Member Spotlight

By Pierre Bush, PhD Moderator | 06 Mar, 2017

Dear Colleagues, we are delighted to introduce Mr. Gordon Cressman in our member spotlight event that is taking place during this week of 03/06/17-03/10/17. Mr. Gordon Cressman is the Senior Director of the international Information and Communication Technologies (ICT) program, Research Computing Division, RTI International. His Bio is found at: https://www.ghdonline.org/users/gordon-cressman/. We will have questions for Mr. Cressman, and members are welcome to join the discussion.

Replies

 

Pierre Bush, PhD Moderator Replied at 8:12 AM, 6 Mar 2017

Dear Gordon,
Here is the first question for our discussion:

 Would you tell us what inspired you to be interested in malaria elimination?

Gordon Cressman Replied at 6:05 PM, 6 Mar 2017

Over the past 34 years I’ve worked in some 32 countries to help them improve information systems for governance, education, and health. I became interested in malaria elimination through our (RTI International) work in Zanzibar. In 2008, we launched a simple mobile phone reporting system in Zanzibar. It enabled public health clinics to report aggregate case and supplies (rapid diagnostic tests, ACT pharmaceuticals) weekly. I had nothing to do with designing or developing this system, but I was excited to see the results. This was a major advance for the Zanzibar Malaria Control Programme, and enabled them to detect outbreaks in less than 2 weeks of onset. This system is still running, 9 years later.

As the use of ACTs and major vector control (indoor residual spraying, treated bed nets) brought the malaria positivity rate down from over 40% in 1999 to less than 5% in 2006, Zanzibar reoriented its program from control to elimination, and then to elimination. Mike McKay, one of my team members based in Nairobi, Kenya, began working with the Zanzibar Malaria Elimination Programme (ZAMEP) to develop mobile software to support reactive case detection (RACD), which they began in June/July of 2012. For the first time, ZAMEP could see and analyze malaria cases as the data were synchronized in real time from Android tablets all over Zanzibar. Now they had near real-time geolocated case data, and a system to help them to monitor RACD, to identify hot spots and transmission patters, and to help target interventions.

That same year, 2012, Cohen et al published a systematic review and assessment of 75 malaria resurgence events. [1] The conclusions showed just how fragile this progress could be. It emphasized to me the critical importance of tools and techniques to target what I anticipate being increasingly limited resources effectively to sustain the gains and prevent the resurgence of malaria. It was disturbing to see the enormous waste of resources when previous malaria control efforts wained, only to see malaria rebound.

By 2015, the Gates Annual Letter put a stake in the ground, “In 15 years, we'll be poised to send malaria the way of smallpox and polio.” The 2015 WHO World Malaria Report showed the dramatic impact of large scale investments in malaria control programs, and demonstrated that Millennium Development Goal 6C, “to have halted by 2015 and begun to reverse the incidence of malaria…,” had been met. The Bill & Melinda Gates Foundation, WHO, Global Fund, Malaria No More, and the President’s Malaria Initiative all set targets for eliminating malaria.

2015 also saw the release of a Bill & Melinda Gates Foundation funded study of information systems to support surveillance for malaria elimination. [2] This study and my subsequent research suggested that, in Zanzibar, we may have a tool that could be used to help other countries sustain gains and perhaps achieve malaria elimination. It was exciting to realize that we may have developed a unique and badly needed tool. I have continued to learn everything I can about challenges and research in eliminating malaria. It is an exciting and evolving field.

An incredible amount of ingenuity and effort is being invested in new methods for vector control, personal protection, testing, treatment, surveillance, and response. But I’ve already seen skepticism and cracks in the funding dike. As indicated by Cohen, I fear that the tremendous progress of the past 15 years could disappear, unless we develop tools and techniques to sustain the gains and achieve elimination with fewer resources. I’m excited about contributing to that.

1. Cohen JM, Smith DL, Cotter C, Ward A, Yamey G, Sabot OJ, Moonen B: Malaria resurgence: a systematic review and assessment of its causes. Malar J 2012, 11:122.
2. Ohrt C, Roberts KW, Sturrock HJ, Wegbreit J, Lee BY, Gosling RD: Information Systems to Support Surveillance for Malaria Elimination. Am J Trop Med Hyg 2015.

