Suppression of the plasma HIV RNA level is a key goal of medical care for persons living with HIV/AIDS (PLWHA) and is increasingly a focus of public health efforts to prevent HIV transmission.1 National estimates suggest that only 19%–28% of PLWHA in the United States have achieved viral suppression2–5 due to compounding breakdowns at each step in the HIV care cascade. However, such estimates are imprecise, reflecting incomplete nationwide surveillance data on viral suppression and the inability of surveillance systems to ascertain antiretroviral therapy (ART) use. Although public health authorities can estimate levels of viral suppression using laboratory reporting of HIV RNA (viral load [VL]) test results, these estimates are limited in many areas by incomplete reporting and by uncertainty about which persons determined to be HIV-infected in a surveillance area continue to reside in that area.6 The Medical Monitoring Project (MMP) of the Centers for Disease Control and Prevention (CDC) estimates ART use with probability-based sampling of HIV care facilities and patients in care during the first 4 months of the year, but to date, MMP data have been limited by incomplete provider and patient participation.7
In addition to the importance of contemporary data on ART use and viral suppression, a clearer understanding of how ART uptake varies among PLWHA in different nadir CD4 count groups is important for monitoring the effect of changes in treatment guidelines and the impact of reductions in state AIDS Drug Assistance Programs, identifying disparities in treatment access, and projecting resources required to expand ART use among PLWHA in the United States. Little is known about the size and composition of the population of persons in HIV care who have nadir CD4 counts >500 cells per cubic millimeter, a group of particular interest in the context of revised guidelines for ART initiation and increasing optimism about the potential for “test and treat” strategies to decrease the size of the HIV epidemic. Furthermore, contemporary estimates of ART use and viral suppression among persons in HIV care can aid studies to assess the impact of expanded health insurance coverage among PLWHA in the United States under the Affordable Care Act.
The primary goal of this study was to estimate the proportion of persons who have initiated ART, are on continuing ART, and have viral suppression in a nationally distributed cohort of persons in HIV care in the United States Our secondary objectives were to compare ART use and viral suppression by nadir CD4 count group and to determine factors associated with being ART naive, discontinuing ART, and/or having detectable viremia despite continuing ART use, stratified by nadir CD4 count.
We conducted a cross-sectional study of patients in the Centers for AIDS Research (CFAR) Network of Integrated Clinical Systems (CNICS) cohort, which includes more than 25,000 HIV-infected adults in care dating back to 1995 at 8 university-affiliated CFARs. The methodology and characteristics of the CNICS cohort are described in detail elsewhere.8 Briefly, CNICS is a dynamic observational cohort with approximately 1400 new patients enrolling and 13% of existing patients leaving care each year. The CNICS data repository systematically capture information from electronic health records and pharmacy systems at each CNICS site. Validation of data occurs in 3 phases: at the individual sites before data transmission, at the time of data submission to the Data Management Core before insertion into the central repository, and through the use of automated validation procedures at the time of data integration. Data then undergo an extensive set of quality assurance procedures and data quality issues are reported to CNICS sites to investigate and correct.
Our objective was to assess historical and current ART use among patients in care at the end of 2010, 1 year after the publication of HIV treatment guidelines recommending ART initiation before progression of the CD4 count to <500 cells per cubic millimeter. The study population included patients who were in care (≥1 medical visit) and had ≥1 VL measurement in 2010. To focus on a population in which clinicians and patients had sufficient time to carry out treatment decisions in response to CD4 counts <500 cells per cubic millimeter, we limited the study population to patients who entered the cohort ≥9 months before the end of 2010. We excluded patients who entered care before the combination ART era, defined as before January 1, 1997 (N = 1311); women who appeared to be receiving ART for perinatal treatment at the time of their most recent regimen (N = 20); and patients from 1 CNICS clinic site due to incomplete data (N = 914).
