The impact of same-day antiretroviral therapy initiation under the WHO Treat-All policy

Bernhard Kerschberger, Andrew Boulle, Rudo Kuwengwa, Iza Ciglenecki, Michael Schomaker
doi: https://doi.org/10.1101/2020.05.04.20090068

ABSTRACT

Background The World Health Organization recommends rapid initiation of antiretroviral therapy (ART) for people living with HIV, with the option to start treatment on the day of diagnosis (same-day-ART). However, the effect of same-day-ART remains unknown in realistic public sector settings.

Methods We established a cohort of ≥16-year-old patients who initiated first-line ART under Treat-All in Eswatini between 2014-2016, either on the day of HIV care enrolment (same-day-ART) or 1–14 days thereafter (early-ART). Directed acyclic graphs, flexible parametric survival analysis and targeted maximum likelihood estimation (TMLE) were used to estimate the effect of same-day ART initiation on the composite unfavourable treatment outcome (loss to follow-up;death;viral failure).

Results Of 1328 patients, 839 (63.2%) initiated same-day ART. The adjusted hazard ratio of the unfavourable outcome was increased by 48% (95% CI:1.16–1.89) for same-day ART compared with early ART. TMLE suggested that after 1 year, 28.9% of patients would experience the unfavourable outcome under same-day ART compared with 21.2% under early ART (difference: 7.7%; 1.3–14.1%). This effect varied over time in regression analysis and TMLE, with a higher hazard during the first year after HIV care enrolment and a similar hazard thereafter.

Conclusions Caution is needed in implementing same-day ART in routine settings.

BACKGROUND

The World Health Organization (WHO) policy for Treat-All recommends lifelong antiretroviral therapy (ART) for all people living with HIV (PLHIV) at the time of diagnosis, irrespective of immunological criteria.1 Despite high uptake of this policy in Africa,2 of 20.6 million PLHIV in Eastern and Southern Africa, treatment coverage (67%) and viral suppression (58%) remained below the UNAIDS targets in 2018, with an additional 3.0 million PLHIV needing to access treatment and achieve viral suppression.3,4

Accelerated ART initiation has been proposed to overcome some of these gaps.5,6 A systematic review found that ART initiation on the same day as HIV diagnosis or the day of treatment eligibility improved treatment uptake, HIV care retention and viral suppression.7 Based on this evidence, WHO released guidelines in 2017 recommending ART initiation within 7 days of HIV diagnosis (rapid ART), with the possibility of initiating treatment on the same day as HIV diagnosis (same-day ART) for patients ready to start.8

As HIV programmes allow for accelerated ART initiation under Treat-All and most treatment initiations already occur quickly, within 14–30 days after HIV diagnosis or care enrolment,9–13 the question increasingly shifts to how much faster ART can be initiated in routine resource-limited settings (RLS). This question has also been raised recently in public HIV treatment programmes in high-income countries.14 Concerns were specifically raised about the feasibility of same-day ART initiation in realistic public sector settings because of lack of real-world evidence and practical limitations. Firstly, evidence of the benefits of accelerated ART mainly originated from randomized trials.7 These trials often applied additional procedures not routinely available in RLS (e.g. accelerated counselling protocols, application of treatment readiness survey), used treatment eligibility criteria in use before Treat-All, restricted ART interventions to specific patient groups (e.g. non-pregnant adults) or few facilities, or applied different definitions of same-day ART.7,15–17 In contrast, benefits of same-day ART initiation remained uncertain in observational studies from RLS.7 Secondly, real-world effectiveness may be compromised because of pre-existing constraints in the public sector, such as resource limitation (e.g. human resources), overburdened health facilities and suboptimal quality of care.18–21

Treat-All has been implemented in a public sector setting in southern Eswatini (formerly Swaziland) since 2014, with same-day ART initiation increasingly practised.12 Therefore, this setting provides a unique opportunity to better understand how much faster ART should be started in a context where it is already started relatively quickly. We aimed to answer the following questions: 1) how is same-day ART already being implemented in a public sector programme applying the Treat-All approach, and 2) what is the effect of same-day ART initiation compared with early ART initiation (1–14 days after HIV care enrolment) on treatment outcomes for patients starting treatment quickly.

METHODS

Setting

Details of the study setting were described elsewhere.12,22 In brief, Eswatini has an HIV prevalence of 32% among adults aged 18–49 years, and annual TB incidence was 308 cases per 100,000 population with 75% HIV co-infection in 2017.23,24 Treat-All was piloted in eight primary and one secondary care public sector facilities in the predominantly rural Nhlangano health zone of the Shiselweni region. ART initiation was possible in the absence of baseline CD4 cell counts and biochemistry results.25 ART initiation on the day of facility-based HIV care enrolment was policy for pregnant/lactating women and encouraged for other patients in the absence of (presumptive) opportunistic infections.25,26 Without specific standard operating procedures in place for same-day ART initiation under Treat-All at that time, same-day ART was practised at the clinician’s discretion based on the clinical and psychological pre-treatment assessment and depended on the patient’s self-perceived readiness. Led by lay counsellors, one group-counselling session and at least one individual-counselling session were recommended, and both could happen on the same day as HIV diagnosis, care enrolment and ART initiation. Adherence counselling support continued thereafter as per patients’ needs. Routine follow-up visits were scheduled at 2, 4 and 12 weeks after ART initiation and 3-monthly thereafter. Routine viral load monitoring was performed 6 and 12 months after ART initiation and annually thereafter. Patients with viral loads >1000 copies/mL received enhanced adherence counselling over 3 months and were switched to second-line ART in case of viral failure (two consecutive viral load measurements >1000 copies/mL).27 Patients who missed their clinical appointment for medication refills received phonic defaulter tracing with the possibility of home visits.

