A Gentle Introduction to the Comparison Between Null Hypothesis Testing and Bayesian Analysis: Reanalysis of Two Randomized Controlled Trials

Effects of a waiting list control design on alcohol consumption among online help-seekers: A randomised controlled trial

BACKGROUND

Indirect evidence suggests that using waiting list control designs in behavioural research may have unintended consequences. The aim of this study was to estimate the effects of a waiting list design on alcohol consumption among individuals who had looked online for help.

METHODS

A two-arm randomised controlled trial was employed. The intervention group was informed that they belonged to the intervention group and would receive immediate access to a digital alcohol intervention. The waiting list control group was informed that they belonged to the group that had to wait four weeks to be given access to the intervention and in the meantime, they would be given a summary of their drinking. However, both groups received immediate access to the same digital alcohol intervention; the experimental contrast was thus between being told to wait or not.

RESULTS

We randomised 3388 participants (intervention: 1692, waiting list: 1696). Data were available for 954 participants at 1-month follow-up. We found no strong evidence that alcohol consumption differed between groups, but the evidence pointed towards the intervention group reporting lowering weekly alcohol consumption compared to the waiting list control group (IRR = 0.95, 95 % CI = 0.83; 1.08, probability of effect = 78.8 %).

CONCLUSION

We found no strong evidence that being informed that access to an intervention would be delayed produced differential self-reported alcohol consumption compared to being informed that access would be immediate. We did find a difference in engagement with the intervention materials, indicating that the experimental manipulation was successful.

Mediated effects of a randomised control trial for a text messaging smoking cessation intervention for online help-seekers and primary care visitors

BACKGROUND AND AIMS

Digital smoking cessation interventions have been shown to be effective in helping individuals achieve prolonged smoking abstinence. Nonetheless, the mechanisms that drive such effects are unclear. The current study aimed to estimate a digital smoking cessation intervention's natural direct and indirect effects.

METHODS

This secondary analysis of mediated effects uses data from a randomised controlled trial which included participants who smoked at least one cigarette a week, had access to a mobile phone, and were 18 years or older. The comparator was existing smoking cessation support available to all members of the Swedish public. Primary outcomes were prolonged smoking abstinence and point prevalence of smoking abstinence, measured at 3- and 6-months post-randomisation. A counterfactual framework was used to estimate three hypothesised mediators of the intervention's effects: importance, knowledge of how to change (know-how), and confidence.

RESULTS

Between 18/09/20 and 16/06/22, 1012 participants were randomised. The intervention led to improved confidence and know-how, which both partially mediated the effects of the digital intervention on smoking abstinence at 3- and 6 months post-randomisation.

CONCLUSIONS

A digital smoking cessation intervention was found to partially affect smoking abstinence by improving individuals' confidence in their ability to quit smoking and developing knowledge on how to quit. Face-value single-item mediator measures, lack of blinding, and attrition limit the study. Future studies should address these limitations and assess additional mechanisms mediating intervention effects.

TRIAL REGISTRATION

ISRCTN13455271.

Still just a matter of taste? Sensorial appreciation of seafood is associated with more frequent and diverse consumption

Improving health and sustainability outcomes in WEIRD (Western, Educated, Industrial, Rich, Democratic) nations necessitates a reduction in red meat consumption. Seafood is often overlooked in achieving this goal. However, simply consuming more of familiar fish species places high stress on production of these species. For this reason, diversification of seafood consumption is also critical. Here the motives for seafood consumption (frequency and diversity) are investigated across two studies by adapting the 4Ns survey to the seafood category. This 16-item survey measures four factors underpinning meat consumption: namely that it is 'Natural', 'Necessary', 'Normal' and 'Nice'. Swedish consumers' hedonic and sensory expectations of two herring concepts (traditional pickled contra novel minced and presented as a burger) are also evaluated in relation to the 4Ns. Study 1 (N = 304) revealed that the seafood 4Ns scale had a similar underlying structure to that of meat and had good test-retest reliability. Study 2 (N = 514) showed that consumers expected to like the pickled herring (associated with being 'seasoned', 'salty', 'sweet', 'firm', 'juicy', 'chewy', and 'slimy') more than the minced herring (associated with being 'mushy', 'fishy', 'grainy', 'dry' and having 'small bones'), and that 'Nice' scores affected expectations of both herring concepts. Food neophobia correlated inversely with seafood consumption frequency, expected liking, the 'Nice' subscale, and food agency. Critically, in both studies, enjoyment of seafood (higher 'Nice' scores) predicted more frequent and diverse seafood consumption, whilst agreeing that seafood is 'Necessary' for health predicted only consumption frequency, not diversity. Communicating the positive sensory attributes of seafood and developing novel product concepts in ways that disconfirm sceptical consumers' negative sensory expectations may increase acceptance of both familiar and unfamiliar seafood concepts.

Randomised study of the effects of sense of entitlement and conflict of interest contrarianism on researcher decision-making to work with the alcohol industry

BACKGROUND

It is well established that the tobacco industry used research funding as a deliberate tactic to subvert science. There has been little wider attention to how researchers think about accepting industry funding. We developed, then tested, hypotheses about two psychological constructs, namely, entitlement and conflict of interest contrarianism (CoI-C) among alcohol researchers who had previously received industry funding.

METHODS

A mixed-methods pilot study involved construct and instrument development, followed by an online survey and nested 3-arm randomised trial. We randomly allocated alcohol industry funding recipients to one of three conditions. In two experimental conditions we asked participants questions to remind them (and thus increase the salience) of their sense of entitlement or CoI-C. We compared these groups with a control group who did not receive any reminder. The outcome was a composite measure of openness to working with the alcohol industry.

RESULTS

133 researchers were randomised of whom 79 completed the experiment. The posterior distribution over effect estimates revealed that there was a 94.8% probability that reminding researchers of their CoI-C led them to self-report being more receptive to industry funding, whereas the probability was 68.1% that reminding them of their sense of entitlement did so. Biomedical researchers reported being more open to working with industry than did psychosocial researchers.

CONCLUSION

Holding contrarian views on conflict of interest could make researchers more open to working with industry. This study shows how it is possible to study researcher decision-making using quantitative experimental methods.

Effects of a single session low-threshold digital intervention for procrastination behaviors among university students (Focus): Findings from a randomized controlled trial

Introduction

Procrastination behaviors are common among university students, and have been found to be associated with stress, symptoms of depression, anxiety, and poorer academic performance. There is a need for interventions that can reach students at scale, and therefore this study aimed to estimate the effects of a single session low-threshold digital intervention (Focus) for procrastination behaviors among university students in Sweden.

Methods and analysis

A two-arm, parallel groups (1:1), single blind randomized controlled trial was conducted between February 8 to April 26, 2023. The study used email to invite university students across Sweden to participate in the trial. Both the intervention and the control group were invited to assess their current procrastination behaviors using the Pure Procrastination Scale (PPS). The intervention group immediately received feedback and behavior change advice by means of an interactive website, while the control group was shown their total PPS score without any further feedback. Students were included in the study if they scored 20 points or more on the PPS. Our primary outcome was procrastination behavior measured at 2 months post-randomization. Analyses were conducted using multilevel regression models estimated with Bayesian inference.

Results

A total of 2209 participants (intervention: 1109, control: 1100) were randomized. The average age of participants was 26.4 years (SD = 7.8) and 65 % were women (n = 1442). The mean PPS score at baseline was 35.6 points (of a maximum of 60). Primary outcome data were available for 45 % (n = 498) of the intervention group and 55 % (n = 601) of the control group. The evidence suggested no marked difference between groups regarding any of the outcomes, although there was weak evidence of lower physical activity in the intervention group. Qualitative findings from open-ended responses uncovered a variety of views on procrastination and perceived problems that may follow. Those not feeling supported by Focus explained having troubles adopting the advice given and converting their intentions into action without more continuous support.

Conclusions

Access to a single session of feedback and behavior change advice by means of an interactive website did not produce differential self-reported procrastination among university students who took the opportunity to self-assess their behaviors. The findings are limited by assessment reactivity due to screening at baseline and attrition to follow-up.

The prevalence of neutropenia and association with infections in patients with systemic lupus erythematosus: a Swedish single-center study conducted over 14 years

Hematologic abnormalities are common manifestations of SLE, although neutropenia is observed less frequently and is not included in the classification criteria. Nonetheless, neutropenia is a risk factor for infections, especially those caused by bacteria or fungi. We aimed to evaluate the impact of neutropenia in SLE through a systematic investigation of all infections in a large cohort of well-characterized patients, focusing on neutropenia, lymphopenia, and hypocomplementemia. Longitudinal clinical and laboratory parameters obtained at visits to the Rheumatology Unit, Linköping University Hospital, and linked data on all forms of healthcare utilization for all the subjects included in our regional SLE register during 2008-2022 were assessed. Data regarding confirmed infections were retrieved from the medical records. Overall, 333 patients were included and monitored during 3,088 visits to a rheumatologist during the study period. In total, 918 infections were identified, and 94 occasions of neutropenia (ANC < 1.5 × 109/L) were detected in 40 subjects (12%). Thirty neutropenic episodes in 15 patients occurred in association with infections, of which 13 (43%) required in-hospital care, 4 (13%) needed intensive care, and 1 (3%) resulted in death. Bayesian analysis showed that patients with ≥ 1 occasion of neutropenia were more likely to experience one or more infections (OR = 2.05; probability of association [POA] = 96%). Both invasiveness (OR = 7.08; POA = 98%) and severity (OR = 2.85; POA = 96%) of the infections were significantly associated with the present neutropenia. Infections are common among Swedish SLE patients, 12% of whom show neutropenia over time. Importantly, neutropenia is linked to both the invasiveness and severity of infections. Awareness of the risks of severe infections in neutropenic patients is crucial to tailor therapies to prevent severe illness and death.

