Private Sector HIV SelfTesting in Kenya Insights from

Private Sector HIV Self-Testing in Kenya: Insights from a Mystery Shopper Study Kristen M. Little, Christine Odour, Heather Awsumb, Hildah Essendi 22 July 2019

Conflicts of Interest § I have no conflicts of interest to declare page 2

HIVST Demonstration Project 1. Introduce HIV self-testing to private sector in Kenya • Private pharmacies & clinics • Provided self-testing service delivery training to all outlets • Nairobi and Mombasa • Blood-based & oral-fluid test kits 2. Understand elements needed for a sustainable private sector self-test market 3. Ability to reach targeted end-user groups with HIVST in the private sector • Adolescents • Young adult men and women page 3

Mystery Shopper Study Background § Research Objective: To understand the private sector’s ability to reach adolescents and young people with high quality HIV self-testing (HIVST) services § We conducted a mystery shopper study at pharmacies and private health facilities participating in an HIVST demonstration project page 4

Methods § Outlets carrying HIVST kits as part of the demonstration project were randomly selected for the study § Facility owners provided verbal consent – Were not informed about the date or time of visits § Mystery shoppers ages 18 -30 recruited for study – Provided a two day training and pilot testing of survey instrument § Visited facilities and attempted to purchase a qualityassured HIVST kit page 5

Methods § Used one of 14 pre-defined mystery shopper scenarios – – Shopper’s age (range: 16 -24 years) Reason for testing Type of kit to be purchased Questions to ask the provider. § Immediately after the visit, shoppers were interviewed about their experiences using a structured guide administered by a trained data collector – Quantitative and some qualitative data collected § Analyzed using Stata 15. 0 page 6

Results

Majority of visits made to pharmacies Mystery Client Visits by County & Facility Type, Sept 2018 60 Pharmacy 50 Health Facility § Visits evenly split between facilities in Nairobi (n=27) and Mombasa (n=28) 40 30 75% 20 86% 10 63% 0 Nairobi page 8 Mombasa Total § Most visits (n=41) were to pharmacies, though some visits (n=14) to private health facilities

Self-test kits frequently stored out of sight Where were the self-tests kept in the outlet? Other In the lab On top of the counter None in facility On the shelf Under the counter* Store room/backroom page 9 § Test kits most commonly stored in storage or back rooms in both pharmacies (41%) and health facilities (57%) § Less than a quarter of the pharmacies had test kits displayed on the shelves (24%) § 7 facilities (13%) did not have any quality-assured HIVST kits in stock at the time of the visit 7% 0% Health Facility Pharmacy 0% 0% 21% 10% 7% 24% 0% 24% 57% 41% 0% 10% 20% 30% 40% 50% 60%

Non-quality assured HIVST also distributed § 8% (n=3) of HIVST kits given to mystery shoppers in pharmacies were not QA’d kits – Repackaged professional use tests – Non PQ’d kits manufactured in China & South Africa page 10 Which HIV Self-Test Were You Given/Tested With? 70% Health Facility (N=11) 60% Pharmacy (N=37) 50% 40% 30% 20% 10% 0% 64% 62% Ora. Quick 36% 30% INSTI 0% 8% Other

Most providers are explaining the testing process, but many need client prompting page 11 Did the Provider Explain the Testing Process to You? 70% 60% Pharmacy Health Facility 50% 40% 30% No Yes, Client-Initiated 36% 39% 64% 0% 0% 10% 44% 20% 17% § While roughly 1/3 of providers explained the testing process without prompting, this was not standard § Few clients received no explanation at all Yes, Provider-Initiated

Most information is about test use & interpretation Were you provided an explanation of: Other Additional sources of information Where to seek support for a positive result Things to do to maintain status if negative result Seeking confirmatory testing for a positive result What to do with an invalid result? How to interpret the results Step-by-step guidance on conducting the test Counseling before doing the test page 12 § Females more likely than males to receive all types of self-test information 0% 10% Female 5% Male 60% 43% 35% 33% 65% 52% § Most information about test use, interpretation, and post-test actions 50% 43% 90% 76% 95% 76% 45% 33% 0% 20% 40% 60% 80% 100%