Pierre Bush, PhD Moderator Replied at 8:07 PM, 6 Mar 2017

Hello Gordon,
Thank you very for accepting our invitation for this event. I appreciate this elaborate answer to the my first question. Your work in Zanzibar is definitely commendable. As you mentioned the work performed by RTI International and ZAMEP made a big difference both in malaria surveillance processes and advancing the country towards malaria elimination. Also, thank you for providing the two resources that you referenced in your response. I just provide the links below. My second question will be posted tomorrow.

Attached resources:

Gonzo Manyasi Replied at 4:18 AM, 7 Mar 2017

Thank you Pierre for introducing Gordon - I find his story very exciting, especially being an East African native and resident. My question is whether the ZAMEP model has spread to other countries. This I'd hoped to confirm through the second link (on info systems) you've kindly provided but unfortunately it says the "page is not available" and so I'll limit myself to what Gordon has stated above in the meantime before I search further afield.
My two questions to Gordon Cressman are: One, is the Mike Mckay team in Nairobi still continuing with the project in Zanzibar and has it been spread to say Kenya? This could be extremely useful, especially with the Highland malaria outbreaks? Secondly, do we or could we also have similar simple mobile phone info reporting systems for other diseases? About the waning of control efforts, part if not most of it is due the lack of a coordinated approach among the many players in the funded projects tackling the issue, especialyy government commitment. It would appear to be mostly about going in, carrying out your work as per your mandate, collect data, write reports, maybe even publish the results, got to a conference or two and that's it. Several times, down the line during my teaching of graduate students, I have asked about one project or the other only to be told that it ended with the research team going away upon "completing their work" with nothing to show in terms of the much touted sustainability (on the host government's part). So I totally agree with Gordon's sentiments that we need to develop tools and techniques to sustain the gains and achieve elimination with the fewer resources available. Thank you again.

Gordon Cressman Replied at 10:16 AM, 7 Mar 2017

I believe that the second link above has expired. Try the link below, which is to the publication abstract. I've also added a link to the 2014 background paper that is the basis for that publication. The Bill & Melinda Gates Foundation is currently supporting an updated assessment of malaria surveillance systems. That work is underway now. I don't know their schedule, but I hope we have access to the findings later this year. Internally, we continue to research approaches in other malaria elimination contexts as we continue working to improve Coconut Surveillance.

Q: "is the Mike Mckay team in Nairobi still continuing with the project in Zanzibar and has it been spread to say Kenya?"

A: Mike McKay as well as our team members in Tanzania and the US are continuing to work with ZAMEP to support and improve the system. The PMI Vector Control and Scale-up Project in Tanzania ended last September. Over the years ZAMEP developed strong analytical capacity and the capacity to manage and use the surveillance system. We are continuing to provide some technical support from our own resources. The surveillance system has not been spread to Kenya, but it could be useful in other malaria pre-elimination and elimination settings. We have explored this through internal studies of Zimbabwe, Uganda, Cambodia, and Haiti. It is very important to assess each environment carefully. This includes human resources, transport, supply chains, mobile device management, connectivity, costs, stakeholders, local ownership, and political will. All this needs to be projected at the scale necessary to cover the targeted area. Of course, this is true of any intervention.

Q: "...do we or could we also have similar simple mobile phone info reporting systems for other diseases?"

A: There are many mobile phone information reporting systems for other diseases. These include text messaging systems using Unstructured Supplementary Service Data (USSD) and SMS. One example of a USSD system is the Malaria Early Epidemic Detection System (MEEDS) in Zanzibar, which is used by healthcare facilities to report aggregate malaria data weekly, as well as to notify the Coconut Surveillance system of each new malaria case. Another example is the eIDSR (electronic Integrated Disease Surveillance and Response) system on Mainland Tanzania, which is used to submit aggregate data for all reportable diseases as well as for notifiable disease alerts. Data from this system feed into the national health information system, which is based on DHIS2. One example of an SMS system is the national Weekly Disease Surveillance System (WDSS) in Zimbabwe, which uses Frontline SMS. This is used by more than 1,600 facilities for IDSR aggregate and notifiable disease reporting, and has sustained (some 3 years) reporting compliance of nearly 100%. Data from this system feed into the national health information system, which is based on DHIS2. Sierra Leone is rolling out Rapid Pro, an SMS system supported by UNICEF, for ISDR reporting. Rapid Pro is used in several countries. In Uganda it is called mTrac, and is used for IDSR reporting.