Study design

This is a nested, retrospectively established cohort of adults ≥16 years old initiating standard first-line ART under the Treat-All programmatic approach in Nhlangano health zone either on the day of facility-based HIV care enrolment (same-day ART) or 1–14 days after HIV care enrolment (early ART), between 10 October 2014 and 31 March 2016. A standard first-line treatment regimen contained lamivudine, and tenofovir or zidovudine, and efavirenz or nevirapine. A patient was considered enrolled into HIV care and initiated on ART if a paper and/or electronic patient record was created. In this setting, we considered early ART as a relevant comparison group to same-day ART because this was the national policy in place at the time of the study.

Analyses and main definitions

Analyses were performed with Stata 14.1 and R. Firstly, baseline characteristics were described with frequency statistics and proportions. The Pearson’s chi-squared and Mann-Whitney U test were used to compare differences in categorical and continues variables. We used multiple imputation by chained equations28 to deal with missing values of the measured pre-treatment variables.

Secondly, we assessed predictors of same-day ART initiation compared with early ART by using multivariable logistic regression models including all variables measured before treatment initiation (listed in Table 1 and Figure 1).

View this table:
Table 1:

Baseline characteristics of patients initiated on ART under same-day and early ART, and predictors of same-day ART initiation

Figure 1:
Figure 1:

Directed acyclic graph (DAG) showing structural assumptions about the data-generating process

Thirdly, we emulated a Target Trial29–31 of HIV-infected patients aged ≥16 years already initiated on ART within 14 days of facility-based HIV care enrolment to estimate the causal effect32 of same-day ART (vs early ART) on the composite unfavourable treatment outcome of death, loss to follow-up (LTFU), viral failure and treatment switching to a second-line ART in the absence of documented viral failure. Time zero was the date of ART initiation because some captured outcomes (viral failure, treatment switch) could only have happened after ART initiation and the outcomes of death and LTFU before ART initiation were not well defined (e.g. pre-treatment visits were not recorded after care enrolment, which may lead to possible misclassification of deaths as LTFU). Therefore, our target population excluded patients starting treatment >14 days after care enrolment and patients never starting treatment for any reason (including deaths within 14 days of care enrolment).

Viral failure was defined as two consecutive viral load measurements >1000 copies/mL measured at least 5 months after ART initiation and 1.5 months apart. The composite endpoint was chosen to reflect the UNAIDS 90-90-90 cascade goals of keeping patients on effective (virally suppressed) ART. Minimum follow-up time before database closure was 7 months. Patients were censored when they were transferred out and at database closure (31 October 2017). LTFU was defined as no-show to the facility for ≥6 months measured at the last clinic visit.

We summarized our assumptions about the data-generating process in a directed acyclic graph (DAG, Figure 1); see Supplementary Material (Technical Appendix) for a detailed explanation. Briefly, treatment assignment is based on various factors, including pregnancy, clinician’s preference in each facility, temporal trends, the patient’s perceived readiness and the impact of counselling, and clinical assessment including CD4 count and co-morbidities. Timing of treatment initiation may affect the composite outcome in different ways: firstly, biologically, if treatment delay would affect viral suppression and thus the development of co-morbidities and negative outcomes; secondly, earlier treatment may have a psychological impact on patients. If they do not feel ready for ART and are possibly coerced into treatment, adherence to therapy may be suboptimal and treatment may be interrupted. The DAG shows that inclusion of all visualized pre-treatment variables, and exclusion of all post-treatment variables (e.g. suppression during follow-up, ART regimen), is sufficient to identify the desired total causal effect (because all back-door paths are blocked and no mediators are being conditioned on).33 However, as treatment readiness and counselling, as well as some baseline co-morbidities (e.g. cryptococcal meningitis), are unmeasured, some remaining unmeasured confounding may persist in our analysis.

Based on the above assumptions, we included all measured pre-treatment variables in an adjusted flexible parametric survival analysis (Royston-Parmar models)34,35 to estimate the effect of same-day ART initiation on the hazard of the unfavourable outcome. We visualized the results of this model using averaged failure and hazard difference curves to compare the time to the composite unfavourable outcome between same-day and early ART.34,35 We allowed the effect of same-day ART to vary with respect to time.

Then, we used targeted maximum likelihood estimation (TMLE)36,37 to estimate the probability of experiencing the unfavourable outcome 12 months, 18 months and 24 months after ART initiation under same-day and early ART, and under no censoring, using all measured pre-treatment variables. TMLE requires estimation of the expected outcome, treatment assignment and censoring processes, given the measured covariates. We facilitated this step using extensive super learning to avoid model mis-specification (see Supplementary Table 1 for details).38,39

Several supplementary analyses were performed. We compared same-day ART with rapid ART initiation defined as ART initiation 1–7 days after HIV care enrolment (rather than early ART) as per WHO recommendations. Then, the composite unfavourable outcome was decomposed to all-cause attrition (death and LTFU combined). Finally, time zero was defined as the date of HIV care enrolment (instead of ART initiation).

Ethics

This retrospective analysis was nested within a prospective cohort study assessing the feasibility of Treat-All12 and was approved by the MSF ethics review board, the Eswatini National Health Research Review Board and the Human Research Ethics Committee of the University of Cape Town.

RESULTS

Figure 2 shows the study flow chart. Of 1899 patients initiating ART, 1341 (70.6%) started treatment within 14 days after facility-based HIV care enrolment. Thirteen (1.0%) patients were removed from the analysis, as study eligibility remained unclear. Of 1328 patients remaining, 839 (63.2%) started ART on the same day as HIV care enrolment.

