Abstract
BACKGROUND: Medication adherence plays an important role for patients living with HIV and achieving the treatment goal of viral suppression. A goal adherence rate of at least 90% has been previously cited and endorsed; however, studies have demonstrated that lower rates of adherence may still lead to high rates of viral suppression. Adherence rates are increasingly being used by payers to assess pharmacy performance.
OBJECTIVE: To determine if there is a difference in the odds of achieving viral suppression with a proportion of days covered (PDC) at least 90% compared with patients with lower PDC levels. Additionally, to determine if demographic factors, including age, ethnicity, sex, primary antiretroviral regimen type, payer type, primary pharmacy location, and refill assistance program enrollment, impact the odds of achieving viral suppression.
METHODS: This retrospective observational study included patients who were aged 18 years or older; were diagnosed with HIV; had at least 2 occurrences of dispensed antiretrovirals between July 1, 2020, and June 30, 2021, within the health system; and had at least 1 HIV-RNA viral load recorded between these dates. PDC was calculated at the generic product identifier (GPI) level. For patients receiving multiple GPIs in this period, a weighted average PDC was calculated. A logistic regression analysis was performed, and odds ratios were calculated with 95% confidence for each demographic factor to determine correlation with viral suppression.
RESULTS: 1,629 patients were included. Overall, 1,516 (93.1%) patients were virally suppressed. 106 (6.5%) patients had a PDC lower than 50% and 639 (39.2%) had a PDC of at least 90%. Of the patients with a PDC lower than 50%, 80 (75.5%) achieved viral suppression as did 617 (96.6%) patients with a PDC of at least 90%. Age and insurance type significantly impacted viral suppression. No statistically significant difference was found between the odds of achieving viral suppression until PDC was below 75%. Patients with a PDC of less than 50% or a PDC of 50% to less than 75% were less likely to achieve viral suppression than patients with a PDC of at least 90% (P < 0.001).
CONCLUSIONS: Patients with adherence rates above 75% achieve similar results compared with patients with adherence rates above 90%. High population viral suppression may be achieved with as few as 39.2% of patients achieving a PDC greater than 90%. Using these results, the Pharmacy Quality Alliance and other guidance setting entities should consider lowering the at least 90% threshold as well as providing further guidance on how payers should use results and network benchmarking when creating pharmacy quality performance measures.
Plain language summary
Taking medication is an important part of reaching treatment goals for people with HIV. We studied the relationship between taking medication and people reaching the goal of having low levels of virus in their blood (viral suppression). Overall, 93% of the people achieved viral suppression. We found that people who filled their medication as directed at least 75% of the time had similar results as people who did so more than 90% of the time.
Implications for managed care pharmacy
The use of pharmacy performance measures has grown dramatically in recent years. Although a variety of measures have been endorsed, how to use them to evaluate pharmacy performance remains variable. Managed care organizations should consider what performance goals they are using to evaluate pharmacy performance for medication adherence in HIV. Lower thresholds than what are often used may be sufficient in achieving favorable real-world outcomes.
According to the Centers for Disease Control and Prevention (CDC), of the estimated 1.2 million people living with HIV in the United States in 2019, only 57% were virally suppressed with an even smaller number maintaining suppression.1 Furthermore, the CDC reports that, although 87% of these patients received a formal diagnosis, only 66% received medical care for their HIV and 50% were lost to follow up.1 People living with HIV should be initiated on antiretroviral therapy (ART) at the time of diagnosis, regardless of immune function, to achieve and maintain viral suppression, reduce mortality, and improve quality of life.2 The CDC defines viral suppression as HIV-RNA levels of less than 200 copies per milliliter.
The likelihood of achieving viral suppression increases for patients who receive medical care. Viral suppression was achieved in 86.2% of patients receiving any HIV medical care in 2019 (eg, CD4 test or viral load test) and 89.5% of patients with at least 1 viral load test.1 Access to clinics focused on the comprehensive treatment of HIV, such as those commonly created within health systems and those funded by the Ryan White HIV/AIDS Program (RWHAP), may help keep patients connected to care and engaged in their goals to achieve viral suppression.3 Integrated health system HIV specialty pharmacy programs have demonstrated the ability to promote high levels of medication adherence and viral suppression.4 The RWHAP was established in 1990 to financially aid both direct health care services and support services (eg, housing, transportation, and food banks) for people living with HIV. The RWHAP aims to improve early HIV diagnosis, connection and retention to care, affordable and clinically appropriate medical treatment, and sustained viral suppression. Overall, 89.4% of the 357,914 patients seen for care within RWHAP centers during 2020 achieved viral suppression at their most recent evaluation.3 This percentage has increased from 84.9% in 2016.3
Medication adherence plays a key role in patients achieving viral suppression. A common method for approximating medication adherence is through the calculation of the proportion of days covered (PDC) using pharmacy claims data. Early ART regimens were often protease inhibitor (PI)–based, poorly tolerated, and complex because of patients needing to take multiple products and multiple doses per day. These regimens were also believed to require very high levels of adherence (eg, PDC > 95%) to achieve and maintain viral suppression. With the development of newer, more potent medications and new mechanisms of action as well as the reduced pill burden offered by many modern dosage forms, it is unclear whether such high levels of medication adherence remain necessary to achieve and maintain viral suppression.5 The Pharmacy Quality Alliance (PQA) endorses a goal of a PDC greater than 90% for at least 3 antiretroviral (ARV) medications for an individual patient to be considered adherent. The PQA also supports the use of the percentage of patients with a PDC greater than 90% as a quality measure to assess pharmacy performance.6 Recently, additional treatment options involving only 2 ARV medications have been brought to market and should be accounted for when assessing viral suppression within patient populations.
