While HAART has led to dramatic decreases in illness and death of patients with HIV, it does so at a price: the drug therapy averages $10,000-$15,000 a year for a single patient. To gauge whether HAART was cost-effective in treating HIV-infected patients, researchers previously relied on modeled estimates of health care expenditures. Now, using actual healthcare utilization data, researchers at the University of Alabama at Birmingham (UAB) have analyzed a year's worth of total health care expenditures for 635 HIV-infected patients, including HAART and non-HAART medications, hospitalization, outpatient clinic visits, and more.
According to the study, as long as a patient sticks to HAART treatment, "the happy paradox is that paying high costs for antiretrovirals decreases overall cost of care," said senior author Michael Saag, MD. The HAART costs were about the same for patients who were sicker ?that is, who had lower CD4 cell counts ?as for healthier patients with higher CD4 cell counts. But the sicker patients ran up much higher bills for other care: Expenditures were almost six times greater for hospitalizations, and almost eight times greater for non-HAART medications, for the sicker patients than for the healthier ones.
On average, the sicker HIV-infected patients have total annual health care expenditures 2.5 times higher than healthier ones -- about $36,500 a year compared with about $14,000 a year, the study found.
"This study confirms the remarkable cost effectiveness of HAART," said Dr. Saag. But it also underscores a grave threat to HIV treatment across the board: the inadequacy of reimbursement to the treating physicians and clinics.
In the study, medications were the single costliest component, comprising $71 to $84 of every $100 spent, regardless of stage of disease. In sharp contrast, expenditures for physicians' services and clinic fees accounted for less than $2 of every $100, or on average, $360 per patient per year. This is far less than the providers' actual costs, which average approximately $1,500 to 2,000 per patient per year. The study assumed all patients had Medicare health insurance and the clinics were reimbursed 100 percent. "However, the majority of people who are infected with HIV are either uninsured or underinsured" and unable to pay for their care, said Dr. Saag. "So in reality, the actual reimbursement to clinics is markedly less than $360 per patient per year, making the situation even more dire," he added.
As a consequence, said Dr. Saag, "We are approaching a potential crisis in the availability of physicians to provide care for patients? Many private-practice HIV clinics are closing. Most academic institutions within the United States, including ours, are absorbing the cost of care for HIV patients" ?a burden that can range into the millions of dollars, he said. These treatment providers' necessary reliance on federal assistance "underscores the tremendous need for reauthorization of the Ryan White CARE Act with increased resources allocated for reimbursement of medical care, which is now under discussion in Congress," Dr. Saag said.
Although institutions are developing ways to bring down HIV treatment costs, they won't be able to cover those costs themselves indefinitely, Dr. Saag warned. Notwithstanding the promising findings on HAART's cost-effectiveness, "Providing drugs alone is not enough," he said. "We must focus now on creating appropriate incentives for providers to care for HIV patients. The current pool of providers is struggling to make ends meet. As they begin to retire, where will the new wave of physicians come from to replace them?"
A commentary accompanying the HIV expenditure article, by Kenneth Mayer, MD, and Sreekanth Chaguturu, MD, of Miriam Hospital and Brown University, says the article "demonstrates some good news and some worrisome findings about the current state of HIV clinical care." Learning that people with HIV can significantly reduce their treatment costs by getting healthier is encouraging; but having the funding to provide the treatment in the first place is the problem.