Increasing Take‐Up of Social Benefits: A Meta‐Analysis of Field Experiments

Intracoronary autologous bone marrow cell transfer after myocardial infarction: the BOOST-2 randomised placebo-controlled clinical trial.

Randomized controlled trials are considered the golden standard for estimating treatment effect but are costly to perform and not always possible. Observational data, although readily available, is sensitive to biases such as confounding by indication. Structure learning algorithms for Bayesian Networks (BNs) can be used to discover the underlying model from data. This enables identification of confounders through graph analysis, although the model might contain noncausal edges. We propose using a blacklist to aid structure learning in finding causal relationships. This is illustrated by an analysis into the effect of active treatment (v observation) in localized prostate cancer. In total, 4,121 prostate cancer records were obtained from the Netherlands Cancer Registry. Subsequently, we developed a (causal) BN using structure learning while precluding noncausal relations. Additionally, we created several Cox proportional hazards models, each correcting for a different set of potential confounders (including propensity scores). Model predictions for overall survival were compared with expected survival on the basis of the general population using data from Statistics Netherlands (Centraal Bureau voor de Statistiek). Structure learning precluding noncausal relations resulted in a causal graph but did not identify significant edges toward treatment; they were added manually. Graph analysis identified year of diagnosis and age as confounders. The BN predicted a treatment effect of 1 percentage point at 10 years. Chi-squared analysis found significant associations between year of diagnosis, age, stage, and treatment. Propensity score correction was successful. Adjusted Cox models predicted significant treatment effect around 3 percentage points at 10 years. A blacklist in conjunction with structure learning can result in a causal BN that can be used for confounder identification. Treatment effect found here is close to the 5 percentage point found in the literature.

A low-key social insurance reform—effects of multidisciplinary outpatient treatment for back pain patients in Norway

Estimating Treatment Effects in the Presence of Noncompliance and Nonresponse: The Generalized Endogenous Treatment Model

SummaryBackgroundRecipients of health services value not only convenience but also respectful, kind, and helpful providers. To date, research to improve person-centred HIV treatment has focused on making services easier to access (eg, differentiated service delivery) rather than the interpersonal experience of care. We developed and evaluated a Person-Centred Care (PCC) intervention targeting healthcare worker practices.MethodsUsing a stepped-wedge, cluster-randomised design, we randomly allocated 24 HIV clinics stratified by size in Zambia into four groups and introduced a PCC intervention that targeted caring aspects of the behaviour of health-care workers in one group every 6 months. The intervention entailed training and coaching for health-care workers on PCC practices (to capacitate), client experience assessment with feedback to facilities (to create opportunities), and small performance-based incentives (to motivate). In a probability sample of clients who were pre-trained on a client experience exit survey and masked to facility intervention status, we evaluated effects on client experience by use of mean score change and also proportion with poor encounters (≤8 on 12-point survey instrument). We examined effects on missed visits (ie, >30 days late for next scheduled encounter) in all groups and retention in care at 15 months in group 1 and group 4 by use of electronic health records. We assessed effects on treatment success at 15 months (i.e., HIV RNA concentration <400 copies per mL or adjudicated care status) in a prospectively enrolled subset of clients from group 1 and group 4. We estimated treatment effects with mixed-effects logistic regression, adjusting for sex, age, and baseline care status. This trial is registered at the Pan-African Clinical Trials Registry (202101847907585) and is completed.ResultsBetween Aug 12, 2019, and Nov 30, 2021, 177 543 unique clients living with HIV made at least one visit to one of the 24 study clinics. The PCC intervention reduced the proportion of poor visits based on exit surveys from 147 (23·3%) of 632 during control periods to 33 (13·3%) of 249 during the first 6 months of intervention, and then to eight (3·5%) of 230 after 6 months (adjusted risk difference [aRD] for control vs ≥6 months intervention −16.9 percentage points, 95% CI −24·8 to −8·9 Among all adult scheduled appointments, the PCC intervention reduced the proportion of missed visits from 90 593 (25·3%) of 358 741 during control periods to 40 380 (22·6%) of 178 523 in the first 6 months, and then 52 288 (21·5%) of 243 350 after 6 months (aRD for control vs the intervention −4·2 percentage points, 95% CI −4·8 to −3·7). 15-month retention improved from 33 668 (80·2%) of 41 998 in control to 35 959 (83·6%) of 43 005 during intervention (aRD 5·9 percentage points, 95% CI 0·6 to 11·2), with larger effects in clients newly starting treatment (aRD 12·7 percentage points, 1·4 to 23·9). We found no effect on treatment success (based on viral load) in a nested subcohort (379 [83·7%] of 453 in the control phase vs 402 [83·8%] of 480 in the intervention phase; aRD 0·9 percentage points, −5·4 to 7·2).InterpretationImproving the caring aspects of health-care worker behaviour is feasible in public health settings, enhances client experience, reduces missed appointments and increases retention.FundingThe Bill and Melinda Gates Foundation.