Recent research on the correlation between PDC and viral suppression has demonstrated that the odds of viral suppression are not significantly improved when comparing patients with a PDC greater than 90% with those between 80% and less than 90%5,7 Byrd et al found no statistically significant difference in the odds of achieving viral suppression for patients with a PDC greater than 90% compared with those with a PDC of 80% to less than 85% (adjusted odds ratio [aOR] = 0.49 [95% CI = 0.23-1.04]; P = 0.0627) or a PDC of 85% to less than 90% (aOR = 0.96 [95% CI = 0.49-1.90]; P = 0.9138) when controlling for other factors, such as age, race, sex, regimen, and type of insurance. Furthermore, they estimated that the adherence level necessary to achieve viral suppression in 90% of viral load tests was 82%.5 Numerous studies have found that moderate levels of ARV medication adherence, especially adherence in the range of 80% to 90%, do not appear to significantly impact the odds of achieving viral suppression compared with historically high adherence targets (ie, ≥90% or ≥95%) when contemporary ART regimens are used.5,8-14
Pharmacy performance measures that use the percentage of patients achieving a goal PDC threshold are being increasingly used to adjust reimbursements to pharmacies by third party payers. Performance measures are used within the Medicare Part D Direct and Indirect Remuneration fees as well as by commercial payers who have implemented performance requirements within pharmacy networks. According to the Centers for Medicare and Medicaid Services, Direct and Indirect Remuneration fees grew by a staggering 107,400% between 2010 and 2020.15 Both Medicare and commercial plans commonly use or adapt PQA-endorsed measures when creating pharmacy performance measures for their network. Payers including adherence to ART within their pharmacy performance measures are likely to use the PQA-endorsed measure of the percentage of patients with a PDC greater than 90%; however, no guidance is provided to payers by the PQA on what percentage of patients achieving this threshold should constitute poor, good, or excellent performance. This lack of guidance leaves payers to decide what goal to set, which has resulted in either (1) adoption of a network benchmarking process or, of a more concerning nature, (2) development of arbitrary performance goals (eg, 98% of patients must achieve PDC > 90% for the pharmacy to demonstrate good performance). Specific to the use of PDC for ART as a pharmacy performance measure, the high PDC threshold endorsed by the PQA combined with this risk of arbitrary goal setting by payers creates the potential for pharmacies to be punished for “poor” performance despite their patients having clinically reasonable rates of medication adherence and high rates of viral suppression.
To further evaluate the correlation between medication adherence and viral suppression, this study assessed the response to treatment for patients receiving care within the Johns Hopkins Health System who had ART prescriptions dispensed from any of the outpatient and specialty pharmacies operated by the health system. Johns Hopkins Outpatient Pharmacies (JHOP) operates a total of 11 outpatient and specialty pharmacies, all of which may dispense antiretrovirals. The JHOP at the John G. Bartlett Specialty Practice (Bartlett Pharmacy) Clinic is a pharmacy physically located within the Johns Hopkins Hospital infectious disease clinic. This pharmacy services most people living with HIV who fill their prescriptions within Johns Hopkins Health System and offers a refill assistance program (RAP) to assist complex patients with adherence as well as clinical pharmacy services within an integrated care model. Patients enrolled in the RAP often have developed drug resistance or have adherence barriers including but not limited to accessing or organizing medications without monthly refill assistance from the Bartlett Pharmacy team. Although patients are aware that adherence is important to achieve viral suppression, barriers and circumstances, such as lack of access, treatment fatigue, and comorbid conditions, may be present.13 A shared electronic medical record (EMR) is used across the health system to document patient care activities and interventions. Within an integrated health system environment, this study intends to (1) determine if there is a difference in the odds of achieving viral suppression with a PDC greater than 90% compared with patients with lower levels of PDC and (2) determine if demographic factors, including age, ethnicity, sex, primary ART regimen, payer type, primary pharmacy location, and RAP enrollment, impact the odds of achieving viral suppression. This study was determined to qualify as exempt research by the Johns Hopkins Medicine Institutional Review Board and, thus, did not require informed consent or ongoing Institutional Review Board oversight, as it was determined to present no risk or minimal risk to participants.
