Multiple terms, oftentimes expressing different constructs yet with some overlap, have been used to explain how well a patient takes his/her medications. These terms, such as adherence, compliance, persistence, and durability, are often used interchangeably, sometimes lead to inaccurate or imprecise interpretation of patient behavior, and may result in incorrect conclusions about which intervention should be appropriately deployed. Much of the early research and attention to medical management of HIV/AIDS has been focused on the construct of adherence. Especially early in the HIV/AIDS epidemic when regimens were often complicated, had large pill burdens, complex dosing schedules, and low genetic barriers to resistance, adherence played an important role in HIV treatment. Indeed, a large body of literature has repeatedly demonstrated the importance of adherence, wherein a high level of adherence to antiretroviral therapy was associated with successful viral suppression and decreased morbidity. Treatment strategies, however, have changed. In recognition of the changing HIV epidemic and the various subpopulations globally who will eventually access contemporary combination antiretroviral therapy (cART), we provide insight into additional considerations and define the construct of persistence and its operationalization for clinical care and research in the context of HIV treatment. Additionally, we summarize methods for determination of persistence and current literature on persistence in HIV treatment.
The dynamics of HIV replication and impact on treatment
Decreased adherence to or discontinuation of a prescribed therapy is likely to result in unfavorable health outcomes in chronic diseases such as hypertension, hyperlipidemia, congestive heart failure, or chronic obstructive pulmonary disease. Management of HIV differs from other chronic diseases in that successful control of the disease is more complex and also the consequences of failure may be greater. First, suppression of viral replication requires lifelong retention on cART, often consisting of three or more medications. Second, unlike in hypertension or diabetes, inconsistent use of medications leads to development of resistance to one or more medications in the regimen, thereby limiting future treatment options and complicating therapy. Last, failure to suppress viral replication not only affects the health of the patient, but also increases the risk of HIV transmission to others who engage in high-risk behaviors, thereby posing greater public health concerns than with noncommunicable diseases.
Management of HIV has also changed dramatically in recent years. In the Strategies for Management of Antiretroviral Therapy (SMART) trial, participants randomized to discontinue cART until prespecified CD4 cell count thresholds were met had increased non-HIV and HIV-related morbidity and mortality compared to those who continuously remained on therapy . This finding emphasizes more than ever that ‘remaining’ on therapy is crucial, and so is the need for interventions that allow treatment to ‘persist over time.’
Antiretroviral treatment itself has also changed. Development of highly potent, once-daily, low pill burden, and more tolerable cART has greatly improved and altered the landscape of contemporary HIV management . Contemporary regimens are affected less by perturbations in adherence, in part due to the high genetic barrier to resistance by some medications (e.g., ritonavir-boosted protease inhibitors) and long half lives of others (e.g., nonnucleoside reverse transcriptase inhibitor, NNRTI). For example, nevirapine, a NNRTI, has a serum half-life of 48 h  and its serum level remains detectable 1 week after discontinuation . On the other hand, abacavir and lamivudine, nucleoside reverse transcriptase inhibitors (NRTIs), have much shorter serum half-lives of 1.5 and 6 h, respectively [5,6] and the intracellular concentration of the active abacavir metabolite becomes undetectable within 72 h after the last dose . Therefore, discontinuation of a multidrug regimen such as the one above for 3 days or more may result in extended nevirapine monotherapy, during which selective pressure on viral replication may result in development of resistance due to the low genetic barrier to resistance by NNRTIs . In the case of nevirapine administered even as a single-dose monotherapy regimen to prevent mother-to-child-transmission of HIV, high rates of NNRTI resistance have been observed among mothers and infants [9,10]. Efavirenz, a preferred NNRTI component of contemporary regimens, also has a prolonged serum half-life of 40–55 h, and like nevirapine, can remain at a therapeutic dose for longer than 21 days after discontinuation . NNRTIs such as efavirenz and nevirapine have the benefit of continuing to suppress HIV replication when medications are stopped briefly, however, are more likely to develop resistance when they are stopped for longer periods due to their low genetic barrier to resistance.
Ritonavir-boosted protease inhibitor regimens, also part of preferred cART regimens, do not have prolonged serum half-lives like NNRTIs. These antiretroviral agents, however, have other pharmacological advantages due to their high genetic barrier to resistance such that even prolonged periods of monotherapy potently suppress HIV replication, yet seldom result in the development of resistance mutations [12,13]. As such, they pose few problems when there is poor adherence or decreased persistence.
In a proof-of-concept pilot study, participants who were on a suppressive once-daily antiretroviral regimen were changed to 5 days on treatment and then provided with a two-day holiday (72 h since last dose); all participants on an efavirenz-containing regimen and 90% of those on a nevirapine-containing regimen had suppressed HIV-1 RNA levels at 48 weeks suggesting that intervals less than 72 h do not negatively impact clinical outcomes . On the other hand, two out of nine participants on a protease inhibitor-based regimen had experienced virological rebound by 48 weeks. In another trial, in which antiretroviral therapy was interrupted every other week in a ‘1-week-on-1-week-off’ strategy, one of eight patients on an efavirenz-based regimen experienced virological failure with a resultant new resistance mutation to efavirenz. In contrast, 11 of 17 patients on a regimen containing ritonavir-boosted saquinavir experienced virological failure; however, none developed a new resistance mutation to a protease inhibitor . These studies suggest that the duration for which therapy may be discontinued without expecting an adverse outcome differs depending on antiretroviral composition of a regimen. This ‘permissible gap’ (see below for definition) is most likely to be between 2 and 7 days for a regimen containing efavirenz, but further investigations are needed to better characterize the permissible gap that results in adverse consequences.
