The Role of Retail Providers in Antimicrobial Resistance: Experimental Evidence from Kenya

How to cite

Dinerstein, Michael; Karing, Anne; Yoo, Younggeun; Yan, Runfei. 2024. "The Role of Retail Providers in Antimicrobial Resistance: Experimental Evidence from Kenya." AEA RCT Registry. AEARCTR-0014333. https://doi.org/10.1257/rct.14333-2.1

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Michael Dinerstein* , Anne Karing , Younggeun Yoo , Runfei Yan

* Primary investigator

Registry ID
AEARCTR-0014333
DOI
https://doi.org/10.1257/rct.14333-2.1
First registered
Last updated
Status
Completed
Trial dates
2024-07-01 – 2025-12-31
Countries
Kenya (Kisumu, Busia, Siaya, Kakamega, Homa Bay)
Registry record
http://www.socialscienceregistry.org/trials/14333

Abstract

We study how pharmacies and clinics in Western Kenya make decisions that are critical to local health outcomes and antimicrobial resistance, such as the appropriateness of provided medications, diagnostic efforts, diagnostic accuracy, and the provision of partial dosages. We send standardized patients in multiple visits to a sample of 200 providers where we randomly vary the standardized patient's reported symptoms, willingness to provide symptoms, drug requests, and expressed price sensitivity. We plan to assess whether the actions a standardized patient takes influences the provider's willingness to ask diagnostic questions, the recommended drug, and the transaction price.

Intervention

We send trained standardized patients (SPs) to pharmacies and clinics and record data associated with each part of the interaction. We randomly vary four elements of the interaction: the SP's described symptoms, the drug the SP requests (if any), the order in which the SP describes symptoms or makes requests, and whether the SP presents as price sensitive.

The specific elements are:

Described symptoms (associated disease)
1. Acute Rhinitis
2. Cold
3. UTI
4. Typhoid
5. Bacterial Pneumonia

Drug request:
1. An antibiotic
2. A specific antibiotic
3. A non-antibiotic

Order in which describing symptoms or making requests:
1. Request of drug immediately
2. Asking for a price of a drug immediately
3. Presenting symptoms and then requesting a drug
4. Just presenting symptoms

Extra element:
1. Asking what is the most important drug
2. Requesting price discount
3. Requesting partial dosage
4. None

Our interventions combine an element from each of these four categories. We vary these combinations across facilities and rounds.

Experimental design

We visit providers over the course of multiple rounds, where a round describes a period of time when each provider receives exactly 1 SP visit.

Our rounds are structured as follows:
1. Acute rhinitis, describe symptoms only, no drug request, ask for the most important drug if offered multiple
2-3. (Acute rhinitis / cold), (describe symptoms only / describe symptoms then request drug), (none / "an antibiotic"), ask for the most important drug if offered multiple
4. (Acute rhinitis / cold), describe symptoms then request drug, (cheap non-antibiotic / expensive non-antibiotic), none
5-6. (UTI / typhoid), (describe symptoms only / describe symptoms then request drug), (none / drug with established resistance / drug with growing resistance), (none / request price discount)
7. Bacterial pneumonia, (immediate request / immediate price request), Amoxicillin, partial dosage request
8. (UTI / typhoid), (immediate request / immediate price request), (drug with established resistance / drug with growing resistance), request price discount

Rounds 2-3: Providers will receive the same case (symptoms) but across rounds, (describe symptoms only / describe symptoms then request drug) varies. Half of providers receive "describe symptoms only" in round 2 while the other half receive it in round 3.

Round 4: Providers receive the same case they received in rounds 2-3.

Round 5-6: Providers will receive the same case (symptoms) but across rounds, (describe symptoms only / describe symptoms then request drug) varies. Half of providers receive "describe symptoms only" in round 5 while the other half receive it in round 6. Price discount request occurs in round 6 only.

Round 8: Providers receive the same drug request they had in rounds 5-6.

Design details

Randomization method
Randomization by a computer
Randomization unit
Randomization at the level of the provider. Providers will receive different treatments within rounds and different sequences of treatments across rounds.
Clusters
200 Facilities (3 Hospitals, 63 Clinics, 134 Pharmacies)
Observations
1,600 SP Visits (8 rounds of visits for each of 200 facilities in the sample)
Arms
Varies by round, but generally 50 facilities into each of 4 disease case/requested drug combinations.
Primary outcomes
Whether the purchased drug is appropriate for the SP's described symptoms
Secondary outcomes
Diagnostic efforts (whether the provider asks diagnostic questions or suggests a diagnostic step like a test) Diagnostic accuracy Whether the provider sells a partial dose of an antibiotic Characteristics of transacted products (price, number)

Cited by

No citation data available for this registration yet.

Keywords