Team Incentives, Social Cohesion, and Performance: A Natural Field Experiment

Aims/hypothesisThe aim of this work was to assess the effect of liraglutide on ectopic fat accumulation in individuals with type 2 diabetes mellitus.MethodsThis study is a pre-specified subanalysis of the MAGNetic resonance Assessment of VICTOza efficacy in the Regression of cardiovascular dysfunction In type 2 diAbetes mellitus (MAGNA VICTORIA) study, with primary endpoints being the effects of liraglutide on left ventricular diastolic and systolic function. The MAGNA VICTORIA study was a single-centre, parallel-group trial in 50 individuals with type 2 diabetes mellitus (BMI >25 kg/m2) who were randomly assigned (1:1, stratified for sex and insulin use) to receive liraglutide 1.8 mg once daily or placebo for 26 weeks, added to standard care. Participants, study personnel and outcome assessors were blinded to treatment allocation. The secondary endpoints of visceral adipose tissue (VAT), abdominal subcutaneous adipose tissue (SAT) and epicardial fat were measured with MRI. Hepatic triacylglycerol content (HTGC) and myocardial triacylglycerol content (MTGC) were quantified with proton MR spectroscopy. Between-group differences (change from baseline) were tested for significance using ANCOVA. Mean differences with 95% CIs were reported.ResultsThe trial was completed in 2016. Twenty-four participants were randomised to receive liraglutide and 26 to receive placebo. One patient in the liraglutide group withdrew consent before having received the study drug and was not included in the intention-to-treat analysis. Liraglutide (n = 23) vs placebo (n = 26) significantly reduced body weight (liraglutide 98.4 ± 13.8 kg to 94.3 ± 14.9 kg; placebo 94.5 ± 13.1 kg to 93.9 ± 13.2 kg; estimated treatment effect −4.5 [95% CI −6.4, −2.6] kg). HbA1c declined in both groups without a significant treatment effect of liraglutide vs placebo (liraglutide 66.7 ± 11.5 mmol/mol to 55.0 ± 13.2 mmol/mol [8.4 ± 1.1% to 7.3 ± 1.2%]; placebo 64.7 ± 10.2 mmol/mol to 56.9 ± 6.9 mmol/mol [8.2 ± 1.0% to 7.5 ± 0.7%]; estimated treatment effect −2.9 [95% CI −8.1, 2.3] mmol/mol or −0.3 [95% CI −0.8, 0.2]%). VAT did not change significantly between groups (liraglutide 207 ± 87 cm2 to 203 ± 88 cm2; placebo 204 ± 63 cm2 to 200 ± 55 cm2; estimated treatment effect −7 [95% CI −24, 10] cm2), while SAT was reduced by a significantly greater extent with liraglutide than with placebo (liraglutide 361 ± 142 cm2 to 339 ± 131 cm2; placebo 329 ± 107 cm2 to 333 ± 125 cm2; estimated treatment effect −29 [95% CI −51, −8] cm2). Epicardial fat did not change significantly between groups (liraglutide 8.9 ± 4.3 cm2 to 9.1 ± 4.7 cm2; placebo 9.6 ± 4.1 cm2 to 9.6 ± 4.6 cm2; estimated treatment effect 0.2 [95% CI −1.5, 1.8] cm2). Change in HTGC was not different between groups (liraglutide 18.1 ± 11.2% to 12.0 ± 7.7%; placebo 18.4 ± 9.4% to 14.7 ± 10.0%; estimated treatment effect −2.1 [95% CI −5.3, 1.0]%). MTGC was not different after treatment with liraglutide (1.5 ± 0.6% to 1.2 ± 0.6%) vs placebo (1.3 ± 0.5% to 1.2 ± 0.6%), with an estimated treatment effect of −0.1 (95% CI −0.4, 0.2)%. There were no adjudicated serious adverse events.Conclusions/interpretationCompared with placebo, liraglutide-treated participants lost significantly more body weight. Liraglutide primarily reduced subcutaneous fat but not visceral, hepatic, myocardial or epicardial fat. Future larger studies are needed to confirm the results of this secondary endpoint study.Trial registrationClinicalTrials.gov NCT01761318.FundingThis study was funded by Novo Nordisk A/S (Bagsvaerd, Denmark).

