2017-10-14

2018-03-10

The primary outcome is the number of preventative care services selected by the participants at our free clinic.

The preventative services vary from non-invasive to more-invasive - (body mass index, blood pressure, cholesterol screen via fingerprick, diabetes screen via fingerprick and flu shot).

The secondary outcomes is whether the subject would refer their peers to the clinic and /or doctor. We are also interested in the primary outcome dis-aggregated by mistrust and other background characteristics as discussed in the analysis plan.

The experimental design was as follows:
(1) Recruitment: Adult men who self-identified as African American were recruited at community sites in the Oakland, CA area. Most recruitment sites were black-owned barber shops, though some men were recruited from flea markets, churches and schools. Those who chose to enroll in the study completed a baseline survey regarding demographics, health status, history of health care utilization and trust. Survey completion was incentivized by haircut vouchers at the barbershops or $10–$20 in cash, depending on the venue. The participants received coupons for free cardiovascular health screenings to redeem at a clinic that was run on Saturdays specifically for the purposes of the study. It was emphasized that the doctors would not be able to address other health concerns. Uber donated ride-sharing services to transport individuals to and from the clinic.
(2) Presentation and ex-ante demand: Subjects who presented at the clinic were randomized to a male study doctor and an incentive amount. After check-in, the subject was placed in a private patient room. At that point, the subject was given an unconditional cash payment for presentation at the clinic ($50). Afterwards, the subject was introduced to their assigned doctor through a tablet device, which included a generic description of the doctor’s background and his actual photo. Subjects were then asked to select which of five preventative health care services they would like to receive from their assigned doctor. One of the services — a flu shot — was a surprise service the subjects learned about on presentation, at which point they were randomly offered a financial incentive ($0, $5, $10) conditional on selecting to receive the vaccination.
(3) Ex-post demand: After selecting the desired services, the assigned doctor met with the subject. Study doctors were provided with a list of each subject’s ex-ante demand; however, they were instructed to encourage participants to obtain all of the services. The doctors were informed that, due to restrictions on the malpractice coverage purchased for the study, they were not able to provide services outside those in the screening. The ex-post demand (the demand after interacting with the assigned doctor) was also recorded. Subjects were asked to fill out a satisfaction survey after the encounter.

We are using computer randomization via Survey CTO.

The unit of randomization is the individual.

No

Not applicable.

1,500 individuals (African American men) will be recruited and 750 will be randomized.

1,500 African American men will be randomly assigned to black or non-black doctors with a 50% chance of being assigned to each type of doctor. We expect 750 subjects to present at the clinic, these 750 will be randomized.

Pre-Analysis Plan Worksheet: Oakland Men's Health Disparities Study

/versions/27023/docs/version/document

Pre-Analysis Plan: Oakland Men's Health Disparities Study