Moca Mobile Overview

PDF Document

“Last year, my father had chest pains and needed to see a specialist in Manila. We sold our cows and our land so he can go. I had to take leave from my job to accompany him. It turned out to be pneumonia, which could have been diagnosed here.”
- a local from Batanes, the northernmost region of the Philippines, to Moca team members, June 2009

A Complete Health System: Our Innovative Approach
One of the largest problems facing resource-poor regions is a lack of trained medical expertise. While there is not a shortage of untrained or semi-trained workforce, health workers in many developing nations are not able to dispense adequate care due to a lack of training or expert consultation. Moreover, inefficiencies in the way medical data is collected, stored and transmitted leads to a paucity of information and long delays in treatments which result in medical errors and exacerbated conditions. The telecommunications revolution has the potential to transform the health care divide for under-served populations. Moca is the first end-to-end digital health solution designed for resource-poor settings (www.mocamobile.org). The main interface is an open-source and intuitive form-based system that the user can tailor to any clinical application. The phone allows rapid uploading of any type of medical data to a central server which can then automatically alert remote medical experts to examine the medical information and compare it to a patient's history. The impression and recommendation are then sent instantly back to the cell phone for same-day treatment or referral. In doing so, we can reduce delay in diagnosis and treatment from weeks to just hours, and thereby minimize the multiple impacts of poor health (by reducing the delay in treatment and number of missed follow-ups), costs to the health care provider (by decreasing the administrative overheads) and the costs to the patient (due to lost work when traveling to the clinic, seeking alternative stop-gap solutions and requiring more heroic treatments for exacerbated conditions).
A key challenge facing remote diagnostic platforms that utilize cellular networks in developing nations is the issue of connectivity. In order for a system to be useful, it must be robust in resource-poor areas. Moca is specifically designed to work in bandwidth- and resource-constrained rural areas, and uses synchronization, packetization, and multimodal transfers to ensure reliable, low-cost data transfers. Moca’s strength is the use of the cell phone as both the diagnostic sensor suite as well as the interface to the medical record.
We are aware that the implementation of innovation is far more challenging than the design of the innovation itself, especially if it involves process re-design. To this end, we are assisting our partner organizations with value chain analysis, aligning of incentives among the stakeholders, workflow re-engineering, operations management and creation of a learning system. These are crucial ingredients for successful adoption of our clinical information system.

Our Technology
Moca is a reliable tool that allows health workers to transmit any medical file, whether text notes, audio, photo, video or even information rich x-rays and ultrasound images through a cell phone 1) to a central server for archiving and incorporation into an electronic medical record, and 2) to a remote specialist for real-time decision support. When strong cellular connectivity is not available, Moca solves the problem using packetization, synchronization and multi-modal transfer. Moca integrates with OpenMRS, an open-source electronic medical records system used in developing countries to facilitate longitudinal care and patient tracking. The platform enables collaborative quality assessment among partner organizations, which will be crucial in setting the standards for telemedicine technology and medical care in resource-poor settings. A short demonstration video on the web is found at http://vimeo.com/6675001.
Using mobile technologies for tele-health in developing nations requires the health worker to deal with multiple standards, different networks, and local challenges ranging from intermittent connectivity to cultural differences. Moca realizes that few instantiations of remote medical diagnostics are the same. Unlike hodge-podge solutions, Moca is highly customizable, allowing clinicians to create their own workflows for common procedures (Fig 1).