Transdiagnostic and tailored internet intervention to improve mental health among university students: Research protocol for a randomized controlled trial

BACKGROUND

Emerging adulthood is often associated with mental health problems. About one in three university students report symptoms of depression and anxiety that can negatively affect their developmental trajectory concerning work, intimate relationships, and health. This can interfere with academic performance, as mood and anxiety disorders are key predictors of dropout from higher education. A treatment gap exists, where a considerable proportion of students do not seek help for mood and anxiety symptoms. Offering internet interventions to students with mental health problems could reduce the treatment gap, increase mental health, and improve academic performance. A meta-analysis on internet interventions for university students showed small effects for depression and none for anxiety. Larger trials are recommended to further explore effects of guidance, transdiagnostic approaches, and individual treatment components.

METHODS

This study will offer 1200 university students in Sweden participation in a three-armed randomized controlled trial (RCT) evaluating a guided or unguided transdiagnostic internet intervention for mild to moderate depression and anxiety, where the waitlist control group accesses the intervention at 6-month follow-up. Students reporting suicidal ideation/behaviors will be excluded and referred to treatment within the existing healthcare system. An embedded study within the trial (SWAT) will assess at week 3 of 8 whether participants in the guided and unguided groups are at higher risk of failing to benefit from treatment. Those at risk will be randomized to an adaptive treatment strategy, or to continue the treatment as originally randomized. Primary outcomes are symptoms of depression and anxiety. Follow-ups will occur at post-treatment and at 6-, 12-, and 24-month post-randomization. Between-group outcome analyses will be reported, and qualitative interviews about treatment experiences are planned.

DISCUSSION

This study investigates the effects of a transdiagnostic internet intervention among university students in Sweden, with an adaptive treatment strategy employed during the course of treatment to minimize the risk of treatment failure. The study will contribute knowledge about longitudinal trajectories of mental health and well-being following treatment, taking into account possible gender differences in responsiveness to treatment. With time, effective internet interventions could make treatment for mental health issues more widely accessible to the student group.

Restrictive versus standard IV fluid therapy in adult ICU patients with septic shock-Bayesian analyses of the CLASSIC trial

BACKGROUND

The CLASSIC trial assessed the effects of restrictive versus standard intravenous (IV) fluid therapy in adult intensive care unit (ICU) patients with septic shock. This pre-planned study provides a probabilistic interpretation and evaluates heterogeneity in treatment effects (HTE).

METHODS

We analysed mortality, serious adverse events (SAEs), serious adverse reactions (SARs) and days alive without life-support within 90 days using Bayesian models with weakly informative priors. HTE on mortality was assessed according to five baseline variables: disease severity, vasopressor dose, lactate levels, creatinine values and IV fluid volumes given before randomisation.

RESULTS

The absolute difference in mortality was 0.2%-points (95% credible interval: -5.0 to 5.4; 47% posterior probability of benefit [risk difference <0.0%-points]) with restrictive IV fluid. The posterior probabilities of benefits with restrictive IV fluid were 72% for SAEs, 52% for SARs and 61% for days alive without life-support. The posterior probabilities of no clinically important differences (absolute risk difference ≤2%-points) between the groups were 56% for mortality, 49% for SAEs, 90% for SARs and 38% for days alive without life-support. There was 97% probability of HTE for previous IV fluid volumes analysed continuously, that is, potentially relatively lower mortality of restrictive IV fluids with higher previous IV fluids. No substantial evidence of HTE was found in the other analyses.

CONCLUSION

We could not rule out clinically important effects of restrictive IV fluid therapy on mortality, SAEs or days alive without life-support, but substantial effects on SARs were unlikely. IV fluids given before randomisation might interact with IV fluid strategy.

Effects of a text messaging smoking cessation intervention amongst online help-seekers and primary health care visitors: findings from a randomised controlled trial

BACKGROUND

Smoking continues to be a leading risk factor for several diseases globally. We hypothesised that an intervention delivered via text messages could help individuals who were looking to quit.

METHODS

A two-arm, parallel-groups, randomised controlled trial was employed. Both groups received treatment as usual, with the intervention group also receiving a 12-week text messaging intervention. Participants were adult, weekly or more frequent smokers, recruited online and in primary health care centres. Research personnel were blinded, while participants were not. Primary outcomes were prolonged abstinence and point prevalence of abstinence, 3 and 6 months post-randomisation. All randomised participants were included in analyses.

RESULTS

Between 18 September 2020 and 16 June 2022, we randomised 1012 participants (intervention: 505, control: 507). Outcome data was available for 67% (n = 682) of participants at 3 months and 64% (n = 643) at 6 months. At 3 months, the odds ratio (OR) of prolonged abstinence was 2.15 (95% compatibility interval [CoI] = 1.51; 3.06, probability of effect [POE] > 99.9%, p < 0.0001), and for point prevalence of abstinence, it was 1.70 (95% CoI = 1.18; 2.44, POE = 99.8%, p = 0.0034) in favour of the text messaging intervention. At 6 months, the OR of prolonged abstinence was 2.38 (95% CoI = 1.62; 3.57, POE > 99.9%, p =  < 0.0001), and for point prevalence, it was 1.49 (95% CoI = 1.03; 2.14, POE = 98.3%, p = 0.0349) in favour of the text messaging intervention. Analyses with imputed data were not markedly different.

CONCLUSIONS

Amongst general population help-seekers-who on average had smoked for 25 years-access to a 12-week text messaging intervention produced higher rates of self-reported smoking abstinence in comparison to treatment as usual only. The intervention could be part of the societal response to the burden which smoking causes; however, findings are limited by risk of bias due to attrition, self-reported outcomes, and lack of blinding.

TRIAL REGISTRATION

The trial was preregistered in the ISRCTN registry on 27/07/2020 (ISRCTN13455271).

Mediators of effects of a digital alcohol intervention for online help-seekers: Findings from an effectiveness trial

BACKGROUND

Digital alcohol interventions have been shown to exert effects in helping individuals reduce their drinking. However, little is known about the mechanisms which mediate such effects. The objective of this study was to estimate natural direct and indirect effects of a digital alcohol intervention.

METHODS

This secondary analysis of mediated effects used data from a randomised controlled trial which included individuals with unhealthy alcohol use with access to a mobile phone aged 18 years or older in Sweden. The comparator was basic alcohol and health information. The digital intervention was centrally designed around weekly monitoring of consumption followed by feedback and tools to support behaviour change. Mediated effects were estimated using measures from 1-, 2-, and 4-months post-randomisation. Primary outcomes were total weekly consumption (TWC) and frequency of heavy episodic drinking (HED). A counterfactual framework was used to estimate three hypothesised mediators: importance, knowledge of how to change (know-how), and confidence.

RESULTS

Between 25/04/2019 and 26/11/2020, 2129 participants were randomised. The intervention improved know-how and confidence, which in turn mediated the effects on TWC and HED at 2- and 4-months. Analyses with imputed data were not markedly different.

CONCLUSIONS

A digital alcohol intervention was found to exert effects in reducing consumption by means of improving individuals' knowledge of how to reduce their consumption and confidence in their ability to reduce. The use of face-valid single item measures is a study limitation notwithstanding observed findings, as is attrition and lack of blinding of participants.

Patient participation in end-stage kidney disease care: variation over time and effects of staff-directed interventions - a quasi-experimental study

BACKGROUND

Among those elements establishing decent quality of care from a patient perspective, opportunities to participate in accord with one's individual needs and preferences are central. To date, little is known the extent of preference-based patient participation in kidney care, and what facilitates optimal conditions. This study investigated i) preference-based patient participation in kidney care over time, and ii) the effects of interventions designed to enhance person-centred patient participation.

METHODS

A quasi-experimental study was conducted across nine kidney care sites in southeast Sweden. A cohort of 358 patients with stage IV chronic kidney disease (eGRF 15-19 ml/min) or V (eGRF < 15 mL/min) entered the study. Of these, 245 patients (with kidney replacement therapy or intermittent outpatient visits only) completed a survey on patient participation at four time points: every six months from August 2019 to May 2021, patients reported their preferences for and experiences of participation using the validated Patient Preferences for Patient Participation tool, the 4Ps. Between the first and second data collection points, interventions were provided for designated staff to facilitate person-centred participation, using two strategies for two subgroups at three sites each: the managers receiving a bundle of information via e-mail on patient participation in a standard dissemination procedure (three sites), or an additional half-year support program for implementation offered to 1-2 staff per site (three sites), with no intervention for a control group (three sites). The differences in 4Ps data between groups were analysed using multilevel ordinal regression.

RESULTS

Over time and across all sites, most patients' experiences of participation fully or almost fully matched their engagement preferences (57%-90%). Still, up to 12% of patient reports indicated that their preferences and experiences were insufficiently matched: in these cases, the patients had preferred to be more involved than they had experienced, for example, in making healthcare plans and setting health-related goals. The interventions did not affect the levels of preference-based participation, but patients in the control group sites had slightly more consistent matches.

CONCLUSIONS

Living with kidney failure necessitates patient engagement, but opportunities to participate in accordance with one's preferences are not fully provided for all patients. Additional efforts to support a common understanding and to ensure person-centred patient participation is still needed.

Spontaneous Resolution of Uncomplicated Appendicitis may Explain Increase in Proportion of Complicated Appendicitis During Covid-19 Pandemic: a Systematic Review and Meta-analysis

BACKGROUND

Reports of an increased proportion of complicated appendicitis during the Covid-19 pandemic suggest a worse outcome due to delay secondary to the restrained access to health care, but may be explained by a concomitant decrease in uncomplicated appendicitis. We analyze the impact of the pandemic on the incidences of complicated and uncomplicated appendicitis.

METHOD

We did a systematic literature search in the PubMed, Embase and Web Of Science databases on December 21, 2022 with the search terms (appendicitis OR appendectomy) AND ("COVID" OR SARS-Cov2 OR "coronavirus"). Studies reporting the number of complicated and uncomplicated appendicitis during identical calendar periods in 2020 and the pre-pandemic year(s) were included. Reports with indications suggesting a change in how the patients were diagnosed and managed between the two periods were excluded. No protocol was prepared in advance. We did random effects meta-analysis of the change in proportion of complicated appendicitis, expressed as the risk ratio (RR), and of the change in number of patients with complicated and uncomplicated appendicitis during the pandemic compared with pre-pandemic periods, expressed as the incidence ratio (IR). We did separate analyses for studies based on single- and multi-center and regional data, age-categories and prehospital delay.