Figure 2:
Figure 2:

Study flow chart

Predictors of same-day ART initiation

Table 1 shows baseline characteristics of patients starting ART same-day and early. In multivariable analysis (Table 1), the odds of same-day ART initiation were higher for the year 2016 (vs 2014) (adjusted odds ratio [aOR] 2.85; 95% CI: 1.67–4.87), 6 of 8 primary care clinics (vs secondary care clinic) with aORs ranging from 2.88 to 31.91, patients diagnosed ≥90 days before facility-based HIV care enrolment (aOR 4.03; 1.97–8.23) (vs diagnosed on the same day as HIV care enrolment), men (aOR 1.56; 1.10–2.21) and pregnant women (aOR 4.08; 2.80–5.95) (vs non-pregnant women), body mass index ≥25 (aOR 1.44; 1.06–1. 95) and low haemoglobin ≤9 g/dL (aOR 1.61; 1.09–2.39). The aOR tended to be lower for CD4 cell count ≤100 cells/mm3 (aOR 0.63; 0.40–1.00) (vs 201-350), elevated creatinine ≥121 μmol/L (aOR 0.42; 0.17–1.00) and patients co-infected with TB (aOR 0.54; 0.29–1.01).

The effect of same-day ART initiation

Descriptive analyses

Figure 2 shows frequencies of the outcome and follow-up time per intervention. Under same-day ART, retention on ART was lower at the end of the observation period (60.9% vs 70.0%) and LTFU was higher (31.6% vs 24.7%), while other crude outcomes appeared similar (viral failure, ART switch, deaths).

The crude cumulative hazard of remaining on effective first-line ART (not experiencing the unfavourable outcome) was lower for same-day ART (vs early ART) during the first 2 years after ART initiation but comparable at 3 years (see Supplementary Figure 1). For same-day ART, it was 72% (95% CI: 68–74%) (vs early ART: 81%; 77–84%) at 1 year and 62% (59–66%) (vs early ART: 69%; 63–73%) at 3 years. The likelihood of experiencing the unfavourable outcome was high immediately after ART initiation, with 3.7% (95% CI: 2.3–5.8%) and 8.7% (95% CI: 7.0–10.8%) of patients under early and same-day ART never returning to care.

Relative impact of same-day ART on the unfavourable outcome

Multiple imputation of missing values was successful, with good convergence of the imputation algorithm and good other diagnostics (see Supplementary Figure 2 and 3).

The hazard of the unfavourable treatment outcome was increased for same day-ART by 39% in univariate analysis (crude hazard ratio [cHR] 1.39; 95% CI: 1.14–1.70) and by 48% in multivariable analysis (adjusted hazard ratio [aHR] 1.48; 95% CI: 1.16–1.89) (Figure 3a) (see Supplementary Table 2 for the full model). The effect varied over time, with a higher hazard during the first year after ART initiation and a similar hazard thereafter (Figure 3b).

Figure 3:
Figure 3:

Averaged cumulative hazard (a) and averaged difference in hazard rate (b) of the unfavourable outcome for time from ART initiation to unfavourable outcome for patients initiating same-day ART vs early ART

Absolute difference in unfavourable outcomes comparing same-day ART with early ART

Using TMLE, we estimated that 28.9% (95% CI: 25.4–32.3%) of patients would have experienced an unfavourable outcome after 12 months if they had received same-day ART compared with 21.2% (15.8–26.6%) if they had received early ART, which corresponds to a difference of 7.7% (1.3–14.1%) and an odds ratio of 1.51 (1.05–2.17). Differences between the two treatment strategies were also observed for 2 and 3 years of follow-up, though less pronounced than in the first year (see Figure 4). Diagnostics of the TMLE approach were satisfactory, with no truncation of estimated probabilities of treatment assignment, small maximum clever covariates and a broad selection of learning algorithms (see Supplementary Table 1).

Figure 4:
Figure 4:

Main results: estimated effect of same-day ART initiation on the unfavourable outcome using TMLE (panels A and B) and using flexible parametric survival models (panel B)

Supplementary analyses

Table 2 presents crude and adjusted hazard ratios for different assumptions. Changing time zero to the date of care enrolment, changing the unfavourable composite outcome to all-cause attrition and comparing same-day ART with rapid ART did not change findings overall, with adjusted hazard ratios ranging from 1.43 to 1.83.

View this table:
Table 2:

Effect estimates of combination of supplementary analyses with different assumptions

DISCUSSION

This is to our knowledge the first study evaluating faster ART initiation in a routine programmatic HIV-care setting applying the Treat-All policy. Patients initiating ART on the day of facility-based HIV care enrolment had inferior treatment outcomes compared with patients starting treatment 1–14 days thereafter or starting treatment within 1–7 days. The effect was accrued during the first year of therapy.

Interpretation of findings

Predictors of same-day ART

The main predictors of same-day ART initiation were related to policy and programmatic factors. Pregnant women and the calendar year 2016 were associated with increased same-day ART initiation, coinciding with the same-day ART policy under PMTCTB+ and its wider introduction in the latter study period. Facility-level factors also played a role, with almost all primary care facilities providing more same-day ART than the secondary care facility. This may be because primary care facilities had point-of-care biochemistry, haemoglobin and CD4 testing available, thus making baseline results available on the same day for treatment decisions, as opposed to the secondary care facility where results often became available a few days later. Although lack of availability of same-day results should not have delayed prompt ART according to national guidelines,25 clinicians may have felt more comfortable initiating ART with CD4 cell count and biochemistry known. On the other hand, more patients may have had unmeasured co-morbidities at the secondary care facility, necessitating delaying ART initiation.

Clinical factors played some role too. Patients co-infected with TB tended to be less likely to initiate same-day ART, which was in line with clinical guidelines recommending deferral of ART for 2–8 weeks after TB treatment initiation.25,40 Patients with low haemoglobin were more likely to start same-day ART, probably due to greater perceived need for prompt ART initiation, while elevated creatinine level delayed treatment, possibly due to further clinical investigation to rule out kidney impairment before initiation of a tenofovir-based treatment regimen.