Figure 1 Example of a Moca workflow: Surgery follow-up
These workflows can be dynamically loaded onto phones running Moca. The front-end for data and media capture is accessible through a fully-programmable interface. The back-end provides an intuitive user interface for management of medical media. Moca was built for integration with OpenMRS, but can be quickly ported to other commonly used medical record systems. This system design allows for maximum modularity and interoperation.
Using the Moca application, health workers run a procedure and collect patient data on the phone. Moca then uploads the information to OpenMRS for doctor review. After reviewing the case, doctors notify health workers by sending diagnosis and treatment recommendations to the Moca application. Clinicians have the ability to build unique procedures for health care workers and organizations. Procedures are step-by-step workflows, and are at the core of Moca. In most scenarios, a procedure is a set of pages that have questions or prompts. For instance, a page might prompt a user to take a picture or record audio. Other pages in a procedure might ask the user to enter text, check boxes or record video from a connected medical device. Procedures in Moca are defined in an extremely compact XML format. While it is easy to write a procedure in our procedure format, in the future we will develop graphical tools to define procedures. To promote greater interoperability, we plan on integrating the X-Forms implementation, created under OpenROSA, a popular consortium to standardize medical data collection tools. In addition to pages with question-response pairs, a procedure has the ability to branch. Branching is defined in the Procedure XML, and allows for any arbitrary logic to be performed on other previously answered questions. This extremely powerful feature can be used to do everything from skipping certain questions if they are irrelevant given prior responses, to allowing physicians to create complete decision tree diagnosis utilities.
The complete Moca system (Fig 2) consists of at least one (and in most instances several) phones and a web-connected server. The server runs both the medical records system of choice, such as OpenMRS, and the Moca Dispatch Server (MDS). The MDS runs on the server that is responsible for communication to and from phones registered in the system. It takes care of receiving data via the lower-level synchronization and packetization that the Moca-enabled phones perform. In addition, the MDS can be modified to interface with any medical records system. Using our custom MDS plug-in, Moca is fully-compatible with OpenMRS. We have extended OpenMRS to have a queue of pending diagnoses in addition to allowing data such as images to be tagged to a patient record.
A key challenge facing remote diagnostic platforms that utilize cellular networks in developing nations is the lack of adequate connectivity. Moca employs three strategies to ensure reliable, low-cost data transfers:
1.Packetization: Some acquired data are extremely large, such as video and high-resolution images, which require an inordinate amount of time to upload over GPRS. Oftentimes, half-complete transfers are interrupted and fail mid-upload due to poor service, and all data is lost on the receiving end. Using packetization algorithms, Moca uploads large files in smaller chunks so that little bandwidth is wasted in the case of a dropped connection.
2.Synchronization: When a procedure is completed, the medical data is stored in a local, on-phone database. At this point, the procedure can be rerun for the next patient. A background service in Moca is constantly listening for cellular service. As soon as service is available, all completed procedures in the database are uploaded to the server.
3.Multi-modal transfers: Moca has the ability to transfer data over a number of common interfaces, including GPRS, WiFi, SMS and USB tether. The different interfaces are used for specific purposes. Images and audio are relayed via GPRS, WiFi or USB tether, while text can be optionally sent via SMS. This is particularly useful if the phone is outside the coverage area, since the cellular network operator will automatically store and forward the SMS as soon as the phone reenters the service area.