Voluntary land conservation programs depend upon the willingness of land owners to participate. Since participation requires commitment to long‐term contracts, most studies on participation focus on changes to the pecuniary incentives facing land owners. This study presents a large‐scale field experiment within the USDA's Conservation Reserve Program (CRP) that examines whether informational outreach, including behavioral nudges, could improve land owners’ willingness to participate. The experiment evaluates the impact of three types of reminder letters on the rate at which land is offered into the CRP. We find that for the most well‐informed group, farms with expiring CRP contracts, the reminder letters did improve participation. We interpret this result as evidence of inattentive behavior. We do not detect any differences in the estimated treatment effects among the basic reminder letter and the letters augmented with peer comparisons and social norm messaging, nor do we detect any treatment effect among currently unenrolled farms. From a policy perspective, these results imply that the USDA can generate additional CRP offers among farms with expiring contracts at an average cost of $39 per additional offer. Assuming a twenty‐five million acre program, reminder letters sent during every sign‐up period would result in re‐enrollment offers from an additional 420,000 acres. Using simulations based on offers from prior CRP sign‐ups, we estimate that these additional offers in the CRP auction would reduce program costs. Depending on the year of simulation, the outreach effort achieves a benefit‐cost ratio of between 20:1 and 90:1.

BackgroundIn Traumatic Brain Injury (TBI), large between-centre differences in outcome exist and many clinicians believe that such differences influence estimation of the treatment effect in randomized controlled trial (RCTs). The aim of this study was to assess the influence of between-centre differences in outcome on the estimated treatment effect in a large RCT in TBI.MethodsWe used data from the MRC CRASH trial on the efficacy of corticosteroid infusion in patients with TBI. We analyzed the effect of the treatment on 14 day mortality with fixed effect logistic regression. Next we used random effects logistic regression with a random intercept to estimate the treatment effect taking into account between-centre differences in outcome. Between-centre differences in outcome were expressed with a 95% range of odds ratios (OR) for centres compared to the average, based on the variance of the random effects (tau2). A random effects logistic regression model with random slopes was used to allow the treatment effect to vary by centre. The variation in treatment effect between the centres was expressed in a 95% range of the estimated treatment ORs.ResultsIn 9978 patients from 237 centres, 14-day mortality was 19.5%. Mortality was higher in the treatment group (OR = 1.22, p = 0.00010). Using a random effects model showed large between-centre differences in outcome (95% range of centre effects: 0.27- 3.71), but did not substantially change the estimated treatment effect (OR = 1.24, p = 0.00003). There was limited, although statistically significant, between-centre variation in the treatment effect (OR = 1.22, 95% treatment OR range: 1.17-1.26).ConclusionLarge between-centre differences in outcome do not necessarily affect the estimated treatment effect in RCTs, in contrast to current beliefs in the clinical area of TBI.

Over the past decade, permanent absentee voting, or “Vote-by-Mail” (VBM), has become increasingly popular. More than half the U.S. states offer their citizens the option to vote by mail. Yet until now, we have known relatively little about which voters are most likely to use this option when it is available, and what (if anything) can be done to convert voters effectively to permanent VBM status. In this paper, we address both these issues. We conduct a field experiment with all registered voters in California’s San Joaquin County as our subjects, where a randomly chosen treatment group receives postcards – a low-cost conversion option – offering them the chance to easily obtain permanent VBM status. We supplement this experiment with a survey of a subset of these voters, allowing identification of what types of voters are most likely to convert to VBM. Though the low-cost postcard treatment did significantly increase conversion to permanent VBM status, this effect shows up at disproportionately high rates among groups that already vote at higher rates.

