Effectiveness of intensive adherence counselling in achieving an undetectable viral load among people on antiretroviral therapy with low-level viraemia in Uganda

Abstract

Introduction

Uganda was using a threshold of 1,000 copies/ml to determine viral non-suppression for antiretroviral therapy (ART) monitoring among people living with HIV (PLHIV), prior to this study. It was not clear if PLHIV with low-level viraemia (LLV, ≥50 to <1,000 copies/ml) would benefit from intensive adherence counselling (IAC). The purpose of this study was to determine the effectiveness of IAC among PLHIV on ART with LLV in Uganda, to guide key policy decisions in HIV care including the review of the viral load (VL) testing algorithm.

Methods

This cluster-randomized clinical trial comprised of adult PLHIV on ART who had recent VL results with LLV (tested from July 2022 to October 2022) at 8 HIV clinics. Participants in the intervention arm clinics received three monthly sessions of IAC, while those in the comparison non-intervention arm clinics received the standard of care. At the end of the study, all the participants in both arms were tested again for VL, to determine the proportions of those who had acquired an undetectable VL (below 50 copies/ml). Statistical association between cross-tabulated variables was assessed using Fisher’s exact test. Modified Poisson regression was then used.

Results

A total of 136 participants were enrolled into the study at 8 HIV clinics. All the 68 participants in the intervention arm completed all the sessions of IAC. Only one participant in the comparison non-intervention arm was lost-to-follow-up. The average follow-up time was 3.7 months (SD – 0.2) and 3.5 months (SD – 0.1) in the intervention and comparison non-intervention arms respectively.

A total of 59 PLHIV (43.7%) out of 135 PLHIV achieved an undetectable VL during the study follow-up period. The effect of IAC on attaining an undetectable VL among LLV clients was nearly twice as high in the intervention arm (57.4%), as compared to the comparison non-intervention arm (29.9%), adj. RR=1.9 (1.0, 3.5), p=0.037.

Conclusion

IAC doubled the likelihood to achieve a non-detectable VL among PLHIV with LLV. Hence IAC has been instituted as an intervention to manage PLHIV with LLV in Uganda, and this should also be adopted in other Sub-Sahara African countries with similar settings.

Introduction

Uganda is progressing towards achieving the global targets of ending the HIV/AIDS epidemic by 2030 (1, 2), ensuring that more than 1.2 million (out of 1.4 million) people living with HIV (PLHIV) are on antiretroviral therapy (ART) (3), with implementation of a comprehensive viral load (VL) testing program to monitor the efficacy of ART (4). However prior to this study, Uganda was using a threshold of 1,000 copies/ml to determine viral non-suppression (5), and VL results below 1,000 copies/ml were considered to be suppressed. Hence PLHIV with low-level viraemia (≥50 to <1,000 copies/ml) were being offered no intervention. Currently, Uganda is implementing and rolling out new HIV treatment consolidated guidelines, which stipulate that IAC should be offered to PLHIV with LLV (6). This study aimed to determine whether intensive adherence counselling (IAC) is effective in achieving an undetectable VL among PLHIV on ART with low-level viraemia (LLV) in Uganda.

The use of highly active antiretroviral therapy (HAART) has yielded multiple gains in the treatment and control of the HIV epidemic over the years, leading to viral suppression and increased body immunity function (7). VL testing is recommended by the World Health Organisation (WHO) to monitor the efficacy of ART among PLHIV, and a threshold of 1,000 copies/ml to determine viral non-suppression is recommended (8). In Uganda, prior to this study, VL results below 1,000 copies/ml were considered a suppressed VL, and no intervention was offered, other than routine adherence counselling (5).

PLHIV with a VL of 50 copies/ml but below 1,000 copies/ml (≥50 to <1,000 copies/ml) are considered to have LLV (9, 10). The number of PLHIV with LLV in Uganda increased from 2.0% in 2016 to 8.6% in 2020, and LLV was associated with being of lower age, male sex and second line regimen (11). Furthermore, LLV among PLHIV has been associated with ART duration longer than 5 years and initial CD4 count below 200 cells/mm³ (9, 12). In one Chinese study, associations with subtype B infections were observed among PLHIV with LLV (12).