Our Competitive Advantage
A fundamental goal of all tele-health focused organizations is to introduce an application that will enable a clinical transition in health systems to one that incorporates mobile innovation and ultimately reduces negative health outcomes in resource-poor developing countries. The shift towards the adoption of emerging tele-medicine applications in resource-poor developing countries is happening in tandem with advancing mobile technology in the developed world. Where phones with GPS and web-browsing are commonplace in the developed world, the developing world is using the more basic SMS and MMS technologies in enterprising ways. Given the number of people living in range of a cell tower coupled with the efforts of national and grassroots-level organizations working to provide access to smartphones and cutting-edge technologies has resulted in the rapid uptake of mobile technologies by the general population in developing countries.
Within the health sector, mobile communication is able to retrieve and transmit a wide array of information which can be divided into the following user-specific categories: education and awareness, remote data collection, remote monitoring and diagnostics and treatment support in health systems management. On the front of education and awareness, several groups are using SMS technology to coordinate information and for interactive text messaging. One such example is Project Masiluleke, which transmits over 1 million text messages throughout South Africa with information encouraging users to be tested and treated for HIV/AIDS. Frontline SMS is free and open-source software that enables users to create communication hubs in the developing world by connecting a laptop to a cell phone for larger-scale text messaging. Rapid Android is similar software that uses Google's open-source Android platform to create a mini-server for data analysis directly from your cell phone.
In the scope of remote data collection, smartphones and PDAs are being used on the ground to collect data and health information to reduce negative health outcomes. EpiSurveyor is an open-source cell phone-based application at the forefront of mobile data collection. EpiSurveyor is designed to create and deploy field-based survey forms from a cell phone or laptop to a remote server where all data is backed up. Another front-runner in mobile data collection is the Android-based Open Data Kit system, which integrates with Google's App Engine to collect and store a variety of data types. JavaROSA is an X-Forms-based open-source application developed by OpenROSA which runs on J2ME software, allowing it to integrate with a wide array of mobile devices.
Mobile communication has also taken on the challenge of medication compliance in patients. In the developing world, this is especially important as patients often travel large distances just to receive an initial check-up. Health workers can now use software such as CommCare, which assists them in managing in-home visits and scheduling. A health worker can also use CommCare to collect and report data to help monitor and evaluate community health programs. Similarly, GeoChat is an open-source technology that allows users to geospatially monitor activity over any phone and on any platform. By maintaining a system for patient follow-up which accounts for time and location, groups such as these ensure that appointments will not be missed and medication regimens will be strictly followed. In Thailand, the TB-DOTS program has seen a jump from 30% to 90% of medication compliance thanks to applications like these.
The application of mobile technologies in delivering or enabling distance health care, known as mHealth, is not an entirely new concept. The final division under the umbrella of tele-medicine, and also the most relevant to Moca's mission and goals, is that of mobile diagnostics and treatment support. A number of techniques adapted for awareness campaigns, remote surveying and tracking, treatment recommendation and administration reminders, remote diagnosis support and more have been trialed around the world, mainly for research purposes. This includes applications which are designed to communicate medical records and diagnostic information between remote health workers and remote physicians. One of the front-runners in mobile diagnostics is the group Click Diagnostics. This is a team that shares Moca's mission to provide access to quality medical care for the developing world through tele-medicine. Click Diagnostics provides a proprietary solution to a variety of problems that arise in the management of health systems in the developing world. By using SMS and MMS technology, this team has created a for-profit model for the communication of medical data from the field to a medical database or medical staff. However, so far these trials have been focused on very specific clinical applications, in very specific regions. Moca aims to develop an all-encompassing framework for mobile diagnostics that is designed to be extended to a multitude of different clinical applications in any number of regions. It is important to emphasize how we differ from other key players in this field.
In general, all of the above projects are aimed at low-end hardware and SMS-mediated data transfer. The interfaces are platform-dependent, and do not naturally integrate with other systems. Other key differences between the above projects and Moca are as follows.
Moca:
1.can upload any type of data file (formatted or not), and packetizes it to ensure delivery of large files (such as ultrasound videos) even in areas of low bandwidth and poor connectivity.
2.caches the medical record on the phone and dynamically synchronizes with the back-end whenever cellular connectivity is available.
3.is designed to provide feedback to the user concerning the quality of the data they have captured to improve data quality at the source.
4.is designed to adjudicate between several experts and capture the diagnosis and treatment recommendation. The labels can be used to provide quality assessment auditing and to develop intelligent decision software for delivering diagnoses.
5.is designed to provide both patients and health care workers with software to aid medical decision support. We envision our system as an accessible mobile doctor for every family.
6.is both a student and a teacher. It learns from the experts, and also from the changing medical landscape over time. It can provide instruction to sparsely trained health care workers as to why a diagnosis was given. This will allow health care workers (and heads of families) to attain a deeper understanding of the medical decision process. This enables the user to judge when the software may be in error and when to make his or her own decision if the case is unclassifiable by the software.
Our approach is not a ‘one-size-fits-all’ philosophy. By working collaboratively with both local doctors and developers in the target country, we are able to co-create targeted solutions for our partners. Our partners on the ground share and own the system, which motivates its use and enables it to scale in the face of changing needs and technical problems. Additionally, firms and hospitals that currently implement tele-health solutions have internally developed “strung together” systems. In our discussions with these firms, we have learned about several problems plaguing these systems: they were expensive to build, and yet are not robust and flexible; and there is a need for better workflow management as well as analytical tools on the back-end to deliver metrics and quality assurance. In creating Moca, we have had numerous dialogues with potential adopters of our tele-health system to understand and overcome these problems in our system.
Finally, Moca is a collaborative solution operating under the GNU GPL license. We enable entrepreneurs, companies, governments and NGOs to bring down the barriers in deploying a tele-health system while increasing its effectiveness. Moca is in the business of improving the health of its users. It is the combination of Moca’s scalability, reliability in resource poor locations, and cost-effective and collaborative solution that makes Moca the best solution for improving health care systems in developing countries.