BackgroundThe standard methodology for estimating treatment effects in Alzheimer’s disease clinical trials has remained unchanged for decades. With an increasing focus on disease‐modifying treatments, there are opportunities for developing more powerful methods for detecting such types of treatment effects and quantifying them in more interpretable ways.MethodsThe Progression Model for Repeated Measures (PMRM) was developed as an alternative to the Mixed Model for Repeated Measures (MMRM). The assumptions and flexibility of the PMRM are like the MMRM except treatment effects are handled differently (Figure 1). The PMRM estimates treatment effects as time‐delays/slowing of progression relative to the control group while the MMRM estimates treatment effects as point differences/reduction in decline on the outcome scale at a given visit. Because the unit of the treatment effect is shared with the time variable, PMRMs can simultaneously handle multiple outcomes in a natural and efficient manner. The power and interpretability of the PMRM was compared to the MMRM and other methods in a range of scenarios based on simulated and historical clinical trial data.ResultsThe PMRM was found to produce accurate and interpretable estimates of treatment effects across a range of simulated scenarios. Using CDR‐SB as an outcome, the PMRM produced the overall highest power to detect treatment effects across all simulated scenarios compared to all studied approaches. When treatment effects were simulated as time delays or slowing of progression, the PMRM offered substantial power gains compared to the MMRM. Using CDR‐SB, ADAS‐cog and ADCS‐ADL as multivariate outcomes, the PMRM was found to outperform models that combined the information in the outcome using the global test statistic and iADRS.ConclusionThe PMRM offers advantages in terms of interpretation and meaningful quantification of treatment effects, also in situations where the average decline observed in the control group is small. PMRMs can seamlessly include multiple outcome measures to estimate a single treatment effect and it was found that PMRMs offered substantially increased power to detect slowing of disease progression compared to conventional methods using both single and multiple outcome measures.

The role of welfare regimes on socioeconomic inequalities in edentulism: a cross-national analysis of 40 countries

Aim To evaluate the association between welfare regimes with edentulism and to investigate if it modifies the magnitude of socioeconomic inequalities in edentulism. Methods We analysed data from 40 high-, middle- and low-income countries collected between 2007 and 2018, encompassing 117397 individuals 20 years or older. The main outcome was edentulism and welfare regimes were the main exposure and effect modifier. Individual level variables were sex, age and quintiles of the wealth score. Multilevel logistic regression was used to estimate the odds of being edentulous and cross-level interaction terms between individual wealth and country factors were tested. Findings Individuals at the lowest wealth quintile had a consistently higher prevalence of edentulism. The highest age-sex standardised prevalence of edentulism was found in Eastern European countries (9.9%) followed by Informal Security (9.4%), while the lowest was among the Insecurity regime (1.4%) followed by the Social Democratic regime (4.7%). Liberal countries presented the highest magnitude of absolute and relative inequalities, where the lowest quintile had OR = 16.7 (95%CI: 10.5-26.5) times more chances of being edentulous and 14.4 percentage points higher prevalence. Low-income countries in the Insecurity Regime in social and labour market presented the lowest level of inequalities. Among HIC and upper-middle income countries, Social Democratic regime had the lowest absolute inequalities (4.0 percentage points difference between highest and lowest quintiles) and the Informal Security regime had the lowest relative differences between the highest and lowest quintiles (OR = 3.21, 95%CI: 2.83-3.63). Interpretation Our findings suggest that policies in some welfare regimes might improve oral health while reducing inequalities. High and upper-middle-income countries presented a high level of edentulism and inequalities. Key messages • Larger socioeconomic inequalities in edentulism were found in high-income countries with liberal policies. • A possible benefit of welfare policies on prevalence and inequalities in edentulism may be only evident among industrialised countries.

Team Incentives, Social Cohesion, and Performance: A Natural Field Experiment

We conducted a field experiment in a Dutch retail chain of 122 stores to study the interaction between team incentives, team social cohesion, and team performance. Theory predicts that the effect of team incentives on team performance increases with the team’s social cohesion because social cohesion reduces free-riding behavior. In addition, team incentives may lead to more coworker support or to higher peer pressure and thereby, can affect the team’s social cohesion. We introduced short-term team incentives in a randomly selected subset of stores and measured for all stores, both before and after the intervention, the team’s sales performance and the team’s social cohesion as well as coworker support and peer pressure. The average treatment effect of the team incentive on sales is 1.5 percentage points, which does not differ significantly from zero. In line with theory, the estimated treatment effect increases with social cohesion as measured before the intervention. Social cohesion itself is not affected by the team incentives. This paper was accepted by Yan Chen, decision analysis.