Part III-Your Work is Never Finished: Why Iterating after You Launch Should Be Part of Your Plan

~Written by Lauren Spigel, Monitoring and Evaluation Coordinator (Contact: lauren.spigel@vaxtrac.com; Twitter: @vaxtrac)

Also published on VaxTrac blog

If you missed our last three blog posts in this series on human centered design, you can learn about what human centered design is here, read a case example of how we build empathy with health works in Nepal, and see how we’ve used prototyping to test a new monitoring and evaluation dashboard in Benin.

Our final post is going to explore the concept that your project is never complete; even after you launch, it’s important to continue to get feedback. In this post we’ll share an example of how we’ve iterated on our software based on feedback from health workers and ministry officials in Benin.

View fullsize

A health worker sorts through paper records. Source: VaxTrac.com

The Problem
The World Health Organization (WHO) and the Benin Ministry of Health issue a country-wide immunization schedule that recommends when children should receive their immunizations, beginning from the moment they are born and lasting through the first year of life. Health workers in urban clinics manage hundreds of children’s schedules using paper records. Keeping track of which children are due for which vaccine during any given vaccination session is a time-consuming task.

Parents of these children lead busy lives and often live far from the clinic. It takes hours out of their day to bring their child to vaccination sessions. While parents value vaccines, health workers don’t always communicate clearly to parents about when to come back to the clinic for the child’s next vaccination. A direct consequence of this is that children often miss their appointments.


Our Solution: Callback List 1.0
Clinics that use VaxTrac to record childhood immunization data have an advantage: as long as children are registered in the VaxTrac system, the system can automatically generate a list of children that are due for upcoming appointments. This can save health workers from several hours of paperwork each week.

Our team of software engineers saw this as an opportunity and developed a basic callback list. The first version of the callback list pulled a list of children that were due for an upcoming appointment along with basic information, such as date of birth, village and contact information. But health workers weren’t using it. We wanted to know why.

Back to the Drawing Board: Stay Responsive to User Needs through Iteration
Technology is meant to change over time. VaxTrac’s software engineers like to remind our team that the software is never finished; it’s constantly evolving and adapting to user needs.

When we began developing our mobile (Android)-based system, we brought health workers together to get their feedback on what they wanted from a callback list feature. We used a number of human centered design methods to elicit feedback, such as prototyping [link to prototype blog], brainstorming lists of what they do during vaccination sessions, and breaking into groups to sketch out what they wanted the callback list to look like.

View fullsize

Meredith leads a focus group with health workers. Source: VaxTrac

We asked them to create the callback list over again from scratch. We asked guided questions: Is any of the information available on the callback list useful? If so, what is it used for? What other information should it include? How would they like to see the callback list organized? What rules should the callback list follow? How long should a child stay on the list?

Having health workers sketch their answers to these questions helped all of us think through these abstract questions together.

We found key insights:

• Health workers were using the callback list, but not in the way that we had originally intended. Instead of using it to contact the parents of children who were due for upcoming vaccinations, they used it to track down children who had missed an appointment.
• Phone numbers change often so we needed to make it easier for them to update parents’ contact information.
• In order to be more useful, the callback list would need to be interactive, allowing health workers to sort the information in a variety of ways.

Hearing the health workers’ perspectives helped us rethink the purpose of the callback list and how to redesign it.

Build, Do, Learn, Repeat: VaxTrac’s Philosophy on Iteration
Build: Our software engineers took our learnings from the user feedback session and went to work on building a new and improved callback list. In addition to the callback list, we created a defaulter list, providing health workers with a list of patients that have missed their appointment. We also made both lists sortable by any category (village, date of birth, sex, date of appointment, etc.) And lastly, we made it easier to update contact information.

View fullsize

Callback List. Source: VaxTract

Do: After our software engineers updated the callback and defaulter lists, we made sure health workers received adequate training on how to use it. Our Benin-based team visited clinics for additional training.

View fullsize

VaxTrac staff training health workers. Source: VaxTrac

Learn: After a few months of using the new callback list, we held a focus group with health workers to learn more about what they thought of the different VaxTrac features, including the callback list. We learned that health workers would like to be able to sort by the mother’s name in addition to sorting by the other categories. We also learned that health workers would like a way for the callback list to help them contact parents of children who are due for upcoming vaccination sessions.