Patients who knew their HIV-positive status for ≥90 days were more likely to initiate ART on the same day. They likely comprised two groups. Firstly, patients may have been transferred in from community HIV testing sites and other facilities. Given more time between testing and care enrolment, they may have come to terms with life-long therapy and therefore been ready to start same-day treatment. Secondly, treatment interruptions are frequent in routine settings,41 and these patients may have been treatment interrupters re-initiating ART without disclosing prior treatment.

Lastly, social factors such as level of education and marital status appeared not to play a major role in quicker ART initiation. This may indicate that same-day ART initiation was driven by health policy and programmatic factors as indicated in our analysis rather than by patients’ preferences, knowledge and life circumstances.

Effect of same-day ART

Same-day ART initiation had a higher hazard of the unfavourable treatment outcome than early and rapid ART. This effect was time-varying, with increased hazard during the first year of treatment and similar hazards thereafter.

We provide several explanations. Firstly, same-day ART may not address patients’ concerns about expedited ART initiation in the context of universal ART, and not give enough time to accept an HIV-positive status and to conceptualize lifelong therapy.42–47 This may have contributed to immediate disengagement from care after treatment initiation. In this study, 9% of patients under same-day ART never returned for a follow-up visit.

Secondly, estimates may be affected by unmeasured confounding. For instance, treatment readiness may predict assignment to the intervention and is also likely associated with the outcome (through the factor of adherence). In addition, we could not measure all possible baseline and time-updated co-morbidities that may predict the intervention and the outcome. For instance, cryptococcal meningitis – a main cause of death among PLHIV48 – may have been unevenly distributed in the groups and affect early death and loss to care differently.

Thirdly, health systems factors may have played a role. The clinical and counselling tools used to assess treatment readiness may have been inappropriate to identify patients ready for same-day ART, as the very same tools were used before same-day ART initiation was an option. Contextualized screening tools for expedited ART initiation adapted to different populations (e.g. pregnant women) and settings may be needed to reliably assess patients’ readiness for same-day treatment. For instance, one randomized trial used a treatment readiness survey to identify patients not ready for same-day ART initiation and excluded them from expedited treatment.16 In addition, training related to expedited counselling protocols and same-day ART for health workers during the early implementation period was lacking. This may have left health workers poorly equipped for effective implementation of same-day ART at scale. Lastly, counselling support after same-day ART initiation may have been de-prioritized in this busy public sector setting with competing activities, thus providing insufficient adherence support early during treatment.

Findings in context

The definition of same-day ART – and, as such, intervention exposure – differs across studies, compromising comparison of findings. Definitions include treatment initiation on the day of HIV diagnosis, day of treatment eligibility, day of HIV care enrolment, or a combination of them.15–17,49–51 The same-day ART intervention group often consisted of patients initiating treatment days after the offer of same-day treatment,15,17,50 so that studies evaluated the intention to initiate same-day treatment rather than actual same-day treatment initiation.15–17 The offer of same-day ART was often combined with additional interventions (e.g. point-of-care CD4 and biochemistry testing),52 and restriction of the patient sample to healthier individuals16 and non-pregnant adults15–17 may make findings less applicable to routine public sector settings. Streamlining definitions of same-day ART initiation and clarity of what and who is evaluated are warranted.

While same-day ART initiation improves treatment uptake, it may downshift loss to care to the time of treatment.15,52 Treatment interruptions were already common in routine HIV programmes before the introduction of the rapid ART policy,41,53 and are associated with acquired drug resistance.54 Balancing of patient-level and public health benefits and risks (e.g. possible higher (temporary) disengagement from treatment) is required to make an informed health policy decision.

More emphasis may be needed on a differentiated approach to ART initiation adapted to the patient’s needs, with clinical and programmatic (e.g. logistical) constraints taken into consideration, than on choosing between same-day and rapid/early ART initiation as a blanket approach. In fragile health systems, hasty low quality and possibly coerced ART initiation may occur if HIV programmes and funding organizations prioritize achieving targets related to numbers of same-day ART initiations instead of differentiated patient-centred rapid ART initiation with the option of quality same-day treatment for patients ready to start.

Importantly, this study did not assess the impact of a policy of same-day ART initiation for all PLHIV, as this was not feasible with the observational study design and available data. Thus, findings are not directly comparable to randomized trials evaluating the offer of same-day treatment to treatment-eligible patients. Our research, however, intends to estimate the risks and benefits of same-day ART initiation for patients with the ability to start treatment early. If there is a causal relationship between same-day ART and unfavourable treatment outcome, then deferral of treatment initiation should be considered for these patients. For patients wishing to start treatment the same day, enhanced support should be provided early during treatment. However, more research into the methods may be required to address questions of frequency, intensity, content and minimum quality of early adherence support in routine public sector settings.

Limitations and strengths

Firstly, this study assessed outcomes of patients successfully initiated on ART soon after facility-based HIV care enrolment. Therefore, it differs from previous randomized controlled trials and observational studies.15–17,50,51 Restriction allowed the establishment of two potentially comparable groups in the context of an observational study design but limits direct comparison with settings where most patients initiate ART 2 weeks after care enrolment. It was beyond the scope of this study to assess outcomes of patients starting treatment late or never, and they may differ in their characteristics and risks for an adverse outcome.

Secondly, we did not account for loss between the diagnosis of HIV and care enrolment. This may have caused selection bias in our analysis because only patients successfully linked to facility-based HIV care are considered. Specifically, loss between community-based HIV diagnosis and facility-based enrolment can be high.17,55 Therefore, findings should not be generalized to predominantly community settings but rather to settings similar to ours where most HIV diagnosis happens at facility-level.56

Thirdly, we could not adjust for all possible confounding factors identified in the DAG (e.g. co-morbidities and treatment readiness), possibly biasing the effect estimate in either direction.