Adherence, rather than persistence, has been the center of focus in research of medication-taking behaviors among HIV-infected patients (see below for definition). Contrary to the pervasive view that low adherence leads to development of resistance, data suggest that the relationship between adherence and resistance may be more complex , particularly when contemporary preferred regimens are prescribed. For example, imperfect adherence of many ritonavir-boosted protease inhibitor-containing regimens does not result in significant levels of resistance .
In sum, adherence alone when using contemporary treatment regimens may be insufficiently predictive of clinical outcomes in HIV-infected patients and is dependent on the type of regimen prescribed. Clinical care and research in the management of HIV would, therefore, benefit from an additional ‘time-dependent’ measure of medication-taking behaviors.
Adherence, the most frequent medication construct for HIV treatment, is defined as ‘the extent to which a patient's behavior corresponds with the recommendations of a healthcare provider’, and is often synonymous with compliance . In this article, we will use adherence to represent this construct. When adherence refers to taking medication, it generally quantifies the extent to which a patient acts in accordance with a prescribed interval and doses of a prescribed regimen within a given time period . By definition, it is expressed as a percentage of correctly timed doses (doses taken/doses prescribed × 100) . Researchers and clinicians alike have tried to quantify the optimal level of adherence (e.g., greater than 95%) needed to simultaneously suppress viral replication and avoid development of resistance, yet these binary definitions have not been borne to be equally predictive for differing cART regimens . Moreover, adherence thresholds have been plagued by measurement problems (e.g., self-report, electronic monitoring, pharmacy refills) and with quantifying exactly how much adherence is enough [16,20].
Medication persistence, on the other hand, is also a medication-taking construct defined as ‘the duration of time from initiation to discontinuation of therapy’ . By definition, it is expressed solely as a function of time or the number of days (or months) on treatment. Alternatively, persistence can be expressed as a binary variable (persistent or nonpersistent), measured at the end of a prespecified time period. Similar to adherence, defining persistence as a binary variable is challenging due to the ‘permissible gap’ in time that is allowed to pass after discontinuation of a prescribed regimen that is associated with a poor treatment outcome. Permissible gaps, however, are likely to differ based on the type of cART regimen prescribed.
As with adherence, the concept of persistence may be applied to a variety of situations, including taking medications, following dietary advice, and changing health habits. Because persistence emphasizes the concept of continuous therapy, a permissible gap, the maximum number of consecutive doses that a patient can miss without expecting a reduced or suboptimal outcome, should be prespecified in any assessment of persistence. Such permissible gaps are ones that should have no negative clinical consequences for patients.
Medication persistence and adherence are similar in that they both measure the extent to which a patient's behavior agrees with recommendations of a healthcare provider. They differ, however, in the dimensions of this agreement. With regard to taking medications, adherence measures the proportion of times that a patient takes medication as prescribed within a given interval, whereas persistence measures the duration of time that a patient continuously adheres to a prescribed regimen. In other words, adherence measures ‘how often’, whereas persistence measures ‘for how long.’ As such, these constructs are complimentary but distinct.
With regard to cART for the treatment of HIV, medication persistence merits further categorization, including patient persistence and regimen persistence. Distinction between these constructs is described further.
According to the stringent definition of persistence as a continuous therapy, a patient is persistent in adhering to the prescribed regimen as long as the permissible gap is not exceeded. A permissible gap can be defined as the maximum duration for which a patient may discontinue medication without experiencing a suboptimal outcome or adverse consequence. Due to the high replication capacity of HIV, the permissible gap for HIV treatment is likely to be on the magnitude of days rather than weeks (see below); however, as cART consists of multiple medications, often with different pharmacokinetic profiles, the permissible gap may vary depending on individual medications within a prescribed regimen. In addition, the duration of medication discontinuation necessary for development of a suboptimal outcome may also vary depending on the adverse consequence of interest (e.g., incomplete viral suppression, development of resistance, development of an adverse clinical event). In the case of viral suppression, a permissible gap may be on the order of days, whereas the time to a clinical event (e.g., myocardial infarction or opportunistic infection) is likely to be on the order of months to years.
If a patient discontinues medication for a period that exceeds the permissible gap, the patient is no longer persistent with the HIV treatment. The duration of nonpersistence is equivalent to the time lapsed between the first missed dose and re-initiation of therapy. A patient's persistence with medication is expressed as a continuous variable in days (or weeks) with the goal being a lifetime.
The concept of persistence may be extended beyond that of the individual patient and be applied to an entire antiretroviral regimen. This concept is most pertinent in resource-poor regions where available cART regimens are limited and unlike contemporary regimens in resource-rich settings, are less tolerable, have higher pill burdens, and are pharmacologically inferior to newer regimens. We define regimen persistence as ‘the duration between the initiation and discontinuation of a specified antiretroviral regimen as agreed upon by the patient and the healthcare provider.’ Using this definition, any change in any part of a regimen, for any reason, would result in the regimen being nonpersistent at the time of regimen discontinuation or modification.
In contrast to patient persistence – a measure of a patient's continued taking of cART, irrespective of the individual medications contained within the regimen – regimen persistence measures duration of a particular cART regimen as a means to suppress viral replication. Regimen persistence depends on factors both intrinsic and extrinsic to the regimen. Intrinsic factors include adverse side effects, pill burden, underlying levels of resistance to one or more components of the regimen, and cost. Extrinsic factors, those that contribute to a change in the components of the regimen, include new findings from clinical trials, new treatment guidelines, and availability of antiretroviral medications in a particular region. In contrast to patient persistence, the concept of a permissible gap is not applicable in the definition of regimen persistence; a regimen is persistent as long as it has not been explicitly modified or discontinued by either the patient or the healthcare provider.