Repeat: Each time we add new features and users, we get new perspectives. All of the feedback that we’ve gotten from health workers have helped us make our callback list and defaulter list more user-friendly. But we’re not done yet! We are currently conducting a study to assess the possibility of incorporating an appointment reminder feature to the callback and defaults lists, so health workers can use the system to contact parents directly, possibly through SMS or Interactive Voice Response (IVR).

Once we learn everything we can, our cycle will repeat again.

The more we iterate, the stronger our product becomes because it’s based on feedback from the people who use our system. While it can be daunting to go back to the drawing board, adding iteration into your project plan from the beginning can save you time and resources down the line.

We’ve embraced the philosophy that our software is never finished. And we can’t wait to see how far it’ll take us.

________________________________________
To learn more about incorporating design thinking into your projects, contact Lauren at lauren.spigel@vaxtrac.com or check out IDEO’s resources[link: http://www.designkit.org/resources/1.

How to Incorporate Design Thinking into your ICT4D Projects: A Blog Series

~Written by Lauren Spigel, Monitoring and Evaluation Coordinator (Contact: lauren.spigel@vaxtrac.com; Twitter: @vaxtrac)

Also published on the VaxTrac blog

View fullsize

2015 was a year abuzz with talk about human centered design in the international development community. Words like “human centered” and “design thinking” may sound like international development buzzwords on par with “sustainability” and “capacity building” (thanks Devex), but behind the words are tangible methods you can use to elicit feedback from the people your project serves and use their insights to build better, smarter solutions.

If we break down the jargon, human centered design means to listen to the people that you are building a program for – before, during and after implementation. You listen to them because they are the experts. If you are designing an app for health workers to register patient data, you listen to health workers because they are experts on their workflow and needs. If you are designing an educational SMS system for youth, youth are experts in understanding what they want to know and how they want to discover that information. Simply put, human centered design is a series of methods implementers can use to engage with users throughout a project’s lifespan.

The human centered design process walks us through methods we can use during three key phases

1. Inspiration: defining the problem, the audience, understanding facilitators and barriers
2. Ideation: brainstorming ideas, testing out prototypes, finding the best solution
3. Implementation: choosing the best idea and implementing it, while still getting feedback and iterating

While human centered design is typically thought about in terms of technology projects, in recent years, the concept has been applied more broadly to solve complex global health challenges. In an interview with WIRED magazine, Melinda Gates, Co-Chair and Trustee of the Bill and Melinda Gates Foundation, described human centered design as “meeting people where they are and really taking their needs and feedback into account.” It sounds intuitive that we would want to understand the needs of the people we serve, but we often lack the tools and resources to do this well.

We are writing a three-part blog series to share our own experiences designing a mobile vaccine registry system for health workers. Our blog series will give case examples of the methods we used at various stages of the human design process. The first post will focus on how we are building empathy with health workers in Nepal to improve our user interface and workflow. The second post will look at how we’ve been prototyping a monitoring and evaluation dashboard with our team in Benin. Lastly, our final post will emphasize the importance of iterating after implementation by sharing our experience customizing our software, based on user feedback in Benin and Nepal.
We hope you will be able to use our experience to incorporate human centered design into your own projects.

To learn more about incorporating design thinking into your projects, contact Lauren at lauren.spigel@vaxtrac.com or check out IDEO’s resources

Technology is Changing the Way Infectious Diseases are Tracked

~Written by Theresa Majeski (Contact: theresa.majeski@gmail.com; Twitter: @theresamajeski)

Technology is progressively becoming a bigger part of our lives. This holds true in high-income countries and in low- and middle-income countries. By 2012, three quarters of the world’s population had gained access to mobile phones, pushing mobile communications to a new level. Of the over 6 billion mobile subscriptions in use worldwide in 2012, 5 billion of them were in developing countries. The Pew Research Center’s Spring 2014 Global Attitudes survey indicated that 84% of people owned a mobile phone in the 32 emerging and developing nations polled. Internet access is also increasing in low- and middle-income countries. The 2014 Pew Research Center survey indicated that the Internet was at least occasionally used by a median of 44% of people living in the polled countries.