We analyzed the proportion of patients ever on ART, on continuing ART, and with viral suppression at the end of 2010. We classified subjects as “ever on ART” if they had initiated ART on or before December 31, 2010 and as “on continuing ART” if they had initiated ART and had no documentation of discontinuing all ART as of December 31, 2010. The latter group included patients who discontinued then restarted ART as long as they continued a regimen through the end of 2010. ART initiation and cessation dates defined by the treating provider in the electronic health record and in pharmacy prescription fill/refill data were validated at the individual sites and centrally through chart review and applications to monitor data quality. During our analysis, the Data Management Core sent lists of patients who had undetectable VLs but no corresponding ART data to each CNICS site for investigation. The sites responded with corrected ART data or confirmation of elite controller status for each case.
For analysis of viral suppression, we identified the last recorded VL for each patient and defined suppression as any result below the level of quantitation of the assay. We focused on undetectable VL as the primary outcome for viral suppression because it is the primary clinical goal of ART and the Institute of Medicine (IOM) recommended indicator.9 We separately report the proportion of patients with VL ≤200 copies per milliliter to facilitate comparison with other reports.4,5,10 We did not exclude patients who had viral suppression in the absence of ART initiation from this analysis because our goal was to assess viral suppression in the entire study population, regardless of the underlying mechanism of suppression.
We conducted a subset analysis stratifying each outcome (ever on ART, continuing ART, and viral suppression) by nadir CD4 count. We grouped subjects by nadir CD4 count ≤350, 351–500, or >500 cells per cubic millimeter based on the relevance of these categories to US HIV treatment guidelines for ART initiation. We limited analysis of nadir CD4 count to patients who were ART naive at the time of entry into the CNICS cohort because pretreatment CD4 counts are not uniformly available in patients who were treatment experienced at entry. We identified the nadir CD4 count for each patient as of March 31, 2010. We conducted a repeat cross-sectional analysis to compare patterns of ART use and viral suppression in 2009 and 2010, using the same methods as those described above to assess the outcomes at the end of 2009 among patients who had entered the cohort before March 31, 2009.
For bivariate comparisons of ART use and viral suppression by patient subpopulation, we used Pearson chi-square tests. We selected independent variables a priori, based on previous demonstrations of association with ART initiation, adherence, and/or viral suppression,11–14 including age, sex, race/ethnicity, HIV risk group, engagement in HIV care, psychiatric disorders, and substance abuse. We categorized HIV risk group as men who have sex with men (MSM) (including injection drug using MSM), injection drug user (IDU), non-IDU heterosexual, and other. We defined engagement in continuous care as ≥2 visits ≥3 months apart in 2010, the measure recommended by IOM9 and the US National HIV/AIDS Strategy.15 We analyzed psychiatric and substance use disorders diagnosed by treating providers. These data are not available uniformly from all CNICS sites, and we limited analysis of these variables to patients from reporting sites. We used the methodology validated by Tegger et al16 to categorize psychiatric comorbidities and substance use into separate, mutually exclusive, hierarchical groups: (1) psychotic disorder, bipolar disorder, and/or personality disorders with or without depression and anxiety; (2) depression and/or anxiety only; and (3) no mental illness. We categorized substance use into 5 mutually exclusive, hierarchical groups: (1) opiates with or without other drugs, (2) amphetamines with or without other drugs, (3) cocaine with or without alcohol, (4) alcohol only, or (5) no substance use.16 Patients at CNICS sites that report these data who did not have a diagnosis were classified as having “none of the above” diagnoses. We were unable to distinguish treated from untreated psychiatric disorders or active from past substance use.
We used Poisson regression clustered by clinic site for multivariate analysis of 3 outcome measures, stratified by nadir CD4 count. For these models, we defined the outcomes as “failure” isolated to individual steps in the care cascade: ART noninitiation, ART discontinuation, and detectable viremia despite continuing ART use. Poisson regression allows for calculation of estimated relative risks instead of odds ratios in cohort data.17 We estimated prevalence ratios (PR) of each outcome at the end of 2010. Clustered analysis adjusts for correlation between results from individual sites, thereby accounting for unmeasured factors associated with potentially different ART prescribing practices. We constructed a model for each of the 3 outcome measures for each of the nadir CD4 groups. Our initial models included all the independent variables we defined a priori, except for psychiatric and substance abuse because not all sites reported this information, and we retained all variables in the final model. We then explored the potential effect of length of time in care on ART initiation, and because this was significantly associated with ART initiation in 1 nadir CD4 group, the final models include adjustment for years in care at a CNICS site. We used Stata 10.1 for all analyses (StataCorp, College Station, TX). Institutional review boards (IRB) at each university approved data collection procedures, and the University of Washington IRB judged this analysis to be exempt from review due to the use of de-identified data.