The Pilot Project: mRadiology in Batanes, Philippines
Through our partnerships and volunteer work, we identified the Philippines as an ideal target market for tele-health. We envision Moca to serve as the mobile clinical information platform for the entire country. With funding from MIT, several Moca team members were able to visit the Philippines on three occasions this past year to perform pre-implementation research. We have partnered with the National Tele-health Center (NThC) of the University of the Philippines for the pilot implementation of our clinical information system. NThC was established in June 1998 through a University of the Philippines Board of Regents Resolution with the mandate to increase access to health information and services through information and communications technology. The NThC has all the footprints that Moca requires of its partner organizations. We are aware that the success of Moca relies on our partner organizations’ ability to implement and sustain the projects.
Over the past year, we have been working with all the key stakeholders including the national and local government, the medical sector, and the telecom industry. We sent members to assist the National Tele-health Center in performing pre-implementation research that will also be useful in drafting policies surrounding the delivery of tele-medicine. We are aware that implementation of innovation, especially one that could potentially transform health care delivery, can be far more challenging than designing the innovative system itself. It requires meticulous value chain analysis and alignment of incentives among the key stakeholders. Our most important role is to serve as a catalyst for change among our partner organizations. The National Tele-health Center must take ownership of the project and accountability for its success. We will provide assistance in workflow re-engineering, process re-design and clinical outcomes tracking that are crucial to achieve their goals.
The Moca platform promises to deliver quality healthcare to significant segments of the rural, developing world. As a large archipelago of more than 7000 islands, the large rural population and an existing ICT infrastructure, the Philippines is an ideal test site for a Moca pilot deployment. Potential unique contributions include:
1.Moca will be the first open-source, complete end-to-end (from data capture for longitudinal medical record to point-of-care decision support) diagnostic mobile health information system designed for the developing world
2.Implementation of a permanent electronic medical record for rural patients
3.Assessment of the processes and barriers encountered in implementation
4.Evaluation of the workflow modification introduced by the new information management system
We have chosen Batanes to implement a mRadiology project using Moca (Fig 3). Batanes is a remote province with mostly general practitioners manning the health care system. These general practitioners have expressed their need to access to specialists' opinion for a subset of their cases. Although the province has a radiology machine, it does not have a radiologist available. When an x-ray of a patient is taken, the hospital sends the plates to the nearest hospital in the Luzon mainland via courier services. To maximize the courier service, they try to accumulate a certain number of x-ray plates before sending them. When the mainland radiologist reads the plate, they return the plates and results via courier services. In effect, the turn-around time for this process is two to three weeks. Although wireless internet is already present in the provincial capital, the farther municipalities within the province only have GPRS connectivity at most. A short video describing the pilot project on the web is found at http://vimeo.com/6265089. 

Figure 3 The Batanes mRadiology Project

There are very limited studies available evaluating the economic impact of tele-health and tele-medicine projects in the literature. The project would benefit from a prospective clinical study that will provide descriptive data and empiric evidence that will be used to prove the value of the platform, evaluate areas for improvement in the workflow and refine the clinical information system. Data obtained from the pilot study will be used to develop a Markov model to generate the denominator of the cost-effective analysis, and the cost-savings in the cost benefit analysis. This is what is truly required to demonstrate the value of this mobile technology-enabled clinical information system. These analyses will assess cost-effectiveness of tele-health and tele-medicine and guide governments and NGOs as regards funding allocation.

Standard Mobile Tele-health Platform in Southeast Asia
Moca has a memorandum of agreement with the Centers for e-Health and Tele-medicine in Southeast Asia (CETSEA), a consortium of medical informatics practitioners from the Philippines, Vietnam, Laos, Indonesia and Malaysia. Our current efforts are focused on a successful implementation in the Philippines so that we can replicate the process in the rest of Southeast Asia. We are in the process of planning a CETSEA conference on mHealth to be held in Manila in March 2010 in partnership with the National Tele-health Center. The goal is not only to standardize the technology platform for mobile health, but to standardize the process of implementation as well.