The increase in Internet and mobile phone access has significant implications for how infectious diseases can be better tracked around the world. Although robust and validated traditional methods of data collection rely on established sources like governments, hospitals, environmental, or census data and thus suffer from limitations such as latency, high cost and financial barriers to care. An example of a traditional infectious disease data collection method is the US Centers for Disease Control and Prevention’s (CDC) influenza-like illness (ILI) surveillance system. This system has been the primary method of measuring national influenza activity for decades but suffers from limitations such as differences in laboratory practices, and patient populations seen by different providers, making straightforward comparisons between regions challenging. On an international scale, the WHO receives infectious disease reports from its technical institutions and organizations. However, these data are limited to areas within the WHO’s reach and may not capture outbreaks until they reach a large enough scale.

View fullsize

Figure 1. CDC Flu View Interactive dashboard: http://gis.cdc.gov/grasp/fluview/fluportaldashboard.html

Compared to traditional global infectious diseases data collection methods, crowdsourcing data allows researchers to gather data in near real-time, as individuals are diagnosed or even before diagnosis in some instances. Furthermore, getting individuals involved in infectious disease reporting helps people become more aware of and involved in their own health. Crowdsourcing infectious disease data provides previously hard to gather information about disease dynamics such as contact patterns and the impact of the social environment. Crowd-sourced data does have some limitations, including data validation and low specificity.

Internet-based applications have resulted in new crowd-sourced infectious disease tracking websites. One example is HealthMap. HealthMap is a freely available website (and mobile app) developed by Boston Children’s Hospital which brings together informal online sources of infectious disease monitoring and surveillance. HealthMap crowd-sources data from libraries, governments, international travelers, online news aggregators, eyewitness reports, expert-curated discussions, and validated official reports to generate a comprehensive worldwide view of global infectious diseases. With HealthMap you can get a worldwide view of what is happening and also sort by twelve disease categories to see what is happening within your local area. 

View fullsize

Figure 2. HealthMap. http://www.healthmap.org/en/

Another crowd-sourced infectious disease tracking platform was Google’s Flu Trends, and also their Dengue Trends. Google was using search pattern data to estimate incidence of influenza and dengue in various parts of the world. Google’s Flu Trends was designed to be a syndromic influenza surveillance system acting complementary to established methods, such as CDC’s surveillance. Google shut down Flu Trends after 2014 due to various concerns about the validity of the data. As an initial venture into using big data to predict infectious diseases, Flu (and Dengue) Trends have provided information that researchers can use to improve future big data efforts. 

With the increase of mobile phone access around the world, organizations have started using short message service (SMS), also known as text messaging, as a method of infectious disease reporting and surveillance. Text messaging can be used for infectious disease reporting and surveillance in emergency situations where regular communication channels may have been disrupted. After a 2009 earthquake in Sichuan province, China, regular public health communication channels were damaged. The Chinese Center for Disease Control and Prevention distributed solar powered mobile phones to local health-care agencies in affected areas. The phones were pre-loaded with necessary software and one week after delivery, the number of reports being filed returned to pre-earthquake levels. Mobile phone reporting accounted for as much as 52.9% of total cases reported in the affected areas during about a two-month time period after the earthquake. 

Text message infectious disease reporting and surveillance is also useful in non-emergency settings. In many malaria-endemic areas of Africa, health system infrastructure is poor which results in a communication gap between health services managers, health care workers, and patients. With the rapid expansion and affordability of mobile phone services, using text-messaging systems can improve malaria control. Text messages containing surveillance information, supply tracking information and information on patients’ proper use of antimalarial medications can be sent from malaria control managers out in the field to health system managers. Text messaging can also be sent by health workers to patients to remind them of medication adherence and for post-treatment review. Many text message based interventions exist, but there is a current lack of peer-reviewed studies to determine the true efficacy of text message based intervention programs.

Increasing global access to the Internet and mobile phones is changing the way infectious diseases are reported and how surveillance is conducted. Moving towards crowd-sourced infectious disease reporting allows for a wider geographical reach to underserved populations that may encounter outbreaks, which go undetected for a delayed period. While crowdsourcing such data does have limitations, more companies than ever are working on using big data and crowd-sourced data in a reliable way to inform the world about the presence of infectious diseases.

Empowerment is Key to Improving Health Infrastructure in Developing Countries

~Written by Kathleen Lee, MPH Epidemiology, Vanderbilt University Medical Center (Contact: kathleen.g.lee@vanderbilt.edu)

Providing greater health access and more efficient health care delivery, especially for vulnerable populations, are priorities for anyone involved in public health. Poor health systems in developing countries mean a shortage of trained health care workers, inconsistent inventory of medical supplies, and inadequate surveillance systems. This list is not exhaustive, but we can start here. Building a better health infrastructure, like many public health priorities, requires multi-level coordination. Empowerment has to spread out from the government to the community and to the individual.