There are also several mobile applications for Android devices that are being used for IDSR and single disease reporting. In Guinea, our team has developed an IDSR module for DHIS2, and is working towards introducing the use of the Tracker Capture mobile software for reporting from healthcare facilities. This will take time, as local capacity is developed to use and support these systems. Zimbabwe has pilot tested the use of Tracker Capture for malaria case reporting into DHIS2. Other countries are using CommCare to enable community health workers to report, and CommCare was used in the Conakry area of Guinea for Ebola case reporting and contact tracing.

All of the other examples provided above are designed as reporting systems. Most are used for aggregate data reporting only, but a few are now being used for individual case reporting. In collaboration with PMI and ZAMEP, we took a different approach. We used technology that enables the system to guide health workers through their tasks in the field, even as response protocols change. At the same time, it also provides near real-time geo-located data to supervisors, program managers, and researchers. This is made possible by a robust record synchronization system that manages versioning and conflicts automatically. This system recognizes that workers in the field may have only occasional access to network connectivity. Therefore, the information on each mobile device is current as of the most recent synchronization with the shared database. This makes it possible for workers in the field to be guided according to the latest response protocol and transmission risk stratification. They can also collaborate on responding to individual cases, and can transfer cases from one area (e.g. district) to another. They also have access, on the mobile device, to all case data within their assigned area (e.g. district). This meets the particular needs of reactive case detection for malaria elimination. It could also be useful for guiding the rapid response to other notifiable diseases. For example, in 2016 we also adapted the Coconut software platform for Ebola contact tracing as a proof of concept. It's software architecture makes it fairly easy to adapt to other diseases, such as Yellow Fever. We are now seeking opportunities to test this.

Attached resources:

Pierre Bush, PhD Moderator Replied at 11:15 AM, 7 Mar 2017

Thank you Gordon and Gonzo. The work RTI International is doing is very impressive. It must be quite challenging to operate these systems in areas where there is limited internet access. I would like to follow up with the following question:

 Last year during the discussion on the importance of surveillance for malaria elimination, you introduced Coconut malaria surveillance system that is used mainly in Zanzibar, what have been the results of its use thus far?

Gordon Cressman Replied at 5:59 PM, 7 Mar 2017

For more than 3 years, district malaria surveillance officers in Zanzibar, equipped with motorbikes, inexpensive Android tablets, and Coconut Surveillance have helped Zanzibar to sustain its gains against malaria. Since 2012, Coconut Surveillance has helped malaria surveillance officers in Zanzibar respond to more than 9,500 reported cases of malaria, complete nearly 10,500 household visits, test more than 40,000 household members, and identify and treat 2,100 previously unknown cases. Of course, these are only outputs. While we believe that this system has been instrumental in helping ZAMEP to avoid a resurgence, there is no counterfactual to separate the impact from the many other factors at play. Zanzibar has implemented its malaria program across its entire territory.

I have attached a presentation given last Wednesday to the Malaria Branch of the U.S. Centers for Disease Control and Prevention in Atlanta, Georgia. Look at Slide #2 and Slide #3. While the seasonal peaks in malaria transmission appear to smooth out somewhat after the start of RACD, this is very subjective. The fact that there continue to be seasonal peaks suggests that there may be opportunities for improving vector control measures, and perhaps RACD. Possibilities that we've considered include targeting spray campaigns and bed redistribution in perennial hot spots late in the dry season to try to depress the mosquito population further, before the beginning of the wet season. A high proportion of cases are also imported from coastal areas of Mainland Tanzania, where malaria is endemic. We've also considered offering free screening and treatment at ferry ports to help determine the proportion of cases arriving by informal boat traffic. Some research also suggests that it may be more effective to simply treat all residents in a focal area (mass drug administration) rather than expending the time to test and then treat only those who test positive. This also recognizes the limitations of RDTs. Finally, there have been some issues recently with maintenance and fuel for the motorbikes and other procurement issues that have reduced the rate of case follow-up. We are optimistic that this situation will improve soon, but it emphasizes the importance of an effective logistics system to support RACD.