A strength of this study is that we applied different analytical approaches, including state-of-the-art methods (TMLE), all of which concurred in their main findings. We also included a wide range of patients as found in other HIV programmes implementing the Treat-All programmatic approach, so findings may be generalizable to other similar settings in rural Sub-Saharan Africa. This study also discussed potential shortfalls in programmatic implementation of Treat-All related to contextualized screening tools and training provided, thus drawing more attention to the method and quality of implementation.

Conclusions

Facility, programmatic and clinical factors were the main predictors of same-day ART initiation. The failure to see an obvious benefit of same-day ART in patients who started treatment on the day of care enrolment suggests that caution is needed in implementing such an approach, although more data from different settings may be needed to confirm our findings.

Data Availability

The software code and data can be obtained from the authors. New analyses on the data require approval from the government of Eswatini and MSF.

Supplementary Material

Supplementary Material, Technical appendix: Some explanation of the developed directed acyclic graph (DAG)

Treatment assignment, such as the decision to initiate ART on the same day, was based on the following considerations. Firstly, ART initiation on the day of facility-based HIV care enrolment was policy for pregnant/lactating women and encouraged for other patients in the absence of (presumptive) opportunistic infections. Secondly, same-day ART was practised at the clinician’s discretion based on the clinical and psychological pre-treatment assessment and depended on the patient’s self-perceived readiness, where the clinician’s advice may vary by facility. Thirdly, baseline CD4 cell count may also predict timing of ART initiation (e.g. patients with low CD4 count and no symptoms suggestive of opportunistic infection may initiate ART on the same day, patients with low CD4 cell count and symptoms suggestive of opportunistic infection may delay ART initiation and patients with high CD4 count may initiate immediately or delay). Fourthly, treatment readiness of the patient with regards to psychological preparedness (and not clinical readiness) at baseline could be a main factor predicting same-day ART initiation, as clinicians would not enforce same-day ART if the patient communicated not being ready for it. Fifthly, a patient’s marital status may also affect the decision to start treatment on the same day; married patients may favour deferring the decision on when to start treatment, as they may first wish to consult with their partner. However, higher education level may accelerate initiation, as knowledge of HIV and the benefit of treatment may be increased. Sixthly, the decision to start on the same day may relate to calendar year, as practising clinicians may change advice with increasing experience. Lastly, the time from HIV diagnosis to care enrolment may influence same-day ART, with patients knowing their HIV status for longer possibly being more ready to start treatment immediately. For these reasons, we considered sex, pregnancy status, age, comorbidities at enrolment, laboratory values at enrolment, education, marital status, temporal trends (calendar year), time of HIV diagnosis, facility, treatment counselling and treatment readiness to be relevant to deciding on the timing of treatment assignment. Consequently, these variables lie on back-door paths from treatment to the outcome.1

There are two ways through which timing of ART initiation could affect the composite outcome. The first way is biologically, if treatment delay would affect viral suppression and thus the development of co-morbidities and negative outcomes. Secondly, earlier treatment may have a psychological impact on patients, who, if they did not feel ready for ART and were pushed into treatment, could stop being adherent to their treatment or even be lost to the programme.

Note that all post-treatment variables, which are visualized in the DAG, are mediators on the path from the intervention to the outcome and should thus not be conditioned upon.1 However, those variables that determine treatment assignment, as described above, are crucial to block back-door paths from the intervention to the outcome. While we have measured most of these variables, treatment readiness and counselling are unmeasured, as are some baseline comorbidities; this suggests the possibility of some unmeasured confounding.

View this table:
Supplementary Table 1: Details and diagnostics for the TMLE analysis
View this table:
Supplementary Table 2: Univariate and multivariate analysis of time to unfavourable treatment outcome for patients initiating same-day ART vs early ART
Supplementary Figure 1:
Supplementary Figure 1: Cumulative hazard of favourable treatment outcome for patients initiating same-day ART vs early ART
Supplementary Figure 2:
Supplementary Figure 2: Trace plots of imputed data for all covariates with missing values
Supplementary Figure 3:
Supplementary Figure 3: Kernel density plots for imputed CD4 cell count and haemoglobin for all imputed datasets as an example using the midiagplots command in Stata

Acknowledgements

We thank all the patients and healthcare workers who were involved in piloting the Treat-All approach and accelerated ART initiation in the Shiselweni region, and specifically the patients in Nhlangano health zone. In addition, we thank all the MSF teams involved in data collection and data cleaning.

