Assessing the feasibility of collecting mortality-related survey data by mobile phone

Last registered on March 08, 2022
View Trial History

Pre-Trial

Trial Information

General Information

Title
Assessing the feasibility of collecting mortality-related survey data by mobile phone
RCT ID
AEARCTR-0008065
Initial registration date
September 14, 2021

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
September 14, 2021, 6:19 PM EDT

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

Last updated
March 08, 2022, 1:31 AM EST

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

Locations

Country
Malawi
Region

Primary Investigator

Name
Stéphane Helleringer
Affiliation
New York University
Contact Primary Investigator

Other Primary Investigator(s)

PI Name
Boniface Dulani
PI Affiliation
University of Malawi
Contact Investigator
PI Name
Michael Chasukwa
PI Affiliation
University of Malawi
Contact Investigator
PI Name
Augustine Choko
PI Affiliation
Malawi-Liverpool Wellcome Trust Clinical Research Programme
Contact Investigator
PI Name
Mathero Nkhalamba
PI Affiliation
University of Malawi
Contact Investigator

Additional Trial Information

Status
Completed
Start date
2021-09-22
End date
2021-10-27
Keywords
Education, Health, Other
Additional Keywords
Data collection, survey data, demography, mortality
JEL code(s)
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
In many low-income countries, few deaths are recorded with civil registration systems. Instead, vital statistics about mortality are computed from data collected every 3-5 years during large-scale household surveys. Unfortunately, such surveys are too infrequent to detect short-term fluctuations in mortality. They are also often postponed or interrupted during epidemics because they present transmission risks (e.g., during in-person interviews, large-scale trainings or cross-country travel). In this project, we will test whether survey data on mortality can be collected remotely by mobile phone. Surveys of mobile phone subscribers are increasingly common in low-income countries. They can be conducted more frequently than household surveys. They also present more limited transmission risks during epidemics. However, surveys of mobile subscribers have seldom included questions about mortality, in large part due to concerns that participants might not be willing to discuss recent deaths in their family during a phone call, or that the survey questions might generate negative feelings (e.g., sadness, distress). We will work in Malawi, a low-income country in southeastern Africa. We will recruit a random sample of mobile phone subscribers aged 18-64 years old. Among those, some participants will be randomly allocated to an interview about family deaths, whereas others will be asked to complete a commonly used questionnaire on economic activity. We will then test whether participation and completion rates are lower when the interview focuses on mortality, and we will assess whether respondents experience more negative feelings when completing an interview about family deaths. Results from this study will help develop a new method for tracking mortality in real-time in low-income countries with limited data collection systems
External Link(s)

Registration Citation

Citation
Chasukwa, Michael et al. 2022. "Assessing the feasibility of collecting mortality-related survey data by mobile phone." AEA RCT Registry. March 08. https://doi.org/10.1257/rct.8065-1.1
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)
Our study is a methodological study, which aims to document the feasibility of asking questions about mortality and recent family deaths during mobile phone interviews. Our intervention thus consists of exposing selected mobile subscribers to the possibility of participating in a mobile phone interview about recent family deaths. In the control group, mobile subscribers are offered the possibility of participating in a commonly administered questionnaire about recent economic activity and household livelihood.
Intervention Start Date
2021-09-22
Intervention End Date
2021-10-27

Primary Outcomes

Primary Outcomes (end points)
The primary outcome is the cooperation rate, i.e., the number of completed interviews divided by the sum of completed interviews, partial interviews and refusals.
Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)
Secondary outcomes include a) the discontinuation rate, i.e. the number of partial interviews divided by the sum of completed interviews and partial interviews and b) the proportion of respondents who state their willingness to participate in future interviews of varying durations. We will also investigate the occurrence of negative reactions to the interview. We will thus measure c) the proportion of respondents who self-report that some of the survey questions made them upset or generated negative feelings, and d) the proportion of respondents who displayed signs of emotional reactions/negative feelings during the interview according to the interviewer.
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
Our study is a non-inferiority trial of a mortality questionnaire vs. a commonly accepted questionnaire on economic activity. We will enroll a sample of mobile subscribers using random digit dialing. Those who are eligible to participate will be randomly allocated either to an interview about recent family deaths (treatment group) or to an interview about economic activity and household livelihood (control group). Within the treatment group, respondents will be randomly assigned to either a short or an extended questionnaire on deaths. The short questionnaire will only ascertain age at death and time since the death, as well as registration status of the death. The extended questionnaire will also include potentially more sensitive questions about symptoms, circumstances and causes of deaths.
Experimental Design Details
Randomization Method
The randomization will be done in office using random numbers generated using Stata 15.1.
Randomization Unit
The randomization unit will be the individual (mobile subscriber).
Was the treatment clustered?
No