Methods
This study was retrospective and observational in nature. Patients were included if they (1) were aged 18 years or older; (2) were diagnosed with HIV; (3) had at least 2 ART prescriptions dispensed by JHOP between July 1, 2020, and June 30, 2021; and (4) had at least 1 HIV-1 RNA viral load recorded within the EMR during the same time frame. We elected to include patients with at least 2 antiretrovirals instead of the PQA recommendation of at least 3 antiretrovirals, as recently approved dual medication ART regimens are now available. Patients were excluded if they were using longitudinal ART for indications other than HIV treatment, such as hepatitis B virus or pre-exposure prophylaxis. This study identified patients being treated for hepatitis B virus if they were prescribed monotherapy with lamivudine and identified patients being treated for pre-exposure prophylaxis if their refill history consisted only of single-tablet regimens of either emtricitabine-tenofovir alafenamide fumarate or emtricitabine-tenofovir disoproxil fumarate.
Patient demographic factors (sex, ethnicity, primary ART regimen backbone, payer type, primary dispensing pharmacy location, and patient enrollment in an RAP) were collected from a combination of pharmacy dispensing data and the EMR. Laboratory data to assess viral suppression were collected from the EMR (most recent HIV-1 RNA lab result). Pharmacy dispensing claims data were used to calculate the PDC. A variable PDC was calculated at the level of generic product identifier (GPI) for each prescription fill during the study period to estimate the proportion of days in which a patient was in possession of their medication regimen. Days covered were identified from the date of first fill to the date of the last fill within the study period. A fill was defined as the sold date within pharmacy dispensing software. An excess supply of medication was shifted forward, never backward. Any excess supply extending beyond the end of the period was truncated. Patients with dispenses of more than 1 GPI within the period had a weighted average PDC calculated to yield a single summary PDC for each patient. This weighted average was based on the number of fills for each GPI. For example, a patient with 6 total fills of zidovudine with a PDC of 0.91 and 9 total fills of tenofovir disoproxil fumarate/emtricitabine with a PDC of 0.87 would have a weighted PDC of 0.886. Each patient was grouped into 1 of 6 buckets based on a range of PDCs (<50%, 50% to <75%, 75% to <80%, 80% to <85%, 85% to <90%, and ≥90%). Patients who switched insurance coverage were categorized according to the insurance type they had for a longer duration throughout the study period. Similarly, patients who filled ART prescriptions at multiple JHOP locations within the study period were assigned to the pharmacy that was used for most of their fills. In the event of a tie, patients were assigned based on the pharmacy or insurance type that was used for their most recent dispense. Patients were grouped into 1 of 4 categories for ART regimen: (1) integrase inhibitor (INSTI)–based, (2) nonnucleoside reverse transcriptase inhibitor (NNRTI)–based, (3) PI-based, and (4) all other regimens, which included any regimen unable to be categorized in one of the preceding 3 groups. The INSTI-based, NNRTI-based, and PI-based regimen groupings included 1 medication of the named mechanism of action (eg, 1 NNRTI medication) along with 2 nucleoside reverse transcriptase inhibitors (NRTIs). If patients switched therapies throughout the study period, they were assigned to the all other regimens category.
Viral suppression was defined as HIV-1 RNA less than 200 copies/mL at the most recent draw in the 12-month study period. If more than 1 HIV-1 RNA viral load was drawn during the study period, the most recent draw was used. All demographic factors and PDC were assessed to determine correlation with viral suppression. Descriptive statistics as well as bivariable and multivariable logistic regression was performed using Stata (StataCorp LLC, 2021; Stata Statistical Software: release 17). ORs for achieving viral suppression were calculated with 95% confidence and evaluated at a 0.05 level of significance. All demographic factors found to have a statistically significant impact on the odds of achieving viral suppression were analyzed for potential inclusion in the final multivariable regression analysis. These variables were added to the multivariable regression in a stepwise fashion to assess impact. During the analysis and summary of findings, we selected certain demographic categories to serve as a comparative reference with other variables within the category based on their prevalence or impact (eg, Medicaid compared with all other payers).