REFERENCES

  1. 1.↵
    World Health Organization, Department of HIV/AIDS. Consolidated Guidelines on the Use of Antiretroviral Drugs for Treating and Preventing HIV Infection: Recommendations for a Public Health Approach. Geneva, Switzerland: World Health Organization; 2016. http://apps.who.int/iris/bitstream/10665/208825/1/9789241549684_eng.pdf. Accessed July 31, 2016.
  2. 2.↵
    WHO HIV Policy Adoption and Implementation Status in Countries. Geneva, Switzerland: World Health Organization; 2019. http://www.who.int/hiv/pub/arv/treat-all-uptake/en/. Accessed January 8, 2019.
  3. 3.↵
    UNAIDS Data 2019. Geneva, Switzerland: Joint United Nations Programme on HIV/AIDS (UNAIDS); 2019.
  4. 4.↵
    90-90-90 An Ambitious Treatment Target to Help End the AIDS Epidemic. Geneva, Switzerland: UNAIDS; 2014.
  5. 5.↵
    Geng EH, Havlir DV. The science of rapid start—From the when to the how of antiretroviral initiation. PLoS Med. 2017;14(7). doi:10.1371/journal.pmed.1002358
    OpenUrlCrossRef
  6. 6.↵
    Phanuphak N, Seekaew P, Phanuphak P. Optimising treatment in the test-and-treat strategy: what are we waiting for? - PubMed - NCBI. Lancet HIV. 2019;6:e715–22.
    OpenUrl
  7. 7.↵
    Ford N, Migone C, Calmy A, et al. Benefits and risks of rapid initiation of antiretroviral therapy. AIDS. 2018;32(1):17. doi:10.1097/QAD.0000000000001671
    OpenUrlCrossRefPubMed
  8. 8.↵
    Guidelines for Managing Advanced HIV Disease and Rapid Initiation of Antiretroviral Therapy, July 2017. Geneva, Switzerland: World Health Organization; 2017.
  9. 9.↵
    Swaziland Integrated HIV Management Guidelines. Mbabane, Eswatini: Ministry of Health of Swaziland; 2018.
  10. 10.
    Brazier E, Maruri F, Duda SN, et al. Implementation of “Treat-all” at adult HIV care and treatment sites in the Global IeDEA Consortium: results from the Site Assessment Survey. J Int AIDS Soc. 2019;22(7):e25331. doi:10.1002/jia2.25331
    OpenUrlCrossRef
  11. 11.
    Tymejczyk O, Brazier E, Yiannoutsos CT, et al. Changes in rapid HIV treatment initiation after national “treat all” policy adoption in 6 sub-Saharan African countries: Regression discontinuity analysis. PLOS Med. 2019;16(6):e1002822. doi:10.1371/journal.pmed.1002822
    OpenUrlCrossRefPubMed
  12. 12.↵
    Kerschberger B, Jobanputra K, Schomaker M, et al. Feasibility of antiretroviral therapy initiation under the treat-all policy under routine conditions: a prospective cohort study from Eswatini. J Int AIDS Soc. 2019;22(10):e25401. doi:10.1002/jia2.25401
    OpenUrlCrossRef
  13. 13.↵
    Ross J, Sinayobye J d’Amour, Yotebieng M, et al. Early outcomes after implementation of treat all in Rwanda: an interrupted time series study. J Int AIDS Soc. 2019;22(4):e25279. doi:10.1002/jia2.25279
    OpenUrlCrossRef
  14. 14.↵
    Cuzin L, Cotte L, Delpierre C, et al. Too fast to stay on track? Shorter time to first antiretroviral regimen is not associated with better retention in care in the French Dat’AIDS cohort. PLoS ONE. 2019;14(9). doi:10.1371/journal.pone.0222067
    OpenUrlCrossRef
  15. 15.↵
    Rosen S, Maskew M, Fox MP, et al. Initiating antiretroviral therapy for HIV at a patient’s first clinic visit: The RapIT randomized controlled trial. PLoS Med. 2016;13(5):e1002015. doi:10.1371/journal.pmed.1002015
    OpenUrlCrossRefPubMed
  16. 16.↵
    Koenig SP, Dorvil N, Dévieux JG, et al. Same-day HIV testing with initiation of antiretroviral therapy versus standard care for persons living with HIV: A randomized unblinded trial. PLoS Med. 2017;14(7). doi:10.1371/journal.pmed.1002357
    OpenUrlCrossRef
  17. 17.↵
    Labhardt ND, Ringera I, Lejone TI, et al. Effect of offering same-day ART vs usual health facility referral during home-based HIV testing on linkage to care and viral suppression among adults with HIV in Lesotho: The CASCADE randomized clinical trial. JAMA. 2018;319(11):1103–1112. doi:10.1001/jama.2018.1818
    OpenUrlCrossRef
  18. 18.↵
    Nglazi MD, Lawn SD, Kaplan R, et al. Changes in programmatic outcomes during 7 years of scale-up at a community-based antiretroviral treatment service in South Africa. JAIDS J Acquir Immune Defic Syndr. 2011;56(1):e1–e8.
    OpenUrl
  19. 19.
    Grimsrud A, Balkan S, Casas EC, et al. Outcomes of antiretroviral therapy over a 10-year period of expansion: a multicohort analysis of African and Asian HIV programs. J Acquir Immune Defic Syndr 1999. 2014;67(2):e55–66. doi:10.1097/QAI.0000000000000268
    OpenUrlCrossRefPubMed
  20. 20.
    Cornell M, Grimsrud A, Fairall L, et al. Temporal changes in programme outcomes among adult patients initiating antiretroviral therapy across South Africa, 2002-2007. AIDS Lond Engl. 2010;24(14):2263–2270. doi:10.1097/QAD.0b013e32833d45c5
    OpenUrlCrossRefPubMedWeb of Science
  21. 21.↵
    Lynch S, Ford N, Cutsem G van, et al. Getting HIV treatment to the most people. Science. 2012;337(6092):298–300. doi:10.1126/science.1225702
    OpenUrlAbstract/FREE Full Text
  22. 22.↵
    Kerschberger B, Schomaker M, Ciglenecki I, et al. Programmatic outcomes and impact of rapid public sector antiretroviral therapy expansion in adults prior to introduction of the WHO treat-all approach in rural Eswatini. Trop Med Int Health. 2019;24(6):701–714. doi: 10.1111/tmi.13234
    OpenUrlCrossRef
  23. 23.↵
    Bicego GT, Nkambule R, Peterson I, et al. Recent patterns in population-based HIV prevalence in Swaziland. PLoS ONE. 2013;8(10):e77101. doi:10.1371/journal.pone.0077101
    OpenUrlCrossRefPubMed
  24. 24.↵
    Organization WH. Global Tuberculosis Report 2018. World Health Organization; 2018.
  25. 25.↵
    Swaziland Integrated HIV Management Guidelines. Mbabane, Swaziland: Ministry of Health Swaziland; 2015.
  26. 26.↵
    Parker LA, Jobanputra K, Okello V, et al. Barriers and facilitators to combined ART initiation in pregnant women with HIV: lessons learnt from a PMTCT B+ pilot program in Swaziland. JAIDS J Acquir Immune Defic Syndr. January 2015:1. doi:10.1097/QAI.0000000000000537
    OpenUrlCrossRefPubMed
  27. 27.↵
    Jobanputra K, Parker LA, Azih C, et al. Impact and programmatic implications of routine viral load monitoring in Swaziland. J Acquir Immune Defic Syndr 1999. 2014;67(1):45–51. doi:10.1097/QAI.0000000000000224
    OpenUrlCrossRefPubMed
  28. 28.↵
    White IR, Royston P, Wood AM. Multiple imputation using chained equations: Issues and guidance for practice. Stat Med. 2011;30(4):377–399. doi:10.1002/sim.4067
    OpenUrlCrossRefPubMed
  29. 29.↵
    Hernán MA, Robins JM. Using Big Data to Emulate a Target Trial When a Randomized Trial Is Not Available. Am J Epidemiol. 2016;183(8):758–764. doi:10.1093/aje/kwv254
