With aging populations and rising health care costs, many high-income countries are exploring mobile computing technologies to improve the efficiency and effectiveness of health care provision. These technologies, which underpin the field of pervasive computing, introduce a new model of human–computer interaction. Instead of the scenario where a single user interacts with a desk-bound “personal” computer, pervasive computing envisions a world embedded with small, inexpensive, portable networked devices able to communicate seamlessly with each other. In common with resource-rich countries, the field of pervasive computing has the potential to promote and support healthy population development in middle and low-income countries, and this, therefore, has relevance for South Africa. Current estimates suggest that there are between 28 and 32 million mobile phones in South Africa. This means that around 60% of all South Africans own, or have access to, mobile telecommunication. Over 900 000 km2 of the country is covered by the GSM (Global System for Mobile Communication) network of Vodacom, the largest telecommunications company in the country. Over 90% of South Africa is provided with access to mobile connectivity through shared agreements between the country’s major telecommunications networks.
Aims
The ubiquity of mobile phones has resulted in their receiving increasing attention from public health researchers. Yet a better understanding of how mobile phones could support health research in South Africa is still an emerging field with many unanswered
questions. This thesis attempts to fill some of these gaps in our current knowledge. In particular, the primary aim of this work is to implement and evaluate the use of mobile phones as instruments with which to collect and analyse information for monitoring, evaluation and research in low-resource rural African settings.
Methods
To investigate this aim, data were gathered from the development, implementation and evaluation of four health surveys in South Africa. Two surveys were conducted with Birth to Twenty, a birth cohort of South African young adults living in Greater Johannesburg. These data were used to better understand the feasibility and data-quality implications of using mobile phones as a tool for the administration of ‘self-administered’ surveys. Two additional surveys, completed in KwaZulu-Natal province, evaluated the same themes of feasibility, acceptability and impact of data quality in mobile-phone-assisted personal (face-to-face) interviews (MPAPI). The first, conducted with 500 HIV-positive pregnant women in eight primary health clinics and 12 interviewers trained to use the mobile-phone survey software, was used to assess the feasibility and acceptability of MPAPI. The final survey compared the difference in data quality achieved by 100 interviewers using either pen and paper, or mobile phones to conduct a short health survey. De Leeuw's conceptual model was used to frame how mode characteristics influence data quality.
Results
Mobile-phone-assisted interviewing was found to have an impact on the data quality, feasibility and acceptability of health surveys. MPAPI was found to be similar in terms of accuracy and cost to small-scale paper-and-pen interviewing (PAPI) surveys. Time lines
and accessibility were improved by the use of MPAPI. Mobile-phone-assisted self-interviewing (MPASI) surveys were found to have a lower survey response but a higher item-completion rate. Acceptability was found to be moderated by technological familiarity and the use patterns of mobile-phone features. Finally, conducting health research using mobile-phone interviews in South Africa was found to be feasible; to reduce the loss of questionnaires, and photocopying and data-entry costs; and to improve the speed at which data becomes available for analysis. Factors that mediated feasibility included the technical expertise of the project management and field staff, the technological know-how of participants, the comprehensiveness of the interviewer training, the mobile communication channel used (e.g., handset-agnostic SMS) and the presence or absence of an interviewer.
Conclusion
Under the right conditions, mobile-phone-assisted interviewing appears to be a feasible and practical tool for the rapid collection of health information, with data accuracy being the same or better than pen-and-paper interviews. It is argued that these benefits increase as the scale of the survey increases. Improved data can positively influence population health by providing decision makers with more rapid access to accurate data with which to monitor large-scale health systems. Small projects that do not require the rapid availability of data or where staff do not have the appropriate technical proficiencies would be better suited at present to more traditional survey data-collection techniques.
Keywords: mobile phones; pervasive computing; mHealth; data collection; survey error
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:wits/oai:wiredspace.wits.ac.za:10539/13860 |
| Date | 18 February 2014 |
| Creators | Van Heerden, Alastair |
| Source Sets | South African National ETD Portal |
| Language | English |
| Detected Language | English |
| Type | Thesis |
| Format | application/pdf |
Page generated in 0.0019 seconds