A comparison of AIDS-defining events and subsequent CDC stage IV events in IDUs and gay men

Survival by AIDS defining condition in rural Uganda

OBJECTIVES

To report the initial AIDS defining conditions, CD4 lymphocyte counts around the time of AIDS diagnosis, and survival by AIDS defining condition in a population based cohort in rural Uganda.

METHODS

Participants in an HIV natural history cohort in rural Uganda were reviewed every 3 months at routine visits and at other times when they were ill. The date and nature of the first AIDS defining condition in participants developing AIDS during follow up between the start of the cohort in 1990 and the end of 1998 were noted. The CD4 count at, or within, 3 months of this time was recorded for those participants who developed AIDS (WHO stage 4) after 1993.

RESULTS

The median survival from developing AIDS to death was 9.3 months and the median CD4 lymphocyte count around the time of developing AIDS was 150 cells x 10(6)/l. The most frequent AIDS defining conditions were wasting syndrome, oesophageal candidiasis, and mucocutaneous herpes simplex virus infection (HSV) for more than 1 month. The median survival with wasting syndrome, Kaposi's sarcoma, and oesophageal candidiasis was less than 3.5 months; however, survival with cryptosporidial diarrhoea, chronic HSV, and extrapulmonary tuberculosis was greater than 20 months. There was little relation between CD4 count around the time of development of the AIDS defining condition and the median survival with that condition.

CONCLUSIONS

The survival for most AIDS defining conditions was generally shorter and the median CD4 lymphocyte count higher than studies reported from developed countries. However, the conditions with the longest survival (cryptosporidial diarrhoea, chronic HSV, and extrapulmonary tuberculosis) had a similar survival to that in developed countries and these conditions have a high background level in this population.

Monitoring cytomegalovirus retinitis prevalence in an HIV-seropositive cohort: the assessment of improvements observed following the introduction of highly active antiretroviral triple therapy

This paper concerns the ophthalmic assessment of patients with acquired immunodeficiency syndrome (AIDS) for a number of eye conditions and in particular cytomegalovirus (CMV) retinitis. CMV has been the most common opportunistic infection associated with AIDS and the leading cause of blindness among AIDS patients. There have been early indications of a widespread fall in CMV prevalence internationally following the introduction of a new highly active antiretroviral triple (HAART) therapy. Our study sought to assess the position for Ireland. Our cohort was the entire population of stage IV AIDS patients attending the country's leading referral centre. The total number of patients examined was 167 and the period of examination was 1 May 1995 to 30 April 1997. HAART was introduced in March 1996, so the data permitted a 'before and after' comparison of various clinical findings. The incidence of new CMV cases was found to be 4 among the 102 patients examined in the first 12-month period and one among 107 patients examined in the second 12-month period. There were accompanying declines in HIV-related noninfectious retinal vasculopathy (HIVR), keratitis and other conditions. The findings are promising, but we argue that caution is needed in assessing long-term trends. In the paper we discuss a number of methodological issues in the collection and analysis of the clinical data and in the interpretation of results.

Epidemiology and outcome of HIV infection in North-East Scotland (1985-1997)

OBJECTIVE

to assess the epidemiology of HIV infection in North-East Scotland.

METHODS

retrospective casenote review of all HIV-infected patients who have had contact with the Infection Unit in Aberdeen.

RESULTS

one hundred and forty-two HIV-infected patients were treated between April 1985 and December 1997. The risk behaviour related to the acquisition of the HIV infection was: 56 (39%) homosexually infected, 45 (32%) heterosexually-infected, 34 (24%) injecting drug users (IDUs), and seven (5%) blood products or not known. Sixteen of the 45 (36%) heterosexually-infected patients were native to Africa and 16 of the 34 (31%) IDUs were prisoners in Peterhead prison at the time of referral. Fifty-two (37%) of the cohort continue to attend the Infection Unit, 41 (29%) have relocated, 40 (28%) have died and nine (6%) have been lost to follow-up. The ratio of heterosexual:homosexual men:IDUs changed significantly between the first 7 years (12:21:25) and the second 6 years (33:35:9) of the review, with significantly more patients being infected through heterosexual contact and fewer infected by IDU in the second period-P<0.001. The median AIDS survival was 17 months. Survival was significantly longer in those patients who took anti-retroviral therapy (median = 20 months) than in the patients who opted not to take anti-retroviral therapy (median = 11 months)-P<0.01.

CONCLUSIONS

Although homosexual contact represents the commonest risk group for HIV infection in this region, the number of heterosexually-infected patients has increased significantly in the last 5 years. Temporary residents account for one-third of the HIV-infected population cared for in NE Scotland. Almost half of those lost to follow-up have returned to Africa or been released from prison. The introduction of anti-retroviral therapy has resulted in a dramatic improvement in AIDS survival in our cohort as it has done elsewhere.

Statistical assessment of ordinal outcomes in comparative studies

Ordinal regression is a relatively new statistical method developed for analyzing ranked outcomes. In the past, ranked scales have often been analyzed without making full use of the ordinality of the data or, alternatively, by assigning arbitrary numerical scores to the ranks. While ordinal regression models are now available to make full use of ranked data, they are not used widely. This article, directed to clinical researchers and epidemiologists, provides a description of the properties of these methods. Using ordinal measures of back pain in a follow-up study of adolescent idiopathic scoliosis, we illustrate the advantages of those methods and describe how to interpret the estimated parameters. Comparisons with binary logistic regression are made to show why a single dichotomization of the ordinal scale may lead to incorrect inferences. Two ordinal models (the proportional odds and the continuation ratio models) are discussed, and the goodness-of-fit of these models is examined. We conclude that ordinal regression is a tool that is powerful, simple to use, and produces an interpretable parameter that summarizes the effect between groups over all levels of the outcome.