Clinical outcomes on B/F/TAF and dolutegravir-based regimens at 12 months following regimen switch: an observational cohort study

Abstract

Background

This retrospective observational study evaluated the clinical use and treatment outcomes in virologically suppressed people with HIV (PWH) switching to either bictegravir/emtricitabine/tenofovir alafenamide (B/F/TAF) or dolutegravir (DTG)-based regimens (single- and multi-tablet formulations (STR and MTR)).

Methods

We analyzed electronic medical and dispensing records from Trio Health HIV Research Network for treatment-experienced PWH ≥ 18 years suppressed (viral load [VL] < 200 copies/mL) at switch to B/F/TAF, DTG STR (DTG/3TC, DTG/RPV, DTG/3TC/ABC) or most common DTG MTRs with VL at 12 months (+/-3) since switch between April 2019 and December 2024. Univariate comparisons: chi-square or t-test; characteristics associated with virologic suppression at 12 months: multivariable logistic regression [LR], controlling for age, gender, race, baseline CD4 count, regimen, and adherence (proportion days covered [PDC] ≥ 80%).

Results

Of 3141 PWH, 66% switched to B/F/TAF, 27% to DTG STR, and 7% to DTG MTRs. Baseline characteristics differed significantly between groups: B/F/TAF group had a higher proportion of males (81% vs. 67% DTG MTR) and Black individuals (40% vs. 31% DTG STR), while the DTG STR and MTR groups had higher proportions of PWH with cardiovascular disease (35% and 39% respectively vs. 32% on B/F/TAF) and eGFR < 60 mL/min/1.73 m² (18% and 22% respectively vs. 8% on B/F/TAF). At 12 months post switch, proportion of individuals with PDC ≥ 80% was higher in the DTG STR group (76%) compared to DTG MTR (67%) and B/F/TAF (69%) with similar virologic suppression across groups (98%, 98%, and 97% respectively). In multivariable LR, there was an association of White race, PDC ≥ 80%, and CD4 ≥ 200 cells/mm³ with greater probability of suppression at 12 months.

Conclusions

Despite differences in baseline characteristics and regimen adherence between individuals switched to B/F/TAF, DTG STRs, and MTRs, both B/F/TAF and DTG-based regimens were associated with high rates of virologic suppression. This data strongly supports the inclusion of these simple and safe regimens as switch options in virologically suppressed PWH. Baseline differences in patient demographics and characteristics may have impacted the adherence on B/F/TAF compared to DTG STR, however virologic outcomes were preserved, demonstrating the forgiveness of B/F/TAF in populations potentially facing adherence challenges.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12981-025-00789-7.

Background

Integrase strand transfer inhibitors (INSTI) used in combination with two nucleoside transcriptase inhibitors (NRTIs) are recommended first-line antiretroviral agents for people with HIV (PWH) [1, 2]. In addition to a favorable safety profile, dolutegravir (DTG) and bictegravir are highly potent and are less susceptible to the emergence of drug resistance than their earlier generation integrase inhibitor counterparts, raltegravir and elvitegravir [3–5]. Furthermore, fixed-dose co-formulations of dolutegravir and bictegravir provide a simple treatment option that may improve medication adherence and the likelihood of achieving and maintaining viral suppression. Clinical trial data demonstrate that bictegravir/emtricitabine/tenofovir alafenamide (B/F/TAF) is non-inferior to DTG-containing regimens for treatment-naïve PWH [6–9].

The efficacy of regimens based on these contemporary INSTIs as a switch strategy has also been demonstrated through multiple studies. These studies show that switching to B/F/TAF or DTG-based therapy maintains viral suppression similarly to remaining on an alternative regimen [10–14]. Based on these data, B/F/TAF, DTG/3TC, and DTG/RPV are guidelines-recommended single-tablet regimens (STR) for modifying treatment in virologically suppressed patients. DTG-based multi-tablet regimens (MTR) are also used in certain settings, for example, when a single-tablet regimen is not suitable due to comorbidities, potential drug interactions, or resistance profile.

