Home-Based Counseling and Testing with ODK Collect
AMPATH has an extensive home-based and counseling program where community health workers (CHWs) go house to house to identify and enroll persons in need of care (i.e., pregnant women not in antenatal care, orphaned children, persons at high risk for tuberculosis infection). The workers need mobile data collection to document socio-economic data (including GPS location of household) and to implement the counseling and testing protocol.

Before using ODK, AMPATH used Palm TX devices running Pendragon Forms. GPS information was collected using eTrex devices. Problems with this approach were outlined by Rajput et al. in their Evaluation of an Android-based mHealth System for Population Surveillance in Developing Countries (AMIA 2011) paper.

They write, "First, although costs were significantly lower than paper-based data collection methods the costs were still substantial. Second, the data collected could not be directly integrated into the electronic medical record system [OpenMRS] which was already in use at the AMPATH clinics -- integration required dedicated time by several experienced data managers. Third, the cable connection between the PDA and GPS devices was not always reliable, and GPS information had to occasionally be entered manually into the PDA devices. Fourth, the use of the proprietary Pendragon Forms Software on the PDA devices limited flexibility to incorporate some functionality into the data collection software –- some of these functionality included advanced barcode scanning and check-digit algorithms."

In late 2009, we provided an early version of ODK Collect to AMPATH (on an HTC Dream, the first Android device). We spent a week with them in Kenya piloting the system, and based on the available functionality, AMPATH decided to switch to Open Data Kit. Below is a picture of one of the CHWs scanning a patient barcode with the phone.

In early 2010, AMPATH finalized the HCT form, made minor changes to the user interface, acquired Android devices, and started to scale up. It is important to note that ODK's core developers were not involved at this stage.

Rajput et al. evaluated the ODK implementation at AMPATH a few months after and showed that "Users of the system felt it was easy to use, and facilitated their home visits. It is more cost effective than pen-and-paper alternatives. Additionally, electronic data collection facilitated earlier reporting. We have implemented a viable solution at scale for collecting electronic data during household visits."

More importantly, that study also found that for the 63,000 persons encountered, "the direct capture of electronic records greatly facilitated the expeditious performance of initial analyses and reports prior to the conclusion of the three year HCT program. Our work has highlighted ... most notably that only 28% of persons we are identifying as infected with HIV are presenting for follow-up care." AMPATH has acted on this data by launching programs to improve follow-up.

It has been almost two years since AMPATH started using ODK Collect. We recently touched base with the HCT program to see how their use of technology has scaled. Since early 2010, ODK has been used by a few hundred CHWs in over 650,000 patient encounters! The HCT program has a greater than 98% rate of acceptance into the homes it visits, and with the help of technology, has been able to lower mother-to-child transmission of HIV/AIDS to lower than 3%.

Mobile Clinical Decision Support with ODK Clinic
We have continued our collaboration with AMPATH, and this year, we focused on tools for clinicians. We have created a new version of ODK Clinic, an Android-based application that downloads patient data like demographics, disease history, lab results, and recent medications from OpenMRS (and the Clinical Summary Module). We re-designed the entire application from our early v1 release and added features to enable correction of serious mistakes in the patient record. We also added decision support so clinicians receive patient-specific reminders when the system notices that sub-standard care is being offered. All this in near real time and at the point of care -- a major improvement over AMPATH's existing paper-based summaries.

If, for example, the system notices a scheduled lab test is overdue or a patient's health indicators have dropped dramatically, a reminder is inserted into the patient's record and displayed to the clinician. ODK Clinic can also enforce compliance with the reminders and can help compliance by wirelessly printing lab order test requisitions complete with all necessary patient data. We detailed our findings designing the system in our Design of a Phone-Based Clinical Decision Support System for Resource-Limited Settings (ICTD 2012) paper. Below is a video demo the system.

Since that paper was written, the system has been used with about 7,500 patients at two adult HIV clinics. We are still evaluating the results of that deployment in a controlled trial, but very early results seem to show that clinicians using ODK Clinic deliver a higher standard of care. Moreover, clinicians enjoy using the system! As a few told us, "[We] can't see a patient without this phone."

We believe that technology only magnifies existing human intent and capacity and so, sustainable improvements require organizations who are dedicated to the communities they serve. With HCT and within HIV clinics, ODK and OpenMRS have helped make health care providers more efficient and through that efficiency, magnified their impact. We believe this is one of the reasons why this work was featured as one of the mHealth Alliance and Rockefeller Foundation's Top 11 in 2011 Innovators Challenge.

Our work at AMPATH has been in deep collaboration with many people. We want to thank Martin Were, Nyoman Ribeka, Sam Mbugua, Zeshan Rajput for their help. We also thank all our colleagues at OpenMRS and Regenstrief for their support, and Abbott Fund and Google for funding much of this work. Finally, we thank the CHWs and clinicians for their participation in our research. It is their hard work that is making a difference in the lives of the underserved.

As part of those explorations, Chris found Open Data Kit and recruited School of Public Health (SPH) students Marion McNabb and Laura Khurana to train fellow BUSPH students on how to use ODK. The students have put some of the discussions they had, forms they created, and feedback they had online. If you are thinking about implementing ODK for a global health research project, it is a great to start.

Below is video with some of their thoughts on using ODK.

