Uncover your risk! Using Facebook to increase the awareness of diabetes in Indonesia

We will distribute different information ads via the Facebook Advertisement function, which shall make the targeted individuals aware about the risk of diabetes in Indonesia. These ads will address different aspects of the disease and how it can affect individuals and families. Specifically, the different channels/aspects that will be addressed are:

1) Information – ads related to this topic will show an informative statement about the risk factors and potential health consequences of the diseases

2) Social and family aspects – ads related to this topic will show a picture related to a family aspect and contain a statement that refers to social aspects of the disease

3) Religion – ads related to this topic will show a religious picture and contain a health-related statement from the Quran.

4) Healthy lifestyle – ads related to this topic will show a picture related to a healthy lifestyle (work outs, healthy diet) and contain a statement that links a healthy lifestyle to a reduced risk of diabetes.

5) Location specifics – ads related to this topic will contain information about diabetes that are relevant for certain cities or regions. For example, we show an ad with the city-specific prevalence rates of diabetes.

6) Shocking channel – ads related to this topic will contain a slightly shocking picture/ message.

All ads carry a statement about diabetes and encourage the viewer to visit the campaign website (adagulaadadiabetes.com). Randomizing these different ads allows to identify the most effective channel, i.e. to test which kind of approach would be well suited in the context of Indonesia to raise awareness about a health concern and which approach can effectively encourage individuals to seek for more information.

After clicking on one of the ads in Facebook, individuals will be redirected to the landing page, i.e. the campaign website. This website will contain factual information on diabetes in Indonesia, for example, distribution of prevalence rates across the country and individual risk factors. It will also contain information about how diabetes is correctly diagnosed and how it is treated. Moreover, the website offers the opportunity to conduct a 1-minute self-screening test, based on the diabetes Findrisk questionnaire, which has scientifically been validated to diagnose the risk of diabetes. The Findrisk questionnaire has been adapted for Asian populations and consists of eleven questions that can be answered within one minute. It provides a test result in form of a score between zero and 20 points that indicates the degree of being at risk for diabetes. After the test, the participant will receive a summary of the score and a personal message on the risk that was calculated. Additionally, this message contains recommendations to visit a health center or a general practitioner, based on the calculated risk score.

2022-04-01

2022-04-30

Number of views per ad, clicks per ad, views of information page per ad, number of tests started, number of tests completed, costs per view, costs per click, costs per started screening questionnaire and cost per completed questionnaire.

We will also assess:
- Survey participation (in the addtional survey after the test) as a function of ads
- Survey participation (in the addtional survey after the test) as a function of find risk score
- Findrisk score as a function of ads (should not have an effect)
- Findrisk score as a function of find risk survey characteristics and characteristics observed in larger survey characteristics

The experiment will be designed in the form of an online health awareness campaign implemented via the social media platform Facebook. We will randomize ads that differ in their design and content. Specifically, we will randomize different groups of ads, with each group representing a different channel (i.e. social or information) or aspect of diabetes. Each ad groups consists of 2-3 different ads.

The target group of the study includes all Indonesian male and female Facebook users above the age of 21, which live in either Jakarta or Yogyakarta – the two cities with the highest diabetes prevalence rates in the country.

The experiment will be designed in the form of an online health awareness campaign implemented via the social media platform Facebook. We will randomize the ads that differ in their design and content. Specifically, we will randomize different groups of ads, with each group representing one of the channels outlined above. Each ad groups consists of 2-3 different ads.

We will use Facebook’s A/B-Test to implement the randomization procedure. This A/B-Test design allows us to specify that the exact same budget amount is allocated to each ad group and to thereby circumvent that Facebook’s algorithms determine the budget allocation towards the ad groups. Each ad group receives a budget of €10 per day, summing to a total budget of €60 per day for a total of 6 different ad groups. The awareness campaign will for the maximum period possible for the A/B test, which is currently restricted to one month. The planned period for the final campaign – after running a short pilot – is April 2022.

The randomization is taking place at the individual level, no clustering is applied. There is no pre-defined number of individuals/observations for this study, but we aim to reach at least 200 users per ad-group that click on the ads to be redirected to the website and to participate in the self-screening. If after the period of one month the minimum number of participants has not been reached, we plan to extent the campaign for another month.

Randomization done automatically via Facebook

The randomization will take place at the individual level

No

(No clustering) - We aim to reach as many individuals as possible, but at least 1200 in total.

We aim to reach as many individuals as possible, but at least 1200 in total.

Per treatment (ad-group) we aim to reach 200 individuals.

PAP_Facebook_Diabetes_January 2022

Appendix_PAP_Facebook_July2023

Nudging individuals without obvious symptoms of non-communicable diseases (NCDs) to undergo a health screening remains a challenge, especially in middle-income countries, where NCD awareness is low but the incidence is high. We assess whether an awareness campaign implemented on Facebook can encourage individuals in Indonesia to undergo an online diabetes self-screening. We use Facebook’s advertisement function to randomly distribute graphical ads related to the risk and consequences of diabetes. Depending on their risk score, participants receive a recommendation to undergo a professional screening. We were able to reach almost 300,000 individuals in only three weeks. More than 1400 individuals completed the screening, inducing costs of about US$0.75 per person. The two ads labeled “diabetes consequences” and “shock” outperform all other ads. A follow-up survey shows that many high-risk respondents have scheduled a professional screening. A cost-effectiveness analysis suggests that our campaign can diagnose an additional person with diabetes for about US$9.

Fritz, M., Grimm, M., Weber, I. et al. Can social media encourage diabetes self-screenings? A randomized controlled trial with Indonesian Facebook users. npj Digit. Med. 7, 245 (2024). https://doi.org/10.1038/s41746-024-01246-x.

https://www.nature.com/articles/s41746-024-01246-x