    OpenUrlCrossRefPubMed
  30. 30.
    Lodi S, Phillips A, Lundgren J, et al. Effect Estimates in Randomized Trials and Observational Studies: Comparing Apples With Apples. Am J Epidemiol. 2019;188(8):1569–1577. doi:10.1093/aje/kwz100
    OpenUrlCrossRef
  31. 31.↵
    Schomaker M, Kühne F, Siebert U. Re: “Effect Estimates in Randomized Trials and Observational Studies: Comparing Apples with Apples.” Am J Epidemiol. September 2019. doi:10.1093/aje/kwz194
    OpenUrlCrossRef
  32. 32.↵
    Hernán MA. The C-Word: Scientific Euphemisms Do Not Improve Causal Inference From Observational Data. Am J Public Health. 2018;108(5):616–619. doi:10.2105/AJPH.2018.304337
    OpenUrlCrossRefPubMed
  33. 33.↵
    Pearl J. An Introduction to Causal Inference. Int J Biostat. 2010;6(2). doi:10.2202/15574679.1203
    OpenUrlCrossRef
  34. 34.↵
    Royston P, Lambert PC. Flexible Parametric Survival Analysis Using Stata: Beyond the Cox Model. Texas, US: StataCorp LP; 2011.
  35. 35.↵
    Lambert PC, Royston P. Further development of flexible parametric models for survival analysis. Stata J. 2009;9(2):265–290.
    OpenUrlCrossRefWeb of Science
  36. 36.↵
    Luque-Fernandez MA, Schomaker M, Rachet B, Schnitzer ME. Targeted maximum likelihood estimation for a binary treatment: A tutorial. Stat Med. 2018;37(16):2530–2546. doi:10.1002/sim.7628
    OpenUrlCrossRef
  37. 37.↵
    Lendle SD, Schwab J, Petersen ML, Laan MJ van der. ltmle: An R package implementing targeted minimum loss-based estimation for longitudinal data. J Stat Softw. 2017;81(1):1–21. doi:10.18637/jss.v081.i01
    OpenUrlCrossRef
  38. 38.↵
    Laan MJ van der, Rose S. Targeted Learning in Data Science: Causal Inference for Complex Longitudinal Studies. Springer International Publishing; 2018. https://www.springer.com/gp/book/9783319653037. Accessed March 23, 2020.
  39. 39.↵
    Schomaker M, Luque-Fernandez MA, Leroy V, Davies MA. Using longitudinal targeted maximum likelihood estimation in complex settings with dynamic interventions. Stat Med. 2019;38(24):4888–4911. doi:10.1002/sim.8340
    OpenUrlCrossRef
  40. 40.↵
    World Health Organization. Consolidated Guidelines on the Use of Antiretroviral Drugs for Treating and Preventing HIV Infection: Recommendations for a Public Health Approach. Geneva, Switzerland: World Health Organization; 2013. http://www.ncbi.nlm.nih.gov/books/NBK195400/. Accessed December 1, 2014.
  41. 41.↵
    Kranzer K, Ford N. Unstructured treatment interruption of antiretroviral therapy in clinical practice: a systematic review. Trop Med Int Health TM IH. 2011;16(10):1297–1313. doi:10.1111/j.1365-3156.2011.02828.x
    OpenUrlCrossRefPubMed
  42. 42.↵
    Horter S, Thabede Z, Dlamini V, et al. “Life is so easy on ART, once you accept it”: Acceptance, denial and linkage to HIV care in Shiselweni, Swaziland. Soc Sci Med 1982. 2017;176:52–59. doi:10.1016/j.socscimed.2017.01.006
    OpenUrlCrossRef
  43. 43.
    Mamba HT, Hlongwana KW. Deterrents to Immediate Antiretroviral Therapy Initiation by Pregnant Women Living with HIV in Hhohho Region, Swaziland. Afr J Reprod Health. 2018;22(4):72-.
    OpenUrl
  44. 44.
    Stinson K, Myer L. Barriers to initiating antiretroviral therapy during pregnancy: a qualitative study of women attending services in Cape Town, South Africa. Afr J AIDS Res. 2012;11(1):65–73. doi:10.2989/16085906.2012.671263
    OpenUrlCrossRefPubMed
  45. 45.
    Horter S, Bernays S, Thabede Z, et al. “I don’t want them to know”: how stigma creates dilemmas for engagement with Treat-all HIV care for people living with HIV in Eswatini. Afr J AIDS Res. 2019;0(0):1–11. doi:10.2989/16085906.2018.1552163
    OpenUrlCrossRef
  46. 46.
    Pell C, Vernooij E, Masilela N, Simelane N, Shabalala F, Reis R. False starts in “test and start”: a qualitative study of reasons for delayed antiretroviral therapy in Swaziland. Int Health. 2018;10(2):78–83. doi:10.1093/inthealth/ihx065
    OpenUrlCrossRef
  47. 47.↵
    Katirayi L, Namadingo H, Phiri M, et al. HIV-positive pregnant and postpartum women’s perspectives about Option B+ in Malawi: a qualitative study. J Int AIDS Soc. 2016;19(1). doi:10.7448/IAS.19.1.20919
    OpenUrlCrossRef
  48. 48.↵
    Lawrence DS, Boyer-Chammard T, Jarvis JN. Emerging concepts in HIV-associated cryptococcal meningitis. Curr Opin Infect Dis. 2019;32(1):16–23. doi:10.1097/QCO.0000000000000514
    OpenUrlCrossRef
  49. 49.↵
    MacKellar D, Williams D, Bhembe B, et al. Peer-delivered linkage case management and same-day ART initiation for men and young Persons with HIV infection — Eswatini, 2015-2017. Morb Mortal Wkly Rep. 2018;67(23):663–667. doi:10.15585/mmwr.mm6723a3
    OpenUrlCrossRef
  50. 50.↵
    Langwenya N, Phillips TK, Brittain K, Zerbe A, Abrams EJ, Myer L. Same-day antiretroviral therapy (ART) initiation in pregnancy is not associated with viral suppression or engagement in care: A cohort study. J Int AIDS Soc. 2018;21(6). doi:10.1002/jia2.25133
    OpenUrlCrossRef
  51. 51.↵
    Domercant JW, Puttkammer N, Young P, et al. Attrition from antiretroviral treatment services among pregnant and non-pregnant patients following adoption of Option B+ in Haiti. Glob Health Action. 2017;10(1). doi:10.1080/16549716.2017.1330915
    OpenUrlCrossRef
  52. 52.↵
    Stevens WS, Gous NM, MacLeod WB, et al. Multidisciplinary point-of-care testing in South African primary health care clinics accelerates HIV ART Initiation but does not alter retention in care. JAIDS J Acquir Immune Defic Syndr. 2017;76(1):65. doi:10.1097/QA1.0000000000001456
    OpenUrlCrossRef
  53. 53.↵
    Kaplan SR, Oosthuizen C, Stinson K, et al. Contemporary disengagement from antiretroviral therapy in Khayelitsha, South Africa: A cohort study. PLoS Med. 2017; 14(11). doi:10.1371/journal.pmed.1002407
    OpenUrlCrossRef
  54. 54.↵
    Oyugi JH, Byakika-Tusiime J, Ragland K, et al. Treatment interruptions predict resistance in HIV-positive individuals purchasing fixed-dose combination antiretroviral therapy in Kampala, Uganda. AIDS. 2007;21(8):965. doi:10.1097/QAD.0b013e32802e6bfa
    OpenUrlCrossRefPubMedWeb of Science
  55. 55.↵
    Iwuji CC, Orne-Gliemann J, Larmarange J, et al. Uptake of Home-Based HIV Testing, Linkage to Care, and Community Attitudes about ART in Rural KwaZulu-Natal, South Africa: Descriptive Results from the First Phase of the ANRS 12249 TasP ClusterRandomised Trial. PLoS Med. 2016;13(8). doi:10.1371/journal.pmed.1002107
    OpenUrlCrossRef
  56. 56.↵
    HIV Annual Program Report 2018. Mbabane, Eswatini: Ministry of Health of Eswatini; 2019.