Patient persistence, regimen persistence, and adherence
Although both are important, persistence and adherence are different, but interrelated, as illustrated in Fig. 1. Adherence levels are indicated as a solid line, and HIV-1 RNA levels are shown as a dashed line. Optimal HIV viral suppression is observed when HIV-1 RNA levels fall below the dotted line. In this example, a patient initiates an NNRTI-based regimen and initially achieves virological suppression. Later, however, medications are discontinued completely (patient nonpersistence or 0% adherence), resulting in virological rebound or replication to a detectable level. Alternatively, if the provider discontinued medications for any reason (low supply, too costly, etc.), the patient would be persistent, the regimen would be nonpersistent, and the patient would be 100% adherent because the patient did what the clinician recommended. As the patient has discontinued medications for a period exceeding the permissible gap, he/she is no longer persistent with his/her original regimen; the patient is nonpersistent with medication for the duration between the first missed dose and the next dose of medication he takes (patient nonpersistence in Fig. 1). The duration of patient persistence, in this case, is defined as the time period between the first dose and the last dose of the regimen [patient (1) in Fig. 1].
On the other hand, the regimen is persistent as long as it has not been explicitly modified or discontinued through agreement between the patient and the provider. Therefore, in this case, regimen persistence continues until the provider changes the NNRTI-based regimen to a boosted protease inhibitor-based one [regimen (1) in Fig. 1]. Patient persistence is not affected by the regimen modification as long as the patient continuously adheres to medication without exceeding the permissible gap; the second phase of patient persistence [patient (2) in Fig. 1], which began with re-initiation of medication, continues despite the modification in regimen.
As demonstrated in this example, a regimen may be persistent while the patient is nonpersistent. Likewise, a patient may be persistent with medication-taking even when his/her regimen is changed as long as he/she continues to take his/her prescribed medications. Finally, a patient may be persistent while achieving a low level of adherence (sometimes defined as nonadherence). For example, if a patient is prescribed medication twice daily, but takes the regimen once daily everyday, he would be persistent but would maintain at a 50% adherence level.
Methods for determining persistence
Measurement of persistence and methods to collect persistence data have been summarized previously [21,22], and several methods may be used to determine persistence in HIV treatment (see Table 1). Patient persistence can be determined through measurements of a patient's pill-taking history. These may include direct observation, Medication Event Monitoring System (MEMS), patient self-report recall, review of pharmacy refill, and medical records. Regimen persistence may also be determined using many of the same methods as in patient persistence and a regimen change form on a study instrument in a prospective study. Reasons for patient or regimen nonpersistence can be measured via a study instrument that assesses these constructs or by review of medical records.
Methods that yield high granularity of data, such as direct observation or MEMS, often require prospective studies and are likely to be expensive. Therefore, it is often impractical to gather this level of granularity in large retrospective studies. On the other hand, pharmacy refill records can be obtained with less cost and effort compared to other methods. Pharmacy refill data, however, lack sufficient detail to adequately measure patient persistence, cannot accurately measure small permissible gaps in treatment, or answer why a regimen is no longer persistent, but are often satisfactory to measure regimen persistence per se.
Impact of persistence on clinical outcomes
Patient nonpersistence in HIV treatment has been insufficiently assessed in current research and is associated with adverse clinical outcomes. In a Spanish cohort study in which the median duration of follow-up was 8.3 years, 43% of patients had a treatment interruption longer than 3 days, and these patients had a higher risk of treatment failure . In a Ugandan study with the median time on therapy of 38 weeks, 23% of patients had a history of treatment interruption greater than 4 days, which was significantly associated with virological failure . Among injection drug users in Baltimore, 78% of individuals had one or more treatment discontinuations, and 20% of the study population never resumed cART .
In a study in which pill-taking history was measured using MEMS, 65% of patients had a treatment interruption longer than 48 h in the 24 weeks of observation period and were more likely to develop drug resistance than those without an interruption . Similarly, patients with a history of more than one drug holiday (patient nonpersistence) lasting at least 48 h were more likely to fail therapy and develop a resistance to NNRTI-containing regimens compared to those with one or less drug holiday . In another study, intermittent use of cART in the first year of therapy was significantly associated with increased mortality .
Results from prospective randomized controlled trials on structured treatment interruptions confirm that patient nonpersistence adversely affects clinical outcomes in HIV-infected patients eligible for cART. In one trial, scheduled treatment interruptions exceeding 4 weeks were associated with development of resistance . In another study, a structured ‘1-week-on-1-week-off’ treatment strategy using cART was associated with increased likelihood of virological failure, and development of resistance among the patients taking an NNRTI-based regimen ; these findings suggest that even missing 1 week of therapy has significant adverse consequences. In the randomized controlled SMART trial, planned cART discontinuation using a priori CD4 cell guidance thresholds was associated with increased rates of HIV-associated and non-HIV-associated morbidity, decreased levels of HIV suppression, and lower CD4 cell counts, when compared to those who persisted with therapy .
On the other hand, in a randomized controlled treatment strategy of a ‘five-days-on, two-days-off’ schedule of therapy in patients with durable virological suppression, 11–22% experienced virological failure among patients on a nevirapine-based or protease inhibitor-based regimen; however, no failure was observed among those on efavirenz-based regimens over 48 weeks of observation . Parienti et al. [30,31], using an observational analytical approach, demonstrated that frequent and longer duration of treatment interruption (nonpersistence) were better predictors of virological rebound (i.e., failure) among patients on an NNRTI-based regimen. According to their logistic model, a treatment interruption of 15 days was associated with a 50% probability of virological rebound among those on an NNRTI-based regimen. On the other hand, higher average adherence rates overall appeared to be a better correlate of virological suppression among those on a boosted-protease inhibitor-based regimen. These last two studies suggest that there may be unique properties of one or more of the components of the cART regimen that contributes to different permissible gaps in treatment interruptions that affect adverse clinical consequences.