RESULTS

The meta-analysis of 100,059 patients in 63 reports from 25 countries shows an increase in the proportion of complicated appendicitis during the pandemic period (RR 1.39, 95% confidence interval (95% CI 1.25, 1.53). This was mainly explained by a decreased incidence of uncomplicated appendicitis (incidence ratio (IR) 0.66, 95% CI 0.59, 0.73). No increase in complicated appendicitis was seen in multi-center and regional reports combined (IR 0.98, 95% CI 0.90, 1.07).

CONCLUSION

The increased proportion of complicated appendicitis during Covid-19 is explained by a decrease in the incidence of uncomplicated appendicitis, whereas the incidence of complicated appendicitis remained stable. This result is more evident in the multi-center and regional based reports. This suggests an increase in spontaneously resolving appendicitis due to the restrained access to health care. This has important principal implications for the management of patients with suspected appendicitis.

Evaluating the effectiveness of a brief digital procrastination intervention targeting university students in Sweden: study protocol for the Focus randomised controlled trial

INTRODUCTION

The concept of procrastination can be described as a conscious, yet irrational, postponement of important tasks or decisions-despite awareness that the delay may lead to negative consequences. Procrastination behaviours are common among university students and is often described as a failure of self-regulation, and the behaviour is associated with stress, symptoms of depression and anxiety, poorer academic performance and negative effects on overall health and well-being.

METHODS AND ANALYSIS

A two-arm, parallel groups (1:1), single-blind randomised controlled trial will be conducted to assess the effectiveness of a brief digital procrastination intervention (Focus) among university students in Sweden. The intervention consists of a screening and feedback component based on Pure Procrastination Scale (PPS) score, allowing intervention participants to assess their current procrastination behaviours and receive behaviour change advice. Participants in the control group will be shown their total PPS score without any further feedback. Monte Carlo simulations (assuming a standardised effect of 0.35 Cohen's d of the intervention on the primary outcome, to at least 80% of the time estimate a posterior probability of effect of at least 95%) indicated that data from 1000 participants are required for analysis, meaning that 2000 participants are required to be randomised when assuming a 50% attrition rate. The primary outcome will be procrastination behaviour measured at 2 months postrandomisation. Secondary outcomes will be anxiety and stress symptoms and lifestyle behaviours. Outcomes will be analysed using multilevel regression models estimated using Bayesian inference.

ETHICS AND DISSEMINATION

The study was approved by the Swedish Ethical Review Authority on 2022-08-24 (dnr 2022-00353). Students will be asked to give informed consent before participation and after having been given information about the study. The results will be submitted for publication in open access, peer-reviewed journals.

TRIAL REGISTRATION NUMBER

ISRCTN13533793.

Haloperidol vs. placebo for the treatment of delirium in ICU patients: a pre-planned, secondary Bayesian analysis of the AID–ICU trial

PURPOSE

The AID-ICU trial was a randomised, blinded, placebo-controlled trial investigating effects of haloperidol versus placebo in acutely admitted, adult patients admitted in intensive care unit (ICU) with delirium. This pre-planned Bayesian analysis facilitates probabilistic interpretation of the AID-ICU trial results.

METHODS

We used adjusted Bayesian linear and logistic regression models with weakly informative priors to analyse all primary and secondary outcomes reported up to day 90, and with sensitivity analyses using other priors. The probabilities for any benefit/harm, clinically important benefit/harm, and no clinically important differences with haloperidol treatment according to pre-defined thresholds are presented for all outcomes.

RESULTS

The mean difference for days alive and out of hospital to day 90 (primary outcome) was 2.9 days (95% credible interval (CrI) - 1.1 to 6.9) with probabilities of 92% for any benefit and 82% for clinically important benefit. The risk difference for mortality was - 6.8 percentage points (95% CrI - 12.8 to - 0.8) with probabilities of 99% for any benefit and 94% for clinically important benefit. The adjusted risk difference for serious adverse reactions was 0.3 percentage points (95% CrI - 1.3 to 1.9) with 98% probability of no clinically important difference. Results were consistent across sensitivity analyses using different priors, with more than 83% probability of benefit and less than 17% probability of harm with haloperidol treatment.

CONCLUSIONS

We found high probabilities of benefits and low probabilities of harm with haloperidol treatment compared with placebo in acutely admitted, adult ICU patients with delirium for the primary and most secondary outcomes.

Effectiveness of a Smartphone App (MINISTOP 2.0) integrated in primary child health care to promote healthy diet and physical activity behaviors and prevent obesity in preschool-aged children: randomized controlled trial

BACKGROUND

Childhood overweight and obesity is a public health priority. We have previously reported the efficacy of a parent-oriented mobile health (mHealth) app-based intervention (MINISTOP 1.0) which showed improvements in healthy lifestyle behaviors. However, the effectiveness of the MINISTOP app in real-world conditions needs to be established.

OBJECTIVE

To evaluate the real-world effectiveness of a 6-month mHealth intervention (MINISTOP 2.0 app) on children's intake of fruits, vegetables, sweet and savory treats, sweet drinks, moderate-to-vigorous physical activity, and screen time (primary outcomes), and on parental self-efficacy (PSE) for promoting healthy lifestyle behaviors, and children's body mass index (BMI) (secondary outcomes).

METHODS

A hybrid type 1 effectiveness-implementation design was utilized. For the effectiveness outcomes, a two-arm, individually randomized controlled trial was conducted. Parents (n = 552) of 2.5-to-3-year-old children were recruited from 19 child health care centers across Sweden, and, randomized to either a control (standard care) or intervention group (MINISTOP 2.0 app). The 2.0 version was adapted and translated into English, Somali and Arabic to increase reach. All recruitment and data collection were conducted by the nurses. Outcomes were assessed at baseline and after six months, using standardized measures (BMI) and a questionnaire (health behaviors, PSE).

RESULTS

Among the participating parents (n = 552, age: 34.1 ± 5.0 years), 79% were mothers and 62% had a university degree. Twenty-four percent (n = 132) of children had two foreign-born parents. At follow-up, parents in the intervention group reported lower intakes of sweet and savory treats (-6.97 g/day; p = 0.001), sweet drinks (-31.52 g/day; p < 0.001), and screen time (-7.00 min/day; p = 0.012) in their children compared to the control group. The intervention group reported higher total PSE (0.91; p = 0.006), PSE for promoting healthy diet (0.34; p = 0.008) and PSE for promoting physical activity behaviors (0.31; p = 0.009) compared to controls. No statistically significant effect was observed for children's BMI z-score. Overall, parents reported high satisfaction with the app, and 54% reported using the app at least once a week.

CONCLUSION

Children in the intervention group had lower intakes of sweet and savory treats, sweet drinks, less screen time (primary outcomes) and their parents reported higher PSE for promoting healthy lifestyle behaviors. Our results from this real-world effectiveness trial support the implementation of the MINISTOP 2.0 app within Swedish child health care.

TRIAL REGISTRATION

Clinicaltrials.gov NCT04147039; https://clinicaltrials.gov/ct2/show/NCT04147039.

Reactions to being allocated to a waiting list control group in a digital alcohol intervention trial

OBJECTIVE

To study reactions of control group participants allocated to two different presentations of basic health information in a digital alcohol intervention trial.

METHOD

Control participants were randomised to wait with one of two different presentations of basic health information. Multiple choice questions and free-text comments assessed reactions, four months post randomisation. Effects of differential health information on responses were estimated, as were associations between responses, baseline characteristics and change in alcohol consumption.

RESULT

Of 1066 control group participants, 572 (54%) responded to the questionnaire. Contrasting two different presentations of basic health information revealed no statistically significant differences. Responses revealed that 38% were interested sufficiently to look at the information while 42% felt frustration, irritation, or disappointment about having to wait. Approximately 55% responded that they decided to reduce their drinking whilst 17% stated that they continued to drink as usual, and 11% gave up on the idea of reducing their drinking. The two latter groups reported markedly higher alcohol consumption at follow-up in comparison to the former (probability of association >99.9%).

CONCLUSION

Being made to wait may invite negative research participation effects.

PRACTICE IMPLICATION

Comparator guidance should be updated to reflect the potentially negative consequences which are under researched.

The effect of question order on outcomes in the orbital core outcome set for alcohol brief interventions among online help-seekers (QOBCOS): Findings from a randomised factorial trial

Objective

A core outcome set (COS) has been developed in alcohol brief intervention (ABI) research through international consensus. This study aimed to estimate order effects among questions in the COS.

Methods

Individuals aged 18 or older who searched online for alcohol-related help were invited to complete the COS. The order of questions was randomised following a factorial design. Primary outcomes were order effects among the COS items and patterns of attrition.

Results

Between 21/10/2020 and 26/11/2020, we randomised 7334 participants, of which 5256 responded to at least one question and were available for analyses. Current non-drinkers were excluded. We found evidence of higher self-reported average consumption and odds of harmful and hazardous drinking was found among those who first answered questions on recent consumption and impact of alcohol use. Lower self-reported recent consumption was found among those first asked about average consumption. Quality of life (QoL) was reported lower among those who first responded to when questions on impact of alcohol use were asked first, which in turn was lower among those who first answered question on when average consumption and QoL were asked first. Attrition was lowest when average consumption was asked first, and highest when QoL or impact of alcohol use was asked first. Median completion time for the COS was 4.3 min.

Conclusions

Question order affects outcomes and attrition. If the aim is to minimize attrition, consumption measures should be asked before QoL and impact of alcohol use; however, this order impacts self-reported alcohol consumption and so researchers should be guided by study priorities. At a minimum, all participants should be asked the same questions in the same order.

Trial registration

The trial was prospectively registered (ISRCTN17954645).

