Patients are presenting later for HIV testing and treatment today than they were in 1990, according to the results of a study from a large HIV clinic in Baltimore, Maryland, published in the November 15th edition of the journal, Clinical Infectious Diseases. An accompanying editorial calls for universal access to healthcare in the United States alongside universal HIV testing, noting that “despite significant advances in antiretroviral therapy and improved services for persons living with HIV infection, our health care system appears to be less capable of detecting the disease early in its course than it was nearly two decades ago.”
Study background
Despite more than two decades of HIV testing and treatment campaigns funded by the US government – notably the 1990 Ryan White Comprehensive AIDS Resources Emergency (CARE) Act and the 1998 the Minority HIV/AIDS Initiative (MAI) – about one-in-four of the 1.1 million people estimated to be living with HIV remain unaware of their HIV infection.
Consequently, a considerable number of people only discover that they are HIV-positive when they are hospitalised with a serious HIV-related illness, and this late diagnosis results in an increased risk of death in the short-term. Previous studies have found that Black or Hispanic men, and older people are the most likely to be diagnosed late.
Given the efforts to improve access to care since 1990, coupled with the major improvements that have occurred in HIV therapy since 1996, investigators at Johns Hopkins University School of Medicine, Baltimore – the largest HIV care provider in the state of Maryland – wanted to assess whether patients had presented for HIV care earlier in the course of their HIV infection over time.
The Johns Hopkins HIV Clinical Cohort is a longitudinal observational study established in 1990. All patients attending the main HIV clinic at Johns Hopkins and six others operated by Johns Hopkins throughout Maryland are enrolled in the Johns Hopkins HIV Clinical Cohort on the first HIV clinic visit. A total of 3348 antiretroviral naive patients were enrolled in the cohort from January 1990 until June 2006, and were included in the analysis.
Patients were stratified into four groups based on the year of enrollment: 1990–1994, 1995–1998, 1999–2002, and 2003–2006. The investigators then analysed the cohort as a whole – which was predominately male and of black ethnicity – for the median absolute CD4 cell count at presentation and the number of days between diagnosis of HIV-positive status (based on patient self-report) and presentation at the clinic.
Later presentation seen over time
Surprisingly, the investigators found that patients were attending for care at a later stage in their HIV disease over time, and saw a significant decrease in the median CD4 cell count between 1990-1994 and 2003-2006. (p
The type of transmission risk also changed over time, with a significant decrease in the proportion of men who have sex with men (MSM) and patients reporting IDU, and an almost threefold increase in patients reporting heterosexual sex as a risk factor during 2003–2006 compared with 1990-1994 (p
Separate analyses for gender, ethnicity, and transmission risk group found that median CD4 cell counts at presentation significantly decreased over time for women, men, black patients, and patients who reported IDU and heterosexual exposures as HIV risk factors.
In fact, MSM were the only risk group whose median CD4 cell counts increased over time, from 280 cells/mm3 in 1990–1994, to 332 cells/ mm3 in 2003–2006. MSM also had the shortest time interval between HIV diagnosis and presentation at clinic, with a significant decrease between 1990 and1994 (215 days) and 2003–2006 (87 days; p
Multivariate analysis found that men, and individuals of black ethnicity, were more likely to have lower CD4 cell counts than the cohort as whole throughout all time periods (p
It also confirmed that the three time periods following 1990-1994 were significantly associated with a lower presenting CD4 cell count (p
The investigators note that their findings contradict their initial hypothesis that, with a focus on increased testing and better outcomes from HIV treatment and care post-1996, “patients would present earlier in the course of disease. Instead, our data indicate that patients are presenting later for care than in earlier years, with lower CD4+ cell counts, a small increase in those who have AIDS, and no improvement in the time between HIV diagnosis and presentation for care for women or those with heterosexual transmission.”
The authors conclude, “during the past 16 years in Maryland, there has been no improvement in getting patients into care earlier in the course of their disease... These data provide compelling evidence that implementation of new strategies for HIV testing and entry to care, recommended by the CDC, are urgently needed.”
Can these results be generalised to the entire US?
In an accompanying editorial, Doctors Miguel Goicoechea and Davey Smith of the Department of Medicine, University of California, San Diego, suggest that the decrease in presenting CD4 counts over time are “probably not” being driven by changes in patient demographics (i.e. proportionally fewer MSM and greater heterosexuals).
They note that “although MSM were the only subjects to have improved CD4+ T-cell counts over time, these patients initially had the most advanced HIV disease at presentation, with only a modest gain of 52 cells/mm3 by 2003–2006. Conversely, all other groups had much higher presenting CD4+ T-cell counts in 1990–1994, with large decreases over time, ranging from 76 to 174 cells/mm3.”
The study’s authors say that they “suspect” that their findings may “generalise to other locations in the United States or to other industrialized countries” but add, “our results are most likely to generalize to HIV-infected populations in which the demographic characteristics are most similar.”
Goicoechea and Smith suggest that the Johns Hopkins patient population bucks increased HIV testing trends seen throughout the US in that it “contains mainly underinsured persons who likely lacked general health care before their diagnosis of HIV infection.” They note that most HIV testing takes place within the private health care sector, which can only be accessed by individuals with health insurance. “Potentially,” they suggest, “the Johns Hopkins cohort today contains fewer insured patients and serves a more disenfranchised population, with fewer opportunities for HIV testing, than was the case in the early 1990s.”
“Because underinsured persons most often gain access to the healthcare system through public health clinics, emergency departments, and urgent care facilities,” they continue, “and these settings often have lower HIV testing rates than do private physician offices and managed care organisations, it is plausible that the Johns Hopkins HIV patient population is receiving diagnoses at a later stage simply because of the health care systems that are available to them.”
Is universal HIV testing enough?
Goicoechea and Smith agree with the study's authors that “these data support the argument for mainstream HIV testing,” but add that, “they also highlight the issue of universal health care coverage.”
“HIV disease is a disease of poverty," they note. "In the United States, HIV infection disproportionately affects uninsured, low-income persons, who constitute a vulnerable population that often has multiple health care needs.”
They conclude by arguing that, "‘universal’ HIV testing also requires ‘universal’ health care for there to be a significant impact on diagnosing HIV infection at the earliest stage possible. As the United States and other resource-wealthy countries move forward to build health care infrastructure and scale-up antiretroviral therapy in resource-limited settings, it is a shameful commentary on our own health care system that the average CD4+ T-cell count before the initiation of antiretroviral therapy in North America is similar to that of some underdeveloped countries in Africa.”
Keruly JC and Moore RD. Immune Status at Presentation to Care Did Not Improve among Antiretroviral-Naive Persons from 1990 to 2006. Clin Inf Dis 45 (10): epub 11th Oct, 2007.
Goicoechea M and Smith DM. Universal HIV Testing: Is It Enough? Clin Inf Dis 45 (10): epub 11th Oct, 2007.