In sum, these data highlight that permissible gaps in HIV treatment may be as short as a few days and also vary depending on the unique pharmacokinetic and genetic barrier to resistance profiles of the various components of a cART regimen.
Though reported measurements have been imperfect, persistence of an initial antiretroviral regimen ranges from 11.8  to 34.3 months  with the trend toward longer regimen persistence in more recent years. Though newer salvage regimens have resulted in markedly improved levels of viral suppression, it remains true that maintaining a patient on the initial regimen is likely to result in the greatest likelihood of virological suppression. Compared to the initial regimen, the second and the third regimens have significantly lower probability of achieving virological suppression (adjusted odds ratio 0.49 and 0.22, respectively, and P < 0.02 for both) . Furthermore, each modification is associated with a more complex dosing schedule, a less favorable toxicity profile, and also decreased persistence of the subsequent regimen . In another study, patients who started on a persistent, NNRTI-based regimen were less likely to experience subsequent regimen modifications and a three-class regimen, compared to a less-persistent, protease inhibitor-based regimen .
Regimen persistence is a particularly important issue in resource-poor settings, where available antiretroviral choices are limited and the medication alternatives are costly . Virological failure due to resistance to therapy may leave patients with few or no remaining treatment options.
Factors that affect persistence
Adherence, patient persistence, and regimen persistence are intimately interrelated; they may be influenced by not only a similar set of patient, medication, and socioeconomic characteristics, but also by one another. For example, low adherence and frequent patient nonpersistence due to toxicity of a regimen may lead to development of resistance. Subsequent virological failure will eventually result in modification of the patient's regimen, leading to decreased regimen persistence. Existing literature on patient and medication characteristics that impact patient and regimen persistence is summarized below.
In the treatment of HIV infection, many patient characteristics contribute to decreased persistence. Clinical characteristics of patient-associated factors that have been associated with decreased patient persistence include female gender, high HIV RNA level, active substance use disorder , depression, and shorter time on cART . Younger age and black race have also correlated with decreased patient persistence .
Patient characteristics associated with regimen persistence are summarized in Table 2 [37–47]. These include high or increasing HIV RNA levels [36,48–51], low CD4 cell count prior to cART initiation [37,50], current high CD4 cell count [38,39], short duration on therapy , previous cART experience [32,40], history of opportunistic infection , and hepatitis C virus coinfection .
Also, affective mental disorders , depression [36,51], use of alternative medicine, hospitalization , female gender [39,48], men who have sex with men , black or minority race/ethnicity [36,41,54], younger age , low weight , lack of medical coverage [53,55], and incarceration [56–58] have been associated with decreased regimen persistence.
Comorbidities such as mental illness and substance use disorders are common among HIV-infected patients, and these patients frequently transition through correctional facilities [59,60]. Nonpersistence is a great challenge in this population both within the community and upon transition between a correctional and a community setting [57,61,62]. In a study among HIV-infected prisoners, 95% of released inmates failed to fill their cART prescription within 10 days of release (the time period for which medications were provided upon release), and patients, therefore, presumably did not take HIV medications beyond this period . Others have confirmed that HIV-1 RNA levels increase during this postrelease period; the finding that the HIV-1 RNA levels return to their pretreatment levels, and not just a partial increase, suggests that nonpersistence rather than nonadherence is the mechanism of poor treatment outcomes . In another study of jail detainees, only 15% of those who were re-incarcerated repeatedly persisted with their medications. Those who did not persist or who were never prescribed medications had an increased likelihood of having higher HIV-1 RNA levels and decreasing CD4 cell counts . This suggests that patient nonpersistence after release from prison or jail is common and is an important public health concern.
Existing data assessing the characteristics of a specific medication component or entire cART regimen on persistence primarily focus on regimen persistence. Medication characteristics associated with regimen persistence are summarized in Table 2. In western countries, adverse events associated with cART and treatment failure were the two most common reasons for medication discontinuation or modification [48,63]. In addition, a greater number of medications within a regimen  and more frequent dosing  were associated with early regimen discontinuation. In developing countries, in addition to adverse events and treatment failure, high cost and inadequate supply of medications were cited as common reasons for regimen discontinuation among patients [40,42].
Characteristics of individual antiretroviral medications within a cART regimen also influence regimen persistence. Overall, NNRTI-based regimens have been associated with increased persistence, compared to boosted or unboosted protease inhibitor-based, triple-NRTI-based, or triple-class regimens [33,43–45]. Newer generations of NRTIs such as tenofovir, lamivudine, emtricitabine, and abacavir improve persistence compared to zidovudine, stavudine, and didanosine [33,38,39]. Also, efavirenz has been associated with increased regimen persistence compared to nevirapine, and also the protease inhibitors lopinavir, saquinavir, and indinavir [38,42,45,46]. Among protease inhibitors, darunavir and atazanavir were less likely to result in regimen switch due to toxicity [39,64], whereas lopinavir, saquinavir, and ritonavir were associated with discontinuation or modification of therapy [36,39,48].
In a recent trial in which patients were randomized to receive coformulated tenofovir/emtricitabine or abacavir/lamivudine and efavirenz or atazanavir boosted with ritonavir, lower rates of virological failure and increased persistence were observed among the group assigned to tenofovir/emtricitabine compared to abacavir/lamivudine . Adverse consequences associated with abacavir resulted in medication discontinuation at a higher rate than for tenofovir.