Moca’s Vision: Diagnostic Support for All
The business model of health care delivery in richer regions is predicated on a return model which generates the vast majority of the revenue through support contracts, cabling and disposable sensors. But delivery of such medical products and accessories is problematic outside of the intended market. Even when equipment is donated, as is the usually the case for third world countries, it can rapidly become obsolete and cost the recipient money to store or dispose of the equipment, creating additional burdens from a well-intentioned gift. For many years, non-profit organizations have wondered why Coca Cola can deliver to almost any remote place on the planet and yet we are unable to deliver health care in these same locations. Finally, an industry has appeared that accomplishes this: the telecommunications industry. The user-driven demand for cheap communication perhaps overrides many of our other impulses, and the poorest sections of humanity are reallocating their small amounts of disposable income to cell phones and air time. This has led to a formal and informal infrastructure that maintains and delivers communication facilities around the globe. By ‘piggy-backing’ on this system and leveraging the desire for humans to communicate we have an advantageous setting to deliver health care at a vastly reduced cost.
Another cell phone trend which is making this idea a reality is the rapidly increasing quality and variety of sensors which come with the phone. Doctors can use the camera for tele-dermatology or tele-radiology, the audio microphone to record heart and lung sounds, the accelerometer to record physical activity and compliance post hospitalization and the GPS, WiFi or GSM signal to provide location triangulation functionality. Furthermore, cell phones are more frequently being made with standard robust connectors (such as USB ports and audio jacks) and therefore they can connect with many standard peripheral devices to provide an almost limitless extension of monitoring capabilities. Some examples are blood glucose levels, medication compliance pill boxes, asthma or respiration tracking, heart rate, weight and oxygen saturation. We are currently beginning two programs to extend the diagnostic capabilities of the cell phone in this manner (see next section).The fact that cell phones have several means of transmitting and receiving data (GPRS, Bluetooth, WiFi, radio, USB port, etc) also means that they can provide the seamless ability to connect with remote databases to provide real-time support and data back-up in a highly redundant manner.
An additional advantage to using cell phones for data entry and retrieval is that, unlike laptops, cell phones provide familiar interfaces to almost anyone, are cheap and quick to replace, run on batteries for an entire day, are easy to charge (from the car cigarette lighter for example) and are more difficult (and less attractive) to steal since, for instance, users can more easily hide them in their pockets. Moreover, cell phones are the classic data encryption paradigm: something you have and something you know. By requiring passwords, authenticated from a pre-approved phone, it is easy to maintain data privacy while monitoring and managing access to medical data.
With the back-end in the cloud, Moca can quickly and easily scale up operations, and can benefit from the power of these computing systems. The key to cloud computing is to have the right partners. We have begun talks with Google.org in Cambridge, MA, USA, to determine if we can use the back-end computational and storage facilities of Google to store (de-identified or encrypted) patient data as a back-up to local hospitals and medical centers. This will also provide a strong computational platform to begin data mining to spot anomalies in the data, predict adverse events (such as disease outbreaks or drug stock-outs) and to train algorithms to repeat the expertise of human experts. It will be key to split the data between de-identified data (such as history of weight, glucose levels etc), and identifiable data (such as location, name etc). The latter will be encrypted, and a key will only be available to the hospital and perhaps one or two other nominated parties such as the patient and the government.
Moca aims to provide a solution to the shortage of trained doctors by recognizing the massive opportunity that cell phones present is their ability to provide medical decision support. Moca has the potential to expand its infrastructure in the Philippines to provide health care to more citizens than was previously possible. Moca will not only be a tool for maximizing health care delivery, but can improve the health of the Philippine nation as a whole. We have the health of the developing world in mind and are working hard to address this huge unmet need, beginning with rural health care.

References:
Celi LA, Sarmenta L, Rotberg J, Marcelo A, Clifford G, for the Moca Team. Mobile care (Moca) for remote diagnosis and screening. Journal of Health Informatics in Developing Countries 2009; 3(1): 17-21.
Bohmer, Richard. Designing Care: Aligning the Nature and Management of Health Care. Boston: Harvard Business School Publishing, 2009.