Are untimed antiretroviral drug levels useful predictors of adherence behavior?

Liechty, Cheryl Aa,b,c; Alexander, Christopher Sf; Harrigan, P Richardf; Guzman, J Davidc; Charlebois, Edwin Dc; Moss, Andrew Rd; Bangsberg, David Ra,c,d,e

Research Letters
Author Information

aDivision of Infectious Diseases, bCenter for AIDS Prevention Studies, cEpidemiology and Prevention Interventions Center, dDepartment of Epidemiology and Biostatistics, and eSan Francisco General AIDS Program, University of California San Francisco, San Francisco, CA, USA; and fBritish Columbia Center for Excellence in HIV/AIDS, University of British Columbia, Vancouver, BC, Canada.

Sponsorship: This study was funded by NIMH grants 54907 and 63011. Dr Bangsberg received additional funding from The Doris Duke Charitable Foundation.

Received: 18 June 2003; revised: 1 July 2003; accepted: 15 July 2003.

Article Outline

We examined cross-sectionally the relationship between untimed drug levels and adherence in 83 individuals. Abnormally low untimed antiretroviral drug levels were sensitive in identifying individuals adherent to 60% or less of medication doses over a 3–5-week period. An abnormally low drug level was associated with a higher viral load. A single abnormally low untimed antiretroviral drug level can identify an individual with very low adherence at high risk of HIV disease progression and death.

Adherence to HIV antiretroviral therapy is a strong predictor of virological suppression, disease progression, and death [1–6]. However, currently available approaches to measuring adherence have notable limitations [7–9], and individual patient assessments by medical providers do not accurately predict adherence [10–12].

Some investigators have assessed the antiretroviral drug level as a measure of adherence [13–16]. With one exception, those studies were limited by a reliance on self-reported measures of adherence [14]. The interpretation of random drug levels is complicated, however, because concentrations reflect drug absorption and metabolism as well as adherence to recent doses.

To determine the extent to which untimed drug levels are associated with adherence behavior, we measured these levels in a subset of participants from the REACH Cohort of homeless and marginally housed HIV-positive adults in San Francisco. This analysis included 83 individuals with complete pill count adherence data on combination antiretroviral regimens. Individuals on ritonavir-boosted regimens were excluded. All procedures were approved by the University of California, San Francisco Committee on Human Subjects Research.

Adherence assessments for the REACH Cohort have been described previously [17]. We measured adherence at unannounced pill count visits to the subjects’ place of residence every 3–5 weeks. Adherence was the difference between the current and previous pill counts divided by the prescribed number of doses for the same period. We have previously found that this method is closely associated with electronic cap monitoring (r = 0.91, P < 0.001) [18,19].

Procedures for drug level determination and the a priori establishment of concentration cut-offs below which drug levels were classified as ‘abnormally low’ have been described elsewhere [20]. The HIV viral load was measured using the Roche Amplicor assay.

Our sample was 85% male, with a high proportion of past and current injection drug use. A total of 59% of study subjects had undetectable viral loads on a variety of antiretroviral regimens. Mean adherence was 70.6% (SD = 31.4%). Drug levels were abnormally low in 31.3% of subjects.

The untimed drug level was significantly associated with adherence dichotomized at 90%: mean adherence for those with normal levels was 82.6%, versus 44.4% for those with abnormally low levels (P < 0.0001). The proportion of individuals with a normal drug level was significantly associated with the mean for each quintile of adherence values in a linear regression model with a slope 0.96 and y-intercept −1.3 (R2 = 0.94, P = 0.006).

Table 1 summarizes the specificity and sensitivity of the untimed drug level to detect non-adherence based on our data. The sensitivity of abnormally low drug levels to detect adherence of 90% or less was poor, ranging from 31 to 56%. Specificity at this threshold was higher, at least 90% for all drugs except indinavir. Sensitivity improved at lower adherence thresholds: at the 60% or less level, the sensitivity for all drugs pooled was 72%, rising to 83% among those of 50% or less. For efavirenz, a particularly long half-life drug, sensitivity at 90% or less was low at 44%; however, specificity was 100%.

We also assessed the relationship between the drug level and viral suppression. Abnormally low drug levels were found to be significantly associated with a higher log viral load using Wilcoxon rank-sum (P = 0.01).

Our findings support the concept that abnormally low untimed antiretroviral drug levels are associated with very low levels of adherence. The best test performance characteristics for this measure are in the setting of adherence below 60%. In keeping with this concept, we found that abnormally low untimed drug levels were associated with a higher viral load. A single random drug level can identify patients with very low level adherence. However, a normal range random drug level does not reliably signify high-level adherence.

The development of objective measures of pill-taking behavior has been challenging. Drug level assays appeal to the desire for quantitative assessments and can be measured retrospectively on stored specimens in research settings, or at office visits to providers.

