Recruitment, selection effects and external validity in clinical trials

Last registered on April 19, 2023

Pre-Trial

Trial Information

General Information

Title
Recruitment, selection effects and external validity in clinical trials
RCT ID
AEARCTR-0011276
Initial registration date
April 17, 2023

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
April 18, 2023, 5:18 PM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Last updated
April 19, 2023, 2:47 AM EDT

Last updated is the most recent time when changes to the trial's registration were published.

Locations

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Primary Investigator

Affiliation
Bernhard-Nocht Institute fro Tropical Medicine

Other Primary Investigator(s)

PI Affiliation
Bernhard-Nocht Institute for Tropical Medicine
PI Affiliation
Bernhard-Nocht Institute for Tropical Medicine

Additional Trial Information

Status
In development
Start date
2023-04-18
End date
2026-12-31
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
There is a notable lack of knowledge on the external validity of clinical trials. The proposed study aims to fill this gap by providing insights into the magnitude and implications of selection effects into clinical trials. The objective of this study is to identify who participates in clinical trials with a particular focus on people’s socio-economic characteristics, abilities, personality traits, preferences, and behaviors by comparing the trial population to a general population (adults age 18 to 65). We will elaborate on potential biases in the estimation of treatment effects. Additionally, using an RCT compromising a control and two treatment groups the study will investigate to what extent the information provided in the clinical trial recruitment process influences the participation in the clinical trial of various population groups. The control group receives a standard recruitment campaign. Intervention groups will additionally receive an intervention aimed at encouraging persons to participate, whereby intervention T1 will aim to activate altruistic concerns and motives among respondents by emphasizing the positive benefits of clinical trial participation for the society and others (emphasizing altruism) and T2 aims at fostering trust and highlighting the rights of clinical trial participants to reduce anxiety and misperceptions. The study takes place as part of the LoaLoa cinical trials conducted by researchers of CERMEL in Lambaréné, Gabon.
External Link(s)

Registration Citation

Citation
Mbavu, Cédric, Jan Priebe and Kerstin Unfried. 2023. "Recruitment, selection effects and external validity in clinical trials." AEA RCT Registry. April 19. https://doi.org/10.1257/rct.11276-1.1
Sponsors & Partners

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information
Experimental Details

Interventions

Intervention(s)
After a baseline survey, households will be randomly selected into control or treatment groups. There is one control group and two treatment groups (T1 & T2). The control group represents a standard outreach activity as typically implemented by the staff of the research institute CERMEL in order to recruit participants for their clinical trials. The treatment groups comprise alternative elements: T1 will focus on activating altruism related concerns in respondents by emphasizing the importance of diagnostic testing and clinical trials for society. T2 will address possible mistrust and anxiety. In T2 the intervention will strengthen trust and inform respondents about the rights of clinical trial participants. All three recruitment campaigns are expected to take around 10 minutes and will be done by CERMEL staff.
Intervention Start Date
2023-04-18
Intervention End Date
2023-07-15

Primary Outcomes

Primary Outcomes (end points)
Our principal outcome variables (left-hand side variables) are as follows:
• Willingness to participate in LoaLoa pre-screening
• Willingness to participate in LoaLoa clinical trial
• Participation in LoaLoa pre-screening
• Conditional on being LoaLoa positive: Participation in various steps of the clinical trial
Primary Outcomes (explanation)
• Willingness to participate in LoaLoa pre-screening: self-reported intention to participate in blood test; binary indicator
• Willingness to participate in LoaLoa clinical trial: self-reported intention to participate in clinical trial; binary indicator
• Participation in LoaLoa pre-screening: documentation of actual participation in blood test; binary indicator
• Conditional on being LoaLoa positive (positive screening result): Actual clinical trial
o Participation in bus pick-up to CERMEL (binary indicator)
o Start of participation in clinical trial at CERMEL (binary indicator)
o Successful finish of clinical trial participation (binary indicator)

Secondary Outcomes

Secondary Outcomes (end points)
Secondary outcomes include variables that we will use to investigate mechanisms through heterogeneous treatment effects. Secondary outcomes are related to (a) trust (b) altruism and pro-social preferences (c) individual health status (d) perceptions and attitudes toward clinical trials and Loa loa, (e) social norms, and (f) coping capacity of participants.
Secondary Outcomes (explanation)

• Institutional Trust towards CERMEL
- Unincentivized measure: Institutional trust (PCA-based index)
- Incentivized measure: trust game using the strategy method
• Altruism
- Unincentivized measure: altruism (global preference module questions, PCA-based index)
- Incentivized measure: donation task
• Socio-demographic variables
- Age, gender, education levels, education levels
• Other economic preferences
- Risk preferences (simple risk question on the willingness to take risk)
- Time preferences (global preference module questions)
- Positive reciprocity (global preference module questions)
• Cognitive ability
- Reading: Self-reported measure, short reading test
- Working memory capacity: digit span test
- Analytical skills: logic number sequences (test from IFLS)
• Personality traits
- BIG 5 personality traits (BFI-10)
- Optimism (index from Gavrilov-Jerkovic et al. 2014)
• Attitudes towards clinical trials: survey measure (Ohmann and Deimling scale)
• Health status (self-reported)
• Prior exposure and experiences with CERMEL (self-reported)
• Local social norms (self-reported survey measure)
• Emotional arousal (index based on discrete emotions questionnaire)

Experimental Design

Experimental Design
The experimental design involves an RCT in which randomization occurs at the household-level. The RCT involves one control condition and two additional treatment arms (T1, T2). The RCT will involve four stages: Stage 1: the baseline survey, stage 2: the interventions, stage 3: a post-intervention questionnaire and stage 4: Various real-world outcome measures spaced over time.
Experimental Design Details
Not available
Randomization Method
Randomization is done by the survey software and is programmed by the research team.
The sample is not stratified since it involves a complete census of the respective adult population (age 18-65)
Randomization Unit
Randomization of interventions are at the household level.
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
around 900 households
Sample size: planned number of observations
We have 3 conditions (1 control and 2 treatment arms) and assume minimum detectable effect size (MDE) of about 8 to 10 pp. Calculations show that the study will have sufficient statistical power if we recruit a minimum of 2,000 persons (about 670 persons per treatment arm). Out of these 2,000 persons a sub-sample will be potential clinical trial participants.
Sample size (or number of clusters) by treatment arms
We aim to have about 670 persons per treatment arm. Randomization is done at the household level. We expect that per household on average 2 persons take part in the survey, resulting in around 335 clusters per treatment arm.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
We assume a minimum detectable effect size (MDE) of about 8 to 10 pp.
Supporting Documents and Materials

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information
IRB

Institutional Review Boards (IRBs)

IRB Name
Institutional ethic committeee of CERMEL
IRB Approval Date
2022-08-25
IRB Approval Number
2022-18
Analysis Plan

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information