Results
A total of 1,629 patients were analyzed after application of the inclusion and exclusion criteria as outlined in Figure 1. Overall, 1,516 (93.1%) patients were virally suppressed. Demographic factors are summarized in Table 1. The median patient age was 52 years, 64.8% of patients were aged at least 50 years, 82.9% identified as Black non-Hispanic individuals, and 87.4% of patients used Bartlett Pharmacy as their primary JHOP dispensing location. INSTI-based ART regimens were the most commonly prescribed (Table 1). Only 106 (6.5%) patients had a PDC lower than 50%, and 639 (39.2%) had a PDC of at least 90%. Of the patients with a PDC lower than 50%, 80 (75.5%) were virally suppressed, whereas 617 (96.6%) patients with a PDC of at least 90% were virally suppressed. The rate of viral suppression in all other PDC categories ranged from 88.6% to 97.8% (Table 2). A weighted average PDC was calculated for 266 patients (16%) because of their filling of multitablet regimens during the study time frame with discordant individual product PDCs.
FIGURE 1.
Patients Evaluated
TABLE 1.
Baseline Characteristics
| Characteristic | Total (N = 1,629) |
|---|---|
| Age, median (IQR), years | 52 (44-62) |
| Aged at least 50 years, n (%) | 1,056 (64.8) |
| Black non-Hispanic, n (%) | 1,351 (82.9) |
| Sex, n (%) | |
| Male | 989 (60.7) |
| Female | 640 (39.3) |
| Medical insurance, n (%) | |
| Medicare | 465 (28.5) |
| Medicaid | 706 (43.3) |
| Commercial | 459 (28.2) |
| Viral suppression (<200 copies/mL), n (%) | 1,516 (93.1) |
| PDC category, n (%) | |
| Less than 50% | 106 (6.5) |
| 50% to less than 75% | 385 (23.6) |
| 75% to less than 80% | 141 (8.7) |
| 80% to less than 85% | 185 (11.4) |
| 85% to less than 90% | 173 (10.6) |
| At least 90% | 639 (39.2) |
| RAP-enrolled patients, n (%) | 425 (26.1) |
| Primary pharmacy location, n (%) | |
| JHOP Bartlett Pharmacy | 1,423 (87.4) |
| JHOP Bayview Pharmacy | 60 (3.7) |
| JHOP Bayview Annex Pharmacy | 4 (0.2) |
| JHOP Green Spring Station Pharmacy | 18 (1.1) |
| JHOP Discharge Pharmacy | 9 (0.6) |
| JHOP JHOC Pharmacy | 10 (0.6) |
| JHOP Monument Street Pharmacy | 55 (3.8) |
| JHOP Arcade Pharmacy | 43 (2.6) |
| JHOP Viragh Pharmacy | 2 (0.1) |
| JHOP Weinberg Pharmacy | 5 (0.3) |
| ART regimen category, n(%) | |
| INSTI-based | 1,059 (65) |
| NNRTI-based | 70 (4.3) |
| PI-based | 81 (5) |
| All other | 419 (25.7) |
| Top “all other” regimen classifications, n (%) | |
| INSTI/NRTI (2)/PI/BOOSTER | 59 (3.62) |
| PI/BOOSTER/INSTI | 49 (3.01) |
| INSTI/NNRTI/NRTI (2) | 35 (2.15) |
| PI/BOOSTER/INSTI/NNRTI | 32 (1.96) |
| INSTI (2)/NRTI (4) | 24 (1.47) |
| INSTI/NNRTI | 21 (1.29) |
ART = antiretroviral therapy; INSTI = integrase inhibitor; JHOC = Johns Hopkins Outpatient Center; JHOP = Johns Hopkins Outpatient Pharmacies; NNRTI = nonnucleoside reverse transcriptase inhibitor; NRTI = nucleoside reverse transcriptase inhibitor; PI = protease inhibitor; RAP = refill assistance program.
TABLE 2.
Rates of Viral Suppression by PDC Category
| PDC category (N = 1,629) | n (%) | Patients achieving viral suppression (HIV-RNA <200 copies/mL), n (%) |
|---|---|---|
| Less than 50% | 106 (6.5) | 80 (75.5) |
| 50% to less than 75% | 285 (23.6) | 341 (88.6) |
| 75% to less than 80% | 141 (8.7) | 135 (95.7) |
| 80% to less than 85% | 185 (11.4) | 181 (97.8) |
| 85% to less than 90% | 173 (10.6) | 162 (93.6) |
| At least 90% | 639 (39.2) | 617 (96.6) |
| All patients | 1,629 (100) | 1,516 (93.1) |
PDC = proportion of days covered.