Avoiding Under- and Overrecruitment in Behavioral Intervention Trials Using Bayesian Sequential Designs: Tutorial

Reducing research waste and protecting research participants from unnecessary harm should be top priorities for researchers studying interventions. However, the traditional use of fixed sample sizes exposes trials to risks of under- and overrecruitment by requiring that effect sizes be determined a priori. One mitigating approach is to adopt a Bayesian sequential design, which enables evaluation of the available evidence continuously over the trial period to decide when to stop recruitment. Target criteria are defined, which encode researchers' intentions for what is considered findings of interest, and the trial is stopped once the scientific question is sufficiently addressed. In this tutorial, we revisit a trial of a digital alcohol intervention that used a fixed sample size of 2129 participants. We show that had a Bayesian sequential design been used, the trial could have ended after collecting data from approximately 300 participants. This would have meant exposing far fewer individuals to trial procedures, including being allocated to the waiting list control condition, and the evidence from the trial could have been made public sooner.

Frequentist and Bayesian Hypothesis Testing: An Intuitive Guide for Urologists and Clinicians

Given the limitations of frequentist method for null hypothesis significance testing, different authors recommend alternatives such as Bayesian inference. A poor understanding of both statistical frameworks is common among clinicians. The present is a gentle narrative review of the frequentist and Bayesian methods intended for physicians not familiar with mathematics. The frequentist p-value is the probability of finding a value equal to or higher than that observed in a study, assuming that the null hypothesis (H0) is true. The H0 is rejected or not based on a p threshold of 0.05, and this dichotomous approach does not express the probability that the alternative hypothesis (H1) is true. The Bayesian method calculates the probability of H1 and H0 considering prior odds and the Bayes factor (Bf). Prior odds are the researcher's belief about the probability of H1, and the Bf quantifies how consistent the data is concerning H1 and H0. The Bayesian prediction is not dichotomous but is expressed in continuous scales of the Bf and of the posterior odds. The JASP software enables the performance of both frequentist and Bayesian analyses in a friendly and intuitive way, and its application is displayed at the end of the paper. In conclusion, the frequentist method expresses how consistent the data is with H0 in terms of p-values, with no consideration of the probability of H1. The Bayesian model is a more comprehensive prediction because it quantifies in continuous scales the evidence for H1 versus H0 in terms of the Bf and the posterior odds.

Is short-course radiotherapy and total neoadjuvant therapy the new standard of care in locally advanced rectal cancer? A sensitivity analysis of the RAPIDO clinical trial

BACKGROUND

The results of the RAPIDO trial have been accepted as evidence in favour of short-course radiotherapy (SC-RT) followed by chemotherapy before total mesorectal excision in high-risk locally advanced rectal cancer. A noteworthy concern is that the RAPIDO trial did not ensure that all patients in the control arm received adjuvant chemotherapy. This may bias statistical estimates in favour of the experimental arm if adjuvant chemotherapy is active in rectal cancer. Moreover, the 5-year update revealed an increase in the risk of local relapse in the experimental arm.

MATERIALS AND METHODS

We carried out sensitivity analyses to determine how plausible effects of adjuvant chemotherapy, adjusted by the proportion of patients in the standard arm receiving adjuvant treatment, would have influenced the observed treatment effect estimate of the RAPIDO trial. The most plausible values for the benefit of adjuvant chemotherapy were determined by Bayesian re-analysis of a prior meta-analysis.

RESULTS

The meta-analysis suggested that oxaliplatin/fluorouracil-based adjuvant chemotherapy may improve disease-free survival (DFS) in rectal cancer although the signal is weak [hazard ratio (HR) 0.84, 95% credible interval, 0.57-1.15]; probability of benefit (HR <1) was 91.2%. In the sensitivity analysis, the HR for disease-related treatment failure would remain <1, thus favouring total neoadjuvant therapy (TNT), on most occasions, but the null hypothesis would not have been rejected in various credible settings. For the RAPIDO data to be consistent with the null effect, a moderate benefit of adjuvant chemotherapy (HR for DFS between 0.75 and 0.80) and 70%-80% of exposed participants would suffice.

CONCLUSION

The decision to make adjuvant chemotherapy optional in the standard arm may have biased the results in favour of the experimental arm, in a scenario in which TNT does not offset the increase in local recurrences after SC-RT.

Agents intervening against delirium in the intensive care unit trial-Protocol for a secondary Bayesian analysis

BACKGROUND

Delirium is highly prevalent in the intensive care unit (ICU) and is associated with high morbidity and mortality. The antipsychotic haloperidol is the most frequently used agent to treat delirium although this is not supported by solid evidence. The agents intervening against delirium in the intensive care unit (AID-ICU) trial investigates the effects of haloperidol versus placebo for the treatment of delirium in adult ICU patients.

METHODS

This protocol describes the secondary, pre-planned Bayesian analyses of the primary and secondary outcomes up to day 90 of the AID-ICU trial. We will use Bayesian linear regression models for all count outcomes and Bayesian logistic regression models for all dichotomous outcomes. We will adjust for stratification variables (site and delirium subtype) and use weakly informative priors supplemented with sensitivity analyses using sceptical priors. We will present results as absolute differences (mean differences and risk differences) and relative differences (ratios of means and relative risks). Posteriors will be summarised using median values as point estimates and percentile-based 95% credibility intervals. Probabilities of any benefit/harm, clinically important benefit/harm and clinically unimportant differences will be presented for all outcomes.

DISCUSSION

The results of this secondary, pre-planned Bayesian analysis will complement the primary frequentist analysis of the AID-ICU trial and facilitate a nuanced and probabilistic interpretation of the trial results.

Digital multiple health behaviour change intervention targeting online help seekers: protocol for the COACH randomised factorial trial

INTRODUCTION

Unhealthy lifestyle behaviours continue to be highly prevalent, including alcohol consumption, unhealthy diets, insufficient physical activity and smoking. There is a lack of effective interventions which have a large enough reach into the community to improve public health. Additionally, the common co-occurrence of multiple unhealthy behaviours demands investigation of efforts which address more than single behaviours.

METHODS AND ANALYSIS

The effects of six components of a novel digital multiple health behaviour change intervention on alcohol consumption, diet, physical activity and smoking (coprimary outcomes) will be estimated in a factorial randomised trial. The components are designed to facilitate behaviour change, for example, through goal setting or increasing motivation, and are either present or absent depending on allocation (ie, six factors with two levels each). The study population will be those seeking help online, recruited through search engines, social media and lifestyle-related websites. Included will be those who are at least 18 years of age and have at least one unhealthy behaviour. An adaptive design will be used to periodically make decisions to continue or stop recruitment, with simulations suggesting a final sample size between 1500 and 2500 participants. Multilevel regression models will be used to analyse behavioural outcomes collected at 2 months and 4 months postrandomisation.

ETHICS AND DISSEMINATION

Approved by the Swedish Ethical Review Authority on 2021-08-11 (Dnr 2021-02855). Since participation is likely motivated by gaining access to novel support, the main concern is demotivation and opportunity cost if the intervention is found to only exert small effects. Recruitment began on 19 October 2021, with an anticipated recruitment period of 12 months.

TRIAL REGISTRATION NUMBER

ISRCTN16420548.

Effectiveness of a digital intervention versus alcohol information for online help-seekers in Sweden: a randomised controlled trial

BACKGROUND

The ubiquity of Internet connectivity, and widespread unmet needs, requires investigations of digital interventions for people seeking help with their drinking. The objective of this study was to test the effectiveness of a digital alcohol intervention compared to existing online resources for help seekers.

METHODS

This parallel randomised controlled trial included 2129 risky drinkers with access to a mobile phone and aged 18 years or older. Randomised sub-studies investigated consent procedures and control group design. Simple computerised randomisation was used. Participants were aware of allocation after randomisation; research personnel were not. The digital intervention was designed around weekly monitoring of alcohol consumption followed by feedback and tools for behaviour change. Primary outcomes were total weekly consumption (TWC) and frequency of heavy episodic drinking (HED), measured 2 and 4 months post-randomisation.

RESULTS

Between 25/04/2019 and 26/11/2020, 2129 participants were randomised (intervention: 1063, control: 1066). Negative binomial regression was used to contrast groups, with both Bayesian and maximum likelihood inference. The posterior median incidence rate ratio (IRR) of TWC was 0.89 (95% CI = 0.81;0.99, 98.2% probability of effect, P-value = 0.033) at 2 months among 1557 participants and 0.77 (95% CI = 0.69;0.86, > 99.9% probability of effect, P-value < 0.001) at 4 months among 1429 participants. For HED, the IRR was 0.83 (95% CI = 0.75;0.93, > 99.9% probability of effect, P-value = 0.0009) at 2 months among 1548 participants and 0.71 (95% CI = 0.63;0.79, probability of effect > 99.9%, P-value < 0.0001) at 4 months among 1424 participants. Analyses with imputed data were not markedly different.

CONCLUSIONS

A digital alcohol intervention produced self-reported behaviour change among online help seekers in the general population. The internal and external validity of this trial is strong, subject to carefully considered study limitations arguably inherent to trials of this nature. Limitations include higher than anticipated attrition to follow-up and lack of blinding.

TRIAL REGISTRATION

The trial was prospectively registered ( ISRCTN48317451 ).

Effectiveness of a health technician-delivered brief intervention for alcohol: a Bayesian reanalysis of a clinical trial

Objective

To reanalyze a clinical trial on the effectiveness of a Brief Intervention (BI) delivered by non-professionals to reduce risky alcohol drinking. Our previous null-hypothesis test of the effects of the BI yielded a ‘non-significant’ p-value, yet remained uninformative. Here we use the Bayesian paradigm which allows for expressing the probability of different effect sizes to better inform public policy decisions.

Results

The posterior probability of the odds of risky drinking at follow-up favored a marked effect of the BI, with 96% of the probability mass being less than OR = 1, and 84% being less than OR = 0.8. Our findings show that there is a high probability that the BI delivered by health technicians lowered risky alcohol use. The posterior distributions of the BI’s effects are presented to help contextualize the evidence for policy making in Chile.

