Claim Mobile: Engaging Conflicting Stakeholder Requirements in Healthcare in Uganda Melissa R. Ho, Emmanuel K. Owusu, and Paul M. Aoki Abstract—Claim Mobile is a platform designed to support a project that subsidizes healthcare by reimbursing health service providers in Uganda for treatment of patients with sexually transmitted infections. As with many development projects, the Uganda Output-Based Aid (OBA) project involves a number of stakeholders: the service providers, the project implementers, the financiers, and the Ugandan government. Design of an appropriate solution requires meeting the various and conflicting requirements of all of these stakeholders. In this paper we detail the rapid design and testing of a pilot implementation of a mobile and web-based system for processing claims forms, based on two prior field visits to Uganda. Based on a comparative device study, semi-structured interviews, health clinic surveys, and a brief deployment, we affirm the selection of the mobile phone as a platform from the health clinic perspective, and further suggest that effective design for development requires more than addressing requirements of the the “users” of the mobile phones but also all the other stakeholders involved, who may have conflicting requirements. Index Terms—mobile phone, ICTD, health, participatory de- sign, Africa, HCI I. I NTRODUCTION Mobile phones are frequently touted as being the appro- priate and sustainable platform for rural healthcare in Africa. They are relatively cheap, durable, consume less power than laptops and desktops, and incorporate a battery that makes them more amenable to use in places with intermittent or no power. Commonly proposed uses are for data collection [1], [2] and decision support for rural health workers [3], [4]. Some projects also use mobile devices as a platform for information dissemation as well as data gathering [5]. However, these are all generally “closed loop” systems in which researchers are able to control all aspects of the system design and operation, focusing their research primarily on the rural health workers that will be using the mobile phones. Other applications have even more potential for large-scale impact. In the agricultural sector, we have observed how the introduction of transparent market prices and subsequent hiring of “middlemen” to purchase from farmers has reduced Manuscript received October 1, 2009. This work was supported by the Blum Center for Developing Economies and the U.S. National Science Foundation Summer Undergraduate Program in Engineering Research at Berkeley (SUPERB) under Grant No. 0453604. Melissa R. Ho is with the School of Information at the University of California, Berkeley, 94720, USA. (phone: +256 777 723 786; email: mho@ischool.berkeley.edu). Emmanuel K. Owusu is with the Computer Engineering Department at Iowa State University, 50011, USA. (email: kwame@iastate.edu). Paul M. Aoki is with Intel Research, Berkeley, CA, 94704, USA (email: aoki@acm.org) fraud and transformed supply-chain management for the E- choupal project [6]. While health information is critical to the improvement of healthcare in developing regions, financing healthcare also remains a significant unsolved problem. Can we take lessons from e-Choupal and apply them in the healthcare sector? The design of usable, reliable, and fraud- resistant tools for management of these aid flows is an area with potential for very significant impact. However, in the case of healthcare, the financial models are very different from commercial markets – financing of health- care typically comes through transnational aid agencies like the World Bank and International Monetary Fund (IMF), and is implemented by non-governmental organizations (NGOs) and the local government. Since the NGOs are typically experts in health, not technology, data processing is often outsourced to third-party information technology (IT) vendors. Relationships between the vendors, the NGOs, the local governments, and the transnational aid agencies are not always smooth - and limitations in communications infrastructure means that the information flows between them are scattered at best. In this paper we suggest that the “closed loop model” generally used by researchers in deployments of mobile health applications does not map onto the financial and political realities of the mainstream of healthcare provision in Africa, and limits the ability of pilot programs to increase their scale and impact. We describe an innovative, IT-based, NGO-run healthcare access program in Uganda, and our experiences designing and deploying Claim Mobile, a mobile-phone based system intended to address inefficiencies and help the program scale to additional districts. We argue that in addition to addressing the needs of the primary users in the system, the health workers, our design must consider the requirements, motivations and concerns of the other stakeholders: the IT vendors, the NGOs, the government, and the aid agencies. Our designs must consider the larger order ramifications of how we may positively and negatively impact both the “users” who will be generating the data, and the entities that will be engaged in managing and using the information in the resulting database. Just as the e-Choupal project assimilated the mid- dlemen by hiring them as kiosk operators, we propose that we can design applications structured to accommodate conflicting stakeholder requirements, while also alleviating information inequalities resulting from limitations in the system prior to the introduction of the information technology.