Table 3 contains the results of the bivariable logistic regression analyses. The results identified a statistically significant improvement in the odds of viral suppression for those aged at least 50 years (OR = 2.300; 95% CI = 1.565-3.380; P < 0.001) and those with commercial insurance (OR = 2.216; 95% CI = 1.353-3.630; P < 0.002) or Medicare insurance (OR = 2.488; 95% CI = 1.493-4.146; P < 0.001) compared with Medicaid. Patients within the 2 lowest category ranges for PDC (<50% and 50% to <75%) were found to be significantly less likely to achieve viral suppression compared with patients with a PDC of at least 90%, those with a PDC of less than 50% (OR = 0.110; 95% CI = 0.059-0.203; P < 0.001), and those with a PDC of 50% to less than 75% (OR = 0.276; 95% CI = 0.163-0.469; P < 0.001). Compared with those with a PDC of at least 90%, no statistically significant difference in the odds of achieving viral suppression was identified for the PDC categories of 75% to less than 80%, 80% to less than 85%, or 85% to 90%. Male sex, enrollment in the RAP, or Black non-Hispanic race were all found to be nonsignificant factors on the odds of viral suppression (OR = 0.737, 95% CI = 0.491-1.106, P = 0.141; OR = 1.195, 95% CI = 0.760-1.880, P = 0.440; OR = 0.729, 95% CI = 0.417-1.276, P = 0.269, respectively). INSTI-based (OR = 1.459; 95% CI = 0.677-3.141; P = 0.335), NNRTI-based (OR = 2.447; 95% CI = 0.623-9.610; P = 0.200), and all other ART regimens (OR = 1.530; 95% CI = 0.671-3.490; P = 0.312) were also found to be nonsignificant. Although not statistically significant (P = 0.051), there was a trend toward higher odds of viral suppression for patients who filled the majority of their prescriptions at the JHOP at the Bartlett location compared with other JHOP locations (OR = 1.642; 95% CI = 0.9980-2.703; P = 0.051).
TABLE 3.
Bivariate Analysis of Factors Associated With HIV Viral Suppression (HIV-RNA Levels <200 Copies/mL)
| Characteristic | Bivariate odds ratio (95% CI) | P value |
|---|---|---|
| Aged at least 50 y | 2.300 (1.565-3.380) | <0.001 |
| Black non-Hispanic | 0.729 (0.417-1.276) | 0.269 |
| Male | 0.737 (0.491-1.106) | 0.141 |
| Insurance | ||
| Medicare | 2.488 (1.493-4.146) | <0.001 |
| Medicaid | Ref | — |
| Commercial | 2.216 (1.353-3.630) | 0.002 |
| PDC level | ||
| Less than 50% | 0.110 (0.059-0.203) | <0.001 |
| 50% to less than 75% | 0.276 (0.163-0.469) | <0.001 |
| 75% to less than 80% | 0.802 (0.319-2.017) | 0.639 |
| 80% to less than 85% | 1.613 (0.549-4.742) | 0.385 |
| 85% to less than 90% | 0.525 (0.250-1.105) | 0.090 |
| At least 90% | Ref | — |
| Refill assistance program | 1.195 (0.760-1.880) | 0.440 |
| JHOP Bartlett Pharmacy | 1.642 (0.998-2.703) | 0.051 |
| ART regimen | ||
| INSTI-based | 1.459 (0.677-3.141) | 0.335 |
| NNRTI-based | 2.447 (0.623-9.610) | 0.200 |
| PI-based | Ref | — |
| All other | 1.530 (0.671-3.490) | 0.312 |
ART = antiretroviral therapy; INSTI = integrase inhibitor; JHOP = Johns Hopkins Outpatient Pharmacies; NNRTI = nonnucleoside reverse transcriptase inhibitor; NRTI = nucleoside reverse transcriptase inhibitor; PDC = proportion of days covered; PI = protease inhibitor; Ref = reference; y = years.
After the completion of stepwise multivariable logistic regression, the following variables were included in the final model: age, insurance type, PDC, and sex (Table 4). The results of multivariable analysis identified a continued statistically significant difference in the odds of viral suppression for those aged at least 50 years (OR = 1.735; 95% CI = 1.142-2.636; P = 0.010) and those with commercial insurance (OR = 1.975; 95% CI = 1.190-3.277; P = 0.008) or Medicare insurance (OR = 1.832; 95% CI = 1.063-3.159; P = 0.029) compared with Medicaid. Patients within the 2 lowest category ranges for PDC were still found to be significantly less likely to achieve viral suppression compared with patients with a PDC of at least 90%, those with a PDC of less than 50% (OR = 0.135; 95% CI = 0.072-0.253; P < 0.001), and those with a PDC of 50% to less than 75% (OR = 0.295; 95% CI = 0.173-0.503; P < 0.001). Compared with those with a PDC of at least 90% and as seen in the bivariable analysis, no statistically significant difference was identified for those with PDC categories of 75% to less than 80%, 80% to less than 85%, or 85% to 90%. The only demographic factor found to differ in impact compared with the findings observed during bivariable analysis was sex, with male sex found to be less likely to achieve viral suppression during multivariable analysis (OR = 0.573; 95% CI = 0.375-0.878; P = 0.010).