Attached resource:

Pierre Bush, PhD Moderator Replied at 9:08 PM, 7 Mar 2017

Dear Gordon,
Thank you for the response. The slides you used for your presentation at the CDC are very informative and well done. As I can see, the community health workers use motorcycles, and tablets. This requires money for fuel and for maintenance, and internet cost. This brings the question about the affordability. As you mentioned in the previous posting Coconut surveillance system can be used free of charge, and can be used for the surveillance of other diseases. The following is my third question:

 Are there other countries that have used or are using it? What is the cost for implementing the software? Is it affordable for low income countries without foreign aid?

Saad Hashmi Replied at 2:37 AM, 8 Mar 2017

It is quite interesting program.I am thinking if we use this software to prevent outbreak of dangue fever in South East Asia.

Gonzo Manyasi Replied at 7:11 AM, 8 Mar 2017

Thank you Gordon for those very comprehensive answers to both my questions and even Pierre's - I've gained quite a lot from that. Also I managed to get the full article of the problematic link from the abstract page. Thanks Pierre for that questions on the Coconut Surveillance system, especially the last one regarding affordability in the absence of foreign aid. Two questions arising (for Gordon): One, from your statement that you are seeking opportunities to test the adaptation of the Coconut software to other diseases, could I interest you in doing it the western region of Kenya? I would like to be part of it though I would need induction to fully understand the exactly how the surveillance system works - if you don't mind me picking your brains on that one Gordon. Two, when you talk of the possibility of "treat all residents in a focal area (through mass drug administration), is that instead of the traditional prophylaxis method (which I do appreciate was mostly for those coming to such areas)?

Gonzo Manyasi Replied at 7:17 AM, 8 Mar 2017

Please excuse the (apparent grammatical) typos above!! Multitasking isn't male gender-friendly!

Maimunat Alex-Adeomi Moderator Replied at 2:53 PM, 8 Mar 2017

Dear Gordon,

Thank you for this interesting spotlight and your very detailed explanation on the program. I am interest in any challenges you might have faced using the mobile phone technology. Seeing that this is probably dependent on cell phone receptions and often somewhat erratic coverage in certain geographical areas is this something you encountered in your program? If so how did you overcome that to ensure availability and accuracy of data.

Regards
Maimunat

Gordon Cressman Replied at 5:23 PM, 9 Mar 2017

I apologize for being a bit slow to respond to this one, but yesterday and today have been quite busy!

Q: "Are there other countries that have used or are using it? What is the cost for implementing the software? Is it affordable for low income countries without foreign aid?"

A: Coconut Surveillance is currently used only in Zanzibar. We have examined how it might be adapted and integrated into other malaria elimination contexts. These include Zimbabwe, Cambodia, Uganda, and Haiti. In each case we researched the existing malaria pre-elimination or elimination program, malaria case loads, malaria information system, national health information system, mobile network coverage and mobile phone penetration, geography, and other contextual factors.

Costs include healthcare workers (e.g. malaria surveillance officers), transportation (e.g. motorbikes, maintenance, and fuel), mobile devices (Android phones or tablets) and replacement mobile devices, mobile network services (case notification, data synchronization), and server or cloud hosting costs. These costs can vary significantly from one context to another and based on many decisions concerning transportation, types of mobile devices, and other factors. We have a costing model that is a good start at examining these costs. In Zanzibar, annual costs per health care facility (case notification system) are about USD 45, and annual costs for each district malaria surveillance officer for Coconut Surveillance is about USD 510. Those costs do not include salaries or transport costs. Technical support for initial system implementation and continuing technical support will also vary depending on context and implementation decisions. For example, we started with 7-inch Android tablets in Zanzibar, but now nearly every malaria surveillance officer has his/her own Android smart phone. Coconut Surveillance could run on these smart phones. The annual loss/breakage rate for the Android tablets has been between 5% and 8% over the last 4 years. Using existing Android handsets or instituting a rent-to-own system for the devices might reduce this rate.