Healthcare setting characteristics
The organization of healthcare and even pharmacy services for patients impacts persistence. For example, the frequency with which either clinicians choose to follow their patients or even how it is dictated by insurance or managed healthcare providers can negatively influence persistence. Prior authorization for medication changes has been associated with patient nonpersistence , as have copays and requirements for patients to spend down their personal resources before prescription benefits are renewed . Most patients who become nonpersistent do so without their provider actually knowing it until the patients' next scheduled appointment (if the patient manages to return at all). In the case of HIV-infected drug users, the onsite integration of buprenorphine treatment into HIV treatment settings resulted in improved retention in care and continuation of medications compared to those who were referred for treatment for their opioid dependence off-site . Thus, organizational factors that lead to fragmented healthcare or less frequent monitoring, especially early in cART initiation, may disrupt continuity of cART and worsen HIV treatment outcomes.
Much of research on HIV medication-taking behavior continues to be focused on adherence and mistakenly incorporates elements of persistence into its construct. Adherence and persistence are similar in that both constructs measure accordance of patient behavior with a prescribed therapy. In contrast to adherence, persistence is a longitudinal measure of antiretroviral therapy, with the emphasis on continuity rather than frequency.
In this article, we deconstruct persistence into two types: patient persistence and regimen persistence. The former measures continuous adherence to cART without exceeding a permissible gap, and the latter measures duration of a prespecified cART regimen. Measures of persistence can be obtained in various ways, including pharmacy data, medical records, and medical electronic monitoring system. These measures have not, however, been codified in accordance with contemporary cART regimens.
As cART regimens become more tolerable, less complex, and are created with higher barriers to development of resistance, interventions designed to improve medication-taking behaviors need to increasingly focus on nonpersistence in addition to nonadherence. Such interventions will likely need to incorporate measurement of persistence in real-time so that lapses in medication-taking are averted promptly.
As summarized here, patient nonpersistence is associated with adverse clinical outcomes, including higher rates of treatment failure, development of drug resistance, and increased mortality. Importantly, a longer duration and a higher frequency of patient nonpersistence appear to increase the risk of adverse outcomes.
Patient nonpersistence is of a major public health concern because viral resistance may develop during nonpersistent periods and subsequently require a more costly and toxic regimen for viral suppression. Furthermore, a higher viral load observed during nonpersistent periods increases the risk of transmission of a potentially drug-resistant virus, if the patient engages in a risky behavior.
In addition to interventions that address patient characteristics associated with decreased persistence and clinical outcomes, such as substance abuse, incarceration, mental illness, and depression, new approaches may benefit both patients and the public. For example, a system in which failure of a patient to refill his or her medication in a scheduled time period leads to notification of his or her healthcare provider by the pharmacy would enable the physician to address the persistence issue with the patient and prevent the patient from being nonpersistent. Additionally, an improved coordination between a correctional and a community healthcare system would help many recently incarcerated HIV patients to remain persistent with therapy. Education of clinicians and patients of importance of continuous adherence, and impact of nonpersistence and ‘drug holidays’, especially for NNRT-based regimens, may lead to a better decision-making with regard to selection and continuation of therapy. Adherence tools, such as schedules, dosettes, and electronic reminder systems, may also increase both adherence and persistence . Finally, patients with a high risk of nonadherence and nonpersistence may benefit from a directly observed therapy [69–71].
Each regimen change is associated with a diminished chance of viral suppression as well as higher toxicity and cost, and thus regimen persistence is an important issue from both clinical and public health perspectives. As summarized in this article, fewer drugs in a regimen, fixed-dose combinations, newer generations of NRTIs, boosted-protease inhibitors, and efavirenz correlate with greater regimen persistence. Consideration of effects of each antiretroviral on regimen persistence is needed to maximize chances of prolonged viral suppression.
Limited availability of antiretroviral medications makes regimen persistence an especially important issue in resource-limited settings. Additionally, inconsistent supply of drugs may be a hindrance to patient as well as regimen persistence. Unfortunately, newer generations of antiretrovirals associated with increased persistence tend to be more costly and unavailable in developing countries.
Last, the improved adherence with contemporary treatment regimens and data from the SMART trial remind clinicians and researchers that persistence has become the ‘Achilles Heel’ of HIV treatment and interventions that retain patients on effective treatment are urgently needed.
The authors would like to acknowledge the National Institutes on Drug Abuse for research (R01 DA13805 and R01 DA17059–02) and career develop (K24 DA017072) awards for F.L.A., Medical Scholarship Award from Infectious Disease Society of America for J.W.B., and Alicia G. Woods, PhD (Padure Biomedical Consulting) for assistance and advice in the preparation of this article through funding from Gilead Sciences Inc.
J.W.B. and F.L.A. collected all data and drafted the article. All authors edited, commented, and approved the final version of the article.
1. Strategies for Management of Antiretroviral Therapy (SMART) Group. CD4+ count-guided interruption of antiretroviral treatment
. N Engl J Med
3. Mirochnick M, Fenton T, Gagnier P, Pav J, Gwynne M, Siminski S, et al
. Pharmacokinetics of nevirapine in human immunodeficiency virus type 1-infected pregnant women and their neonates. Pediatric AIDS Clinical Trials Group Protocol 250 Team. J Infect Dis 1998; 178:368–374.
4. Mackie NE, Fidler S, Tamm N, Clarke JR, Back D, Weber JN, Taylor GP. Clinical implications of stopping nevirapine-based antiretroviral therapy: relative pharmacokinetics and avoidance of drug resistance. HIV Med 2004; 5:180–184.
5. Kumar PN, Sweet DE, McDowell JA, Symonds W, Lou Y, Hetherington S, LaFon S. Safety and pharmacokinetics of abacavir (1592U89) following oral administration of escalating single doses in human immunodeficiency virus type 1-infected adults. Antimicrob Agents Chemother 1999; 43:603–608.
6. Johnson MA, Moore KH, Yuen GJ, Bye A, Pakes GE. Clinical pharmacokinetics of lamivudine. Clin Pharmacokinet 1999; 36:41–66.
