Explore the vast range of titles available.

Background In a non-inferiority trial, the choice of margin depends on the expected control event risk. If the true risk differs from expected, power and interpretability of results can be affected. A non-inferiority frontier pre-specifies an appropriate non-inferiority margin for each value of control event risk. D3 is a non-inferiority trial comparing two treatment regimens in children living with HIV, designed assuming a control event risk of 12%, a non-inferiority margin of 10%, 80% power and a significance level (α) of 0.025. We consider approaches to choosing and implementing a frontier for this already funded trial, where changing the sample size substantially would be difficult. Methods In D3, we fix the non-inferiority margin at 10%, 8% and 5% for control event risks of ≥9%, 5% and 1%, respectively. We propose four frontiers which fit these fixed points, including a Smooth Away From Expected (SAFE) frontier. Analysis approaches considered are as follows: using the pre-specified significance level ( α =0.025); always using a reduced significance level (to achieve α ≤0.025 across control event risks); reducing significance levels only when the control event risk differs significantly from expected (control event risk <9%); and using a likelihood ratio test. We compare power and type 1 error for SAFE with other frontiers. Results Changing the significance level only when the control event risk is <9% achieves approximately nominal (<3%) type I error rate and maintains reasonable power for control event risks between 1 and 15%. The likelihood ratio test method performs similarly, but the results are more complex to present. Other analysis methods lead to either inflated type 1 error or badly reduced power. The SAFE frontier gives more interpretable results with low control event risks than other frontiers (i.e. it uses more reasonable non-inferiority margins). Other frontiers do not achieve power close (i.e. within 1%) to SAFE across the range of likely control event risks while controlling type I error. Conclusions The SAFE non-inferiority frontier will be used in D3, and the non-inferiority margin and significance level will be modified if the control event risk is lower than expected. This ensures results will remain interpretable if design assumptions are incorrect, while achieving similar power. A similar approach could be considered for other non-inferiority trials where the control event risk is uncertain.

Background Evidence suggests that engagement in care (EIC) may be worse in young people living with perinatal HIV (YPLPHIV) compared to adults or children living with HIV. We took a published EIC algorithm for adults with HIV, which takes patients’ clinical scenarios into account, and adapted it for use in YPLPHIV in England, to measure their EIC. Methods The adult algorithm predicts when in the next 6 months the next clinic visit should be scheduled, based on routinely collected clinical indicators at the current visit. We updated the algorithm based on the latest adult guidelines at the time, and modified it for young people in paediatric care using the latest European paediatric guidelines. Paediatric/adolescent HIV consultants from the UK reviewed and adapted the resulting flowcharts. The adapted algorithm was applied to the Adolescent and Adults Living with Perinatal HIV (AALPHI) cohort in England. Data for 12 months following entry into AALPHI were used to predicted visits which were then compared to appointment attendances, to measure whether young people were in care in each month. Proxy markers (e.g. dates of CD4 counts, viral loads (VL)) were used to indicate appointment attendance. Results Three hundred sixteen patients were in AALPHI, of whom 41% were male, 82% of black African ethnicity and 58% born abroad. At baseline (time of AALPHI interview) median [IQR] age was 17 [15–18] years, median CD4 was 597 [427, 791] cells/µL and 69% had VL ≤50c/mL. 10 patients were dropped due to missing data. 306 YPLPHIV contributed 3,585 person months of follow up across the 12 month study in which a clinic visit was recorded for 1,204 months (38/1204 dropped due to missing data). The remaining 1,166 months were classified into 3 groups: Group-A: on ART, VL ≤ 50c/mL—63%(734/1,166) visit months, Group-B: on ART, VL > 50c/mL—27%(320/1,166) Group-C: not on ART-10%(112/1,166). Most patients were engaged in care with 87% (3,126/3,585) of months fulfilling the definition of engaged in care. Conclusions The adapted algorithm allowed the varying clinical scenarios of YPLPHIV to be taken into account when measuring EIC. However availability of good quality surveillance data is crucial to ensure that EIC can be measured well.