Q: "Seeing that this is probably dependent on cell phone receptions and often somewhat erratic coverage in certain geographical areas is this something you encountered in your program? If so how did you overcome that to ensure availability and accuracy of data."

A: Coconut Surveillance operates offline. Surveillance officers synchronize the case data when they have a mobile network connection or a connection to a Wi-Fi network. (It is also, at least theoretically, to synchronize case data between two mobile devices. We've done this with previous versions, but not with the latest Coconut Surveillance 2.0 software platform.) In Zanzibar mobile network coverage is fairly good, so surveillance officers are able to synchronize case data frequently, providing near real time updates. However, we have used the same software technology in other country contexts where mobile network coverage was much worse. For example, mobile nurse counsellors in Zimbabwe have been using a mobile medical record system that we developed using the same technology. They are using Android tablets that have Wi-Fi, but not mobile data capability. They work offline during the day, and bring the tablets back to a portable Wi-Fi access point / mobile network router at the end of the day to synchronize the case data. We've seen this same strategy work well on a large scale with Tangerine, a mobile early grade reading assessment tool that we also developed using some of the same technology.

The software includes many features designed to make it as easy as possible for mobile users (e.g. surveillance officers) to enter data accurately. These include automated skip and validation logic. This catches errors in data types, ranges, and logic (e.g. pregnant male) during data entry on the mobile device. The software guides the user through correcting any errors in each question set (form).

I hope these answers help! Please ask more questions!

Rachel Daniels Replied at 5:41 PM, 9 Mar 2017

Hello! I’m curious if you feel that Coconut Surveillance could potentially be able to incorporate genetic signals generated from reference labs in order to inform ministries and decision-makers. Data beyond just drug resistance, such as the appearance of clonal parasite types or imported parasites.

Pierre Bush, PhD Moderator Replied at 10:17 PM, 9 Mar 2017

Hello Gordon,
Thank you. Excellent information. The coconut seems to be affordable. As we are approaching the end of the event I would like to ask you the following question:

 What are the challenges that you face in your work and how can Global Health Online Delivery Malaria community team help, especially in disseminating your research findings?
Thank you.
For our colleagues: please see below the link to Coconut surveillance system.

Attached resource:

Gonzo Manyasi Replied at 4:40 AM, 10 Mar 2017

I am extremely grateful to Pierre for introducing Gordon to us. It has been
an exciting and useful discussion that has provided further insight into
the use of mobile technology in various disease situations. I intend to
follow up on a few issues sooner or later.

Gordon Cressman Replied at 9:34 AM, 10 Mar 2017

Q: "I’m curious if you feel that Coconut Surveillance could potentially be able to incorporate genetic signals generated from reference labs in order to inform ministries and decision-makers. Data beyond just drug resistance, such as the appearance of clonal parasite types or imported parasites."

A: Yes, this is possible, but it has not been done. Data from reference labs could be integrated into individual case records electronically. We have established similar links between labs and case record systems in several countries. The technologies and standards used depend on what technical resources exist or are expected to exist at each lab. For example, in Zambia we linked a national lab to the Zambian Electronic Perinatal Record System (ZEPRS) via HL7 middleware, such that CD4 counts were automatically transferred into electronic medical records. Authorized clinicians were alerted to new lab results and had immediate access to them. The key is writing the case ID on the sample container to link the sample, lab data, and record in Coconut Surveillance. This link would have great value in linking microscopy and RDT results, and genetic data for the purposes you mention above.

Q: "What are the challenges that you face in your work and how can Global Health Online Delivery Malaria community team help, especially in disseminating your research findings?"