ART Use and Outcomes Among Patients in Care in 2010
The study population included 8633 persons from 7 CNICS sites (range, 720–1735 patients per site) who met inclusion criteria for active HIV medical care in 2010 (Fig. 1). As of December 31, 2010, these patients had been in care at CNICS sites for a median of 60 months (interquartile range [IQR], 30–100 months). The inclusion criteria of ≥1 visit and ≥1 VL in 2010 excluded 1312 persons (13%) compared with the population of patients with ≥1 visit and ≥1 VL anytime in either 2009 or 2010 (N = 9945).
At the end of 2010, 94% of patients had ever been on ART (range, 92%–95% across sites), 89% were on continuing ART (range, 88%–91%), and 70% had viral suppression (range, 56%–85%) (Table 1). By the less restrictive definition of VL ≤200 copies per milliliter, 79% (range, 71%–86%) had viral suppression. Restricting the analysis to patients who had been on ART ≥6 months, the prevalence of viral suppression was similar in patients on ART <1 year (76%) compared with >1 year (78%; P = 0.45). The distribution of the 555 patients who remained ART naive at the end of 2010 by nadir CD4 count was as follows: 102 (18%) ≤350 cells per cubic millimeter, 177 (32%) 351–500 cells per cubic millimeter, and 276 (50%) >500 cells per cubic millimeter. Approximately 1% of the study population (N = 70) had undetectable VL in the absence of documented ART initiation, of whom 16 (23%) had a nadir CD4 count ≤350 cells per cubic millimeter, 18 (26%) 351–500 cells per cubic millimeter, and 36 (51%) >500 cells per cubic millimeter.
In bivariate analysis, viral suppression was less common among women compared with men (P < 0.001); patients <50 compared with >50 years (P < 0.001 for all comparisons); non-Hispanic blacks compared with non-Hispanic whites (P < 0.001); and patients not engaged compared with those engaged in continuous care (P < 0.001). Viral suppression was more common in MSM than heterosexuals and IDU (P < 0.001). Comparing patient subgroups, differences in the proportion of patients who had initiated ART were generally smaller than differences in the proportion of patients with viral suppression. For example, 92% of non-Hispanic blacks and 94% of non-Hispanic whites had initiated ART, but only 62% of non-Hispanic blacks had viral suppression compared with 73% of non-Hispanic whites.
ART Use and Outcomes in Patients Who Were ART Naive at the Time of Entry Into Care
Table 2 shows ART status and viral suppression, stratified by nadir CD4 count, in the 4637 patients who were ART naive at the time of entry into care at a CNICS site. These patients had been in care at CNICS sites for a median of 60 months (IQR, 31–99). Initial CD4 counts after entry to care were ≤350 cells per cubic millimeter in 49%, 350–500 cells per cubic millimeter in 22%, and >500 cells per cubic millimeter in 28% of patients. At the end of 2010, 88% of patients had initiated ART, 85% were on continuing ART, and 67% had viral suppression (76% had VL ≤ 200 copies/mL). More patients with nadir CD4 counts ≤350 cells per cubic millimeter had initiated ART and were on continuing ART (97% and 94%, respectively) compared with patients with nadir CD4 counts 351–500 cells per cubic millimeter (73% and 70%, respectively) and >500 cells per cubic millimeter (38% and 36%, respectively) (P < 0.001 for all comparisons). Similarly, the proportion of patients with viral suppression, on or off ART, varied by nadir CD4 group (≤350 cells per cubic millimeter, 73%; 351–500 cells per cubic millimeter, 57%; >500 cells per cubic millimeter, 33%; P < 0.001). The association of lower nadir CD4 count with higher levels of ART use and viral suppression was evident across all subgroups presented in Table 2. Among patients on ART for ≥6 months, there was a trend toward higher rates of suppression in patients who initiated ART with higher CD4 counts (≤350 cells per cubic millimeter, 79%; 351–500 cells per cubic millimeter, 82%; >500 cells per cubic millimeter, 86% ; P = 0.095).