In October 2011, staff from EI and United Nations Environment Programme (UNEP) trained a group of agronomists and staff members from CSI on ODK. HRI reports that during "the week of the ORE ict4Ag training, the South Department experienced severe flooding as a result of heavy rains in the region. Anticipating widespread damage to personal property and infrastructure, UNEP asked ORE to undertake a rapid agricultural assessment to gauge the agriculture losses caused by the flooding. ORE agronomists were able to use their new phones and take advantage of their new ODK knowledge to undertake a rapid agricultural assessment."

Congratulations to the ORE team on their quick transition from training to field usage! Find out more about their work at http://haitiregeneration.org.

He notes, "my two main objectives with this paper were to explore the software and steps required for collecting GIS data with mobile devices (smartphones such as iPhone, Android, and Windows Mobile) and for displaying that collected data on the web. I accomplished this by attempting to develop a mobile GIS collection and display solution using each of the options currently available for building mobile apps.

The second part of the paper, detailed instructions, gives detailed step by step instructions on how to create GIS data collection and display solutions for each of the development options found to have the most potential. Each set of detailed instructions covers all of the steps needed by a user with no previous experience to install, configure, build, and deploy the solution."

Download the paper from Doug Browning's site.

Matt Berg writes, "We've incorporated the XLS2XForm library directly into formhub. This allows you to upload an xls into formhub and then be deployed directly to your phone. We also provide a nice excel/csv export that provides support for things like repeats and some very basic mapping visualizations. We are also working on incorporating the idea of form sharing (via xls files) into formhub. This is an idea we are really excited about."

formhub is Python/Django, open source and an absolute delight to use. Find out more in the "Data Gatherers of the World Unite" blog post and sign up for an account at http://formhub.org.

Thanks to the Modi Research Group for their many contributions to the ODK ecosystem and for making data collection a lot easier!

Martin's implementation of ODK Collect has been used by community health workers from AMPATH in Kenya during home visits for HIV counseling and testing for over 650,000 individuals. More recently, he has used ODK Clinic to help clinicians make faster and better decisions about care.

Joaquín's use of ODK Voice has addressed problems associated with non-communicable diseases by connecting patients to clinics via automated phone calls. Like Martin, his work ensures patients receive appropriate and effective care.

Both projects are due to a close collaboration with the University of Washington, and we are excited to see Martin and Joaquín recognized for their work using mobile technology in innovative ways to improve health systems and outcomes. Congrats to Martin and Joaquín and follow the links below to find out more!

Martin Were

• Read More: Regenstrief Investigator a Winner of Top 11 in 2011 Innovator Challenge
• Watch More: Enabling Better Care At AMPATH With ODK Clinic

Joaquín Blaya

• Watch More: MiDoctor - Chronic Diseases
• Read More: eHealth Systems

If you are running the .9x versions of Aggregate and have not upgraded, you are missing out on a simplified installer, a new user interface, submission filtering, better data security, help videos, and a ton more features.

If you are running Aggregate 1.0.x, this upgrade will fix a few bugs and make changes to help keep your App Engine quotas down.

Check out the demo server at https://opendatakit.appspot.com to try things out, read over the release notes and then download the update.

As always, if you have any other problems upgrading or installing this release, please file a detailed bug report.

In this release, we've added multiple questions per screen, secure network connectivity, a few new widgets and a lot of bug fixes that will make data collection a little easier. Find out all the new features on our project roadmap.

If you have the Android Market, download it here or by searching for "ODK Collect". You can also find it on our downloads page.

If you are upgrading from Collect 1.1.5 and use our http://opendatakit.appspot.com demo server, please note that the http:// URL no longer works. Please change it to https://. If you have any other problems upgrading or installing this release, please file a detailed bug report.

If you want to help with the ODK project, go to http://opendatakit.org/participate for ideas (and a donation link). We also appreciate your ratings and comments on the Android Market.

As always, thanks for all your patience and your support. We hope you enjoy this release!

• Derek Treatman: CommCare-Sense
  Dimagi has implemented CommCare on Android using ODK.
• Joaquin Blaya: MiDoctor
  MiDoctor uses ODK Voice for their interactive voice response functionality.
• Julius K.A. Awakame: OpenMRS-based telepsychiatry using mobile phones
  Mobile components of Julius' system will build on ODK.
• Martin Were: AMPATH's mHealth Programs
  AMPATH uses ODK Collect and ODK Clinic in many of their mHealth programs.

HealthUnbound needs your help to get down to the winners. Go to http://healthunbound.org/top11_vote to learn more and vote on your favorite.

Congratulations to all, and please do vote now -- the deadline is this Friday!

In their detailed "PHSRP - Monitoring the Monitors" whitepaper, Program Director Farasat Iqbal says, "Our traditional monitoring system had many weaknesses which we wanted to overcome. The practical situation is that the current system is irregular and unreliable."

Their report goes on to note, "Along with recording indicators like non-functional equipment, non-available medicines and absent staff, the supervisors will take a self-photo that will be submitted along with the data. The submitted data automatically pops up on a map in real-time with GPS accuracy of five meters. This data can be monitored by senior officials and back-end auditors.

With this mobile-based intervention, there is no need to fill in long paper-based forms. Data is validated and entered directly into the database automatically. Secondly, the requisition of taking self-photos ensures that the form is filled in by the person himself. Moreover, the phones have GPS which means that the exact location where the data is entered is recorded along with the data."

The data PHSRP generates is available on their website. They also have a (non-English) demo video of their form. Make sure to read the white paper for more info.