Symptoms of COVID-19 contagion in different social contexts in association to self-reported symptoms, mental health and study capacity in Swedish university students

Objective

The present study investigates if symptoms of COVID-19 contagion in different social contexts (cohabitants, family, acquaintances, and others) are associated with university students' own self-reported symptoms of COVID-19 contagion, mental health, and study capacity. This was investigated by a cross-sectional survey administrated in Sweden during the first wave of the COVID-19 pandemic, at the time when universities were locked down to limit viral spread and contagion.

Results

Mild to moderate symptoms of COVID-19 in cohabitants and family members were associated with student’s self-reported symptoms of contagion, while no associations could be seen in relation to mental health and study capacity. Symptoms of COVID-19 contagion in acquaintances and others were not associated with students’ self-reported symptoms, nor with their mental health and study capacity.

To conclude, during the initial lockdown of universities students’ self-reported symptoms of contagion were mainly associated with cohabitants and family members, while symptoms of contagion in different social contexts were not associated with mental health and study capacity. Findings suggest that lockdown of universities may have contributed to limiting infection pathways, while still allowing students to focus on their studies despite significant contagion among others known to the student.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13104-022-06009-z.

Medicine before and after David Cox

Herein we recount the legacy of Sir David Roxbee Cox (15 July 1924 - 18 January 2022) from the perspective of practicing clinicians. His-pioneering work in developing the logistic and Cox proportional hazard regression models revolutionized the analysis and interpretation of categorical and time-to-event survival outcomes in modern medicine. This legacy is an inspiration for all those who follow on Sir David Cox's path.

Dexamethasone 12 mg versus 6 mg for patients with COVID-19 and severe hypoxaemia: a pre-planned, secondary Bayesian analysis of the COVID STEROID 2 trial

PURPOSE

We compared dexamethasone 12 versus 6 mg daily for up to 10 days in patients with coronavirus disease 2019 (COVID-19) and severe hypoxaemia in the international, randomised, blinded COVID STEROID 2 trial. In the primary, conventional analyses, the predefined statistical significance thresholds were not reached. We conducted a pre-planned Bayesian analysis to facilitate probabilistic interpretation.

METHODS

We analysed outcome data within 90 days in the intention-to-treat population (data available in 967 to 982 patients) using Bayesian models with various sensitivity analyses. Results are presented as median posterior probabilities with 95% credible intervals (CrIs) and probabilities of different effect sizes with 12 mg dexamethasone.

RESULTS

The adjusted mean difference on days alive without life support at day 28 (primary outcome) was 1.3 days (95% CrI -0.3 to 2.9; 94.2% probability of benefit). Adjusted relative risks and probabilities of benefit on serious adverse reactions was 0.85 (0.63 to 1.16; 84.1%) and on mortality 0.87 (0.73 to 1.03; 94.8%) at day 28 and 0.88 (0.75 to 1.02; 95.1%) at day 90. Probabilities of benefit on days alive without life support and days alive out of hospital at day 90 were 85 and 95.7%, respectively. Results were largely consistent across sensitivity analyses, with relatively low probabilities of clinically important harm with 12 mg on all outcomes in all analyses.

CONCLUSION

We found high probabilities of benefit and low probabilities of clinically important harm with dexamethasone 12 mg versus 6 mg daily in patients with COVID-19 and severe hypoxaemia on all outcomes up to 90 days.

Randomised clinical trials in critical care: past, present and future

Randomised clinical trials (RCTs) are the gold standard for providing unbiased evidence of intervention effects. Here, we provide an overview of the history of RCTs and discuss the major challenges and limitations of current critical care RCTs, including overly optimistic effect sizes; unnuanced conclusions based on dichotomization of results; limited focus on patient-centred outcomes other than mortality; lack of flexibility and ability to adapt, increasing the risk of inconclusive results and limiting knowledge gains before trial completion; and inefficiency due to lack of re-use of trial infrastructure. We discuss recent developments in critical care RCTs and novel methods that may provide solutions to some of these challenges, including a research programme approach (consecutive, complementary studies of multiple types rather than individual, independent studies), and novel design and analysis methods. These include standardization of trial protocols; alternative outcome choices and use of core outcome sets; increased acceptance of uncertainty, probabilistic interpretations and use of Bayesian statistics; novel approaches to assessing heterogeneity of treatment effects; adaptation and platform trials; and increased integration between clinical trials and clinical practice. We outline the advantages and discuss the potential methodological and practical disadvantages with these approaches. With this review, we aim to inform clinicians and researchers about conventional and novel RCTs, including the rationale for choosing one or the other methodological approach based on a thorough discussion of pros and cons. Importantly, the most central feature remains the randomisation, which provides unparalleled restriction of confounding compared to non-randomised designs by reducing confounding to chance.

Multiple lifestyle behaviour mHealth intervention targeting Swedish college and university students: protocol for the Buddy randomised factorial trial

Introduction

The time during which many attend college or university is an important period for developing health behaviours, with potentially major implications for future health. Therefore, it is concerning that many Swedish students excessively consume alcohol, have unhealthy diets, are not physical active and smoke. The potential of digital interventions which integrate support for change of all of these behaviours is largely unexplored, as are the dismantled effects of the individual components that make up digital lifestyle behaviour interventions.

Methods and analysis

A factorial randomised trial (six factors with two levels each) will be employed to estimate the effects of the components of a novel mHealth multiple lifestyle intervention on alcohol consumption, diet, physical activity and smoking among Swedish college and university students. A Bayesian group sequential design will be employed to periodically make decisions to continue or stop recruitment, with simulations suggesting that between 1500 and 2500 participants will be required. Multilevel regression models will be used to analyse behavioural outcomes collected at 2 and 4 months postrandomisation.

Ethics and dissemination

The study was approved by the Swedish Ethical Review Authority on 2020-12-15 (Dnr 2020-05496). The main concern is the opportunity cost if the intervention is found to only have small effects. However, considering the lack of a generally available evidence-based multiple lifestyle behaviour support to university and college students, this risk was deemed acceptable given the potential benefits from the study.

Recruitment will begin in March 2021, and it is expected that recruitment will last no more than 24 months. A final data set will, therefore, be available in July 2023, and findings will be reported no later than December 2023.

Trial registration number

ISRCTN23310640; Pre-results.

Causal Considerations Can Inform the Interpretation of Surprising Associations in Medical Registries

An exploratory analysis of registry data from 2437 patients with advanced gastric cancer revealed a surprising association between astrological birth signs and overall survival (OS) with p = 0.01. After dichotomizing or changing the reference sign, p-values <0.05 were observed for several birth signs following adjustments for multiple comparisons. Bayesian models with moderately skeptical priors still pointed to these associations. A more plausible causal model, justified by contextual knowledge, revealed that these associations arose from the astrological sign association with seasonality. This case study illustrates how causal considerations can guide analyses through what would otherwise be a hopeless maze of statistical possibilities.

Six-Month Outcomes from the NEXit Junior Trial of a Text Messaging Smoking Cessation Intervention for High School Students: Randomized Controlled Trial With Bayesian Analysis

Background

The prevalence of daily or occasional smoking among high school students in Sweden was approximately 20% in 2019, which is problematic since lifestyle behaviors are established in adolescence and track into adulthood. The Nicotine Exit (NEXit) Junior trial was conducted in response to a lack of evidence for the effects of text message smoking cessation interventions among high school students in Sweden.

Objective

The aim of this study was to estimate the 3- and 6-month effects of a text messaging intervention among high school students in Sweden on smoking cessation outcomes.

Methods

A 2-arm, single-blind randomized controlled trial was employed to estimate the effects of the intervention on smoking cessation in comparison to treatment as usual. Participants were recruited from high schools in Sweden using advertising and promotion by school staff from January 10, 2018, to January 10, 2019. Weekly or daily smokers who were willing to make a quit attempt were eligible for inclusion. Prolonged abstinence and point prevalence of smoking cessation were measured at 3 and 6 months after randomization.

Results

Complete case analysis was possible on 57.9% (310/535) of the participants at 6 months, with no observed statistically significant effect on 5-month prolonged abstinence (odds ratio [OR] 1.27, 95% CI 0.73-2.20; P=.39) or 4-week smoking cessation (OR 1.42; 95% CI 0.83-2.46; P=.20). Sensitivity analyses using imputation yielded similar findings. Unplanned Bayesian analyses showed that the effects of the intervention were in the anticipated direction. The findings were limited by the risk of bias induced by high attrition (42.1%). The trial recruited high school students in a pragmatic setting and included both weekly and daily smokers; thus, generalization to the target population is more direct compared with findings obtained under more strict study procedures.

Conclusions

Higher than expected attrition rates to follow-up 6 months after randomization led to null hypothesis tests being underpowered; however, unplanned Bayesian analyses found that the effects of the intervention were in the anticipated direction. Future trials of smoking cessation interventions targeting high school students should aim to prepare strategies for increasing retention to mid- and long-term follow-up.

Trial Registration

IRCTN Registry ISRCTN15396225; https://www.isrctn.com/ISRCTN15396225

International Registered Report Identifier (IRRID)

RR2-10.1186/s13063-018-3028-2

Effects of a waiting list control design on alcohol consumption among online help-seekers: protocol for a randomised controlled trial

INTRODUCTION

Sparse attention has been given to the design of control conditions in trials, despite their important role as contrasts for novel treatments, and thus as a key determinant of effect sizes. This undermines valid inferences on effect estimates in trials, which are fundamentally comparative in nature. Such challenges to understanding also makes generalisation of effect estimates complex, for example, it may not be clear to what degree real-world alternatives to the novel treatments in pragmatic trials are similar to the control conditions studied. The present study aims to estimate the effects of being allocated to a waiting list control condition.