We can address the problem first by tackling the shortage of health care workers. Doctors in developing countries are in critically short supply. In 2006, the World Health Organization compiled data on the impact of HIV/AIDS on the health workforce in developing countries. Results showed that while European and North American countries have doctors at a ratio of 160 to 560 per 100,000 people, African countries only have two to sixty doctors for every 100,000. In Malawi, for example, there is one doctor for every 50,000 people. The global shortage of trained hospital and health care staff currently exceeds four million. Training more staff and volunteers is one solution for improving health systems in developing countries. Training other previously unqualified individuals could ameliorate these shortages. Providing incentives for already trained workers to stay in a vulnerable state or country could help build a struggling health system. Having a foundation of trained workers and preventing them from migrating to wealthier countries is an important first step. Empowerment and opportunities to grow and help are at the heart of this strategy.

The second hurdle is maintaining a constant inventory of equipment, medicines, and other health supplies. War, along with political and social unrest, in certain regions further dampens the efforts to provide a steady supply chain. There has to be cooperation between donors and the government to work with the private sector to ensure receipt of necessary health supplies. Partnering with emerging pharmacy chains increases the availability of medicines and drives down the cost for the patients. In the Philippines, Generics Pharmacy has thousands of small storefronts that are widely used by both the rich and poor. Convenience and ease of access are often of paramount interest to every person, regardless of income. The issue of payment is another facet of the supply and demand problem. Corruption that trickles to the local governments, and even the health care workers themselves, leads to some patients having to pay for medicine or services that should have been free. Reforming payment systems to ensure that patients have the medicines delivered before payment is processed directly to the provider will empower the patients and promote compliance. 

Compounding the shortage problems, both of trained workers and supplies, are the inadequate surveillance systems in place. This is the third issue that needs to be addressed, and it is arguably the most crucial. Surveillance is necessary to monitor not only the needs within health facilities, but also within the community and surrounding areas. Without real-time tracking of disease and medical supplies, logisticians, doctors and community health workers are unable to properly estimate need and completely evaluate the effectiveness of their clinic’s efforts. This is where data comes into play. The Novartis Malaria Initiative, under the Roll Back Malaria Partnership, has led SMS for Life, which harnesses mobile phones, internet, and electronic mapping technology to track stock levels for health facilities. Sending SMS messages between health facilities and the district medical officers ensures treatment availability. Stock-outs have been reduced in Tanzania during a six-month pilot program from 79% to 26% in three districts. Not only are these stock-outs reduced, but when they occur, they are also resolved much quicker due to the ease of communication. In areas where internet is unavailable or running inconsistently, Relief Watch has offered a similar solution. It also uses mobile technology, but the application allows workers to not only track supplies but also disease (http://www.reliefwatch.com/). The easy and free setup is invaluable to developing countries that have previously relied on paper spreadsheets and forms. Giving workers data at their fingertips gives them more control over their health facility and their patients. These technological innovations are not only crucial for immediate supply tracking and disease surveillance, but they provide research institutions and governing bodies more accurate data. After all, it is data that public health professionals and policy-makers rely on to make decisions and plan strategies. 

The aforementioned plan to improve health systems is by no means novel. Public health practitioners have stressed the importance of training more workers, creating a steady supply chain of treatments, and addressing surveillance shortcomings for decades. Adhering to these solutions requires cooperation and active coordination that extend from the public to the private sector. This is something that cannot be over-emphasized. Empowerment—of individuals, community health workers, and governing bodies of fragile states—is an important foundation from which a better health infrastructure can grow.


Resources:

The impact of HIV/AIDS on the health workforce in developing countries http://www.who.int/hrh/documents/Impact_of_HIV.pdf
Healthcare logistics: delivering medicines to where they're needed most
http://www.theguardian.com/global-development-professionals-network/2013/jul/29/healthcare-logistics-best-practice
SMS for Life http://www.malaria.novartis.com/innovation/sms-for-life/
Relief Watch http://www.reliefwatch.com/
Avert: Universal access to HIV treatment http://www.avert.org/universal-access-hiv-treatment.htm