined by unlinked anonymous testing programmes is substantial, but could partly be adjusted for with data from the interview study. The estimates of this study are in the same range as those made earlier with this and other modelling strategies.