REFERENCES

  1. 1.↵
    Pearl J. An Introduction to Causal Inference. Int J Biostat. 2010;6(2). doi:10.2202/1557-4679.1203
    OpenUrlCrossRef
  2. 2.
    Lendle SD, Schwab J, Petersen ML, Laan MJ van der. Itmle: An R package implementing targeted minimum loss-based estimation for longitudinal data. J Stat Softw. 2017;81(1):1–21. doi:10.18637/jss.v081.i01
    OpenUrlCrossRef
  3. 3.
    van der Laan MJ, Polley EC, Hubbard AE. Super learner. Stat Appl Genet Mol Biol. 2007;6:Article25. doi:10.2202/1544-6115.1309
    OpenUrlCrossRef
  4. 4.
    van der Laan M, Rose M, Rose E. Targeted Learning. Springer; 2011.
  5. 5.
    Schomaker M. MAMI: Model Averaging (and Model Selection) after Multiple Imputation - R package Version 0.9.12 2018. http://mami.r-forge.r-project.org.
  6. 6.
    Schomaker M. Shrinkage averaging estimation. Stat Pap. 2012;53(4):1015–1034. doi:10.1007/s00362-011-0405-2
    OpenUrlCrossRef
  7. 7.
    Tibshirani R. Regression Shrinkage and Selection via the Lasso. J R Stat Soc Ser B Methodol. 1996;58(1):267–288.
    OpenUrl
  8. 8.
    Heumann C, Schomaker M, Shalabh. Introduction to Statistics and Data Analysis: With Exercises, Solutions and Applications in R. Springer International Publishing; 2016. https://www.springer.com/de/book/9783319461601. Accessed January 11, 2020.
  9. 9.
    Golub GH, Heath M, Wahba G. Generalized Cross-Validation as a Method for Choosing a Good Ridge Parameter. Technometrics. 1979;21(2):215–223. doi:10.2307/1268518
    OpenUrlCrossRefWeb of Science
Back to top
Posted May 08, 2020.
Download PDF
Data/Code
Email
The impact of same-day antiretroviral therapy initiation under the WHO Treat-All policy
Bernhard Kerschberger, Andrew Boulle, Rudo Kuwengwa, Iza Ciglenecki, Michael Schomaker
medRxiv 2020.05.04.20090068; doi: https://doi.org/10.1101/2020.05.04.20090068
Twitter logo Facebook logo LinkedIn logo Mendeley logo
Citation Tools

Subject Area