A: The Global Health Online Delivery Malaria team can help by disseminating this information among your colleagues, by helping us to prioritize enhancements (such as Rachel's question), and by looking for appropriate opportunities to test Coconut Surveillance in other malaria elimination contexts. We are continuing to look for scientific and funding partners to continue to disseminate and improve this tool. We also welcome organizations who can contribute to the software development. All of the software source code is hosted in public GitHub repositories, so that any organization can use and contribute its development. We are also preparing a new Coconut Surveillance website, community discussion forum, and documentation site, and hope to release these soon.

-gmc

Gordon Cressman Replied at 10:25 AM, 10 Mar 2017

Q: "...could I interest you in doing it the western region of Kenya? I would like to be part of it though I would need induction to fully understand the exactly how the surveillance system works - if you don't mind me picking your brains on that one Gordon."

A: According to the President's Malaria Initiative (PMI), Country Operational Plan for Kenya, 2017, malaria prevalence in western Kenya ranges from 5%-20%. This area is in the Highland and Epidemic Prone malaria risk stratification zone, with an estimated 39% of Kenya's population. Malaria transmission in this area is seasonal with great variation from year to year. I wonder whether transmission rates and case loads are low enough in at least a portion of this area to make individual case tracking and follow-up feasible. We should also consider whether it might be useful to test this with a subset of health care facilities as sentinel sites. The National Malaria Control Program receives significant support from PMI and the Global Fund, and Kenya is a PMI Focal Country. The national health information system is based on DHIS 2. The national e-ISDR weekly and notifiable disease reporting system uses simple mobile phone technology, and interoperates with DHIS 2. Dengue and Yellow Fever are notifiable diseases. Malaria is not. Malaria could be added as a notifiable disease in healthcare facilities that are part of an RACD pilot. This would provide new case notifications to Coconut Surveillance, and Coconut Surveillance could synchronize case data with DHIS 2. I would be interested in exploring this, cognizant of existing stakeholders and systems.

Q: "Two, when you talk of the possibility of 'treat all residents in a focal area (through mass drug administration)', is that instead of the traditional prophylaxis method (which I do appreciate was mostly for those coming to such areas)?"

A: Treatment of all residents in a focal area, which in most of Africa would be done using a form of artemisinin combination therapy, might be done to reduce the parasite reservoir. This approach eliminates the resources needed to test residents, and recognizes that rapid diagnostic tests are not always accurate and are not always interpreted accurately. Residents in these areas, at least in Zanzibar, do not take a prophylaxis for reasons that include cost, availability, and side effects from long-term use.

-gmc

Pierre Bush, PhD Moderator Replied at 12:10 PM, 10 Mar 2017

Hello Gordon,
Thank you very much for your outstanding performance and contribution during this event. The community members benefited from your experience, and expertise. As you rightly mentioned Coconut surveillance tool is in the public domain, so that it can be used free of charge, and improved where needed. We are very grateful for making yourself available and answer to the community's questions. I invite you to post any closing remarks and keep on answering any question that will be asked related to this event. Today is the last day of this event.
Once again, thank very much.

Sungano Mharakurwa Moderator Replied at 3:28 PM, 10 Mar 2017

Thank you very much Pierre for hosting Gordon. This has been such an informative and exciting discussion.

Gordon Cressman Replied at 5:40 PM, 10 Mar 2017

Dear Pierre,

Thank you very much for inviting me to be in this Member Spotlight. It has given me a great opportunity to meet some community members, and to respond to some great questions and suggestions. I have greatly enjoyed the chance to share more about our work, and to think carefully about the questions raised by community to members.

I look forward to the continued great discussion on GHDonline. I encourage anyone who wants to learn more about Coconut Surveillance and to discuss how we might work together to introduce it into other malaria elimination contexts to contact me directly.

Best Regards to you All,

-gmc

Maimunat Alex-Adeomi Moderator Replied at 11:12 AM, 11 Mar 2017

Hello all,

Thank you Gordon for this very informative and enlightening spotlight.

Thank you Pierre for introducing Gordon to us.

You both did a phenomenal job.