Previously, we reported an association between adherence and viral suppression on the single-tablet regimens DTG/ABC/3TC, B/F/TAF, and DTG multi-tablet regimens in clinical practice approximately 2 years after the approval of B/F/TAF [15]. In that analysis, PWH who switched to B/F/TAF had an overall higher level of medication adherence in the first six months on the regimen compared to those who switched to DTG/3TC/ABC or DTG MTRs (proportion days covered [PDC] ≥ 80%: 77% vs. 67% vs. 61% respectively). Higher adherence was significantly associated with viral suppression for DTG triple therapy, but not B/F/TAF, which had comparably high viral suppression through several levels of adherence suggesting regimen forgiveness. With the treatment switch options now including the two-drug DTG single-tablet regimens DTG/RPV and DTG/3TC, the primary objective of this study was to assess the clinical usage and treatment outcomes in virally suppressed PWH switching to all currently available options using B/F/TAF and DTG. To generate hypotheses for future research and further characterize DTG STR group, exploratory subgroup analysis included baseline characteristics and outcomes of individuals on DTG/ABC/3TC, DTG/RPV vs. DTG/3TC.

Methods

We performed a retrospective observational study of adult treatment-experienced PWH with confirmed viral suppression at switch (HIV viral load [VL] < 200 copies/ml). The study data was obtained from the Trio Health HIV electronic medical record and dispensing database between April 2019 and December 2024. Trio database collects longitudinal data from geographically distributed federally qualified healthcare centers (FQHCs) in the US. At the time of the analysis the database contained records for over 50,000 people with HIV with prescriptions for antiretroviral therapy; half of them had dispensing data from onsite pharmacies. Eligibility criteria included confirmed HIV diagnosis, age ≥ 18 years, viral suppression at the time of switch to a single-tablet B/F/TAF, DTG/3TC, DTG/RPV, DTG/3TC/ABC or most common DTG multi-tablet regimens (≥ 10 individuals), with viral load measurements at baseline (≤ 12 months prior to switch) and at 12 months (+/−3 months) after ART switch, and excluding individuals with dispenses available for < 30% of the 12-month observation period, to ensure medication dispensation is captured at the practice. A total of 505 individuals were excluded from the analysis due to fill history (proportion days covered) < 30% for a 12-month period indicating insufficient exposure to the prescribed ART. DTG MTR group included individuals dispensed dolutegravir as a separate tablet in combination with other antiretroviral medications administered as distinct pills.

In addition to describing the characteristics of PWH receiving these regimens, we evaluated adherence to ART at 12 months, assessed with mean (median) PDC and adherence thresholds (PDC ≥ 80% and PDC ≥ 95%). Prior studies consistently indicate that achieving at least 80% PDC with ART is strongly associated with high rates of viral suppression in people with HIV [16–18]. PDC was calculated using dispensing data originating solely from 340B onsite pharmacies. The 340B Drug Pricing Program requires pharmaceutical manufacturers to provide discounted outpatient drugs to eligible safety-net healthcare organizations (e.g., federally qualified health centers, Ryan White clinics), enabling them to serve a disproportionate share of low-income and uninsured individuals. All individuals meeting patient definition regardless of income or insurance status can utilize the onsite pharmacies at covered entities. The optimal adherence threshold was defined as PDC ≥ 80% and high adherence threshold was defined as PDC ≥ 95%. The association of adherence and viral suppression was also examined for each regimen group.

Univariate comparisons between regimen groups DTG STR and DTG MTR vs. B/F/TAF and between regimens in DTG STR group (DTG/ABC/3TC and DTG/RPV vs. DTG/3TC) were conducted using chi-squared test for categorical variables, t-test and Mann-Whitney U test for continuous variables as applicable. Classification regression tree analysis (CRT) followed by logistic regression were used to identify characteristics associated with PDC ≥ 80% and, separately, viral suppression at 12 months. Missing values were treated as “unspecified” and were coded accordingly, helping to preserve the integrity of the data by not assuming any particular value for unavailable entries. Characteristics used to assess drivers of PDC ≥ 80% in both CRT and logistic regression were prior regimen PDC ≥ 80%, payer type, gender, race, age, regimen, baseline CD4 count and estimated glomerular filtration rate (eGFR), baseline substance use, and neuropsychiatric disorder status based on ICD-10 codes (World Health Organization, 2019. International Statistical Classification of Diseases and Related Health Problems, 10th Revision). Logistic regression with the outcome virologic suppression at 12 months accounted for age, gender, race, baseline CD4 count, regimen group, and PDC ≥ 80% during 12 months. Interaction between regimen group and PDC ≥ 80% was assessed separately via logistic regression to evaluate if the effect of PDC varies by regimen group.