Previous studies have associated LLV with virolgical failure for instance a study in four African countries including Uganda, Kenya, Tanzania and Nigeria showed that 57.5% of PLHIV with persistent LLV got virologic failure (13), while a South African study showed that LLV was associated with increased hazards ratios of virological failure (9). A study conducted in Uganda showed that participants with LLV had 4.1 times the hazard rate of becoming non-suppressed, when compared to participants with an undetectable VL. Furthermore, numerous studies have associated LLV with HIV drug resistance (13, 14, 15, 16).

The key treatment goals of ART are to maximally achieve VL suppression, restore and preserve immunologic function, reduce morbidity and mortality, and also reduce HIV transmission among PLHIV (17, 18). However these treatment goals can only be achieved if there is HIV treatment adherence (19, 5).

High levels of treatment adherence are necessary to prevent HIV drug resistance (21), to achieve viral load suppression (22), to halt disease progression to AIDS (23), and to reduce HIV related mortality (24); and these high levels of adherence are associated with well-tolerated ART regimens, better socio-economic status, and being aware of one’s HIV status (25).

Poor treatment adherence is associated with treatment failure, emergence of HIV drug resistance, increased disease progression and increased mortality (25, 26, 27, 28); and various factors like poverty, food insecurity, drug side effects, high pill burden, depression, lack of transportation means to the facility, pharmacy stock outs, low literacy levels, non-disclosure, and poor disease insight have been associated with poor treatment adherence (26, 29).

Intensive adherence counselling (IAC) is recommended by WHO to be offered as an intervention to PLHIV on ART with non-suppressed VL (31). IAC has been shown to improve VL suppression in 57 to 84% of PLHIV on ART with VL non-suppression (31, 32). Uganda initiated IAC for PLHIV on ART with non-suppressed VL in 2016 (33, 5). The goal of IAC is to enable PLHIV on ART with non-suppressed VL to identify their individual barriers to ART adherence and to come up with solutions to overcome the barriers and attain optimal HIV treatment adherence with the respective VL suppression (34, 5).

Prior to this study, the consolidated guidelines for prevention and treatment of HIV in Uganda recommended three monthly IAC sessions using the 5 As (Assess, Advise, Agree, Assist and Arrange) to offer IAC and psychosocial support by a multi-disciplinary team comprising of clinicians, nurses, counselors, family members, and peers among others (5). A repeat VL test was done in the fourth month after the three IAC sessions to ascertain whether the patient has achieved VL suppression. Unlike the non-suppressed PLHIV (having a VL of 1,000 copies/ml or above) who were being offered IAC, suppressed PLHIV (with a VL below 1,000 copies/ml) including those with LLV (≥50 to <1,000 copies/ml) were just given routine adherence counselling.

Sub-optimal treatment adherence has been associated with residual LLV (36), hence this indicates that IAC can be an intervention to manage LLV among PLHIV on ART to attain an undetectable VL, which is desirable as per several international guidelines (36, 37). In 2022, WHO recommended the use of IAC in the management of LLV (39), however this has not yet been taken up by most SSA countries. Important to note is that, there was limited data in Uganda and other SSA countries on whether IAC can be effective in achieving an undetectable VL among PLHIV on ART with LLV.

Methods

Study design

This cluster-randomized clinical trial was conducted between September 2022 and February 2023.

Study setting

The study comprised of eight public HIV clinics across the four regions of Uganda. These were high volume HIV clinics with high numbers of PLHIV having LLV. The clinics were matched and grouped into pairs, basing on geographical location and number of PLHIV with LLV, using the Goldilocks approach (weighting) (39, 40). In each pair, one clinic was randomly assigned to either the intervention arm or the comparison non-intervention arm by tossing a coin. Randomization was done in a ratio of 1:1.

Sample size

Based on Sakpal and Sullivan, n = [(Z_α/2+ Z_β)² × {(p₁ (1-p₁) + (p₂ (1-p₂))}]/(p₁ - p₂)² (41, 42); we estimated a sample size of 79 participants per arm. This ensured a power of 90% and a 5% level of significance to detect a 15% difference in PLHIV on ART who get a non-detectable viral load between those offered IAC and those who are not.

IAC is an intervention used to manage PLHIV with a non-suppressed VL and about 70% of these PLHIV attain viral suppression (44). Therefore, we assumed that p₂ was 0.7. PLHIV with LLV have suppressed VL and we assumed IAC to even be more effective at about 90% in creating a non-detectable VL in these PLHIV, hence we assumed p₁ to be 0.9.