TABLE 4.
Comparison of Bivariate and Multivariate Analysis of Factors Associated With HIV Viral Suppression (HIV-RNA Levels <200 Copies/mL)
| Characteristic | Bivariate odds ratio (95% CI) | P value | Multivariable (95% CI) | P value |
|---|---|---|---|---|
| Aged at least 50 y | 2.300 (1.565-3.380) | < 0.001 | 1.735 (1.142-2.636) | 0.010 |
| Males | 0.737 (0.491-1.106) | 0.141 | 0.574 (0.375-0.878) | 0.010 |
| Insurance | ||||
| Medicare | 2.488 (1.493-4.146) | < 0.001 | 1.832 (1.063-3.159) | 0.029 |
| Medicaid | Ref | — | Ref | — |
| Commercial | 2.216 (1.353-3.630) | 0.002 | 1.975 (1.190-3.277) | 0.008 |
| PDC level | ||||
| Less than 50% | 0.110 (0.059-0.203) | < 0.001 | 0.135 (0.072-0.253) | < 0.001 |
| 50% to less than 75% | 0.276 (0.163-0.469) | < 0.001 | 0.295 (0.173-0.503) | < 0.001 |
| 75% to less than 80% | 0.802 (0.319-2.017) | 0.639 | 0.809 (0.320-2.041) | 0.653 |
| 80% to less than 85% | 1.613 (0.549-4.742) | 0.385 | 1.585 (0.538-4.671) | 0.403 |
| 85% to less than 90% | 0.525 (0.250-1.105) | 0.090 | 0.539 (0.255-1.139) | 0.105 |
| At least 90% | Ref | — | Ref | — |
PDC = proportion of days covered; Ref = reference; y = years.
Discussion
This study found that patients with a PDC of greater than 75% did not have a statistically significant difference in their odds of achieving viral suppression compared with patients with a PDC of at least 90%. These findings support the growing literature that the very high levels of adherence currently endorsed by the PQA may no longer be necessary to achieve high levels of population viral suppression rates. To briefly summarize this literature in addition to the previously cited evidence from Byrd et al, a literature review by Kobin and Sheth9 published in 2011 concluded that virologic suppression and reduced disease progression is more likely as adherence rates increase, but they also concluded that patients with moderate adherence levels may also achieve virologic suppression while on potent ARV therapy. Sutton et al12 found no statistical difference in the odds of viral suppression with adherence levels of at least 80% compared with levels of at least 95% in patients receiving single-tablet regimens. Sokpa et al11 found that adherence levels were predictive of future virologic failure (defined as HIV-RNA > 200 copies/mL) when the PDC was no higher than 52%. In a study of patients new to therapy, Rodregues et al10 found that viral suppression (defined as viral load <50 copies/mL) was achieved in 76% of all new start patients, with 79.3% of patients with high adherence (PDC ≥ 90%) achieving viral suppression, whereas 71.4% was achieved by patients with intermediate adherence (PDC 85%-89%) and 45.2% for patients with low adherence (PDC ≤ 84%). Of note, the definition of viral suppression used in that study was stricter than the one used in this study and by the CDC (<200 copies/mL). The authors also noted that patients on the modern combination regimen of dolutegravir/lamivudine/tenofovir had an approximately 90% probability of achieving viral suppression at intermediate adherence levels. Although we believe that pharmacists and other health care professionals should always advocate for patients to take their medications as prescribed, these findings, alongside our own, provide some reassurance that patients do not seem to risk failure of viral suppression because of a small-to-moderate number of missed doses and that a lower goal of PDC (eg, ≥75% or ≥80%) may be clinically acceptable.
In our study, the overall population viral suppression rate was 93%; however, only 39.2% of patients achieved the PQA goal of PDC greater than 90%. This population viral suppression rate exceeds the national average of 86.2% as reported by the CDC and RWHAP for patients who have established any HIV medical care and the 89.5% of patients with at least 1 viral load test.1 These findings support our concern that, without additional guidance and benchmarking practices, payers may select arbitrary thresholds to identify good pharmacy performance based on standard classroom grading scales (eg, the pharmacy must have >90% of patients achieve PDC > 90% because >90% is an “A”) that are unrealistic and do not correlate well with meaningful population health goals. Instead, if the percentage of patients achieving a goal PDC is used as a pharmacy performance measure, we believe that network benchmarking (eg, percentile-based thresholds) or thresholds supported by published literature must be used to promote transparency and avoid unrealistic and unscientific goal setting.