Of 4265 patients who met inclusion criteria for analysis of ART initiation at the end of 2009, 85% had ever been on ART, 80% were on continuing ART, and 65% had viral suppression. The proportions of patients who had ever initiated ART, were on ART and with viral suppression were higher among those with nadir CD4 counts of 351–500 and >500 cells per cubic millimeter in 2010 compared with 2009 (Table 2).
Engagement in Continuous Care
Overall, approximately 16% of the entire study population (N = 8633) and of the patients who were ART naive at cohort entry (N = 4637) were not engaged in continuous care. Among those ART naive at cohort entry, nonengagement was more common among those with higher nadir CD4 counts (≤350 cells per cubic millimeter, 16%; 351–500 cells per cubic millimeter, 18%; >500 cells per cubic millimeter, 21%; P = 0.02). Adjusting for nadir CD4 count, engagement in care was the single factor most strongly and consistently associated with ART use and viral suppression. Younger age was associated with poorer care engagement (age, 18–29: PR 0.94, 95% confidence interval [CI], 0.91 to 0.97; age, 30–39: PR 0.96, 95% CI, 0.94 to 0.97 compared with patients >50 years), as was non-Hispanic black race (PR 0.97, 95% CI, 0.95 to 0.996), controlling for nadir CD4 count.
Factors Associated With Failure to Initiate ART, Continue ART, and Achieve Viral Suppression
Tables 3, 4, and Table S1 (see Supplemental Digital Content, http://links.lww.com/QAI/A421) present the results of multivariate analyses assessing the association of independent variables with ART noninitiation, ART discontinuation among patients who initiated ART, and detectable viremia in patients on continuing ART, stratified by nadir CD4 count, among patients who were ART naive at the time of entry to care (N = 4637). Failure to engage in continuous care was consistently associated with viremia across all nadir CD4 count groups and with lower levels of ART initiation and ongoing use in patients with nadir CD4 counts ≤500 cells per cubic millimeter. Otherwise, associations were heterogeneous across nadir CD4 groups. In the population with the strongest indication for treatment (CD4 counts <350 cells/mm3), failure at each step was generally higher among young patients and African Americans, though not all PR associating these factors with failures were significant.
We found that, among patients in HIV care at 7 CNICS sites during 2010, 94% had ever initiated ART, 89% were on continuing ART, and 70% had viral suppression at the end of 2010. Treatment initiation increased from 2009 to 2010 in patients with a nadir CD4 count >350 cells per cubic millimeter, including those with a nadir CD4 count >500 cells per cubic millimeter. This demonstrates a movement toward earlier ART initiation in practice likely due, in part, to revised treatment guidelines in 2009. ART-naive patients with a nadir CD4 count >500 cells per cubic millimeter accounted for just 3% of the total population, indicating that the most recent revision of US HIV/AIDS treatment guidelines recommending ART initiation regardless of CD4 count likely only slightly enlarged the number of patients for whom ART initiation is recommended.
Our findings are broadly consistent with those of other recent studies of ART use and viral suppression among persons in HIV care in the United States. An analysis from the CDC MMP found that, among a nationwide cohort of persons, 89% were prescribed ART in the past 12 months and 71% had a VL ≤200 copies per milliliter.10 We found 89% ART use and 79% viral suppression using comparable definitions. Marks et al18 reported that 62% of persons retained in care in 6 HIV outpatient clinics achieved viral suppression (<75 copies/mL), and we found 70% viral suppression using a comparable definition. Programmatic data from the In + Care Project show a mean of 69% viral suppression in reporting populations,19 and client-level data from the Ryan White Services Report show that 85% of clients were prescribed ART and 76% had viral suppression.20 Taken together, the consistency of these data strongly suggest that 60%–80% of persons in HIV care have viral suppression and stand in contrast to earlier estimates of 48%–60% viral suppression among persons in care.2,3 In turn, the data suggest that the proportion of persons with viral suppression among all PLWHA in the United States—which is based, in part, on estimates of viral suppression among those in care—likely exceeds or is at the upper end of the range of previously published estimates of 19%–28%.
