Agile Software Development in Healthcare: Practices Applied to Industry-Specific Projects

The Agile methodology that provides the ability to quickly adapt to changes is adopted by many companies across different industries. First introduced in software development, Agile has moved beyond tech to project management, marketing, and other fields. While studies reveal significant improvements in productivity and quality that Agile brings, the healthcare industry is no strong supporter of this approach.

There’s nothing surprising about that. Agile is touted as the startup methodology most suitable for innovative projects that might not have a string course to follow but rather need the chance to correct the course any moment if necessary. Sceptics may joke that Agile teams might not even know what they are going to solve while working. 

In healthcare, with its tedious regulations and the demand for very specific and predictable solutions, software developers mostly rely on the traditional Waterfall approach with fixed, sequential steps and considerable upfront planning.

However, software solutions continue to fail and turn out outdated, defective, or unusable the moment they are released. Maybe, going Agile can be as much helpful for complex healthcare systems as it has proven to be in other industries? 

At MadAppGang, we’ve been developing mobile applications in line with Agile and know the ins and outs of its methods. Based on our experience, we can guide through the scope of Agile software development for healthcare.

In this article, we will examine how the Agile methodology is applied in software healthcare and what improvements it brings.

Developing software, it’s hard to predict how users will interact with it, regardless of how much upfront planning you do. User needs are evolving, in healthcare as well, and Agile seems to fit here: its ‘test and learn’ attitude drives the process of making a functional version of software from the earliest stage of development. But if in other industries, you can market a half-ready, very light-featured product and only after receiving user feedback, make changes and add new features, you can’t market an unfinished solution in healthcare. 

This industry deals with the most serious matter there could be – human health, so in most cases, there’s no room for failing and learning through errors. A healthcare solution needs to work the way it was expected and perform all the functions from the very beginning. It requires compliance with laws and regulations, top-notch security, and stable performance. 

This is not an area where you can release a poor-quality product and continuously improve it – you need your product to be trusted and used from day one. Of course, you can and should make improvements after the release. But when the product doesn’t perform its function from the beginning, it is removed from the market and causes reputational damage for the company.

Nevertheless, the iterative nature of Agile that helps make the product usable from the early stages can be applied to the benefit of healthcare projects. And here is why:

Slow technology adoption, resistance to transformations, and the prevalence of software solutions that rather confuse than help are among the most acute problems of healthcare. With regulatory bodies and volumes of information being involved in the development, the process is usually extremely time-consuming. 

If there are some changes made to regulatory requirements during the development, companies that use Waterfall are obliged to go several steps back and redo some work. If there is an error in processing information identified at the Waterfall’s testing step, developers should shovel up everything they’ve done to fix the problem. Actions like these can take a while and result in the end product that is unusable and fails to solve the problem it was intended to solve. And when the software is unusable, it brings huge operational losses.

But with Agile, the whole team is always up to date with the process which is divided into small, manageable parts. Organizations questioned for a Forbes Insights survey admitted that the application of Agile principles helped them achieve faster time to market, quicker innovation, and improved product quality.

There already are numerous examples of successful implementation of Agile in healthcare. Mayo Clinic has improved the development time and customisation of software thanks to building an Agile IT team. The development of the UTSW Health System’s clinical decision support software has experienced an increase in productivity, reduction in defects reported, and decrease of the need of after-hours work since the introduction of an Agile team. Some decade-old stories of Agile transformation like the one of GE Healthcare still remain absolutely relevant.

Because of its test-driven environment and flexibility, Agile is being adopted outside software as well. More companies consider changing their workflow and core principles to become more flexible in the face of rapid change. The trend of Agile project management in healthcare is also increasing. For example, Agile Psychological Medicine Clinics in Australia put interactions over processes, valuing the collaboration between clinicians and patients and making the care more human-centric.

How Agile principles can be applied to the development of large-scale, complex systems in healthcare? Computech CTO Filippo Morelli argues that the Unified Process can help navigate challenging projects. This is a practice of setting four phases for Agile activities:

• Inception: the team identifies a potential solution and key risks
• Elaboration: the team builds initial software
• Construction: initial software is enhanced through iterations
• Transition: target users validate the product

Let’s get into the details of how particular projects can fit into this logic, how Agile methods are used in the industry, and why Agile is important in healthcare software development.

We will illustrate the major Agile practices and the results they brought to health-related projects.

Iterations

The plan of development is divided into small cycles called iterations. The most representative form of Agile’s iterative planning is Scrum that uses the term ‘sprint’ which, in software development, means week- or two-week-long cycles. The team has to execute a usable version of the product after each sprint. Each development stage is reviewed by a product owner or other person with relevant responsibilities: this is done to receive regular feedback, validate a working product, and request changes if necessary.

Teams that meticulously adopt Scrum introduce a particular role of the Scrum Master. In a book on Scrum, Jeff Sutherland explains that the Scrum Master stimulates the team to review their progress after each sprint and ask themselves what they can change and what is the biggest sticking point. Honest answers to these questions boost productivity and speed up the process.