METHODS AND ANALYSIS

Individuals searching online for help to reduce their drinking will be invited to take part in a study. Individuals aged 18 years or older, who in the past month consumed six or more drinks on one occasion, or consumed 10 or more drinks the past week, will be eligible to participate. Both groups will receive identical feedback and advice on behaviour change; however, one group will be informed that they have to wait 1 month for the intervention materials. One month postrandomisation, participants will receive an email with the follow-up questionnaire measuring the primary outcomes: (1) frequency of heavy episodic drinking (defined as at study entry) in the past month; and (2) overall past week alcohol consumption. Differences between groups will be analysed using negative binomial regression models estimated using Bayesian inference. Recruitment will begin in October 2021. A Bayesian group sequential design will be employed to determine when to end enrolment (expected to be between 500 and 1500 individuals).

ETHICS AND DISSEMINATION

The study was approved by the Swedish Ethical Review Authority on 2021-01-25 (Dnr 2020-06267). Findings will be disseminated in open access peer-reviewed journals no later than 2023.

TRIAL REGISTRATION TRIAL

ISRCTN14959594; Pre-results.

Novel risk factors for glucarpidase use in pediatric acute lymphoblastic leukemia: Hispanic ethnicity, age, and the ABCC4 gene

BACKGROUND

Carboxypeptidase G₂ (CPDG₂ ; glucarpidase) is a rescue drug for patients at risk for kidney injury from high-dose methotrexate (MTX). As there are no strategies for predicting patients who will require CDPG₂ , we evaluated the role of demographic, clinical, and genetic factors for CPDG₂ use.

PROCEDURE

Cases who received CPDG₂ and controls who did not were identified by chart review of acute lymphoblastic leukemia (ALL) patients who received MTX doses between 1000 and 5000 mg/m² between 2010 and 2017. We used multivariable Bayesian logistic regression to evaluate the association of CPDG₂ use with demographic and clinical variables and, on a subset of patients, with genetic ancestry and 49 single nucleotide variants previously associated with MTX toxicity.

RESULTS

We identified 423 patients who received 1592 doses of MTX. Of the 18 patients who received CPDG₂ , 17 (94%) were Hispanic. No patients who received 1000 or 2000 mg/m² of MTX received CPDG₂ . Hispanic ethnicity (odds ratio: 4.68; 95% compatibility interval: 1.63-15.06) and older age (1.87 [1.17-3.17]) were associated with receiving CPDG₂ . Of the 177 patients in the genomic cohort, 11 received CPDG₂ . Each additional G allele of rs7317112 in ABCC4 increased the odds of requiring CPDG₂ (3.10 [1.12-6.75]). Six other loci (NTRK1/rs10908521, TSG1/rs9345389, STT3B/rs1353327, SCLO1B1/rs4149056, GATA3/rs3824662, ARID5B/rs10821936) demonstrated probabilities of association between 88% and 97%.

CONCLUSION

We demonstrated that demographic characteristics, including Hispanic ethnicity and age, are associated with CPDG₂ use. Additionally, we provide evidence that inherited genetic variation is associated with risk of requiring CPDG₂ . If validated in independent populations, this information could be leveraged to develop targeted toxicity prevention strategies for children with ALL.

mHealth intervention for multiple lifestyle behaviour change among high school students in Sweden (LIFE4YOUth): protocol for a randomised controlled trial

Background

National surveys in Sweden demonstrate that the majority of young people do not engage in health promoting behaviours at levels recommended by the Public Health Agency of Sweden. The objective of this study is to estimate the effectiveness of a novel mHealth intervention named LIFE4YOUth, which targets multiple lifestyle behaviours (alcohol, diet, physical activity, and smoking) among high school students in Sweden.

Methods

A 2-arm parallel groups single blind randomised controlled trial (1:1) will be employed to estimate the effectiveness of the novel mHealth intervention. Students will be recruited at high schools throughout Sweden, and will be included if they fulfil one of six criteria relating to unhealthy behaviours with respect to alcohol, diet, physical activity and smoking. Eligible participants will be randomised to either receive the novel intervention immediately, or to be placed on a waiting list for 4 months. The intervention consists of a combination of recurring screening, text messages, and an interactive platform which is adaptable to individual preferences. Outcome measures with respect to alcohol, diet, physical activity and smoking will be assessed through questionnaires at 2 and 4 months post randomisation.

Discussion

The findings of this trial could be generalised to a diverse high-school student population as our recruitment encompass a large proportion of schools throughout Sweden with various educational profiles. Furthermore, if effective, the mHealth intervention has good potential to be able to be scaled up and disseminated at high schools nationally.

Trial registration

Registered prospectively on 2020-05-20 in ISRCTN (ISRCTN34468623).

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-021-11446-9.

Compliance with recommendations limiting COVID-19 contagion among university students in Sweden: associations with self-reported symptoms, mental health and academic self-efficacy

AIMS

The COVID-19 containment strategy in Sweden uses public health recommendations relying on personal responsibility for compliance. Universities were one of few public institutions subject to strict closure, meaning that students had to adapt overnight to online teaching. This study investigates the prevalence of self-reported recommendation compliance and associations with self-reported symptoms of contagion, self-experienced effects on mental health and academic self-efficacy among university students in Sweden in May-June 2020.

METHODS

This was a cross-sectional 23 question online survey in which data were analysed by multinomial regression, taking a Bayesian analysis approach complemented by null hypothesis testing.

RESULTS

A total of 4495 students consented to respond. Recommendation compliance ranged between 70% and 96%. Women and older students reported higher compliance than did men and younger students. Mild to moderate COVID-19 symptoms were reported by 30%, severe symptoms by fewer than 2%; 15% reported being uncertain and half of the participants reported no symptoms. Mental health effects were reported by over 80%, and changes in academic self-efficacy were reported by over 85%; in both these areas negative effects predominated. Self-reported symptoms and uncertainty about contagion were associated with non-compliance, negative mental health effects, and impaired academic self-efficacy.

CONCLUSIONS

Students generally followed public health recommendations during strict closure of universities, but many reported considerable negative consequences related to mental health and academic self-efficacy. Digital interventions should be developed and evaluated to boost coping skills, build resilience and alleviate student suffering during the pandemic and future similar crises.

Higher vs Lower Doses of Dexamethasone in Patients with COVID-19 and Severe Hypoxia (COVID STEROID 2) trial: Protocol for a secondary Bayesian analysis

BACKGROUND

Coronavirus disease 2019 (COVID-19) can lead to severe hypoxic respiratory failure and death. Corticosteroids decrease mortality in severely or critically ill patients with COVID-19. However, the optimal dose remains unresolved. The ongoing randomised COVID STEROID 2 trial investigates the effects of higher vs lower doses of dexamethasone (12 vs 6 mg intravenously daily for up to 10 days) in 1,000 adult patients with COVID-19 and severe hypoxia.

METHODS

This protocol outlines the rationale and statistical methods for a secondary, pre-planned Bayesian analysis of the primary outcome (days alive without life support at day 28) and all secondary outcomes registered up to day 90. We will use hurdle-negative binomial models to estimate the mean number of days alive without life support in each group and present results as mean differences and incidence rate ratios with 95% credibility intervals (CrIs). Additional count outcomes will be analysed similarly and binary outcomes will be analysed using logistic regression models with results presented as probabilities, relative risks and risk differences with 95% CrIs. We will present probabilities of any benefit/harm, clinically important benefit/harm and probabilities of effects smaller than pre-defined clinically minimally important differences for all outcomes analysed. Analyses will be adjusted for stratification variables and conducted using weakly informative priors supplemented by sensitivity analyses using sceptic priors.

DISCUSSION

This secondary, pre-planned Bayesian analysis will supplement the primary, conventional analysis and may help clinicians, researchers and policymakers interpret the results of the COVID STEROID 2 trial while avoiding arbitrarily dichotomised interpretations of the results.

TRIAL REGISTRATION

ClinicalTrials.gov: NCT04509973; EudraCT: 2020-003363-25.

Causal models accounted for research participation effects when estimating effects in a behavioral intervention trial

OBJECTIVE

Participants in intervention studies are asked to take part in activities linked to the conduct of research, including signing consent forms and being assessed. If participants are affected by such activities through mechanisms by which the intervention is intended to work, then there is confounding. We examine how to account for research participation effects analytically.

STUDY DESIGN AND SETTING

Data from a trial of a brief alcohol intervention among Swedish university students is used to show how a proposed causal model can account for assessment effects.

RESULTS

The proposed model can account for research participation effects as long as researchers are willing to use existing data to make assumptions about causal influences, for instance on the magnitude of assessment effects. The model can incorporate several research processes which may introduce bias.

CONCLUSIONS

As our knowledge grows about research participation effects, we may move away from asking if participants are affected by study design, toward rather asking by how much they are affected, by which activities and in which circumstances. The analytic perspective adopted here avoids assuming there are no research participation effects.

Effectiveness of a Smartphone App to Promote Healthy Weight Gain, Diet, and Physical Activity During Pregnancy (HealthyMoms): Randomized Controlled Trial

Background

Excessive gestational weight gain (GWG) during pregnancy is a major public health concern associated with negative health outcomes for both mother and child. Scalable interventions are needed, and digital interventions have the potential to reach many women and promote healthy GWG. Most previous studies of digital interventions have been small pilot studies or have not included women from all BMI categories. We therefore examined the effectiveness of a smartphone app in a large sample (n=305) covering all BMI categories.

Objective

To investigate the effectiveness of a 6-month intervention (the HealthyMoms app) on GWG, body fatness, dietary habits, moderate-to-vigorous physical activity (MVPA), glycemia, and insulin resistance in comparison to standard maternity care.

Methods

A 2-arm parallel randomized controlled trial was conducted. Women in early pregnancy at maternity clinics in Östergötland, Sweden, were recruited. Eligible women who provided written informed consent completed baseline measures, before being randomized in a 1:1 ratio to either an intervention (n=152) or control group (n=153). The control group received standard maternity care while the intervention group received the HealthyMoms smartphone app for 6 months (which includes multiple features, eg, information; push notifications; self-monitoring; and feedback features for GWG, diet, and physical activity) in addition to standard care. Outcome measures were assessed at Linköping University Hospital at baseline (mean 13.9 [SD 0.7] gestational weeks) and follow-up (mean 36.4 [SD 0.4] gestational weeks). The primary outcome was GWG and secondary outcomes were body fatness (Bod Pod), dietary habits (Swedish Healthy Eating Index) using the web-based 3-day dietary record Riksmaten FLEX, MVPA using the ActiGraph wGT3x-BT accelerometer, glycemia, and insulin resistance.

