As noted in Part A of this series, Goldman reported on a paper by clinical psychiatrist Walt Oteds who had referred to the work of HIV/AIDS researchers and epidemiologists most of whom worked with the CDC.
At that time they were looking forward from the vatange point of 1994; for now let's look backward to the period 1994-2005.
While epidemiologists use very sophisticated statistical modelling for reviewing the history of epidemics, and for projecting estimates of the spread of disease, we can also do a very simple back-of-envelope review of the CDC's historical data on the HIV/AIDS epedimic for the period leading up to 2005. We need data broken down by age and category of transmission; and we need to use basic assumptions from sero-surveys and the like. This will give readers a general idea of the available evidence and it will provide a test of the plausibity of the projections cited earlier.
Please note that the epidemiology of HIV/AIDS is documented in peer-reviewed research and this is just a basic simulation of a much more sophisticated process.
Here are a couple of ways to look back and compare with the projection that was made in 1994.
The 1994 projection:
Epidemiologists estimate that 30 percent of all 20-year-old homosexually-active men will be HIV positive or dead of AIDS by the time they are 30.
E. Goldman, “Psychological Factors Generate HIV Resurgence in Young Gay Men,” Clinical Psychiatry News, Oct. 1994.
Scenario 1.
Based on 10-year average of 13,000 diagnosed infections per year.
Range of estimates of new HIV infection. By 1995, there was a wide range of estimates for projecting HIV incidence -- as high as 90,000 per year for all cases. However, modest estimates used a mid-range such as 10,000 infections per year among homosexually-active men (MSM); and another 3,000 per year among homosexually-active men who had used injection drugs (MSM & IDU).
Tracking 20 year-olds. Looking back on 1985-2004, the CDC reported that .05% of all HIV incidences in 1985-1995 were among the under 20 age group, and 13.2% among the 20-29 age group in 1996-2000, and 12% among that group in 2001-2004. This covers the ten year period during which men who were 20 years old in 1995 became 30 years old in 2005.
Undiagnosed cases. Assuming that the homosexually-active subpopulation is 4% the general male population, there were 80,000 homosexually active men of that age in 1995. Among these men, thousands became infected with HIV/AIDS. The CDC has estimated that diagnosed cases represented about 75% of all infections. So thousands more new HIV infections went unreported.
Estimating HIV infection by age 30. Now, using these parameters, we can estimate the number of new HIV infections among homosexually active men and then calculate the infected share of the homosexual men who reached 30 years old by 2005.
Accordingly, in 1995 only a few 20-year-olds had been infected; but in 1996-2000 about 8,600 of that age cohort had been infected; and another 6,200 by 2004. The accumulated total would have been about 14,800 which represented 75% of all infections. This means that by 2005 about 20,000 homosexually-active men had been infected by age 30 years (diagnosed and undiagnosed).
That would be about 25% of the homosexual male population in that age group.
Scenario 2.
Based on counting the newly diganosed infections by interim periods.
The subpopulation size. In 1995 there were 2 million men aged 20 years in the general population. Of these, about 80 thousand (4%) were homosexually-active. By the end of the following decade, thousands of these young homosexually-active men were living with or were dead due to HIV/AIDS.
Proportion of new cases by age. In its 25th year review of the epidemic, the CDC reported that 0.5% of new HIV infections had occured to the under 20 age group by 1995, 13% to the 20-29 age group by 2000, and 12% to the 20-29 group by 2004. The cohort that was 20 in 1995 passed through the decade and accumulated thousands of new diagnosed cases of HIV.
Number of new cases by transmission. Infections transmitted by men who had sex with men (MSM) and MSM who used injected drugs (MSM & IDU) are referred to here as MSM infections. The CDC estimated 325,200 new MSM infections in 1985-1995, another 106,200 in 1996-2000, and another 74,735 in 2001-2004. About 72% of all new infections among men were MSM in 1985-1995 and 60% in 1996-2000 and in 2001-2004.
Estimated number and proportion of homosexually-active men who, by the year 2005, had been infected by age 30. At age 20 about 1,600 had already been infected; five years later another 14,000; and five years later another 9,000 newly infected. About 24,600 HIV infections had accumulated among those who became 30 years old in 2005.
Expressed as a percentage: Of the 80,000 homosexually active men who were 20 years old in 1995, about 24,600 (30%) had been infected with HIV/AIDS by age 30.
Summary.
Of course, epidemiologists use much more sophisticated modelling to compare projections to the history of epidemics; and, also, to construct and to refine models to forecast the near and distant future spread of disease. However, given the CDC's count of diagnosed infections, and the assumptions regarding undiagnosed infections and the size of the MSM subpopulation, it is plausible that the 1994 projection regarding twenty-year-olds was realized.
In 2006 there were over 1 million infected people in the US; and yet, even with all the awareness programs and the identification of risk factors and improvements in testing, that number must be read with the caution that probably another 300 thousand people do not know they are already infected.
Here is a quote from the CDC's review of HIV/AIDS epidemiology 25 years after the outbreak:
"Analysis of data collected by the National HIV Behavioral Surveillance System, which surveys populations at high risk for HIV to assess prevalence and trends in risk behavior, HIV testing, and use of prevention services, revealed that of MSM surveyed in five U.S. cities, 25% were infected with HIV and of those, 48% were unaware of their infection."The cited study's large sample of homosexually-active men has a median age of 32 years; 76% of those already infected were over the age of 30 years; the highest proportion of unrecognized infection (75%) was among those younger than 30.
The CDC's data for 2006 showed that in that year alone almost 40% (11,000) of newly infected homosexually-active men were under the age of thirty. Assuming that represents only 75% of all infections, as per the CDC, then in 2006 there were almost 14,500 new infections -- diagnosed plus undiagnosed -- in the under-30 age group.
The basis for the back-of-envelope scenario #1 above, is that there would be about 13,000 new infections per year, on average, amongs MSM of all ages during 1995-2005. So if the average climbs closer toward 14,5000, then, it could be that more than 30% of twenty-year-olds will be infected by age 30.
How the data might break-down by age group during the next ten-twenty years, time will tell. At some point a saturation of the subpopulation is reached and growth is slowed to the rate at which young people who enter the sexually active population become infected.
For now we rely on the epidemiological modelling that produces the most plausible projections. We can look back in ten years and compare to see how well the prjection did.
References:
Twenty-Five Years of HIV/AIDS - United States, 1981-2006 [PDF]
TABLE. Estimated numbers and percentages of HIV/AIDS and AIDS cases, by year of diagnosis and selected characteristics - United States, 1981-2004.
Figures 1, 2, and 3, page 591.
Department of health and human services, Centers for Disease Control and Prevention.
Morbidity and Mortality Weekly Report. June 2, 2006 / Vol. 55 / No. 21
Glynn M, Rhodes P. What is really happening with HIV trends in the United States? Modeling the national epidemic. In: Proceedings of National HIV Prevention Conference, Atlanta, GA, June 12-15, 2005.
CDC. Trends in HIV/AIDS diagnoses-33 states, 2001-2004. MMWR 2005;54:1149-53.
CDC. HIV prevalence, unrecognized infection, and HIV testing among men who have sex with men-five U.S. cities, June 2004-April 2005. MMWR 2005;54:597-601.
Table: HIV Prevalence in 2 Probability Telephone Samples of Men Who Have Sex With Men: San Francisco, 1997 and 2002.
Return to Table of Contents.