Results

There were 3141 eligible treatment-experienced virologically suppressed PWH who switched from a stable ART regimen to B/F/TAF, DTG STR, or DTG MTR regimens between April 2019 and December 2023 and had at least 12 months of follow-up on the new regimen. Of these, 2,085 (66%) were switched to B/F/TAF, 850 (27%) to DTG STR, and 206 (7%) to DTG MTR regimens. Of 850 PWH on DTG STR regimens, 569 (67%) were on DTG/3TC, 194 (23%) DTG/RPV, and 87 (10%) DTG/ABC/3TC. Of 206 PWH on DTG MTR regimens, 99 (48%) were dispensed DTG + FTC/TAF, 32 (16%) DRV/c + DTG, 26 (13%) DRV/c/FTC/TAF + DTG, 13 (6%) DTG + DOR and the remaining received less frequently used regimens.

Significant differences (p < 0.05) were observed among PWH prescribed B/F/TAF, DTG STR, and DTG MTR (Table 1): baseline age (median 49 years for B/F/TAF vs. 52 for DTG STR and 61 for DTG MTR), male gender (81% B/F/TAF vs. 67% DTG MTR), Black race (40% B/F/TAF vs. 31% DTG STR), baseline CD4 count < 200 cells/mm³ (2% B/F/TAF vs. 0% DTG STR), eGFR < 60 mL/min/1.73m² (8% B/F/TAF vs. 18% DTG STR and 22% DTG MTR), substance use (7% B/F/TAF vs. 3% DTG STR), cardiovascular disease (27% B/F/TAF vs. 35% DTG STR and 39% DTG MTR). Individuals dispensed B/F/TAF were on therapy longer than DTG STR and DTG MTR groups (mean 35.3 months (SD 17) vs. 32.1 (15) DTG STR and 33.1 (16.9) DTG MTR).

Table 1.

Baseline characteristics

n (%) unless specified		Treatment-experienced suppressed at switch with viral load measure at 12 months
		A: DTG MTR n = 206	B: DTG STR* n = 850	C: B/F/TAF n = 2085	A vs. C p-value	B vs. C p-value
Age	18–30	10 (5)	53 (6)	204 (10)	< 0.001	< 0.001
	31–50	82 (40)	337 (40)	932 (45)
	51–64	81 (39)	336 (40)	765 (37)
	65+	33 (16)	124 (15)	184 (9)
Gender	Male	137 (67)	675 (79)	1689 (81)	< 0.001	0.671
	Female	69 (33)	172 (20)	386 (19)
Race	White	90 (44)	461 (54)	964 (46)	0.433	< 0.001
	Black	81 (39)	262 (31)	834 (40)
	Other**	35 (17)	127 (15)	287 (14)
Payer	Commercial	107 (52)	598 (70)	1219 (58)	< 0.001	< 0.001
	Medicare	56 (27)	142 (17)	318 (15)
	Medicaid	36 (17)	83 (10)	401 (19)
	Ryan White, patient assistance, or self-pay	7 (3)	21 (2)	123 (6)
	Unknown	0 (0)	6 (1)	24 (1)
Prior history, years mean (SD)		5 (4.2)	5.4 (5)	4.9 (4.4)	0.850	0.009
Baseline CD4 count < 200 cells/mm3		10 (5)	4 (0)	51 (2)	< 0.001	< 0.001
Baseline CD4 count, mean (STD) cells/mm3		658 (353) n = 131	766 (301) n = 548	726 (317) n = 1569	0.035	0.010
Baseline body mass index (BMI)	Underweight < 18.5 kg/m2	5 (3)	10 (1)	20 (1)	0.027	0.819
	Normal 18.5–24.9	53 (30)	238 (30)	559 (30)
	Overweight 25–30	57 (32)	302 (38)	758 (40)
	Obese > 30	63 (35)	235 (30)	545 (29)
Baseline estimated glomerular filtration rate (eGFR)	< 60 mL/min/1.73m2	32 (22)	126 (18)	129 (8)	< 0.001	< 0.001
	60–89	64 (44)	371 (53)	825 (49)
	≥ 90	51 (35)	203 (29)	740 (44)
Cardiovascular disease		81 (39)	298 (35)	554 (27)	< 0.001	< 0.001
Diabetes		18 (9)	74 (9)	162 (8)	0.622	0.398
Baseline alcohol abuse		5 (2)	8 (1)	37 (2)	0.505	0.096
Baseline substance use		18 (9)	26 (3)	139 (7)	0.262	< 0.001
Baseline smoking		25 (12)	77 (9)	284 (14)	0.552	< 0.001
Neuropsychiatric disorders		64 (31)	244 (29)	647 (31)	0.991	0.214