Then, an equal sample size of 12 participants for each of the 8 clusters was estimated from;

$m=\lceil \frac{n_I\left(1-\rho \right)}{\left(k-n_I\rho \right)}\rceil$, (45) where n_I was the number of participants required for each arm, k was the number of clusters, and p was the intra-cluster correlation estimate, which was 0.0180 (46), thus a total of 96 participants was required. Considering a design effect of 1.126 and an attrition rate of 20%, the resultant total sample size of 136 participants for the study was calculated, hence a total 17 participants was recruited per cluster.

Study Participants

Study participants were adult PLHIV aged 18 years and above on ART, who had recent VL results with LLV (tested from July 2022 to October 2022).

Intervention Arm (IAC arm)

The intervention was three sessions of intensive adherence counselling (IAC) and psychosocial support using the 5 A’s principles for chronic care (which include; Assess, Advise, Agree, Assist and Arrange). IAC sessions which lasted about one hour, were offered at monthly intervals by experienced ART counsellors. At each HIV clinic, IAC sessions were conducted on different days from those of the routine IAC sessions for the non-suppressed PLHIV; and colour-coded IAC forms were used in the study to differentiate the study from routine IAC services.

The study participants in the intervention arm returned to the clinic every month to receive IAC. During each session, an experienced ART counsellor sat with the participant in a quiet convenient room at the clinic and offered him/her an IAC session. IAC was offered throughout the study period of three months. The study participants were given a transport refund every time they returned to the clinic for the IAC sessions. At the completion of the three-monthly sessions, a blood sample was collected for repeat VL testing in the fourth month.

Comparison non-intervention arm (Standard of Care)

The participants in the comparison non-intervention arm clinics received the routine standard of care. These participants received the normal patient education and encouragement to continue with their ART at the start of the study. They were not reviewed monthly and they were not offered any counselling for the entire three months’ period. After three complete months of the study, the participants were called back to the HIV clinic and a blood sample was collected to do a repeat VL test in the fourth month. The participants were given a transport refund when they returned to the clinic to do a repeat VL test.

Outcomes

The primary outcome was the proportions of PLHIV with LLV achieving an undetectable VL (VL below 50 copies/ml) in the intervention and comparison non-intervention arms after the three months of the study period. A repeat VL test was done for all the study participants at the end of the study period.

Statistical Analysis

Descriptive statistics (frequencies and proportions) were populated in one-way tabulations for the covariates that were categorical in nature and mean and standard deviation for continuous variables. At bivariate level of analysis, the study assessed for the statistical association between cross tabulated variables using the Pearson chi-square test. For the sub-analysis done on participants whose viral load had achieved the outcome of interest, Fisher’s exact test was used to assess for the statistical association. Modified poisson regression was used to identify the variables to be included in the multivariable regression model.

All the models fit were adjusted for clustering at the region level. At multivariable level, a stepwise regression model approach was used where variables with p value less than 0.05 level of significance were considered statistically significant and therefore associated with the outcome of interest. The variables which formed the final model were assessed for interaction by including an interaction term (ART group Vs ART Duration and ART Group Vs Marital status). Each of the interaction terms was assessed for its statistical significance. IRR, reported as adjusted risk ratios and their corresponding 95% confidence intervals (CI) were all presented. Analyses were done using Stata^® (version 15).

Ethical Considerations.

Ethical approval to undertake the study was obtained from the Makerere University School of Public Health Higher Degrees, Research and Ethics Committee (Approval number is SPH-2021–144) and Uganda National Council for Science and Technology (Approval number is HS2008ES). Informed consent was obtained from the participants prior to enrolment into the study. All the study data and laboratory test results were only accessed by authorized individuals of the study team. Confidentiality was maintained through the use of unique identifiers for every study participant.

Results

Enrolment characteristics

A total of 136 PLHIV were enrolled into the study at the 8 HIV clinics as shown in Figure 1. The enrollment characteristics of the participants are shown in Table 1.