7. Hawkins T, Veikley W, St Claire RL, Guyer B, Clark N, Kearney BP. Intracellular pharmacokinetics of tenofovir diphosphate, carbovir triphosphate, and lamivudine triphosphate in patients receiving triple-nucleoside regimens. J Acquir Immune Defic Syndr 2005; 39:406–411.
8. Richman DD, Havlir D, Corbeil J, Looney D, Ignacio C, Spector SA, et al
. Nevirapine resistance mutations of human immunodeficiency virus type 1 selected during therapy. J Virol 1994; 68:1660–1666.
9. Kassaye S, Lee E, Kantor R, Johnston E, Winters M, Zijenah L, et al
. Drug resistance in plasma and breast milk after single-dose nevirapine in subtype C HIV type 1: population and clonal sequence analysis. AIDS Res Hum Retroviruses 2007; 23:1055–1061.
10. Eshleman SH, Hoover DR, Chen S, Hudelson SE, Guay LA, Mwatha A, et al
. Resistance after single-dose nevirapine prophylaxis emerges in a high proportion of Malawian newborns. AIDS 2005; 19:2167–2169.
11. Ribaudo HJ, Haas DW, Tierney C, Kim RB, Wilkinson GR, Gulick RM, et al
. Pharmacogenetics of plasma efavirenz exposure after treatment discontinuation: an Adult AIDS Clinical Trials Group Study. Clin Infect Dis 2006; 42:401–407.
12. Arribas JR, Pulido F, Delgado R, Lorenzo A, Miralles P, Arranz A, et al
. Lopinavir/ritonavir as single-drug therapy for maintenance of HIV-1 viral suppression: 48-week results of a randomized, controlled, open-label, proof-of-concept pilot clinical trial (OK Study). J Acquir Immune Defic Syndr 2005; 40:280–287.
13. Swindells S, DiRienzo AG, Wilkin T, Fletcher CV, Margolis DM, Thal GD, et al
. Regimen simplification to atazanavir-ritonavir alone as maintenance antiretroviral therapy after sustained virologic suppression. JAMA 2006; 296:806–814.
14. Cohen CJ, Colson AE, Sheble-Hall AG, McLaughlin KA, Morse GD. Pilot study of a novel short-cycle antiretroviral treatment interruption strategy: 48-week results of the five-days-on, two-days-off (FOTO) study. HIV Clin Trials 2007; 8:19–23.
15. Ananworanich J, Nuesch R, Le Braz M, Chetchotisakd P, Vibhagool A, Wicharuk S, et al
. Failures of 1 week on, 1 week off antiretroviral therapies in a randomized trial. AIDS 2003; 17:F33–37.
16. Bangsberg DR, Moss AR, Deeks SG. Paradoxes of adherence and drug resistance to HIV antiretroviral therapy. J Antimicrob Chemother 2004; 53:696–699.
17. Bangsberg DR, Charlebois ED, Grant RM, Holodniy M, Deeks SG, Perry S, et al
. High levels of adherence do not prevent accumulation of HIV drug resistance mutations. AIDS 2003; 17:1925–1932.
18. Haynes RB. A critical review of the determinants of patient compliance with therapeutic regimens. In: Sackett DL, Haynes RB, editors. Compliance with therapeutic regimens. Baltimore: Johns Hopkins University Press; 1976.
19. Cramer JA, Roy A, Burrell A, Fairchild CJ, Fuldeore MJ, Ollendorf DA, Wong PK. Medication compliance and persistence: terminology and definitions. Value Health 2008; 11:44–47.
20. Bangsberg DR. A paradigm shift to prevent HIV drug resistance. PLoS Med 2008; 5:e111–e1111.
21. Peterson AM, Nau DP, Cramer JA, Benner J, Gwadry-Sridhar F, Nichol M. A checklist for medication compliance and persistence studies using retrospective databases. Value Health 2007; 10:3–12.
22. Gwadry-Sridhar FH, Manias E, Zhang Y, Roy A, Yu-Isenberg K, Hughes DA, Nichol MB. A framework for planning and critiquing medication compliance and persistence research using prospective study designs. Clin Ther 2009; 31:421–435.
23. Knobel H, Urbina O, Gonzalez A, Sorli ML, Montero M, Carmona A, Guelar A. Impact of different patterns of nonadherence on the outcome of highly active antiretroviral therapy in patients with long-term follow-up. HIV Med 2009; 10:364–369.
24. Spacek LA, Shihab HM, Kamya MR, Mwesigire D, Ronald A, Mayanja H, et al
. Response to antiretroviral therapy in HIV-infected patients attending a public, urban clinic in Kampala, Uganda. Clin Infect Dis 2006; 42:252–259.
25. Kavasery R, Galai N, Astemborski J, Lucas GM, Celentano DD, Kirk GD, Mehta SH. Nonstructured treatment interruptions among injection drug users in Baltimore, MD. J Acquir Immune Defic Syndr 2009; 50:360–366.
26. Oyugi JH, Byakika-Tusiime J, Ragland K, Laeyendecker O, Mugerwa R, Kityo C, et al
. Treatment interruptions predict resistance in HIV-positive individuals purchasing fixed-dose combination antiretroviral therapy in Kampala, Uganda. AIDS 2007; 21:965–971.
27. Parienti JJ, Massari V, Descamps D, Vabret A, Bouvet E, Larouze B, Verdon R. Predictors of virologic failure and resistance in HIV-infected patients treated with nevirapine- or efavirenz-based antiretroviral therapy. Clin Infect Dis 2004; 38:1311–1316.