*DTG STR includes DTG/ABC/3TC (n = 87), DTG/RPV (n = 194), DTG/3TC (n = 569)

**Other race includes American Indian or Alaska Native, Native Hawaiian or Other Pacific Islander, Asian

Compared to those on DTG/RPV, DTG/3TC group had significantly higher proportion of individuals with baseline age 31–50 (42% vs. 31%), lower proportion of those 65 and older (13% vs. 21%), higher proportion of individuals with commercial insurance (75% vs. 62%), lower proportion with Medicare (13% vs. 29%), with baseline eGFR < 60 mL/min/1.73m² (12% vs. 37%), with cardiovascular disease (32% vs. 43%), and similar mean regimen duration (31.8 vs. 34.1 months) [Supplementary Table].

Compared to DTG/ABC/3TC, DTG/3TC had a significantly higher proportion of individuals with age 65+ (13% vs. 8%), higher proportion of commercially insured (75% vs. 55%) and lower proportion on Medicaid (8% vs. 26%), longer prior known treatment history (mean 5.7 years vs. 3.4), higher mean baseline CD4 (798 vs. 661 cells/mm³), lower proportion of those with baseline eGFR < 60 mL/min/1.73m² (12% vs. 24%), similar mean regimen duration (29.8 vs. 31.8 months).

At 12 months after the ART switch, PWH on DTG STR were more adherent than on B/F/TAF or DTG MTR: mean (median) PDC was 86% (93%) for DTG STR vs. 81% (91%) for DTG MTR, and 83% (91%) for B/F/TAF (both p < 0.01 for means and medians). Excluding DTG/ABC/3TC, mean (median) PDC for DTG STR was 87% (93%), similar to 86% (93%) observed in DTG STR group without this exclusion. The proportion of PWH with PDC ≥ 80% was 76% for DTG STR vs. 67% for DTG MTR and 69% for B/F/TAF (both p < 0.01). The proportion of PWH with PDC ≥ 95% was 46% for DTG STR vs. 38% for DTG MTR (p = 0.047), and 39% for B/F/TAF (p = 0.002) [Figure 1]. The adherence was similar on B/F/TAF and DTG MTR. Among those on DTG STR, there was no difference in mean/median PDC at 12 months between DTG/3TC and DTG/RPV (87%/94% vs. 87%/93%), however mean PDC was lower on DTG/ABC/3TC (mean 82%/median 93%) vs. DTG/3TC (p = 0.027).

Fig. 1.