Experiment Characteristics

Sample size: planned number of clusters
There are no clusters in the trial. In total, we will include approximately 1,600 mobile subscribers.
Sample size: planned number of observations
There will be approximately 1,600 mobile subscribers included in the trial.
Sample size (or number of clusters) by treatment arms
Participants will be allocated to the treatment vs. control questionnaires in a 3:1 ratio. We assume that the cooperation rate is 0.86 in the control group. We set a non-inferiority margin of 0.05. Our null hypothesis is that the cooperation rate in the treatment group is lower than in the control group by more than 0.05. Our alternative hypothesis is that the cooperation rate in the treatment group is lower than in the control group by less than 0.05. To test this hypothesis with 80% power and alpha = 0.05, we require 1,194 mobile subscribers in the treatment group and 398 mobile subscribers in the control group. Sample size calculations were performed using the SampleSize4ClinicalTrials package in R.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Supporting Documents and Materials
IRB

Institutional Review Boards (IRBs)

IRB Name
University of Malawi
IRB Approval Date
2021-08-12
IRB Approval Number
P.07/21/76
IRB Name
New York University - Abu Dhabi
IRB Approval Date
2021-09-07
IRB Approval Number
HRPP-2021-93
Analysis Plan

Post-Trial

Post Trial Information

Study Withdrawal

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

Request Information

Intervention

Is the intervention completed?
Yes
Intervention Completion Date
October 12, 2021, 12:00 +00:00
Data Collection Complete
Yes
Data Collection Completion Date
October 12, 2021, 12:00 +00:00
Final Sample Size: Number of Clusters (Unit of Randomization)
1,683
Was attrition correlated with treatment status?
No
Final Sample Size: Total Number of Observations
1,683
Final Sample Size (or Number of Clusters) by Treatment Arms
Data Publication

Data Publication

Is public data available?
No

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

Request Information

Program Files

Program Files
No
Reports, Papers & Other Materials

Relevant Paper(s)

Abstract
Introduction Despite the urgent need for timely mortality data in low-income and lower-middle-income countries, mobile phone surveys rarely include questions about recent deaths. There are concerns that such questions might a) be too sensitive, b) prompt negative/adverse reactions among respondents, c) take too long to ask and/or d) generate unreliable data. We assessed the feasibility of mortality data collection during mobile phone surveys.

Methods We conducted a non-inferiority trial among a random sample of mobile phone users in Malawi. Participants were allocated 3:1 to an interview about recent deaths in their family (treatment group) or about their economic activity (control group). In the treatment group, half of the respondents completed a short mortality questionnaire, focused on information necessary to calculate recent mortality rates, whereas the other half completed an extended questionnaire that also included questions about symptoms and healthcare use. The primary trial outcome was the cooperation rate. Secondary outcomes included the completion rate, self-reports of negative feelings and stated intentions to participate in future interviews. We also documented the amount of time required to collect mortality data, and we explored the quality of death reports.

Results The difference in cooperation rates between treatment and control groups was 0.9 percentage points (95% CI = -2.3, 4.1), which satisfied the non-inferiority criterion. Similarly, the mortality questionnaire was non-inferior to the control questionnaire on all secondary outcomes. Collecting mortality data required approximately 2 to 4 additional minutes per reported death, depending on the inclusion of questions about symptoms and healthcare use. More than half of recent deaths elicited during mobile phone interviews had not been reported to the national civil registration system.

Conclusion Including mortality-related questions in mobile phone surveys appears acceptable and feasible. It might help strengthen the surveillance of mortality trends in low-income and lower-middle-income countries with limited civil registration systems.
Citation
Collecting mortality data via mobile phone surveys: a non-inferiority randomized trial in Malawi
URL
https://www.medrxiv.org/content/10.1101/2022.03.02.22271441v1

Reports & Other Materials