28. Hogg RS, Yip B, Chan KJ, Wood E, Craib KJ, O'Shaughnessy MV, Montaner JS. Rates of disease progression by baseline CD4 cell count and viral load after initiating triple-drug therapy. JAMA 2001; 286:2568–2577.
29. Dybul M, Nies-Kraske E, Daucher M, Hertogs K, Hallahan CW, Csako G, et al
. Long-cycle structured intermittent versus continuous highly active antiretroviral therapy for the treatment of chronic infection with human immunodeficiency virus: effects on drug toxicity and on immunologic and virologic parameters. J Infect Dis 2003; 188:388–396.
30. Parienti JJ, Das-Douglas M, Massari V, Guzman D, Deeks SG, Verdon R, Bangsberg DR. Not all missed doses are the same: sustained NNRTI treatment interruptions predict HIV rebound at low-to-moderate adherence levels. PLoS One 2008; 3:e2783–e12783.
31. Parienti JJ, Ragland K, Lucht F, de la Blanchardiere A, Dargere S, Yazdanpanah Y, et al. Average adherence to boosted protease inhibitor therapy, rather than the pattern of missed doses, as a predictor of HIV RNA replication
. Clin Infect Dis
32. Palella FJ Jr, Chmiel JS, Moorman AC, Holmberg SD, Investigators HIVOS. Durability and predictors of success of highly active antiretroviral therapy for ambulatory HIV-infected patients. AIDS 2002; 16:1617–1626.
33. Willig JH, Abroms S, Westfall AO, Routman J, Adusumilli S, Varshney M, et al
. Increased regimen durability in the era of once-daily fixed-dose combination antiretroviral therapy. AIDS 2008; 22:1951–1960.
34. Klein MB, Willemot P, Murphy T, Lalonde RG. The impact of initial highly active antiretroviral therapy on future treatment sequences in HIV infection. AIDS 2004; 18:1895–1904.
35. The ART-LINC Collaboration of the International Databases to Evaluate AIDS (IeDEA). Antiretroviral therapy in resource-limited settings 1996 to 2006: patient characteristics, treatment regimens and monitoring in sub-Saharan Africa, Asia and Latin America
. Trop Med Int Health
36. Li X, Margolick JB, Conover CS, Badri S, Riddler SA, Witt MD, Jacobson LP. Interruption and discontinuation of highly active antiretroviral therapy in the multicenter AIDS cohort study. J Acquir Immune Defic Syndr 2005; 38:320–328.
37. Willig JH, Echevarria J, Westfall AO, Iglesias D, Henostroza G, Seas C, et al
. Durability of initial antiretroviral therapy in a resource constrained setting and the potential need for zidovudine weight-based dosing. J Acquir Immune Defic Syndr 2010; 53:215–221.
38. Vo TT, Ledergerber B, Keiser O, Hirschel B, Furrer H, Battegay M, et al
. Durability and outcome of initial antiretroviral treatments received during 2000–2005 by patients in the Swiss HIV Cohort Study. J Infect Dis 2008; 197:1685–1694.
39. Lodwick RK, Smith CJ, Youle M, Lampe FC, Tyrer M, Bhagani S, et al
. Stability of antiretroviral regimens in patients with viral suppression. AIDS 2008; 22:1039–1046.
40. Kiguba R, Byakika-Tusiime J, Karamagi C, Ssali F, Mugyenyi P, Katabira E. Discontinuation and modification of highly active antiretroviral therapy in HIV-infected Ugandans: prevalence and associated factors. J Acquir Immune Defic Syndr 2007; 45:218–223.
41. Pence BW, Ostermann J, Kumar V, Whetten K, Thielman N, Mugavero MJ. 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.
42. Kumarasamy N, Vallabhaneni S, Cecelia AJ, Yepthomi T, Balakrishnan P, Saghayam S, et al
. Reasons for modification of generic highly active antiretroviral therapeutic regimens among patients in southern India. J Acquir Immune Defic Syndr 2006; 41:53–58.
43. Springer SA, Friedland GH, Doros G, Pesanti E, Altice FL. Antiretroviral treatment regimen outcomes among HIV-infected prisoners. HIV Clin Trials 2007; 8:205–212.
44. MacArthur RD, Novak RM, Peng G, Chen L, Xiang Y, Hullsiek KH, et al
. A comparison of three highly active antiretroviral treatment strategies consisting of nonnucleoside reverse transcriptase inhibitors, protease inhibitors, or both in the presence of nucleoside reverse transcriptase inhibitors as initial therapy (CPCRA 058 FIRST Study): a long-term randomised trial. Lancet 2006; 368:2125–2135.
45. Braithwaite RS, Kozal MJ, Chang CC, Roberts MS, Fultz SL, Goetz MB, et al
. Adherence, virological and immunological outcomes for HIV-infected veterans starting combination antiretroviral therapies. AIDS 2007; 21:1579–1589.
46. Domingo P, Suárez-Lozano I, Torres F, Teira R, Lopez-Aldeguer J, Vidal F, et al
. First-line antiretroviral therapy with efavirenz or lopinavir/ritonavir plus two nucleoside analogues: the SUSKA study, a nonrandomized comparison from the VACH cohort. J Antimicrob Chemother 2008; 61:1348–1358.
47. Sax PE, Tierney C, Collier AC, Fischl MA, Mollan K, Peeples L, et al. Abacavir-lamivudine versus tenofovir-emtricitabine for initial HIV-1 therapy
. N Engl J Med
48. d'Arminio Monforte A, Lepri AC, Rezza G, Pezzotti P, Antinori A, Phillips AN, et al
. Insights into the reasons for discontinuation of the first highly active antiretroviral therapy (HAART) regimen in a cohort of antiretroviral naïve patients. AIDS 2000; 14:499–507.
