The Effects of Bundling Dental and Medical Appointments and Price on Uptake of Preventive Dental Care

Patients in this trial will be randomized in equal numbers to one of five conditions: One Holdout condition where patients do not receive any WhatsApp message or communications about dental check-ups, or one of four treatment conditions.

The treatment conditions comprise a 2×2 factorial design:

1. Bundling: Whether the message emphasizes the option to schedule the dental check-up on the same day as an already scheduled medical appointment
2. Price: Whether the price offered is $990 or $19,990 CLP

All four message conditions will receive a WhatsApp message offering a full preventive dental check-up with X-rays, with the price of the check-up specified according to the assigned price condition. For patients in the bundling frame conditions, the message will also highlight the opportunity to save time by scheduling the dental check-up on the same day as their upcoming medical appointment.

We ask the following research questions:

1. Does sending patients WhatsApp messages encouraging them to schedule preventive dental check-ups increase dental check-up uptake? We hypothesize that sending patients WhatsApp messages will increase uptake of dental checkups relative to not sending them any messages.
2. How does emphasizing the possibility of same-day scheduling (bundling) affect scheduling rates? There are competing predictions. On one hand, highlighting the option to bundle a dental appointment with an already scheduled medical visit can increase the perceived convenience and thus increase uptake. On the other hand, this intervention may narrow patients' consideration set to that single date, leading them to evaluate only the feasibility of attending on that date rather than considering alternative scheduling opportunities, potentially reducing uptake.
3. How responsive are patients to price? We hypothesize that patients who are offered preventive dental check-ups at a lower price will have higher uptake.
4. Are the effects of bundling and price independent, substitutive, or complementary? We explore the interaction effect between bundling and price discounts on the uptake of preventive dental check-ups.

Analysis:

The investigators will run ordinary least squares regressions (OLS) with robust standard errors to predict outcome variables.

1. First, to test whether the WhatsApp messages increase dental check-up uptake, the investigators will compare the holdout condition to the combination of the four treatment arms. The key independent variable in the OLS regressions will be an indicator for whether each patient received (vs. did not receive) a WhatsApp message in the experiment.
2. To test the main effects of bundling, the investigators will compare the scheduling outcome between the two conditions that include the bundling frame and the two conditions that do not (bundling main effect).
3. To assess price sensitivity for preventive dental check-ups, the investigators will compare scheduling outcomes between the $990 and $19,990 price conditions, controlling for bundling (price main effect). Thus, the key independent variables in the OLS regressions for testing RQ2 and RQ3 will be:

• An indicator for whether the message included the bundling frame
• Indicators for whether a discounted price of $990 was being offered, instead of $19,990

1. To examine whether the effects of bundling and price independent, substitutive, or complementary, the investigators will estimate (1) separate bundling effects at each price level by restricting the sample to each price condition and comparing scheduling rates between bundled and non-bundled messages, as well as (2) separate price effects under the bundle vs. non-bundle condition by restricting the sample to the bundle vs. non-bundle condition and comparing scheduling rates across price levels. The investigators will also run a regression that includes the bundling indicator, price indicators, and their interaction. The investigators will test the significance of the interaction term.

These regressions will include the following control variables:

• Time fixed effects
• Participant age
• Participant gender (male, female, other/unknown)
• Whether patients had had at least one dental visit at this hospital in the past
• Whether the pre-existing medical appointment is 1 day away vs. 2-7 days away
• Type of pre-existing medical appointment (specialty category)
• Region or clinic indicators, if available

For robustness, the investigators will conduct logit models as well. For RQ3, the investigators will compute marginal bundling effects at each price level as well as the marginal price effects with or without the bundling prompt.

The investigators will examine whether the effect of the bundling frame, the effect of price level, and their interaction vary across patient subgroups. In particular, the following moderators will be tested:

• Whether the patient had completed any dental visit prior to the experiment (based on the data the hospital can provide)
• Whether the patient had completed any preventive medical exam (e.g., mammogram) prior to the experiment (based on the data the hospital can provide)
• Prior visits to the clinic network that were not for dental purposes (based on the data the hospital can provide)
• Insurance type (public vs. private)
• Whether the patient lives in the same district as the clinic of their preexisting medical appointment
• Patient age
• Patient gender (male vs. female)

Additional information: Given that the study will run over the course of several months, some patients may become eligible more than once if they have multiple non-dental medical appointments during the study window. To preserve internal validity, the investigators will analyze only the first eligible observation per patient. This means that even if a patient receives more than one message over time, only their first assigned message and associated outcomes (e.g., scheduling, attendance) will be used in the main analyses.