Identifying which young people living with perinatally acquired HIV (PHIV) are less likely to engage in care is crucial to allow targeted interventions to support them to attend clinic. We adapted an existing Engagement in Care (EIC) algorithm for adults with HIV in England, for use in young people. We applied it to data from young people with PHIV in the Adolescents and Adults Living with Perinatal HIV (AALPHI) cohort. The algorithm predicts the timing of the next scheduled clinic visit, within 1–6 months of current visit, based on routine clinical data. Follow-up was 12-months from AALPHI baseline interview. Each person-month was classified as engaged in care or not. Logistic regression models (allowing for clustered data) were used to explore baseline characteristics associated with being engaged in care, adjusting for a priori variables (time from interview, sex, age, ethnicity, country of birth). Potential characteristics were across 7 domains: sociodemographic; risk behaviour practices; mental health; cognition; clinic setting; HIV management and experience; and HIV clinical markers. Of 316 young people, 187(59%) were female, 271(86%) of black ethnicity and 184(58%) born abroad. At baseline, median [IQR] age was 17[15–18] years, and 202(69%) had viral load ≤50 copies/ml(c/mL). 87% of 3,585 person-months were classified as engaged in care. Characteristics independently associated with poorer odds of being engaged in care were: Asian/mixed/other ethnicity, vs. black ethnicity (OR 0.44, 95% CI 0.25, 0.78, p = 0.02); ever self-harmed, vs. not (OR 0.55, 95% CI 0.32, 0.95, p = 0.03); on antiretroviral therapy (ART) and self-assessed bad/not so good adherence (OR 0.46, 95% CI 0.25, 0.84) or not on ART (OR 0.64, 95% CI 0.64, 1.21) vs. on ART and good/excellent adherence (p = 0.04)); baseline VL>50c/mL, vs VL≤50c/mL (OR 0.47, 95% CI 0.30, 0.75, p = 0.002). These characteristics can help identify individuals requiring enhanced support to maintain service engagement.

ObjectivesTo examine the voluntariness of consent in paediatric HIV clinical trials and the associated factors.DesignMixed-methods, cross-sectional study combining a quantitative survey conducted concurrently with indepth interviews.Setting and participantsFrom January 2021 to April 2021, we interviewed parents of children on first-line or second-line Anti-retroviral therapy (ART) in two ongoing paediatric HIV clinical trials [CHAPAS-4 (ISRCTN22964075) and ODYSSEY (ISRCTN91737921)] at the Joint Clinical Research Centre Mbarara, Uganda.Outcome measuresThe outcome measures were the proportion of parents with voluntary consent, factors affecting voluntariness and the sources of external influence. Parents rated the voluntariness of their consent on a voluntariness ladder. Indepth interviews described participants’ lived experiences and were aimed at adding context.ResultsAll 151 parents randomly sampled for the survey participated (84% female, median age 40 years). Most (67%) gave a fully voluntary decision, with a score of 10 on the voluntariness ladder, whereas 8% scored 9, 9% scored 8, 6% scored 7, 8% scored 6 and 2.7% scored 4. Trust in medical researchers (adjusted OR 9.90, 95% CI 1.01 to 97.20, p=0.049) and male sex of the parent (adjusted OR 3.66, 95% CI 1.00 to 13.38, p=0.05) were positively associated with voluntariness of consent. Prior research experience (adjusted OR 0.31, 95% CI 0.12 to 0.78, p=0.014) and consulting (adjusted OR 0.25. 95% CI 0.10 to 0.60, p=0.002) were negatively associated with voluntariness. Consultation and advice came from referring health workers (36%), spouses (29%), other family members (27%), friends (15%) and researchers (7%). The indepth interviews (n=14) identified the health condition of the child, advice from referring health workers and the opportunity to access better care as factors affecting the voluntariness of consent.ConclusionsThis study demonstrated a high voluntariness of consent, which was enhanced among male parents and by parents’ trust in medical researchers. Prior research experience of the child and advice from health workers and spouses were negatively associated with the voluntariness of parents’ consent. Female parents and parents of children with prior research experience may benefit from additional interventions to support voluntary participation.