Several investigators have examined the correlation between antiretroviral drug concentrations and self reported adherence. Murri et al. [21] reported that undetectable protease inhibitor levels were strongly correlated with self-reported non-adherence during the previous day [odds ratio (OR) 15.9, 95% confidence interval (CI) 4.9–50.7] [21]. The OR associated with missing doses more than once in the previous 3 days was 4.4 (95% CI 1.7–11.9). These findings support the concept that drug levels for short half-life medications best reflect adherence patterns over a brief period.

In the ATHENA Project, self-reported adherence was compared with antiretroviral drug levels expressed as concentration ratios (CR) [13]. Those reporting any missed medication doses over the previoius 2 days had significantly lower CR than individuals who reported complete adherence. Lower CR have also been correlated with lower levels of protease inhibitor adherence, as measured by electronic medication monitoring [14].

Our results using unannounced pill counts, which assess adherence behavior over a significantly longer interval, are consistent with the observations described above. However, in this case we have used a drug level determination that is not timed with respect to the last antiretroviral dose.

At the population level, we found a linear relationship between drug level and adherence across the full range of our adherence data. This suggests that untimed drug levels may be more useful in evaluating adherence in research cohorts than at the individual patient level.

In this cross-sectional analysis, we could not assess the impact of abnormally low drug levels on outcomes over time; nor could we examine whether the correlation between low drug levels and non-adherence was consistent over time. In addition, the subgroups in our sample size are relatively small, limiting the ability to study the performance characteristics of drug level measurements for specific drugs. However, our sensitivity and specificity calculations were similar for most drugs, with the exception of indinavir, a particularly short half-life drug.

Although adherence in excess of 90% is thought necessary for maximally sustained virological suppression [3], average adherence is considerably less [2,3,5,7, 17,22]. Despite this, dramatic mortality and morbidity benefits have been documented in the era of highly active antiretroviral therapy. Although perfect adherence should be the goal, it may be most important to identify accurately those who are the most non-adherent. Patients in this group are likely to be at the highest risk of HIV disease progression and death. In this regard, even a single untimed drug level of a single agent in a regimen appears to be useful.