For analysis in this study, government insurance type was split into Medicare and Medicaid. Medicaid was selected as the reference value as the most prevalent insurance type in study population. PI-based regimens, although small in this study population, were used as the reference value because of newer medications coming to market that may have less of a need for high levels of adherence. Notably, patients enrolled in the RAP offered by Bartlett Pharmacy did not have significantly better odds of achieving viral suppression. Patients are selected for enrollment in this program based on having complex treatment regimens, difficulty achieving viral suppression, or documented adherence difficulties. Given these factors, we suspect that the patient population enrolled in the RAP program may have had lower baseline odds of achieving viral suppression without such intervention, making comparison of their viral suppression rates compared with unenrolled patients challenging. Additional research evaluating the impact of enrollment in such programs on viral suppression is warranted.
LIMITATIONS
Limitations include (1) not using the exact PDC methods outlined by the PQA, (2) the timing of when the HIV-RNA level was drawn, (3) regimen categorization challenges, and (4) the use of calculation to approximate medication adherence. Because ARV treatment options containing only 2 components now exist, we chose to include patients on those regimens despite the PQA recommendation to evaluate only patients with at least 3 ARV medications. Additionally, we elected to use a weighted average PDC for patients with fills for more than 1 GPI during the study period. Ideally, a PDC would be calculated using a methodology in which a day is only considered covered if the full intended regimen of ARV medications were accounted for on any given day. Developing this methodology proved challenging and prone to error, as it required both breaking down each combination product dispense to individual ARV ingredients per day as well as confidence that the research team always knew the intended full regimen for each day. Given the challenges, we believe that our average PDC methodology is a reasonable estimate of ART adherence. Regarding the use of variable interval PDC, we believe that the use of a variable interval PDC with the denominator defined by the first and last fill of a medication within the period is the most appropriate when using pharmacy dispensing records because of the potential for patients to transfer prescriptions between pharmacies. This decision is in alignment with recommendations from members of the American Society of Health-System Pharmacists Section of Specialty Pharmacy Practitioners Outcomes and Value Section Advisory Group.16
The timing of HIV-RNA levels was another limitation of this study. There was no minimum time requirement instituted between the date of the first fill and the laboratory evaluation of the HIV-RNA levels used to determine viral suppression. Given this, it is possible that some patients could have had laboratory tests evaluated without sufficient time for treatment to take effect.
Patients on complex combinations of products, including changing ART regimens during the study period, introduced another challenge when attempting to categorize ART regimens (eg, INSTI-based regimens). Many patients ended up being classified within the “all other” category for this reason. The classification of ART regimen was based on all the individual ingredients a patient received during the study period, and our definition for identifying a patient in a specific PI-based, INSTI-based, or NNRTI-based category was strict in only allowing an NRTI (2) backbone and a single NNRTI, PI, or INSTI agent. If, for example, a patient had been on an INSTI plus 2 NRTI regimen (INSTI/NRTI[2]) and then were transitioned to a PI/BOOSTER/NRTI(2) regimen during the study period, this patient would be classified as INSTI/NRTI (4)/PI/BOOSTER and would be placed in the “all other” category for analysis.
Additionally, it is worth noting that the use of any adherence measure calculation, such as PDC, carries the limitation of the researcher not knowing with certainty that the patient administered the medication as prescribed. Lastly, our study did not evaluate for differences in adherence and other outcomes by prescriber, clinic of origin, or other factors within the care team that support the patient beyond the pharmacy that may have an impact on patient behavior and engagement in their care.
Conclusions
Highly potent single-tablet ART regimens have become first-line treatment options for many people living with HIV. With this shift in the treatment paradigm, our study supports the conclusion that lower levels of medication adherence (eg, PDC > 75%) may be sufficient to achieve high odds of viral suppression both for individual patients and at the population level. Although patients should strive to take their medications as directed, there is comfort knowing that missing a few days of treatment because of unforeseen circumstances, human error, or both are not likely to have a large impact on their odds of achieving viral suppression.
Additionally, the PQA and payers should consider (1) lowering the ART PDC threshold below 90% (we suggest ≥75% or ≥80%) and (2) providing guidance on the use of network benchmarking to assess pharmacy quality performance until specific thresholds that have been shown to correlate with high degrees of viral suppression can be cited.