Figure 1:

Flow of the participants in the study (Study profile)

Table 1:

Characteristics of the study participants according to the randomization arm

Characteristics	Intervention Arm (n = 68)	Comparison non-intervention arm (n = 68)
Mean Age (SD), years	43.4 (11.5)	43.2 (12.5)
Gender, n (%)
Female	41 (60.3)	36 (53.0)
Male	27 (39.7)	32 (47.0)
Mean Duration on ART (SD), years	7.1 (4.6)	7.2 (4.7)
Marital Status, n (%)
Married	41 (60.3)	38 (55.8)
Single	6 (8.8)	5 (7.4)
Divorced	11 (16.2)	14 (20.6)
Widowed	10 (14.7)	11 (16.2)
Mean Level of viraemia (SD), copies/ml	139.1 (124.0)	166.3 (172.5)
Education Level, n (%)
No education	8 (11.8)	7 (10.3)
Primary	48 (70.6)	35 (51.5)
Secondary	10 (14.7)	18 (26.5)
Tertiary	2 (2.9)	8 (11.7)
ART regimen, n (%)
First line regimen	66 (97.1)	68 (100.0)
Second line regimen	2 (2.9)	0 (0.0)

Participant retention

The average follow-up time for PLHIV in the intervention arm was 3.7 months (SD – 0.2) while the average follow-up time in the comparison non-intervention arm was 3.5 months (SD – 0.1). The retention of PLHIV throughout the study was good, with 100% of the participants (68 out of 68 participants) completing all the 3 sessions of the IAC, and doing a repeat VL test in the intervention arm, and 98.5% of the participants (67 out of 68 participants) doing a repeat VL test in the comparison non-intervention arm. Only one participant in the comparison non-intervention arm was lost to follow up, because the participant migrated during the study period and was transferred out of the clinic.

Achieving an undetectable VL

A total of 59 PLHIV (43.7%) out of 135 PLHIV achieved an undetectable VL during follow-up period of the study. A total of 39 PLHIV (57.4%) out of 68 PLHIV in the intervention arm achieved an undetectable VL, as compared to 20 PLHIV (29.9%) out of 67 PLHIV in the comparison non-intervention arm who achieved an undetectable VL during the study follow-up period, and the difference between the two groups was significant (p value = 0.001). The proportions of PLHIV who achieved an undetectable VL by sub-group and study arm are shown in Table 2.

Table 2:

Number (and Percentage) of participants who achieved an undetectable VL at the Follow-up by Sub-group and Study arm (n = 59)

Characteristics	Intervention Arm (n = 39) n (%)	Comparison non-intervention arm (n = 20) n (%)	P value
Gender
Female	25 (64.1)	10 (50.0)	0.402
Male	14 (35.9)	10 (50.0)
Marital Status
Married	23 (58.9)	10 (50.0)	0.857
Single	4 (10.3)	3 (15.0)
Divorced	8 (20.5)	4 (20.0)
Widowed	4 (10.3)	3 (15.0)
Education Level
No education	5 (12.8)	2 (10.0)	0.023
Primary	29 (74.4)	10 (50.0)
Secondary	5 (12.8)	4 (20.0)
Tertiary	0 (0.0)	4 (20.0)
ART regimen
First line regimen	37 (94.9)	20 (100.0)	0.544
Second line regimen	2 (5.1)	0 (0.0)

p - values were based on Fisher exact tests.

PLHIV in the intervention arm were 1.9 times more likely to to achieve an undetectable VL, as compared to PLHIV in the comparison non-intervention arm (adjusted risk ratio [aRR] was 1.9, 95% CI: 1.0 to 3.5, p = 0.037), after adjusting for age, sex, ART duration, and marital status, as shown in Table 3.

Table 3:

Adjusted relative risks of achieving an undetectable VL

	Adjusted RR (95% CI)	p value
Comparison non-intervention arm	1 (Reference)
Intervention Arm (IAC Arm)	1.9 (1.0 – 3.5)	0.037
Age	1.0 (0.9 – 1.0)	0.943
Sex: Male	0.9 (0.8 – 1.2)	0.668
Marital Status (Married)	1 (Reference)
Single	1.5 (0.7 – 3.3)	0.336
Divorced	1.2 (0.8 – 1.6)	0.330
Widowed	0.7 (0.3 – 1.6)	0.409
Duration on ART	1.0 (1.0 – 1.1)	0.025

Discussion

Our study showed that offering IAC as an intervention to PLHIV with LLV is associated with achieving an undetectable viral load, and the findings were statistically significant. Our IAC intervention included offering of three monthly IAC sessions to PLHIV with LLV, assessing the barriers to treatment adherence and emphasizing its rationale. Our findings confirm that sub-optimal treatment adherence is associated with LLV among PLHIV as indicated by a study that determined the association between treatment adherence and LLV (36). Furthermore, the findings are in agreement with several other studies which have shown IAC to be effective in management of sub-optimal treatment adherence, mainly PLHIV with viral non-suppression (41, 42).

