Information Avoidance and Medical Screening: A Field Experiment in China
Last registered on June 08, 2018

Pre-Trial

Trial Information
General Information
Title
Information Avoidance and Medical Screening: A Field Experiment in China
RCT ID
AEARCTR-0002872
Initial registration date
June 07, 2018
Last updated
June 08, 2018 5:35 PM EDT
Location(s)
Region
Primary Investigator
Affiliation
National University of Singapore
Other Primary Investigator(s)
PI Affiliation
Peking University
Additional Trial Information
Status
Completed
Start date
2014-06-12
End date
2018-05-31
Secondary IDs
Abstract
Are high-risk individuals more likely to avoid the disease test because of information avoidance? We conduct a randomized field experiment in rural China to investigate the issue. We vary the price of the diabetes test (price treatments) and offer both the diabetes test and the cancer test (disease treatments) after eliciting participants’ subjective beliefs on the risk of having the corresponding disease. We find that both the low- and high-risk group avoid the test, and this pattern is more salient when test price is higher and the disease is more severe. We derive the new predictions from the optimal expectation model in Oster et al. (2013a) to explain our empirical findings. Structural estimation suggests that individuals attach about half of the weight to anticipatory utility compared to the consumption utility, leading to information avoidance. Simulation also suggests that neoclassical view systematically underestimates the importance of subsidy or mandate policies.
External Link(s)
Registration Citation
Citation
Meng, Juanjuan and Changcheng Song. 2018. "Information Avoidance and Medical Screening: A Field Experiment in China." AEA RCT Registry. June 08. https://www.socialscienceregistry.org/trials/2872/history/30570
Experimental Details
Interventions
Intervention(s)
The field experiment has two designs: price treatments and disease treatments.

In the price treatments, we vary the price of diabetes test. Individuals were randomly assigned to one of the three groups: the free group (T0), the 10 RMB group (T10), and the 30 RMB group (T30).

In the disease treatments, individuals were randomly assigned to one of the two groups: the diabetes group and the cancer group. We provided the disease test for free after bloods were drawn for another free blood test (so there was no additional cost of taking the test), but varies the type of the disease to be tested: diabetes or cancer.
Intervention Start Date
2015-01-16
Intervention End Date
2016-04-10
Primary Outcomes
Primary Outcomes (end points)
We are interested in the impact of different treatments on the take-up of screening test, and who are selected to screen under different treatments.
Primary Outcomes (explanation)
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
We collaborate with a large local hospital in one rural county in Beijing, China to study the demand for the disease test. In 2014, 10 villages were randomly selected in the county. We first collected administrative data of all individuals in the sample villages from the local government. The administrative data includes names, gender, birth date and address. We asked village leaders to inform all individuals who do not have diabetes to come to the village office on the date of our study. Thus we conduct survey for full sample if they are eligible. On the day of study, we asked households to fill out a survey in a separate room. We provide free basic medical examination for all individuals after the survey, which includes height, weight, and blood pressure.

We design two experiments to test what determine the demand for diabetes screening: price treatment and disease treatment. We conduct the price treatment in 5 villages and the disease treatment in the other villages. The randomization is at the individual level to increase the power.

In the price treatment, we vary the price of diabetes test. When individuals come to our study, enumerators first conduct surveys. Individuals were randomly assigned to one of the three groups after the surveys: the free group (T0), the 10 RMB group (T10), the 30 RMB group (T30). Individuals draw one out of three sealed envelopes from enumerators, and the voucher inside the envelope indicates what price they need to pay to conduct diabetes test. The price to conduct the diabetes test in the same hospital is 30 RMB. Then we ask whether they would like to take a diabetes test after the survey. If they would like to do the test, nurses from the local hospital would take their blood after the physical examination. We choose the diabetes tests using veinal blood to measure blood sugar and glycated hemoglobin that needs laboratory analysis to deliver result several days later. If individuals had their breakfast before taking the blood test, we took the blood once and measure the random blood sugar level. If individuals fast before taking the blood test, we conduct the fating blood sugar test or the oral glucose tolerance test (OGTT).

In the disease treatment, we vary the diseases of test after bloods are drawn. Village leaders inform all individuals that there would be a free blood test for basic blood count and they should fast before coming to our study. When individuals come, nurses first take veinal blood from all individuals and enumerators conduct surveys. Individuals were randomly assigned to one of the two groups: the diabetes group and the cancer group. The randomization was conducted by the researcher using computers and individuals do not aware about the assignment. In the diabetes group, we asked whether they would like to use their taken blood to do an additional free diabetes test after the survey. The diabetes test is fasting blood sugar test. In the cancer group, we asked whether they would like to use their taken blood to do an additional free test of cancer risk after the survey. The cancer risk test is Carcinoembryonic antigen (CEA) blood test. The price of the CEA test in the same hospital is 40 RMB. If they would like to do the additional test, nurses will inform the testing results to individuals by text messages after several days.
Experimental Design Details
Randomization Method
Randomization conducted by researcher using computers
Randomization Unit
Individual
Was the treatment clustered?
No
Experiment Characteristics
Sample size: planned number of clusters
1195 individuals
Sample size: planned number of observations
1195 individuals
Sample size (or number of clusters) by treatment arms
219 individuals T0, 216 individuals T10, 229 individuals T30, 255 individuals Diabetes, 276 individuals Cancer
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB
INSTITUTIONAL REVIEW BOARDS (IRBs)
IRB Name
National University of Singapore
IRB Approval Date
2014-04-30
IRB Approval Number
2202
Post-Trial
Post Trial Information
Study Withdrawal
Intervention
Is the intervention completed?
No
Is data collection complete?
Data Publication
Data Publication
Is public data available?
No
Program Files
Program Files
Reports and Papers
Preliminary Reports
Relevant Papers