ART initiation will likely increase in response to current US HIV/AIDS treatment guidelines recommending ART for all PLWHA. However, half of the patients in CNICS who had not initiated ART by the end of 2010 had a nadir CD4 count ≤500 cells per cubic millimeter, the threshold at which initiation was uniformly recommended under treatment guidelines active at the time. The strongest and most consistent predictor of ART noninitiation, ART discontinuation, and failure to achieve viral suppression was a lack of engagement in continuous HIV care. This is consistent with the findings of several previous studies21,22 and reinforces the need to better understand barriers to sustained engagement in care. Younger patients and patients with higher nadir CD4 counts were less likely to be engaged in continuous HIV care. This suggests that patients' perceived lack of need for HIV care may be an important barrier to engagement, but may also reflect provider-related decisions to schedule visits less frequently. Future CNICS studies will include patient-reported measures of mental illness and current substance use to better define the impact of these factors on care engagement and viral suppression.
Although race/ethnicity and female sex were not consistent independent predictors of lower ART use and viral suppression, women and non-Hispanic blacks had lower rates of ART initiation, ART continuation, and viral suppression. These disparities, particularly in the context of disproportionate risk for HIV infection and late HIV diagnosis among non-Hispanic blacks,23,24 highlight the ongoing imperative to address disparities at every step in the HIV care cascade.
The greatest strength of our study is that it provides precise contemporary information on key steps in the HIV care cascade in the United States using data from a large, geographically diverse population of persons in HIV care. CNICS data include visit dates, allowing us to directly ascertain engagement in continuous care, and all CNICS data undergo rigorous quality assurance procedures. However, our study was limited to patients who had ≥1 visit in 2010 and thus does not provide information about persons who are completely out of care. Indeed, 13% of the population in care in 2009 was not seen in 2010, and we do not have data on their care and viral suppression status in 2010. The CNICS cohort is composed of patients receiving care at university-affiliated HIV clinics that may not be representative of HIV care throughout the United States, but the similarity of our findings to those of MMP, which samples patients from a diverse array of HIV clinics, suggests that this is not a major limitation. We may have misclassified ART use by assuming continued therapy in the absence of documented discontinuation, but such misclassification would not have affected our estimates of virologic suppression.
In conclusion, we found that the vast majority of persons in active HIV medical care in a nationally distributed cohort in 2010 have initiated ART and that approximately 70% have viral suppression. Along with other recent reports, these findings suggest that viral suppression may be more common among persons in HIV care compared with earlier studies. Our study indicates that the increase in demand for ART under guidelines recommending ART initiation, regardless of CD4 count, is likely to be relatively small and highlights the need to work toward increasing engagement in HIV care.
The authors thank Dr. Jim Hughes and Ms. Kathy Thomas for statistical advice and all CNICS sites for contributing data to the study.
1. Cohen MS, Chen YQ, McCauley M, et al.. Prevention of HIV-1 infection with early antiretroviral therapy. N Engl J Med. 2011;365:493–505.
2. Gardner EM, McLees MP, Steiner JF, et al.. The spectrum of engagement in HIV care and its relevance to test-and-treat strategies for prevention of HIV infection. Clin Infect Dis. 2011;52:793–800.
3. Burns DN, Dieffenbach CW, Vermund SH. Rethinking prevention of HIV type 1 infection. Clin Infect Dis. 2010;51:725–731.
4. Hall HI, Frazer EL, Rhodes P, et al.. Continuum of HIV care: differences in care and treatment by sex and race/ethnicity in the United States
. [FRLBX05]. Paper presented at: XIX International AIDS Conference; July 27, 2012; Washington, DC.
5. Centers for Disease Control and Prevention. Vital signs: HIV prevention through care and treatment—United States
. MMWR Morb Mortal Wkly Rep. 2011;60:1618–1621.