Adherence at 12 months

Based on classification regression tree analysis that implicitly controls for other factors and interactions, the most important drivers of PDC ≥ 80% at 12 months were payer type, prior regimen PDC ≥ 80%, and age, followed by baseline eGFR, race, and baseline CD4, but not current regimen group. In logistic regression, accounting for the same characteristics used in classification regression tree, prior regimen PDC ≥ 80%, payer type, race, and age were also associated with PDC ≥ 80% at 12 months. Specifically, individuals with prior regimen PDC ≥ 80% were more likely to have PDC ≥ 80% on their new regimen (adjusted odds ratio [OR] = 5.4 (95% confidence interval [CI] 3.8–7.7), p < 0.001). Compared to individuals aged 18–30 years old at baseline, PWH age > 30 were also more likely to have PDC ≥ 80% on the new regimen (31–50 years: OR = 1.9 (1.3–2.6), 51–64 years: OR = 2.1 (1.4-3), –5 and older: OR = 2.3 (1.4–3.7), p < 0.001 for all). Lastly, compared to Black or African American PWH, White, and other race non-Black PWH were more likely to have higher PDC (White: OR = 1.8 (1.5–2.2), other non-Black: OR = 1.7 (1.3–2.4), p < 0.001). Compared to commercially insured PWH, those on Medicaid and patient assistance or self-pay were less likely to have PDC ≥ 80% (Medicaid: OR = 0.3 (0.3–0.4), Ryan White/self-pay/patient assistance: OR = 0.5 (0.3–0.8), p < 0.001).

Despite differences in baseline characteristics and adherence, overall viral suppression at 12 months was similar between regimens: 97% B/F/TAF, 98% DTG MTR, and 98% DTG STR [Figure 2]. Within DTG STR group suppression on DTG/ABC/3TC was lower: 92% vs. 98% for DTG/3TC (p = 0.003) and 99% for DTG/RPV (p = 0.001). Excluding DTG/ABC/3TC, 98% of DTG STR group remained suppressed at 12 months; this rate was identical to the suppression rate without exclusion.

Fig. 2.

Viral suppression and adherence at 12 months

We also explored the relationship between adherence thresholds at 12 months (PDC < 80% and PDC ≥ 80%) and the proportion of PWH who were virally suppressed at 12 months within each drug group [Figure 2]. For DTG STR and B/F/TAF, individuals were more likely to be suppressed as adherence levels increased: 94% of DTG STR and B/F/TAF were suppressed at PDC < 80% vs. 99% at PDC ≥ 80% (p < 0.001). For DTG MTR the relationship between the higher level of PDC and suppression rate was not statistically different (p = 0.517): 99% (68/69) were suppressed at PDC < 80% and 97% (133/137) at PDC ≥ 80%.

A multivariable logistic regression analysis was conducted to examine the interaction between treatment regimen (DTG MTR, DTG STR, and B/F/TAF) and PDC ≥ 80% on viral suppression. After adjusting for covariates, PDC ≥ 80% was significantly associated with increased odds of viral suppression (adjusted odds ratio [95% CI] 5.6 [3.2–9.9], p < 0.001). The effect of regimen on suppression was not significant (DTG MTR vs. B/F/TAF: 4.6 [0.6–33.8], p = 0.14; DTG STR vs. B/F/TAF: 1.1 [0.6–2.1], p = 0.8). To further explore how predicted probability of suppression changes across different levels of PDC by regimen group, we visualized this relationship via an interaction plot that showed higher levels of adherence led to higher probability of suppression for DTG STR and B/F/TAF but not for DTG MTR [Figure 3]. Interaction term for PDC and regimen was significant for DTG MTR vs. B/F/TAF (p = 0.036), but not overall (p = 0.108), thus interaction term was excluded from the final logistic regression model. Possible explanation for these results may be small sample size of DTG MTR group, complexity of MTR regimens leading to errors in dosing or timing of drug intake, “drug holidays” not reflected by PDC, “stock piling” of medication, medication sharing, or drug resistance. Additionally, PDC only measures whether medications were dispensed, not whether they were actually taken.

Fig. 3.