In our study, the final outcome was proportions of PLHIV achieving an undetectable VL in the intervention and comparison non-intervention arms. 57.4% of the PLHIV in the intervention arm achieved an undetectable VL, as compared to 29.9% of the PLHIV in the comparison non-intervention arm who achieved an undetectable VL. Despite this difference being statistically significant, the level of the effectiveness of IAC is below our expectation because we anticipated the majority of the PLHIV in the intervention arm to achieve an undetectable VL, since these are already suppressed PLHIV. However this finding is still consistent with numerous other studies which have shown low effectiveness of IAC in causing viral re-suppression among non-suppressed PLHIV. For instance; a Ugandan study showed that only 23% of the paediatric PLHIV achieved viral suppression after IAC (49); 48.2% of the PLHIV achieved viral suppression following IAC in another Ugandan study to evaluate the outcome of IAC in improving viral suppression (50); and 31.2% of PLHIV who had a repeat VL after IAC achieved VL suppression in Zimbabwe (51). This low success rate of IAC could be due to various challenges affecting the implementation of IAC including; missing IAC sessions by clients, shortage of food, stigma and discrimination, delayed viral load result, lack of incentives, and heavy workload among others (43, 46). In our study, PLHIV were given an incentive in form of a transport refund everytime they came to the health facility for IAC, and hence all the PLHIV in the intervention arm reported for all their IAC sessions. However there is an inevitable need to do further studies to improve the effectiveness of IAC in HIV care in Uganda and other SSA countries.

Though LLV has been associated with sub-optimal treatment adherence (36), it has also been associated with HIV drug resistance, for instance a study in China showed that 40.5% of PLHIV with LLV had HIV drug resistance (53). Another study from Lesotho showed that 84% of PLHIV with LLV taking a non-nucleoside reverse transcriptase inhibitor (NNRTI) based first line regimen had HIV drug resistance mutations (54). This implies that PLHIV with LLV not achieving an undetectable VL after IAC could be already having HIV drug resistance, and hence might need HIV drug resistance testing. A study that assessed lack of effectiveness of IAC in managing virological failure among PLHIV on ART in rural Uganda showed that 98% of PLHIV who did not get viral suppression after IAC had at least one mutation while 89.5% of the participants had at least two mutations (55). However in this study, we did not test for HIV drug resistance for PLHIV with LLV who did not achieve an undetectable VL due to resource constraints. Furthermore LLV has been associated with low baseline CD4+ T cell counts, very high baseline VL measurement and the use of non-nucleoside reverse transcriptase inhibitors (56, 57). However there was no data on baseline VL measurement for our study participants since this is not routinely done in the national VL program in Uganda (5).

Strengthen and Limitations:

The main strength is that this is the first cluster randomized clinical trial to determine the effectiveness of IAC in achieving an undetectable VL among PLHIV with LLV in Uganda. The major limitation was that the study included a limited number of 8 HIV clinics, due to funds constraints. Furthermore, HIV drug resistance testing for PLHIV with LLV who failed to achieve an undetectable VL was not done due to limited resources.

Conclusion

Our results indicated that 57.4% (39/68) of the PLHIV in the intervention arm achieved an undetectable VL, as compared to 29.9% (20/67) in the comparison non-intervention arm, and the difference between the two groups was significant (p value = 0.001). Furthermore, PLHIV in the intervention arm were 1.9 times more likely to to achieve an undetectable VL, as compared to PLHIV in the comparison non-intervention arm. These findings indicate that IAC is an effective intervention in the management of PLHIV with LLV in Uganda, provided the other factors like HIV drug resistance are constant. However there is still need for further studies to improve on how IAC is offered, to improve its effectiveness in HIV care programs.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.