The impact of nutritional supplements on weight gain in HIV-infected children on antiretroviral treatment (ART) remains uncertain. Starting supplements depends upon current weight-for-age or other acute malnutrition indicators, producing time-dependent confounding. However, weight-for-age at ART initiation may affect subsequent weight gain, independent of supplement use. Implications for marginal structural models (MSMs) with inverse probability of treatment weights (IPTW) are unclear. In the ARROW trial, non-randomised supplement use and weight-for-age were recorded monthly from ART initiation. The effect of supplements on weight-for-age over the first year was estimated using generalised estimating equation MSMs with IPTW, both with and without interaction terms between baseline weight-for-age and time. Separately, data were simulated assuming no supplement effect, with use depending on current weight-for-age, and weight-for-age trajectory depending on baseline weight-for-age to investigate potential bias associated with different MSM specifications. In simulations, despite correctly specifying IPTW, omitting an interaction in the MSM between baseline weight-for-age and time produced increasingly biased estimates as associations between baseline weight-for-age and subsequent weight trajectory increased. Estimates were unbiased when the interaction between baseline weight-for-age and time was included, even if the data were simulated with no such interaction. In ARROW, without an interaction the estimated effect was +0.09 (95%CI +0.02,+0.16) greater weight-for-age gain per month's supplement use; this reduced to +0.03 (-0.04,+0.10) including the interaction. This study highlights a specific situation in which MSM model misspecification can occur and impact the resulting estimate. Since an interaction in the MSM (outcome) model does not bias the estimate of effect if the interaction does not exist, it may be advisable to include such a term when fitting MSMs for repeated measures.

Background Applications of causal inference methods to randomised controlled trial (RCT) data have usually focused on adjusting for compliance with the randomised intervention rather than on using RCT data to address other, non-randomised questions. In this paper we review use of causal inference methods to assess the impact of aspects of patient management other than the randomised intervention in RCTs. Methods We identified papers that used causal inference methodology in RCT data from Medline, Premedline, Embase, Cochrane Library, and Web of Science from 1986 to September 2014, using a forward citation search of five seminal papers, and a keyword search. We did not include studies where inverse probability weighting was used solely to balance baseline characteristics, adjust for loss to follow-up or adjust for non-compliance to randomised treatment. Studies where the exposure could not be assigned were also excluded. Results There were 25 papers identified. Nearly half the papers (11/25) estimated the causal effect of concomitant medication on outcome. The remainder were concerned with post-randomisation treatment regimens (sequential treatments, n =5 ), effects of treatment timing ( n  = 2) and treatment dosing or duration ( n  = 7). Examples were found in cardiovascular disease ( n  = 5), HIV (n = 7), cancer ( n  = 6), mental health ( n  = 4), paediatrics ( n  = 2) and transfusion medicine ( n  = 1). The most common method implemented was a marginal structural model with inverse probability of treatment weighting. Conclusions Examples of studies which exploit RCT data to address non-randomised questions using causal inference methodology remain relatively limited, despite the growth in methodological development and increasing utilisation in observational studies. Further efforts may be needed to promote use of causal methods to address additional clinical questions within RCTs to maximise their value.

Background Clinical trial investigators may need to evaluate treatment effects in a specific subgroup (or subgroups) of participants in addition to reporting results of the entire study population. Such subgroups lack power to detect a treatment effect, but there may be strong justification for borrowing information from a larger patient group within the same trial, while allowing for differences between populations. Our aim was to develop methods for eliciting expert opinions about differences in treatment effect between patient populations, and to incorporate these opinions into a Bayesian analysis. Methods We used an interaction parameter to model the relationship between underlying treatment effects in two subgroups. Elicitation was used to obtain clinical opinions on the likely values of the interaction parameter, since this parameter is poorly informed by the data. Feedback was provided to experts to communicate how uncertainty about the interaction parameter corresponds with relative weights allocated to subgroups in the Bayesian analysis. The impact on the planned analysis was then determined. Results The methods were applied to an ongoing non-inferiority trial designed to compare antiretroviral therapy regimens in 707 children living with HIV and weighing ≥ 14 kg, with an additional group of 85 younger children weighing < 14 kg in whom the treatment effect will be estimated separately. Expert clinical opinion was elicited and demonstrated that substantial borrowing is supported. Clinical experts chose on average to allocate a relative weight of 78% (reduced from 90% based on sample size) to data from children weighing ≥ 14 kg in a Bayesian analysis of the children weighing < 14 kg. The total effective sample size in the Bayesian analysis was 386 children, providing 84% predictive power to exclude a difference of more than 10% between arms, whereas the 85 younger children weighing < 14 kg provided only 20% power in a standalone frequentist analysis. Conclusions Borrowing information from a larger subgroup or subgroups can facilitate estimation of treatment effects in small subgroups within a clinical trial, leading to improved power and precision. Informative prior distributions for interaction parameters are required to inform the degree of borrowing and can be informed by expert opinion. We demonstrated accessible methods for obtaining opinions.

