Proof of Concept Study
Through generous support from Econet Wireless, GSID was able to conduct a proof of concept study of the GSID System in Zimbabwe. In late 2013, we worked with Dimagi, the University of Washington, and the Ministry of Health and Child Care (MOHCC) to deploy our system at five sites in Manicaland province: one provincial hospital, two district hospitals, and two rural health centers. The sites were selected for year-round endemic malaria, as well as high levels of HIV incidence.
To demonstrate our ability to read tests for different diseases and from multiple manufacturers, the GSID System was calibrated to read six tests already approved and distributed by the MOHCC:
- HIV
- SD Bioline HIV-1/2 3.0 (Standard Diagnostics, South Korea)
- First Response HIV 1-2-0 (v.3.0) Cards (Premier Medical Corporation, India)
- Alere Determine HIV-1/2 (Alere, United States)
- Malaria
- SD Bioline Malaria Antigen P.f – One Step Malaria P.f HRP-II Antigen Rapid Test (Standard Diagnostics, South Korea)
- First Response Malaria Antigen P. falciparum (HRP2) card test (Premier Medical Corporation, India)
- Paracheck Pf (Ver. 3) (Orchid Biomedical Systems, India)
Based on preliminary site visits, we learned that most RDTs are performed one-at-a-time, directly in front of patients. To facilitate use of our system, we provided a phone for each department within the hospitals performing tests (e.g. opportunistic infection clinic, family-and-child health clinic, etc.). Overall, we deployed 15 phones across the five participating sites.
Study Results
Training at Hauna District Hospital
During the course of the two month study, we collected over 2,500 results (summarized in Table 1), with ongoing collection of over 4,500 results as of May 2014. The number of results was significantly higher than we initially expected, and they provided a strong basis for our analysis.
| |
|
|
Number |
Positive |
Number |
Positive |
| Male |
214 |
17% |
749 |
24% |
| Female |
648 |
13% |
965 |
20% |
| Total |
862 |
14% |
1714 |
22% |
Table 1
Some of the highlights of the proof of concept include:
- Easy to Learn – Our team trained over 60 health care workers over the course of ten sessions, conducting sessions in both English and Shona. Training on the system took less than an hour, with ample time provided for each health care worker to practice using the system multiple times.
- Easy to Use - Actual data entry required about 90 seconds per result, and interviews with the health care workers indicated that it fit easily into their workflow with the patients present.
- Positive Feedback from Health Care Workers - Based on the interviews, the health care workers enjoyed using the system and strongly believed that it would improve their work, primarily by eliminating paper-reporting requirements at the end of the day, week, and month.
- Use of System for Quality Assurance – Images of the RDT results were monitored to conduct quality assurance, identifying when the nurses were performing RDTs incorrectly. Representatives at the central reference lab in Harare were able to follow up directly with specific sites, reviewing proper procedures for performing RDTs with the specific health care workers.
- Accurate Interpretation of RDT Results - One of the primary goals was to evaluate the equivalency between reading the tests with the software and a visual interpretation by the trained healthcare workers under real-world operating conditions, and the study demonstrated a strong agreement between the software and visual interpretations.
Results from the study were presented at the Mobile HCI Conference in September 2014. The paper (4.6mb)is available here, and a brief slide presentation (15.8mb) is available here.
Our study has also been profiled in the following publications:
- Smartphone Technology to Bolster Infectious Disease Surveillance (PDF) in the November 2014 issue of Lab Culture, the newsletter of the African Society of Laboratory Medicine. Complete issue available here.
- Mobile HIV & Malaria Diagnosis and Reporting System – Improve disease monitoring in Zimbabwe (PDF) in USAID’s mHealth Compendium, Volume 4. Complete publication available here.
Based on the success of our proof of concept study, we are currently seeking to conduct further evaluations, including a scaled-up demonstration in Manicaland Province in Zimbabwe.
If you are interested in learning more about the GSID System, please contact Ian Francis.