Predicted probability of suppression by adherence (PDC ≥ 80%) across regimen groups. The interaction term represents the joint effect of the drug regimen and adherence. It allows us to see if the effect of the drug regimen on suppression is different at different levels of adherence

In multivariable logistic regression accounting for age, gender, race, baseline CD4, PDC on the regimen, and regimen group, White race, PDC ≥ 80%, and baseline CD4 were predictors of suppression at 12 months [Figure 4]. Specifically, odds of suppression were 3 times higher for PWH identifying with White race vs. Black (adjusted odds ratio [95% CI]: 3 [1.7–5.3], p < 0.001) and 4 times higher for those with PDC ≥ 80% (4 [2.5–6.4], p < 0.001). Individuals with baseline CD4 < 200 cells/mm³ had a 60% reduction in odds of suppression compared to those with CD ≥ 200 cells/mm³ (0.4 [0.1–0.9], p = 0.03). Adjusting for other confounders, the regimen group was not a significant predictor of suppression.

Fig. 4.

Predictors of viral suppression at 12 months. Vertical line indicates OR = 1 no effect. Diamonds represent point estimates of ORs. Horizontal lines represent 95% confidence intervals

Discussion

In this analysis of people switching to B/F/TAF and DTG-based regimens in clinical practice, we found overall high medication adherence and viral suppression at 12 months following switch. For those with dispensing histories indicating more than 95% of days covered by treatment, > 97% maintained viral suppression at 12 months. Perhaps more remarkably, suppression was 94% even when PDC was less than 80%. These results corroborate with other clinical data [19, 20] and underscore the high potency, resistance barrier, and forgiveness of regimens using bictegravir and dolutegravir.

Importantly, there were multiple significant differences between those prescribed different treatment options, most likely reflecting underlying comorbidities and the clinicians’ perspectives on the best options for these conditions. Older people, those with impaired renal function, and cardiovascular disease were significantly less likely to receive B/F/TAF than DTG-based regimens, likely reflecting concerns about tenofovir nephrotoxicity. This concern persists despite data showing comparable renal safety of tenofovir alafenamide to dolutegravir-containing regimens [10, 21, 22].

An additional factor is selection of potential candidates for two-drug regimens by clinicians. The pivotal clinical trials of switching to the two-drug regimens DTG/3TC and DTG/RPV included only those with no prior history of treatment failure or known drug resistance to either drug [10, 11, 23]. Given that three-drug treatments have long been standard of care, prescribers may have preferentially given two-drug regimens to their patients with the best potential medication adherence. Additionally, there have been several studies of switching to both B/F/TAF and DTG + tenofovir/3TC in patients with prior treatment failure or resistance [24–26]. These prospective clinical trials have consistently demonstrated high rates of virologic suppression in this population, whereas there are no such studies using DTG/3TC or DTG/RPV. In support of this channeling of individuals with the highest medication adherence to the two-drug DTG single-pill options, we found that adherence to these regimens was higher than to DTG multi-tablet regimens and to B/F/TAF. This is further supported by the greater use of B/F/TAF in PWH who may face challenges with maintaining high ART adherence levels due to factors like substance use disorder, younger age, systemic racism, and other socio-economic indicators. Nonetheless, despite these adverse factors, virologic suppression with B/F/TAF was quite high overall. Multivariable logistic regression findings accounting for key baseline demographic and clinical characteristics suggest association of White race vs. Black (possibly as proxy to socio-economic status and access to care), baseline immunologic status, and adherence on the regimen with higher probability of viral suppression at 12 months. These findings highlight the complex interplay of social determinants of health, individual-specific factors, and treatment adherence in achieving optimal virologic outcomes. Interestingly, regimen type (DTG MTR, DTG STR vs. B/F/TAF) was not identified as an independent predictor of viral suppression. This finding further reenforces high efficacy of modern INSTI-based regimens, suggesting that once individual-specific factors are considered, the choice between these effective options hinges on individual tolerability or preference, drug-drug interactions, baseline comorbidities, and anticipated adherence.