Introduction The COVID-19 pandemic has globally impacted health service access, delivery and resources. There are limited data regarding the impact on the prevention of mother to child transmission (PMTCT) service delivery in low-resource settings. Neotree ( www.neotree.org ) combines data collection, clinical decision support and education to improve care for neonates. Here we evaluate impacts of COVID-19 on care for HIV-exposed neonates. Methods Data on HIV-exposed neonates admitted to the neonatal unit (NNU) at Sally Mugabe Central Hospital, Zimbabwe, between 01/06/2019 and 31/12/2021 were analysed, with pandemic start defined as 21/03/2020 and periods of industrial action (doctors (September 2019-January 2020) and nurses (June 2020-September 2020)) included, resulting in modelling during six time periods: pre-doctors’ strike (baseline); doctors’ strike; post-doctors’ strike and pre-COVID; COVID and pre-nurses’ strike; nurses’ strike; post nurses’ strike. Interrupted time series models were used to explore changes in indicators over time. Results Of 8,333 neonates admitted to the NNU, 904 (11%) were HIV-exposed. Mothers of 706/765 (92%) HIV-exposed neonates reported receipt of antiretroviral therapy (ART) during pregnancy. Compared to the baseline period when average admissions were 78 per week (95% confidence interval (CI) 70–87), significantly fewer neonates were admitted during all subsequent periods until after the nurses’ strike, with the lowest average number during the nurses’ strike (28, 95% CI 23–34, p < 0.001). Across all time periods excluding the nurses strike, average mortality was 20% (95% CI 18–21), but rose to 34% (95% CI 25, 46) during the nurses’ strike. There was no evidence for heterogeneity (p > 0.22) in numbers of admissions or mortality by HIV exposure status. Fewer HIV-exposed neonates received a PCR test during the pandemic (23%) compared to the pre-pandemic periods (40%) (RR 0.59, 95% CI 0.41–0.84, p < 0.001). The proportion of HIV-exposed neonates who received antiretroviral prophylaxis during admission was high throughout, averaging between 84% and 95% in each time-period. Conclusion While antiretroviral prophylaxis for HIV-exposed neonates remained high throughout, concerning data on low admissions and increased mortality, similar in HIV-exposed and unexposed neonates, and reduced HIV testing, suggest some aspects of care may have been compromised due to indirect effects of the pandemic.

Interventions based around objective measurement of adherence to antiretroviral drugs for HIV have potential to improve adherence and to enable differentiation of care such that clinical visits are reduced in those with high adherence. It would be useful to understand the approximate upper limit of cost that could be considered for such interventions of a given effectiveness in order to be cost effective. Such information can guide whether to implement an intervention in the light of a trial showing a certain effectiveness and cost. An individual-based model, calibrated to Zimbabwe, which incorporates effects of adherence and resistance to antiretroviral therapy, was used to model the potential impact of adherence monitoring-based interventions on viral suppression, death rates, disability adjusted life years and costs. Potential component effects of the intervention were: enhanced average adherence when on ART, reduced risk of ART discontinuation, and reduced risk of resistance acquisition. We considered a situation in which viral load monitoring is not available and one in which it is. In the former case, it was assumed that care would be differentiated based on the adherence level, with fewer clinic visits in those demonstrated to have high adherence. In the latter case, care was assumed to be primarily differentiated according to viral load level. The maximum intervention cost required to be cost effective was calculated based on a cost effectiveness threshold of $500 per DALY averted. In the absence of viral load monitoring, an adherence monitoring-based intervention which results in a durable 6% increase in the proportion of ART experienced people with viral load < 1000 cps/mL was cost effective if it cost up to $50 per person-year on ART, mainly driven by the cost savings of differentiation of care. In the presence of viral load monitoring availability, an intervention with a similar effect on viral load suppression was cost-effective when costing $23-$32 per year, depending on whether the adherence intervention is used to reduce the level of need for viral load measurement. The cost thresholds identified suggest that there is clear scope for adherence monitoring-based interventions to provide net population health gain, with potential cost-effective use in situations where viral load monitoring is or is not available. Our results guide the implementation of future adherence monitoring interventions found in randomized trials to have health benefit.