6. Dombrowski JC, Kent JB, Buskin SE, et al.. Population-based metrics for the timing of HIV diagnosis, engagement in HIV care, and virologic suppression. AIDS. 2012;26:77–86.
7. Blair JM, McNaghten AD, Frazier EL, et al.. Clinical and behavioral characteristics of adults receiving medical care for HIV infection—Medical Monitoring Project, United States
, 2007. MMWR Surveill Summ. 2011;60:1–20.
8. Kitahata MM, Rodriguez B, Haubrich R, et al.. Cohort profile: the Centers for AIDS research Network of Integrated clinical systems. Int J Epidemiol. 2008;37:948–955.
9. Institute of Medicine. Monitoring HIV Care in the United States
: indicators and Data Systems. Washington, DC: The National Academies Press; 2012.
10. Skarbinski J, Johnson C, Frazier E, et al.. Nationally representative estimates of the number of HIV+ adults who received medical care, were prescribed ART, and achieved viral suppression-medical Monitoring Project, 2009 to 2010: US [Abstract 138]. Paper presented at: 12th Conference on Retroviruses and Opportunistic Infections; March 8, 2012; Seattle, WA.
11. Lillie-Blanton M, Stone VE, Snow Jones A, et al.. Association of race, substance abuse, and health insurance coverage with use of highly active antiretroviral therapy among HIV-infected women, 2005. Am J Public Health. 2009;100:1493–1499.
12. McNaghten AD, Hanson DL, Dworkin MS, et al.. Differences in prescription of antiretroviral therapy in a large cohort of HIV-infected patients. J Acquir Immune Defic Syndr. 2003;32:499–505.
13. Mugavero MJ, Lin HY, Allison JJ, et al.. Racial disparities in HIV virologic failure: do missed visits matter? J Acquir Immune Defic Syndr. 2009;50:100–108.
14. Pence BW, Ostermann J, Kumar V, et al.. The influence of psychosocial characteristics and race/ethnicity on the use, duration, and success of antiretroviral therapy. J Acquir Immune Defic Syndr. 2008;47:194–201.
15. White House Office of National AIDS Policy. Federal Implementation Plan. Washington, DC: The White House; 2010.
16. Tegger MK, Crane HM, Tapia KA, et al.. The effect of mental illness, substance use, and treatment for depression on the initiation of highly active antiretroviral therapy among HIV-infected individuals. AIDS Patient Care STDS. 2008;22:233–243.
17. McNutt LA, Wu C, Xue X, et al.. Estimating the relative risk in cohort studies and clinical trials of common outcomes. Am J Epidemiol. 2003;157:940–943.
18. Marks G, Gardner LI, Craw J, et al.. The spectrum of engagement in HIV care: do more than 19% of HIV-infected persons in the US have undetectable viral load? Clin Infect Dis. 2011;11:1168–1169; author's reply 1169–1170.
19. National Quality Center. Nqc news | in+care campaign update. NQC e-Newsletter. 2012;5. Available at: http://archive.constantcontact.com/fs034/1105430511667/archive/1109781861111.html
. Accessed September 21, 2012.
20. HRSA/HAB. HIV performance measure module update. HAB Inf E-mail. 2012;15. Available at: http://health.groups.yahoo.com/group/hivsnp/message/11499
. Accessed September 21, 2012.
21. Mugavero MJ, Lin HY, Willig JH, et al.. Missed visits and mortality among patients establishing initial outpatient HIV treatment. Clin Infect Dis. 2009;48:248–256.
22. Giordano TP, Gifford AL, White AC Jr, et al.. Retention in care: a challenge to survival with HIV infection. Clin Infect Dis. 2007;44:1493–1499.
23. Centers for Disease Control and Prevention. Disparities in diagnoses of HIV infection between blacks/African Americans and other racial/ethnic populations—37 states, 2005-2008. MMWR Morb Mortal Wkly Rep. 2011;60:93–98.
24. Chen M, Rhodes PH, Hall IH, et al.. Prevalence of undiagnosed HIV infection among persons aged ≥13 years—National HIV Surveillance System, United States
, 2005-2008. MMWR Morb Mortal Wkly Rep. 2012;61(suppl):57–64.