While providing valuable real-world insights into utilization and effectiveness of DTG-based regimens and B/F/TAF in clinical practice, our study is subject to several limitations inherent to its retrospective observational design utilizing electronic medical records and dispensing data from 340B pharmacies. First, dispensing data reflects medication acquisition rather than actual consumption, potentially overestimating adherence. Second, some individuals may have partially filled their prescriptions outside of 340B pharmacies, which would underestimate their adherence. Third, all the data was collected as part of clinical care process and data collection was not tailored to the research objectives of this retrospective study, as such important variables such as reasons for regimen selection or discontinuation were not available. Finally, while we accounted for available confounders, the observational nature of the study precludes definitive causal inference, and unmeasured confounding factors may still influence the observed associations.

Further research is warranted to elucidate the precise mechanisms underlying the uncovered associations and to develop targeted interventions that address disparities in access to care and promote adherence among diverse patient populations.

Conclusions

In summary, we found there were significant differences in demographic and clinical baseline characteristics between those receiving B/F/TAF, DTG single-tablet regimens, and DTG multi-tablet regimens. These differences likely represented underlying comorbidities, prior treatment failure and/or resistance, or real or perceived adherence capabilities. Despite differences in adherence, both B/F/TAF and DTG-based regimens were associated with high rates of virologic suppression. As the most frequently used STR in our cohort, B/F/TAF demonstrated similar rates of viral suppression to DTG STR, despite notable differences in baseline characteristics historically pointing to potential adherence challenges and lower observed adherence in B/F/TAF group. This evidence underscores forgiveness of B/F/TAF regimen, which, coupled with its convenience as a single-tablet regimen, support its use in the population with potential adherence challenges. Understanding regimen robustness can inform treatment selection in challenging real-world situations, however optimal adherence remains crucial for all regimens.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

3TC

Lamivudine

ABC

Abacavir

ART

Antiretroviral therapy

ARV

Antiretroviral

B/F/TAF

Bictegravir/emtricitabine/tenofovir alafenamide

CD4

CD4 + T-cell count

CI

Confidence Interval

CRT

Classification regression tree

DTG

Dolutegravir

eGFR

Estimated glomerular filtration rate

HIV

Human immunodeficiency virus

INSTI

Integrase strand transfer inhibitor

IQR

Interquartile range

MTR

Multi-tablet regimen

OR

Odds ratio

PDC

Proportion days covered

PWH

People with HIV

RPV

Rilpivirine

SD

Standard deviation

STR

Single-tablet regimen

TAF

Tenofovir alafenamide

TDF

Tenofovir disoproxil fumarate

VL

Viral load

Author contributions

PS was the major contributor in writing the manuscript. JR analyzed and interpreted data and summarized study results and methods. MD, JG, NPM, MR, SS, KM, RE provided critical revisions and interpretation of the data. RE served as the principal investigator of the study. All authors read and approved the final manuscript.

Data availability

The data that supports the findings of this study are available from Trio Health, but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are, however, available from the corresponding author upon reasonable request and with the permission of Trio Health and Gilead Sciences.

Declarations

Ethics approval and consent to participate

Using the Department of Health and Human Services regulations found at 45 CFR 46.104(d)(4), the central IRB (Advarra IRB, Advarra, 6100 Merriweather Drive, Suite 600, Columbia, Maryland, USA) determined that the research project was exempt from the IRB oversight. Individual identifying information was not accessed by the investigators during the study, therefore consent was not required.

Consent for publication

Not applicable.

Competing interests

PS is a consultant for Gilead Sciences, ViiV Healthcare, and Merck. He received research grants from Gilead Sciences, Merck, and ViiV Healthcare. JR is an employee of Trio Health. MD, NPM and JG are employees of Gilead Sciences. SS serves on the Medical Advisory Board for Gilead Sciences and is a Speaker for Gilead Sciences. KM advises for ViiV Healthcare, Merck, and Gilead Sciences. He is on the speakers’ bureau for ViiV Healthcare, Merck, and Gilead Sciences. He received research grants from ViiV Healthcare, Merck, and Gilead Sciences. MR is a consultant for ViiV Healthcare, Gilead Sciences, and Merck and received research funding from these companies. RE has received grants from Gilead Sciences and ViiV Healthcare, serves on the Advisory boards for Gilead Sciences and ViiV Healthcare, and is a speaker for Gilead Sciences. He is the Chairman of Trio Health Scientific Steering Committee.