Price of test kits was not always consistent How much did you pay for the HIVST? Pharmacy 90% § Most test kits (~80%) were sold for suggested price of 500 Ksh § Actual price paid ranged from 0 Ksh to 900 Ksh § Four pharmacies sold test kits for prices above 500 Ksh, despite a 500 Ksh price sticker on the box Health Facility 80% 70% 60% 50% 40% Less than 500 Ksh page 13 82% 0% 78% 10% 18% 5% 20% 500 Ksh 0% 16% 30% More than 500 Ksh

Providers Mediate End-User’s Test Kit Choice § “…[the provider] convinced me that it’s cheaper and faster to use the [non-quality assured HIVST] costing 150 Ksh [because] it’s easy to interpret the result” § ”Provider prefers INSTI over Ora. Quick, stating that Ora. Quick is preferred for people with diabetes or those with a blood clotting problem. ” § “The provider preferred that I purchase the INSTI testing kit since it’s blood-based. ” page 14

Conclusions page 15

Conclusions § Mystery shopper survey provided a unique source of information about the quality of HIVST services in the private sector § Identified important quality issues in service provision difficult to collect through routine monitoring systems – – – Inconsistent price Non-quality-assured products Location of test kits Stock-outs Provider mediation of test kit choice § Will be conducted quarterly as a form of monitoring in the second phase of the project § Providers may need more ongoing support to answer client questions confidently and accurately page 16

Acknowledgements § § Colleagues at PSI and PS Kenya Children’s Investment Fund Foundation Mystery shoppers Providers participating in the pilot project

Questions? Comments?

Extra Slides page 19

Limited information was provided about other sources of test support Did the provider refer to: § Most test support information provided was about test performance, result interpretation, and IFU Other HIVST chatbot Female Male Gave HIST helpline # Referred to HIVST twitter handle Referred to HIVST facebook page Referred to beselfsure website § Providers rarely mentioned the website, helpline, or other sources of information page 20 The instruction manual was enclosed How to interpret the results How the test is done 0% 20% 40% 60% 80% 100%

End-user questions were usually answered, but not always satisfactorily page 21 60% Pharmacy 50% Health Facility 40% 30% 20% No Yes, satisfactorily 36% 17% 43% 51% 0% 21% 10% 32% § 49% of mystery shoppers in pharmacies and 57% of those in health facilities either did not have their questions answered, or the questions were not answered satisfactorily Did the provider answer any questions you had regarding use of the kit? Yes, unsatisfactorily

Monitoring data collection was uneven Did the Provider: 90% 80% Pharmacy Health Facility 70% 60% 50% 40% Ask your age** page 22 Ask if you'd ever tested for HIV before* Offer any precounseling 86% 73% 43% 29% 64% 0% 32% 10% 64% 20% 17% 30% Give you information about how to correctly use the kit? § Providers asked to collect data on age, sex, and HIV testing history and to provide direct assistance to adolescents § These activities conducted more frequently at health facilities

Indirect assistance with test performance was typically provided to consumers Did the Provider Assist you in Performing the Self-Test? § Most clients received some form of assistance in both pharmacies and health facilities § Roughly half had to explicitly request that assistance 60% Pharmacy Health Facility 50% 40% 30% 20% No page 23 Yes, Directly 36% 35% 55% 43% 9% 0% 0% 0% 22% 10% Yes, Indirectly (Client. Yes, Indirectly Initiated) (Provider-Initiated)

Presence of HIVST signage varied significantly % of Facilities with HIVST Signage by County and Outlet Type 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% page 24 41% 80% 67% 75% 56% 79% Pharmacy Health Facility Nairobi* Mombasa Total § Facilities provided with demand creation materials to display § Presence of materials more common in health facilities (79%) compared to pharmacies (56%)

Offers of professional use testing common at health facilities Did the provider offer to test you with something other than a self-test? 60% 50% 40% 30% 55% 20% 16% 0% Pharmacy** page 25 Health Facility