REFERENCES
- 1.Center for Disease Control and Prevention. Monitoring selected national HIV prevention and care objectives by using HIV surveillance data—United States and 6 dependent areas, 2019. HIV Surveillance Supplemental Report. 2021;26(No.2). Accessed October 19, 2021. https://www.cdc.gov/hiv/library/reports/hiv-surveillance.html [Google Scholar]
- 2.Panel on Antiretroviral Guidelines for Adults and Adolescents. Guidelines for the Use of Antiretroviral Agents in Adults and Adolescents with HIV. Department of Health and Human Services. Accessed August 19, 2021. https://clinicalinfo.hiv.gov/sites/default/files/guidelines/documents/adult-adolescent-arv/guidelines-adult-adolescent-arv.pdf [Google Scholar]
- 3.Ryan White HIV/AIDS Program. Ryan White HIV/AIDS program annual client-level data report. Health Resources and Services Administration. Accessed March 10, 2022. https://ryanwhite.hrsa.gov/sites/default/files/ryanwhite/data/rwhap-annual-client-level-data-report-2020.pdf
- 4.Barnes E, Zhao J, Giumenta A, Johnson M. The effect of an integrated health system specialty pharmacy on HIV antiretroviral therapy adherence, viral suppression, and CD4 count in an outpatient infectious disease clinic. J Manag Care Spec Pharm. 2020;26(2):95-102. doi:10.18553/jmcp.2020.26.2.95 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Byrd KK, Hou JG, Hazen R, et al. ; Patient-Centered HIV Care Model Team. Antiretroviral adherence level necessary for HIV viral suppression using real-world data. J Acquir Immune Defic Syndr. 2019;82(3):245-51. doi:10.1097/QAI.0000000000002142 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Pharmacy Quality Alliance. Pharmacy Quality Alliance quality measures. Updated March 5, 2021. Accessed March 10, 2022. https://www.pqaalliance.org/pqa-measures [Google Scholar]
- 7.Byrd KK, Hou JG, Bush T, et al. ; Patient-Centered HIV Care Model Team. Adherence and viral suppression among participants of the patient-centered human immunodeficiency virus (HIV) care model project: A collaboration between community-based pharmacists and HIV clinical providers. Clin Infect Dis. 2020;70(5):789-97. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Bangsberg DR. Less than 95% adherence to nonnucleoside reverse-transcriptase inhibitor therapy can lead to viral suppression. Clin Infect Dis. 2006;43(7):939-41. doi:10.1086/507526 [DOI] [PubMed] [Google Scholar]
- 9.Kobin AB, Sheth NU. Levels of adherence required for virologic suppression among newer antiretroviral medications. Ann Pharmacother. 2011;45(3):372-9. doi:10.1345/aph.1P587 [DOI] [PubMed] [Google Scholar]
- 10.Rodrigues VA, Ceccato M das BG, de Oliveria Costa J, Almeida-Brasil CC, Silveria MR, Alfonso Reis E. Levels of adherence to contemporary antiretroviral regimens and the likelihood of viral suppression: A cohort study in a Brazilian metropolis. AIDS Care. 2023;35(7):976-81. doi:10.1080/09540121.2022.2072802 [DOI] [PubMed] [Google Scholar]
- 11.Sokpa D, Lyden E, Fadul N, Bares SH, Havens JP. Antiretroviral refill histories as a predictor of future human immunodeficiency virus viremia. Open Forum Infect Dis. 2022;9(3):ofac024. doi:10.1093/ofid/ofac024 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Sutton SS, Magagnoli J, Hardin JW. Odds of viral suppression by single-tablet regimens, multiple-tablet regimens, and adherence level in HIV/AIDS patients receiving antiretroviral therapy. Pharmacotherapy. 2017;37(2):204-13. doi:10.1002/phar.1889 [DOI] [PubMed] [Google Scholar]
- 13.Rosenblum M, Deeks SG, van der Laan M, Bangsberg DR. The risk of virologic failure decreases with duration of HIV suppression, at greater than 50% adherence to antiretroviral therapy. PLoS One. 2009;4(9):e7196. doi:10.1371/journal.pone.0007196 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Shubber Z, Mills EJ, Nachega JB, et al. Patient-reported barriers to adherence to antiretroviral therapy: a systematic review and meta-analysis. PLoS Med. 2016;13(11):e1002183. doi:10.1371/journal.pmed.1002183 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Medicare Program; contract year 2023 policy and technical changes to the Medicare Advantage and Medicare Prescription Drug Benefit programs; policy and regulatory revisions in response to the COVID-19 public health emergency; additional policy and regulatory revisions in response to the COVID-19 public health emergency. Fed Regist. 2022;87(89):27704-902. To be codified at 42 CFR §417, 422, and 423. [Google Scholar]
- 16.Loucks J, Zuckerman AD, Berni A, Saulles A, Thomas G, Alonzo A. Proportion of days covered as a measure of medication adherence. Am J Health Syst Pharm. 2022;79(6):492-6. doi:10.1093/ajhp/zxab392 [DOI] [PubMed] [Google Scholar]