Back to Top | Article Outline


1. Bangsberg DR, Perry S, Charlebois ED, Clark RA, Robertson M, Zolopa AR, Moss A. Non-adherence to highly active antiretroviral therapy predicts progression to AIDS. AIDS 2001, 15: 1181–1183.
2. Arnsten JH, Demas PA, Farzadegan H, Grant RW, Gourevitch MN, Chang CJ, et al. Antiretroviral therapy adherence and viral suppression in HIV-infected drug users: comparison of self-report and electronic monitoring. Clin Infect Dis 2001, 33:1417–1423.
3. Paterson DL, Swindells S, Mohr J, Brester M, Vergis EN, Squier C, et al. Adherence to protease inhibitor therapy and outcomes in patients with HIV infection. Ann Intern Med 2000, 133:21–30.
4. Hogg RS, Heath K, Bangsberg D, Yip B, Press N, O'Shaughnessy MV, Montaner JSG. Intermittent use of triple-combination therapy is predictive of mortality at baseline and after 1 year of follow-up. AIDS 2002, 16:1051–1058.
5. McNabb J, Ross JW, Abriola K, Turley C, Nightingale CH, Nicolau DP. Adherence to highly active antiretroviral therapy predicts virologic outcome at an inner-city human immunodeficiency virus clinic. Clin Infect Dis 2001, 33:700–705.
6. Garcia de Olalla P, Knobel H, Carmona A, Guelar A, Lopez-Colomes JL, Cayla JA. Impact of adherence and highly active antiretroviral therapy on survival in HIV-infected patients. J Acquir Immune Defic Syndr 2002, 30:105–110.
7. Liu H, Golin CE, Miller LG, Hays RD, Beck CK, Sanandaji S, et al. A comparison study of multiple measures of adherence to HIV protease inhibitors. Ann Intern Med 2001, 134:968–977.
8. Wendel CS, Mohler MJ, Kroesen K, Ampel NM, Gifford AL, Coons SJ. Barriers to use of electronic adherence monitoring in an HIV clinic. Ann Pharmacother 2001, 35:1010–1015.
9. Rudd P, Byyny RL, Zachary V, LoVerde ME, Titus C, Mitchell WD, Marshall G. The natural history of medication compliance in a drug trial: limitations of pill counts. Clin Pharmacol Ther 1989, 46:169–176.
10. Bangsberg DR, Hecht FM, Clague H, Charlebois ED, Ciccarone D, Chesney M, Moss A. Provider assessment of adherence to HIV antiretroviral therapy. J Acquir Immune Defic Syndr 2001, 26:435–442.
11. Gross R, Bilker WB, Friedman HM, Coyne JC, Strom BL. Provider inaccuracy in assessing adherence and outcomes with newly initiated antiretroviral therapy. AIDS 2002, 16:1835–1837.
12. Miller LG, Liu H, Hays RD, Golin CE, Beck CK, Asch SM, et al. How well do clinicians estimate patients’ adherence to combination antiretroviral therapy? J Gen Intern Med 2002, 17:1–11.
13. Nieuwkerk PT, Sprangers MA, Burger DM, Hoetelmans RM, Hugen PW, Danner SA, et al. Limited patient adherence to highly active antiretroviral therapy for HIV-1 infection in an observational cohort study. Arch Intern Med 2001, 161: 1962–1968.
14. Hugen PW, Langebeek N, Burger DM, Zomer B, van Leusen R, Schuurman R, et al. Assessment of adherence to HIV protease inhibitors: comparison and combination of various methods, including MEMS (electronic monitoring), patient and nurse report, and therapeutic drug monitoring. J Acquir Immune Defic Syndr 2002, 30:324–334.
15. Gandhi M, Ameli N, Bacchetti P, Ponath C, Benet LZ, Baluom M, et al. Poor correlation between self-reported adherence to antiretrovirals and plaslasma drug levels. In: XIVth International AIDS Conference. Barcelona, Spain, 2002 [Abstract 5917].
16. van Rossum AM, Bergshoeff AS, Fraaij PL, Hugen PW, Hartwig NG, Geelen SP, et al. Therapeutic drug monitoring of indinavir and nelfinavir to assess adherence to therapy in human immunodeficiency virus-infected children. Pediatr Infect Dis J 2002, 21:743–747.
17. Bangsberg DR, Hecht FM, Charlebois ED, Zolopa AR, Holodniy M, Sheiner L, et al. Adherence to protease inhibitors, HIV-1 viral load, and development of drug resistance in an indigent population. AIDS 2000, 14:357–366.
18. Bangsberg DR, Hecht FM, Charlebois ED, Chesney M, Moss AR. Comparing objective methods of adherence assessment: electronic medication monitoring and unannounced pill counts. AIDS Behav 2001, 5:275–281.
19. Murri R, Antinori A, Ammassari A, Nappa S, Orofino, G, Abrescia N, et al. Physician estimates of adherence and the patient-physician relationship as a setting to improve adherence to antiretroviral therapy. J Acquir Immune Defic Syndr 2002, 31 (Suppl 3):S158–S162.
20. Alexander CS, Asselin JJ, Ting LSL, Montaner JS, Hogg RS, Yip B, et al. Antiretroviral concentrations in untimed plasma samples predict therapy outcome in a population based setting with advanced disease. J Infect Dis 2003, 4:541–548.
21. Murri R, Ammassari A, De Luca A, Cingolani A, Marconi P, Wu AW, Antinori A. Self-reported nonadherence with antiretroviral drugs predicts persistent condition. HIV Clin Trials 2001, 2: 323–329.
22. Golin CE, Liu H, Hays RD, Miller LG, Beck CK, Ickovics J, et al. A prospective study of predictors of adherence to combination antiretroviral medication. J Gen Intern Med 2002, 17:756–765.

Cited By:

This article has been cited 3 time(s).

Adherence–resistance relationships for protease and non-nucleoside reverse transcriptase inhibitors explained by virological fitness
Bangsberg, DR; Acosta, EP; Gupta, R; Guzman, D; Riley, ED; Harrigan, PR; Parkin, N; Deeks, SG
AIDS, 20(2): 223-231.
PDF (225) | CrossRef
JAIDS Journal of Acquired Immune Deficiency Syndromes
Pharmacologic Optimization of Protease Inhibitors and Nonnucleoside Reverse Transcriptase Inhibitors (POPIN)-A Randomized Controlled Trial of Therapeutic Drug Monitoring and Adherence Support
Khoo, SH; Lloyd, J; Dalton, M; Bonington, A; Hart, E; Gibbons, S; Flegg, P; Sweeney, J; Wilkins, EG; Back, DJ
JAIDS Journal of Acquired Immune Deficiency Syndromes, 41(4): 461-467.
PDF (484) | CrossRef
JAIDS Journal of Acquired Immune Deficiency Syndromes
The Relationship Between Resistance and Adherence in Drug-Naive Individuals Initiating HAART Is Specific to Individual Drug Classes
Tam, LW; Chui, CK; Brumme, CJ; Bangsberg, DR; Montaner, JS; Hogg, RS; Harrigan, PR
JAIDS Journal of Acquired Immune Deficiency Syndromes, 49(3): 266-271.
PDF (259) | CrossRef
Back to Top | Article Outline
© 2004 Lippincott Williams & Wilkins, Inc.