49. Mocroft A, Youle M, Moore A, Sabin CA, Madge S, Lepri AC, et al
. Reasons for modification and discontinuation of antiretrovirals: results from a single treatment centre. AIDS 2001; 15:185–194.
50. van Roon EN, Verzijl JM, Juttmann JR, Lenderink AW, Blans MJ, Egberts AC. Incidence of discontinuation of highly active antiretroviral combination therapy (HAART) and its determinants. J Acquir Immune Defic Syndr Hum Retrovirol 1999; 20:290–294.
51. Ahdieh-Grant L, Tarwater PM, Schneider MF, Anastos K, Cohen M, Khalsa A, et al
. Factors and temporal trends associated with highly active antiretroviral therapy discontinuation in the Women's Interagency HIV Study. J Acquir Immune Defic Syndr 2005; 38:500–503.
52. Chen RY, Westfall AO, Mugavero MJ, Cloud GA, Raper JL, Chatham AG, et al
. Duration of highly active antiretroviral therapy regimens. Clin Infect Dis 2003; 37:714–722.
53. Hooshyar D, Napravnik S, Miller WC, Eron JJ Jr. Effect of hepatitis C coinfection on discontinuation and modification of initial HAART in primary HIV care. AIDS 2006; 20:575–583.
54. Crystal S, Sambamoorthi U, Moynihan PJ, McSpiritt E. Initiation and continuation of newer antiretroviral treatments among Medicaid recipients with AIDS. J Gen Intern Med 2001; 16:850–859.
55. Robison LS, Westfall AO, Mugavero MJ, Kempf MC, Cole SR, Allison JJ, et al
. Short-term discontinuation of HAART regimens more common in vulnerable patient populations. AIDS Res Hum Retroviruses 2008; 24:1347–1355.
56. Palepu A, Tyndall MW, Chan K, Wood E, Montaner JS, Hogg RS. Initiating highly active antiretroviral therapy and continuity of HIV care: the impact of incarceration and prison release on adherence and HIV treatment outcomes. Antivir Ther 2004; 9:713–719.
57. Springer SA, Pesanti E, Hodges J, Macura T, Doros G, Altice FL. Effectiveness of antiretroviral therapy among HIV-infected prisoners: reincarceration and the lack of sustained benefit after release to the community. Clin Infect Dis 2004; 38:1754–1760.
58. Kerr T, Marshall A, Walsh J, Palepu A, Tyndall M, Montaner J, et al
. Determinants of HAART discontinuation among injection drug users. AIDS Care 2005; 17:539–549.
59. Zahari MM, Bae WH, Zainal NZ, Habil H, Kamarulzaman A, Altice FL. Psychiatric and substance abuse comorbidity among HIV seropositive and HIV seronegative prisoners in Malaysia. Am J Drug Alcohol Abuse 2010; 36:31–38.
60. Altice FL, Kamarulzaman A, Soriano VV, Schechter M, Friedland GH. Treatment of medical, psychiatric, and substance-use comorbidities in people infected with HIV who use drugs
61. Baillargeon J, Giordano TP, Rich JD, Wu ZH, Wells K, Pollock BH, Paar DP. Accessing antiretroviral therapy following release from prison. JAMA 2009; 301:848–857.
62. Pant Pai N, Estes M, Moodie EE, Reingold AL, Tulsky JP. The impact of antiretroviral therapy in a cohort of HIV infected patients going in and out of the San Francisco County jail. PLoS One 2009; 4:e7115–e17115.
63. O'Brien ME, Clark RA, Besch CL, Myers L, Kissinger P. Patterns and correlates of discontinuation of the initial HAART regimen in an urban outpatient cohort. J Acquir Immune Defic Syndr 2003; 34:407–414.
64. Ortiz R, Dejesus E, Khanlou H, Voronin E, van Lunzen J, Andrade-Villanueva J, et al
. Efficacy and safety of once-daily darunavir/ritonavir versus lopinavir/ritonavir in treatment-naive HIV-1-infected patients at week 48. AIDS 2008; 22:1389–1397.
65. Zhang Y, Adams AS, Ross-Degnan D, Zhang F, Soumerai SB. Effects of prior authorization on medication discontinuation among Medicaid beneficiaries with bipolar disorder. Psychiatr Serv 2009; 60:520–527.
66. Wilson J, Axelsen K, Tang S. Medicaid prescription drug access restrictions: exploring the effect on patient persistence with hypertension medications
. Am J Manag Care
67. Lucas GM, Chaudhry A, Hsu J, Woodson T, Lau B, Olsen Y, et al. Clinic-based treatment of opioid-dependent HIV-infected patients versus referral to an opioid treatment program: a randomized trial
. Ann Intern Med
68. Ostrop NJ, Gill MJ. Antiretroviral medication adherence and persistence with respect to adherence tool usage. AIDS Patient Care STDS 2000; 14:351–358.
69. Altice FL, Maru DS, Bruce RD, Springer SA, Friedland GH. Superiority of directly administered antiretroviral therapy over self-administered therapy among HIV-infected drug users: a prospective, randomized, controlled trial. Clin Infect Dis 2007; 45:770–778.
70. Maru DS, Bruce RD, Walton M, Mezger JA, Springer SA, Shield D, Altice FL. Initiation, adherence, and retention in a randomized controlled trial of directly administered antiretroviral therapy. AIDS Behav 2008; 12:284–293.
71. Maru DS, Bruce RD, Walton M, Springer SA, Altice FL. Persistence of virological benefits following directly administered antiretroviral therapy among drug users: results from a randomized controlled trial. J Acquir Immune Defic Syndr 2009; 50:176–181.
Keywords:© 2011 Lippincott Williams & Wilkins, Inc.
adherence; antiretroviral medication; HIV/AIDS; measurement; persistence; retention in care