Page 190 - HIVMED_v21_i1.indb

Page 3 of 7 Original Research

gender, mobile number, education level and whether they Evaluation of HIV-self-testing and mobile app
had self-tested before. The instructions, which were usage
developed in English, provided the tester with step-by-step
guidance, presented pictorially with simple wording taken Acceptability outcomes
directly from the HIVST kit manufacturer’s instruction sheet, The evaluation of mobile apps may provide challenges to
so that self-testing could be performed independently of a researchers due to the nature of their varied users, objectives,
33
clinical setting. interfaces and mobility. In many cases, app developers and
researchers develop data collection tools that are app-specific,
All data gathered by the app was automatically uploaded via in order to explore concepts exclusive to their app. 34,35 For this
a secure server to the Aspect data management platform pilot study, a survey was developed to advise on the
TM
for viewing and review by the research team. Data collected preliminary scale-up of the app, which looked at general
TM
in Aspect was presented in aggregate form on a data acceptability and asked a set of closed-ended (yes/no) and
dashboard that could be configured to display any relevant open-ended questions, similar to the methodologies found in
statistics for the research team. The app security was other mHealth app evaluations. 21,36 The survey collected
implemented with privacy by design methodology as per participant demographic information and included questions
30
Protection of Personal Information (POPI) guidelines with on whether the app was easy to use; which steps, if any, were
patient data encrypted in transit and at rest, and also followed difficult to understand; would they use the app again; would
best practice guidelines in accordance with General Data they be willing to download this app in the future and if they
Protection Regulation recommendations. 31 had any suggestions to improve the app. The demographic
information collected by the survey and recorded by the
Data collection HCW was also used to reference the accuracy of data capture
on the app.
Trained HCWs obtained voluntary informed consent from
the participant in a private room, then uploaded the Feasibility outcomes
participant’s unique study identification number on the app. Similar to acceptability, there is no universal measure for
Once uploaded, the participant was handed a Samsung J5 determining the feasibility of an app; however, the generally
TM
smartphone, preloaded with the Aspect HIVST app, and accepted formula for feasibility includes three criteria:
an accompanying HIVST kit. The sealed test kit contained an the participant’s acceptance of using the app, the ability of the
English brochure with instructions for use (IFU) as part of
the standard packaging; however, the participant was participant to complete tasks on the app and the ability of the
37
requested to perform the HIVST by following the IFU app to perform the required tasks. These variables inevitably
included in the HIVST kit and the digital version of the IFU change based on the functionality of the app and its intended
provided on the app. Obtaining the sample takes 5–8 min users, and for this pilot the feasibility criteria were as follows:
when using the IFU (either paper or digital), followed by a • User acceptance of the app: The number of participants
20 min incubation period. The OraQuick HIVST kit (Orasure who agreed to use the app.
Technologies Inc., Bethlehem, USA) was used for the study • Successful test completion using the app: The number of
as it had already undergone full evaluation and was participants who completed the testing through the app
32
approved for use in South Africa. In a private room at a without error (i.e. experiencing difficulties or asking the
clinic, participants were asked to navigate the app and HCW for assistance).
perform the HIVST with no assistance, whilst the HCW • Success of data capture through the app: The number of
observed the process and recorded any deviations from the participants who captured their demographic information
app instructions. Following the test, the HCW asked (when compared to the original records collected by the
the participant a number of questions to obtain feedback on HCW), uploaded their interpreted test result and captured
the app design and willingness to use an app for HIVST in their test-strip images correctly.
future.
The final feasibility score is then presented as a percentage of
After the 28 min test was completed, the participant returned the final criteria. 37
the phone to the HCW, who then uploaded their professional
interpretation of the HIVST result on the app. Regardless of Data analysis
the HIVST result, the HCW performed confirmatory testing
using a commercial HIV rapid test (Advanced Quality, InTec All data extracted from the survey questionnaire (paper
Products, Inc., Xiaman, China). If the participant’s self-test based) were entered into an access controlled Excel
and HCW confirmatory tests were discordant, a third test spreadsheet. The quantitative data captured on Aspect
TM
was performed (Abon 1/2/O Tri-line, Abon Biopharm were extracted into a separate access controlled Excel
Hangzhou Co., Hangshou China). The HCW uploaded all spreadsheet. Quality control checks involved a 10%
results, as applicable, on the app for reporting purposes. randomised check comparing paper-based tools against data
Participants with HIV-positive results (based on the on the spreadsheet. This was performed by the quality
confirmatory testing) were referred to a clinic as per standard control officer on a daily basis. All data were coded and then
of care. 7 exported to Stata version 15.1 (StataCorp, USA) for descriptive

http://www.sajhivmed.org.za 182 Open Access