Results

Overall, we found no statistically significant effect on GWG (P=.62); however, the data indicate that the effect of the intervention differed by pre-pregnancy BMI, as women with overweight and obesity before pregnancy gained less weight in the intervention group as compared with the control group in the imputed analyses (–1.33 kg; 95% CI –2.92 to 0.26; P=.10) and completers-only analyses (–1.67 kg; 95% CI –3.26 to –0.09; P=.031]). Bayesian analyses showed that there was a 99% probability of any intervention effect on GWG among women with overweight and obesity, and an 81% probability that this effect was over 1 kg. The intervention group had higher scores for the Swedish Healthy Eating Index at follow-up than the control group (0.27; 95% CI 0.05-0.50; P=.017). We observed no statistically significant differences in body fatness, MVPA, glycemia, and insulin resistance between the intervention and control group at follow up (P≥.21).

Conclusions

Although we found no overall effect on GWG, our results demonstrate the potential of a smartphone app (HealthyMoms) to promote healthy dietary behaviors as well as to decrease weight gain during pregnancy in women with overweight and obesity.

Trial Registration

ClinicalTrials.gov NCT03298555; https://clinicaltrials.gov/ct2/show/NCT03298555

International Registered Report Identifier (IRRID)

RR2-10.2196/13011

External validity of docetaxel triplet trials in advanced gastric cancer: are there patients who still benefit?

BACKGROUND

The purpose of our study was to develop an online calculator to estimate the effect of docetaxel triplets (DPF) in first line of advanced gastric cancer (AGC), and to assess the external validity of docetaxel trials in individual patients.

METHODS

The study includes patients with HER2(-) AGC treated with platin and fluoropyrimidine (PF) or with DPF in first line. Treatment effect and interactions were assessed using Bayesian accelerated failure time models.

RESULT

The series comprises 1376 patients; 238 treated with DPF and 1138 with PF between 2008 and 2019. DPF was associated with increased progression-free survival (PFS) and overall survival (OS) with time ratio (TR) 1.27 (95% credible interval [CrI], 1.15-1.40), and TR 1.19 (95% CrI, 1.09-1.27), respectively. Serious adverse events were more common with DPF, particularly hematological effects (32% vs 22%). Younger participants received greater DPF dose density without achieving greater disease control, while severe toxicity was likewise higher. DPF yielded superior OS in Lauren intestinal (TR 1.27, 95% CrI, 1.08-1.11) vs diffuse subtype (TR 1.17, 95% CrI, 1.09-1.24) and the probability of increasing OS > 15% was 90% vs 67% in each subtype, respectively. The effect dwindles over time, which can be attributed to pathological changes and clinical practice changes.

CONCLUSION

Our study confirms the effect of DPF is highly dependent on several clinical-pathological variables, with discreet and gradually declining benefit over platinum doublets in later years, at the expense of increased toxicity. These results may help to underpin the idea that external validity of AGC trials should be revised regularly.

Effects of a Text Messaging Smoking Cessation Intervention Among Online Help Seekers and Primary Health Care Visitors in Sweden: Protocol for a Randomized Controlled Trial Using a Bayesian Group Sequential Design

BACKGROUND

A steady decline of the smoking prevalence in Sweden has been recorded over the past decade; however, people still start and continue to smoke. There is a need for effective smoking cessation interventions that can scale to a national level and that are designed to reach individuals requiring smoking cessation support in the general population.

OBJECTIVE

Previous randomized controlled trials of smoking cessation interventions among high school and university students in Sweden have found consistent evidence that text messaging interventions are effective in helping students quit smoking. However, there are no studies that investigate the effects of text messaging interventions in a more general population. The objective of this study is to estimate the effects of a text messaging intervention on individuals seeking help to quit online and individuals visiting primary health care units.

METHODS

A 2-arm, parallel-group (1:1), randomized controlled trial will be employed to address the study objectives. The trial will follow a Bayesian group sequential design. Recruitment will be conducted using online advertisement (Google, Bing, and Facebook) and through health care professionals at primary health care units. All participants will receive treatment as usual; however, participants who are allocated to the intervention arm will also be given access to a 12-week text message smoking cessation intervention. Primary outcomes are 8-week prolonged abstinence and 4-week point prevalence, measured 3 months and 6 months postrandomization. Mediator variables (self-efficacy, importance, and know-how) will be measured to estimate causal mediation models.

RESULTS

Recruitment commenced in September 2020 and will not exceed 24 months. This means that a complete dataset will be available at the latest towards the end of 2022. We expect to publish the findings from this trial by June 2023.

CONCLUSIONS

This trial will further our understanding of the effects of text messaging interventions among a more general population than has previously been studied. We also aim to learn about differential effects between those who seek support online and those who are given facilitated support at primary health care units. Trial recruitment is limited to the Swedish population; however, a strength of this study is the pragmatic way in which participants are recruited. Through online advertisements, individuals are recruited in reaction to their own interest in seeking help to quit. At primary health care units, individuals who were not necessarily looking for smoking cessation support are given information about the trial. This closely mimics the way the intervention would be disseminated in a real-world setting and may therefore strengthen the argument of generalizability of findings.

TRIAL REGISTRATION

ISRCTN 13455271; http://www.isrctn.com/ISRCTN13455271.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID)

PRR1-10.2196/23677.

The Effect of Question Order on Outcomes in the Core Outcome Set for Brief Alcohol Interventions Among Online Help-Seekers: Protocol for a Factorial Randomized Trial

Background

A core outcome set (COS) for trials and evaluations of the effectiveness and efficacy of alcohol brief interventions (ABIs) has recently been established through international consensus to address the variability of outcomes evaluated.

Objective

This is a protocol for studies to assess if there are order effects among the questions included in the COS.

Methods

The 10 items of the COS are organized into 4 clusters. A factorial design will be used with 24 arms, where each arm represents 1 order of the 4 clusters. Individuals searching online for help will be asked to complete a questionnaire, and consenting participants will be randomized to 1 of the 24 arms (double-blind with equal allocation). Participants will be included if they are 18 years or older. The primary analyses will (1) estimate how the order of the clusters of outcomes affects how participants respond and (2) investigate patterns of abandonment of the questionnaire.

Results

Data collection is expected to commence in November 2020. A Bayesian group sequential design will be used with interim analyses planned for every 50 participants completing the questionnaire. Data collection will end no more than 24 months after commencement, and the results are expected to be published no later than December 2023.

Conclusions

Homogenizing the outcomes evaluated in studies of ABIs is important to support synthesis, and the COS is an important step toward this goal. Determining whether there may be issues with the COS question order may improve confidence in using it and speed up its dissemination in the research community. We encourage others to adopt the protocol as a study within their trial as they adopt the ORBITAL (Outcome Reporting in Brief Intervention Trials: Alcohol) COS to build a worldwide repository and provide materials to support such analysis.

Trial Registration

ISRCTN Registry ISRCTN17954645; http://www.isrctn.com/ISRCTN17954645

International Registered Report Identifier (IRRID)

PRR1-10.2196/24175

MINISTOP 2.0: a smartphone app integrated in primary child health care to promote healthy diet and physical activity behaviours and prevent obesity in preschool-aged children: protocol for a hybrid design effectiveness-implementation study

Background

Childhood obesity is still a major health problem in many countries, including Sweden. Childhood obesity and obesity-related behaviours in childhood, such as low physical activity and unhealthy eating habits, tend to track into adulthood, which highlights the need for early prevention. Our aims are to evaluate whether a parent-oriented mobile health app (the MINISTOP 2.0 app) integrated into primary child health care can improve diet and physical activity behaviours and reduce the prevalence of overweight and obesity in preschool-aged children as well as to evaluate the implementation among child health care nurses and parents.

Methods

This trial uses a hybrid type 1 effectiveness-implementation design. Families (n = 500) who attend a routine visit to one of 15–20 primary child health care centres throughout Sweden, when their child is 2.5 years, are offered participation in a randomised controlled trial (effectiveness evaluation). After acceptance, families will be randomised (1:1) to control or intervention groups. The intervention group receives a 6-month parent-oriented smartphone intervention aimed at improving the dietary and activity behaviours of their child (the MINISTOP 2.0 app) and the control group receives routine child health care. Dietary habits, physical activity and screen time (primary outcomes), body weight and height in children, and parental self-efficacy (secondary outcomes) are measured at baseline and at 6 months post randomisation. Implementation outcomes (i.e. perceived acceptability, appropriateness, and feasibility) of the intervention will be assessed among primary child health care nurses and parents in the trial through questionnaires and qualitative interviews.

Discussion

This trial will evaluate whether the MINISTOP 2.0 app can be used in primary child health care to improve diet and physical activity behaviours, and prevent overweight and obesity, in preschool-aged children. If effectiveness is proven, and the MINISTOP 2.0 app is considered acceptable, appropriate and feasible, it can be implemented nationally as part of the preventive strategies to combat childhood obesity provided by routine child health care.

Trial registration

The trial was registered at the Clinicaltrials.gov register platform (ID NCT04147039) on 31 October 2019.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-020-09808-w.

The P Value Line Dance: When Does the Music Stop?

