|
Field
Abstract
|
Before
A vast literature in economics documents large differences in health outcomes by gender, race, wealth, and religion. To what extent these differences are driven by biases of health care providers is less understood. In this study, we explore whether health care providers discriminate against people with different religious beliefs using an audit study design (standardized patient method) on the clinical case of tuberculosis, an illness that is still widespread in Indonesia. The audit study is conducted in three provinces of Indonesia – Sumatera Berat, Sulawesi Utara and Jawa Tengah. The study covers 400 doctors at 400 randomly selected primary health care facilities in these three provinces. We conduct 1,600 standardized patient visits in which we vary the religious belief the patient. We complement the audit study with detailed facility- and doctor surveys in order to study the underlying nature of the observed biases.
|
After
A vast literature in economics documents large differences in health outcomes by gender, race, wealth, and religion. To what extent these differences are driven by biases of health care providers is less understood. In this study, we explore whether health care providers discriminate against people with different religious beliefs using an audit study design (standardized patient method) on the clinical case of tuberculosis, an illness that is still widespread in Indonesia. The audit study is conducted in three provinces of Indonesia – Sumatera Barat, Sulawesi Utara and Jawa Tengah. The study covers 400 doctors at 400 randomly selected primary health care facilities in these three provinces. We conduct 1,600 standardized patient visits in which we vary the religious belief the patient. We complement the audit study with detailed facility- and doctor surveys in order to study the underlying nature of the observed biases.
|
|
Field
Last Published
|
Before
December 05, 2022 09:13 AM
|
After
December 06, 2022 06:57 AM
|
|
Field
Intervention End Date
|
Before
January 31, 2023
|
After
February 28, 2023
|
|
Field
Randomization Method
|
Before
The random assignment was be done by PIs with a replicable procedure using statistical software.
|
After
The random assignment was done by PIs with a replicable procedure using statistical software.
|
|
Field
Planned Number of Observations
|
Before
We plan with a minimum sample of 400 primary health centers, across 3 provinces of Indonesia.
|
After
We plan with a minimum sample of 400 primary health centers, across 3 provinces of Indonesia. At the health provider visit level, the total number of observations is 1,600 (4 standardized patient visits x 400 health care providers).
|
|
Field
Sample size (or number of clusters) by treatment arms
|
Before
We will assign 4 standardized patients to each health care provider.
|
After
We will assign 4 standardized patients to each health care provider. At the health provider visit level each treatment arms has 400 observations.
|
|
Field
Intervention (Hidden)
|
Before
We employ an audit study in which standardized patients (SP) – all female – will portray a clinical case of tuberculosis. The audit study adopts a 4 X 2 cross-randomization design. Religion and Religiousness will be expressed in terms of differences in women’s clothing. There will be four religious groups (3 Muslim groups: very conservative, conservative, less conservative) and 1 non-Muslim group: Christian). In addition, we vary the perceived level of wealth (2 groups: rich vs. poor) of the standardized patient by varying the patient’s quality of clothing.
The audit study is complemented by detailed health facility- and doctor surveys. The health facility survey collects information on the staffing, equipment, patients and services offered. The doctor survey includes information on the personal background, work experience and knowledge, religiousness and other preferences.
|
After
We employ an audit study in which standardized patients (SP) – all female – will portray a clinical case of tuberculosis. The audit study adopts a 4 X 2 cross-randomization design. Religion and Religiousness will be expressed in terms of differences in women’s clothing. There will be four religious groups (3 Muslim groups: very conservative, conservative, less conservative) and 1 non-Muslim group: Christian). In addition, we vary the perceived level of wealth (2 groups: rich vs. poor) of the standardized patient by varying the patient’s quality of clothing.
The audit study is complemented by detailed health facility- and doctor surveys. The health facility survey collects information on the staffing, equipment, patients and services offered. The doctor survey includes information on the personal background, work experience and knowledge, religiousness, implicit discriminatory beliefs, and other preferences.
|
|
Field
Secondary Outcomes (End Points)
|
Before
Secondary outcomes include variables that we will use to investigate mechanisms through heterogeneous treatment effects. Secondary outcomes are related to three distinct features of the intervention: (a) patient characteristics such as religion/religiousness and wealth, (b) provider characteristics such as religion/religiousness, wealth, social preferences, time preferences, tolerance, in-group favoritisms, competition, and (c) social norms at the community-level.
|
After
Secondary outcomes include variables that we will use to investigate mechanisms through heterogeneous treatment effects. Secondary outcomes are related to three distinct features of the intervention: (a) patient characteristics such as religion/religiousness and wealth, (b) provider characteristics such as religion/religiousness, implicit attitudes and beliefs towards certain religions, wealth, social preferences, time preferences, tolerance, in-group favoritisms, competition, and (c) social norms at the community-level.
|