TL;DR
This paper introduces a custom-built mobile form software designed for low-resource settings, specifically for conducting population health surveys. The study demonstrates that this mobile solution is not only preferred by local surveyors over traditional paper methods but also enhances data collection efficiency and usability, paving the way for more effective health initiatives in underserved communities.
What Problem Does This Paper Address?
Conducting comprehensive population health surveys is crucial for allocating limited resources effectively, especially in low-resource communities. Traditionally, these surveys rely on pen-and-paper methods, which are labor-intensive for data tabulation and analysis. While digital solutions exist, they often present challenges in low-resource environments due to factors like cost, complexity, and unreliable internet connectivity. This paper tackles the problem of bridging this gap by developing and evaluating a mobile form system that is user-friendly, offline-capable, and specifically tailored to the needs of surveyors in such settings, aiming to improve the efficiency and accessibility of health data collection.
Methods
The researchers developed a custom mobile form software using Next.js and React for the front end, with Redux for state management and Firestore for real-time database functionality. A key innovation is its offline capability, managed through Redux's offline storage, ensuring data persistence even without internet access. The software also includes time-tracking functionality to measure survey completion times and individual question response times.
The study involved a two-phase approach:
- Pilot Interviews: 53 local surveyors in the Philippines were interviewed to gauge their comfort level with technology and their preferences for survey methods.
- Field Test: The developed mobile form software was field-tested in three locations in Metro Manila. 33 surveyors collected 20-question health surveys from 266 participants over three days. Post-field test interviews were conducted with these surveyors to assess their user experience and satisfaction.
Datasets
Not discussed in the paper. The study utilized data collected from surveys administered to local populations in the Philippines.
Key Insights and Ideas
- Preference Shift: A significant finding is the shift in preference towards mobile forms. Initially, 60% of surveyors preferred mobile forms, and this increased to 76% after they had field-tested the system, suggesting that hands-on experience dramatically improves acceptance.
- Offline is Essential: The study highlighted that internet connectivity was unavailable 25% of the time during surveys, underscoring the critical need for offline functionality in low-resource environments.
- Simplicity Drives Adoption: The mobile form was designed with a focus on ease of use, presenting one question per screen and minimizing visual complexity. This approach was crucial for surveyors who had varying levels of technical proficiency.
- Time Efficiency: While not explicitly quantified as a percentage reduction, the results indicate enhanced efficiency in data collection compared to paper-based methods, with surveyors reporting the forms as "faster" and "easier."
- Targeted Design: The custom software addressed limitations of existing tools like REDCap, SurveyMonkey, EpiCollect, Google Forms, and ODK, which were found to be too complex, costly, or lacking essential features like offline capability or time-tracking.
Key Results
- Surveyor Preference: 60% (32/53) of surveyors preferred mobile forms over paper in the pilot phase. This preference rose to 76% (25/33) after field-testing.
- Usability: 70% (23/33) of surveyors described the mobile phone survey as "very easy" to use, with no surveyors finding it "hard" or "very hard."
- Training Time: 96% of surveyors completed training in 10 minutes or less, demonstrating the system's intuitive design.
- Data Collection: 33 surveyors collected 266 patient surveys, each containing 20 questions.
- Typing Speed: Average typing speed was 11.5 words per minute, which is consistent with previously reported speeds for low-resource populations. Multiple-choice questions were answered significantly faster than text-based questions.
Limitations and Future Work
Limitations: * Geographic Scope: The study was limited to Metro Manila, which may not be representative of all low-resource environments due to varying socioeconomic and technological conditions. * Sample Size: The number of surveyors was relatively small (33 for field testing), and future studies could benefit from larger and more diverse participant groups. * Potential Bias: The involvement of surveyors in testing the software might have introduced bias. * Technological Familiarity: The varied technological familiarity among surveyors could have influenced usability testing.
Future Work: * Broaden the geographic scope to include more diverse regions. * Conduct studies with larger and more diverse participant groups. * Consider using independent groups of testers to mitigate potential bias. * Incorporate more comprehensive evaluations of general literacy skills. * Further research could explore more objective performance metrics to strengthen the validation of these findings.
Summary updated