When should a trial stop? Such a seemingly innocent question evokes concerns of type I and II errors among those who believe that certainty can be the product of uncertainty and among researchers who have been told that they need to carefully calculate sample sizes, consider multiplicity, and not spend P values on interim analyses. However, the endeavor to dichotomize evidence into significant and nonsignificant has led to the basic driving force of science, namely uncertainty, to take a back seat. In this viewpoint we discuss that if testing the null hypothesis is the ultimate goal of science, then we need not worry about writing protocols, consider ethics, apply for funding, or run any experiments at all-all null hypotheses will be rejected at some point-everything has an effect. The job of science should be to unearth the uncertainties of the effects of treatments, not to test their difference from zero. We also show the fickleness of P values, how they may one day point to statistically significant results; and after a few more participants have been recruited, the once statistically significant effect suddenly disappears. We show plots which we hope would intuitively highlight that all assessments of evidence will fluctuate over time. Finally, we discuss the remedy in the form of Bayesian methods, where uncertainty leads; and which allows for continuous decision making to stop or continue recruitment, as new data from a trial is accumulated.

Effectiveness of a 3-Month Mobile Phone-Based Behavior Change Program on Active Transportation and Physical Activity in Adults: Randomized Controlled Trial

BACKGROUND

Active transportation (AT; ie, walking and cycling as a mode for transportation) has been associated with decreased morbidity and mortality; however, low-cost and scalable intervention programs are lacking.

OBJECTIVE

The goal of the research was to determine the effectiveness of a 3-month behavior change program delivered via a mobile phone app to promote AT (TravelVu Plus) on time spent in moderate-to-vigorous physical activity (MVPA).

METHODS

For this 2-arm parallel randomized controlled trial, we recruited a population-based sample of 254 adults from Stockholm County who were aged 20 to 65 years and had access to a smartphone. On completion of 1-week baseline measures, the 254 participants were randomized to either the control or intervention group (1:1 ratio). Both groups had access to the standard TravelVu app (Trivector AB) for monitoring their AT for 6 months. The intervention group also received a 3-month behavior change program to promote AT (TravelVu Plus app). Assessors of outcomes were blinded to group allocation. Outcomes were objectively measured MVPA at 3 (primary) and 6 months. Secondary outcomes were AT, attitudes toward AT, and health-related quality of life at 3 and 6 months.

RESULTS

No effect on MVPA was observed after 3 months (P=.29); however, at 6 months the intervention group had a greater improvement in MVPA than the controls (6.05 minutes per day [95% CI 0.36 to 11.74; P=.04]). A Bayesian analyses showed that there was a 98% probability that the intervention had any effect at 6 months, and a 63% probability that this effect was >5 minute MVPA per day.

CONCLUSIONS

No effect on MVPA immediately after the intervention period (at 3 months) was observed; however, there was a delayed effect on MVPA (6 minutes per day) at 6 months, which corresponds to approximately 30% of the weekly MVPA recommendation. Our findings suggest that a behavior change program promoting AT delivered via an app may have a relevant effect on PA.

TRIAL REGISTRATION

ClinicalTrials.gov NCT03086837; https://clinicaltrials.gov/ct2/show/NCT03086837.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID)

RR2-10.1186/s12889-018-5658-4.

The Mobile Health Multiple Lifestyle Behavior Interventions Across the Lifespan (MoBILE) Research Program: Protocol for Development, Evaluation, and Implementation

BACKGROUND

Clustering of multiple lifestyle risk behaviors has been associated with a greater risk of noncommunicable diseases and mortality than one lifestyle risk behavior or no lifestyle risk behaviors. The National Board of Health and Welfare in Sweden reported in 2018 that it is important to provide additional support to individuals with multiple lifestyle risk behaviors, as risks from these behaviors are multiplicative rather than additive. However, the same report emphasized that there is a lack of knowledge regarding interventions that support changes to unhealthy lifestyle behaviors.

OBJECTIVE

The MoBILE (Mobile health Multiple lifestyle Behavior Interventions across the LifEspan) research program has brought together two Swedish research groups supported by international collaborators. Through this collaboration, we aim to design and evaluate a number of novel and tailored mobile health (mHealth) multiple lifestyle behavior interventions across the life span of different health care populations. In addition, the MoBILE research program will extend ongoing research to include mHealth interventions for migrant pregnant women and children.

METHODS

Each project within the MoBILE program will focus on a specific group: pregnant women, preschool children, high school and university students, and adults in primary and clinical care. All the projects will follow the same 4 phases: requirements, development, evaluation, and implementation. During the requirements phase, implementers and end users will aid the design of content and functionality of the interventions. In the development phase, findings from the first phase will be synthesized with expert domain knowledge and theoretical constructs to create interventions tailored to the target groups. The third phase, evaluation, will comprise randomized controlled trials conducted to estimate the effects of the interventions on multiple lifestyle risk behaviors (eg, alcohol, nutrition, physical activity, and smoking). The final phase will investigate how the interventions, if found effective, can be disseminated into different health care contexts.

RESULTS

The research program commenced in 2019, and the first results will be available in 2020. Projects involving pregnant women, preschool children, and high school and university students will be completed in the first 3 years, with the remaining projects being planned for the program's final 3 years.

CONCLUSIONS

The development of evidence-based digital tools is complex, as they should be guided by theoretical frameworks, and requires large interdisciplinary teams with competence in technology, behavioral science, and lifestyle-specific areas. Individual researchers or smaller research groups developing their own tools is not the way forward, as it means reinventing the wheel over and over again. The MoBILE research program therefore aims to join forces and learn from the past 10 years of mHealth research to maximize scientific outcomes, as well as the use of financial resources to expand the growing body of evidence for mHealth lifestyle behavior interventions.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID)

PRR1-10.2196/14894.

An Electronic Screening and Brief Intervention for Hazardous and Harmful Drinking Among Swedish University Students: Reanalysis of Findings From a Randomized Controlled Trial Using a Bayesian Framework

Background

Due to a resurgent debate on the misuse of P values, the Journal of Medical Internet Research is hosting a standing theme issue inviting the reanalysis of (primarily digital health) trial data using a Bayesian framework. This first paper in this series focuses on an electronic screening and brief intervention (eSBI), targeting harmful and hazardous alcohol consumption, which student health care centers across Sweden have routinely administerd to all students during the past decade. The second Alcohol Email Assessment and Feedback Study Dismantling Effectiveness for University Students (AMADEUS-2) trial aimed to assess the effect of the eSBI on alcohol consumption among students who were harmful and hazardous drinkers. A two-arm randomized controlled trial design was employed, randomizing eligible participants to either a waiting list or direct access to an eSBI. Follow-up assessments were conducted 2 months after randomization. Subsequent analysis of the trial followed the conventional null hypothesis approach, and no statistical significance was found between groups at follow-up with respect to the number of standard drinks consumed weekly. However, in an unspecified sensitivity analysis, it was discovered that removing three potential outliers made the difference between the groups significant.

Objective

The objective of this study is to reperform the primary and sensitivity analysis of the AMADEUS-2 trial using a Bayesian framework and to compare the results with those of the original analysis.

Methods

The same regression models used in the original analysis were employed in this reanalysis (negative binomial regression). Model parameters were given uniform priors. Markov chain Monte Carlo was used for Bayesian inference, and posterior probabilities were calculated for prespecified thresholds of interest.

Results

Null hypothesis tests did not identify a statistically significant difference between the intervention and control groups, potentially due to a few extreme data points. The Bayesian analysis indicated a 93.6% probability that there was a difference in grams of alcohol consumed at follow-up between the intervention and control groups and a 71.5% probability that the incidence rate ratio was <0.96. Posterior probabilities increased when excluding three potential outliers, yet such post hoc analyses were not necessary to show the preference toward offering an eSBI to harmful and hazardous drinkers among university students.

Conclusions

The null hypothesis framework relies on point estimates of parameters. P values can therefore swing heavily, depending on a single or few data points alone, casting doubt on the value of the analysis. Bayesian analysis results in a distribution over parameter values and is therefore less sensitive to outliers and extreme values. Results from analyses of trials of interventions where small-to-modest effect sizes are expected can be more robust in a Bayesian framework, making this a potentially better approach for analyzing digital health research data.

Trial Registration

International Standard Randomized Controlled Trial Number (ISRCTN) 02335307; http://www.isrctn.com/ISRCTN02335307

Electronic Screening for Alcohol Use and Brief Intervention by Email for University Students: Reanalysis of Findings From a Randomized Controlled Trial Using a Bayesian Framework

BACKGROUND

Almost a decade ago, Sweden became the first country to implement a national system enabling student health care centers across all universities to routinely administer (via email) an electronic alcohol screening and brief intervention to their students. The Alcohol email assessment and feedback study dismantling effectiveness for university students (AMADEUS-1) trial aimed to assess the effect of the student health care centers' routine practices by exploiting the lack of any standard timing for the email invitation and by masking trial participation from students. The original analyses adopted the conventional null hypothesis framework, and the results were consistently in the expected direction. However, since for some tests the P values did not pass the conventional .05 threshold, some of the analyses were necessarily inconclusive.

OBJECTIVE

The outcomes of the AMADEUS-1 trial were derived from the first 3 items of the Alcohol Use Disorders Identification Test (AUDIT-C). The aim of this paper was to reanalyze the two primary outcomes of the AMADEUS-1 trial (AUDIT-C scores and prevalence of risky drinking), using the same models used in the original publication but applying a Bayesian inference framework and interpretation.

METHODS

The same regression models used in the original analysis were employed in this reanalysis (linear and logistic regression). Model parameters were given uniform priors. Markov chain Monte Carlo was used for Bayesian inference, and posterior probabilities were calculated for prespecified thresholds of interest.

RESULTS

Where the null hypothesis tests showed inconclusive results, the Bayesian analysis showed that offering an intervention at baseline was preferable compared to offering nothing. At follow-up, the probability of a lower AUDIT-C score among those who had been offered an intervention at baseline was greater than 95%, as was the case when comparing the prevalence of risky drinking.

CONCLUSIONS

The Bayesian analysis allows for a more consistent perspective of the data collected in the trial, since dichotomization of evidence is not looked for at some arbitrary threshold. Results are presented that represent the data collected in the trial rather than trying to make conclusions about the existence of a population effect. Thus, policy makers can think about the value of keeping the national system without having to navigate the treacherous landscape of statistical significance.

TRIAL REGISTRATION

ISRCTN Registry ISRCTN28328154; http://www.isrctn.com/ISRCTN28328154.