AIDS Underestimated by 50%, Study Says : CDC Disputes Data Putting Number Infected as High as 3 Million
Twice as many Americans may be infected with the AIDS virus as the government estimates, and the virus may be much more prevalent among heterosexuals than is widely believed, researchers reported Friday.
A “best guess” at the number of AIDS-infected Americans at the end of last year was between 1.9 million and 3 million, with the most probable range being between 2.2 million and 2.6 million people, according to a study by the Hudson Institute, a nonprofit research and public policy center based in Indianapolis.
CDC Sticks to Estimates
But officials at the federal Centers for Disease Control, the government’s primary monitor of the AIDS epidemic, stood by CDC estimates that, at the end of 1987, between 900,000 and 1.4 million Americans were infected with human immunodeficiency virus (HIV), which causes acquired immune deficiency syndrome.
“The Hudson study makes a number of assumptions that can’t be justified,” said W. Meade Morgan, chief of the CDC statistics branch, adding that the results are “very inconsistent” with data that has been widely accepted.
Both Hudson and the CDC conceded that, because of a lack of precise data about AIDS, it is impossible to make any estimates with a high degree of confidence. However, both said that rough estimates indicate the general severity of the nationwide epidemic and are necessary for establishing policy to deal with the growing problem of AIDS.
Faster Spread Seen
Hudson officials, speaking at a news conference, said the institute’s study suggests that AIDS may be spreading much faster than had been suspected and may have much more sweeping policy implications.
“If there is a larger pool of infected people, then there are important new implications for national health policy--not only in terms of the strain on medical treatment resources but also in terms of the potential spread of the disease in coming years,” Hudson Research Director George A. Keyworth said. “And there are equally important implications for individual behavior in the face of this increased hazard.”
Hudson officials said that perhaps the most alarming conclusion of their study is that the virus appears to be much more widespread among heterosexuals than commonly believed--especially those who do not use intravenous drugs.
Many heterosexuals, they said, mistakenly believe they are not susceptible to AIDS and have become complacent in recent months about taking precautions against contagion.
Possibility of ‘Breakout’
“A breakout of the disease into the non-monogamous heterosexual population is more than a theoretical possibility,” Kevin R. Hopkins, director of Hudson’s AIDS research, said. “According to our study, such a breakout into the mainstream heterosexual population is unavoidable unless dramatic behavioral change or medical progress takes place soon.”
The report said that between 850,000 and 1.4 million heterosexuals are infected, including about 330,000 who did not contract the virus through drug use. The government estimates that about 80,000 to 160,000 non-drug-using heterosexuals carry the virus.
Gay men and intravenous drug users are thought to run the highest risk of contracting the virus, which is spread through sexual intercourse and blood-to-blood contact, such as the sharing of contaminated needles.
CDC calculations have been evaluated and confirmed by several outside academic researchers, including mathematical modeling experts from Harvard University, the Massachusetts Institute of Technology and Johns Hopkins University, who found results consistent with CDC’s and much lower than those of the Hudson Institute, said Timothy Dondero, epidemiologist in the AIDS program at CDC.
Survey Methods Differ
The estimates by CDC and the conservative Hudson Institute emerged from different survey methods. CDC relied largely on surveys that tried to measure the percentage of people infected with the virus in different population subgroups, statistician Morgan said. The subtotals were added to derive a national total, he said.
But Hopkins, calling CDC methods “very uncertain,” said that Hudson relied on more sophisticated methodology using more recent data. He said the study used a computer model that took into account changing behavior patterns among high-risk groups, under-reporting and delays in reporting of diagnosed AIDS cases and new information about the rate of progression from HIV to fully developed AIDS.
Methodology Questioned
However, CDC officials said that Hudson made too many questionable assumptions, leading to results sharply at odds with data from studies of closely observable segments of the population.
“If the Hudson figures were accurate, it would mean that 1% of the entire U.S. population had to be infected,” Dondero said.
“Since infection is concentrated in young adults, who make up about half the total population,” he said, “it means you would have to have about 2% of the entire American young adult population infected--and where is the evidence of anything like that? The data that we have already gathered in surveys of child-bearing women, anonymous hospital testing, prisoners, military recruits and blood donors is well below 1%.”
Staff writer Marlene Cimons contributed to this story.