The iterative environment maximises the chances for a fully-featured end product to satisfy a client. Since the earlier problems are detected, the easier and less expensive is to fix them, regular short-term assessments give a valuable benefit to development teams.

Jonathan Bush, CEO of Athenahealth that makes cloud solutions for healthcare providers, finds the use of outdated software one of the biggest problems to fix in healthcare. He’s learned that to craft a transformational solution, he needs to transform the internal approach as well, and transitioning to Agile was that very step that enhanced Athenahealth’s development cycle. Bush claims that weekly reviews of tactical plans allowed them to offer demonstrable progress to clients and as a result, build trust among the clients and attract new ones to the network.

A healthcare company Aetna made their marketing team go Agile and successfully optimised their spend across different media placements. It’s because of the iterative planning they achieved such results: bringing all members into regular communication enabled them to quickly test their ideas and learn from the success of their implementation.

User Stories

Before each sprint, the team gathers to plan the activities needed for this development cycle. To do so, they define how the features they are going to implement should work and set the requirements seen from a user’s perspective. These requirements are called user stories. They are prioritised by the product owner and kept in the backlog.

User stories should be:

• Independent: represent a particular, separate action
• Negotiable: leave space for discussion
• Valuable: have a distinguishable value
• Estimable: allow defining their size and priority level 
• Small: have to be scheduled into a single iteration
• Testable: designed in a way that tests can be defined for them

A study of Agile clinical decision support tool development showed the decrease in the compliance defect rate and the Clinical Disease Activity Index – achievements derived from measurable user stories written from the perspective of the clinician or patient user.

An example of a Homecare system made by Allscripts and SoftServe also proves that Agile development encourages collaboration and gives flexibility in reaction to changes. The product owner was involved in continuous communication with the development team and the Client Focus Group helped prioritise user stories.

Rapid Prototyping

Since the principles of working software and response to change are among the top priorities in Agile, the product needs to be tested in a practical environment and adjusted accordingly to end users’ feedback. Rapid prototyping helps execute an MVP faster than the traditional approach that puts following a plan over regular testing. Check out the story of how our MadAppGang’s team built an app in the record six weeks.

The agency IE Digital explained how their work for Public Health England, aimed to stimulate health apps usage activity, benefited from Agile. After the discovery phase, the agency tested the prototype on end users and received feedback on the general workflow and specific aspects. They also demonstrated their product at relevant conferences to get a broader picture of how it can function in the industry.

A Bain survey of healthcare executives revealed that the majority accept the need to be more Agile but the minority has agreed on the use of Agile methodologies in their teams. By far, one of the biggest obstacles to adopting Agile in a medical project is a failure to actually use its core principles. Sometimes, companies use the language of Agile innovation but don’t change the way they work, which is often described as ‘bad Agile.’

The failure of Healthcare.gov highlights two important aspects related to Agile in healthcare: it is an example of bad Agile and a complex system characteristic of the industry. Here is what happened: the government-backed project had to release a website for patients to compare and choose affordable care plans. On the day of the launch, website demand turned out to be five times more than expected and Healthcare.gov crashed within two hours. 

The project was developed in a traditional Waterfall manner and then transitioned to Agile. But the process didn’t become truly Agile: it missed the major principle of ‘working software’ and failed to coordinate the team work and test the product on users. An Agile version of Healthcare.gov might have released the website with one core component and add new ones after learning how users interact with it. 

The need for doing some work on cross-team projects with non-agile teams is also a big obstacle for healthcare. But the solution lies in the problem itself: since Agile puts a strong focus on team collaboration and makes everyone involved and aware of what’s going on, it’s a perfect methodology for navigating cross-functional teams.

Another significant obstacle to adopting Agile and generally a big challenge for healthcare software development is regulatory compliance. Governments are involved in regulation and specific organisations provide guidance and control over the development to make sure the end product is safe.

Healthcare regulations shouldn’t prevent developers from adopting Agile – on the contrary, Agile’s focus on collaboration across teams is what can boost effective compliance management. Some regulations, for example, of secure messaging in healthcare, can be incorporated into a plan as non-functional requirements written as specific user stories.

Regulatory bodies like the Medical Devices Directive or the Federal Drug Association can guide the development but never enforce to use a specific development cycle. Agile practices are added to the list of the approved standards by many regulators. What you need to do to narrow down the scope of regulations applied to your particular project is to define if the software can be categorised as a medical device and if it use health information that requires certain protection by law.

No matter how conservative healthcare is or how heavy are the regulations imposed on software, the main obstacle to Agile development is the inability to change organisational culture. Companies should learn and understand what the Agile methodology means for their particular projects and adopt it thoroughly: involve every stakeholder and team member in communication, provide iterative planning, and so on.

With the Agile approach, you can navigate a project where not all of the requirements are known from the beginning and where there is a need to have a usable program in a short time. If you have a team and want to transition it to Agile, communicate it through to make sure everybody’s ready for the level of involvement Agile requires. If you’re searching for an Agile development team or have some questions about how it can work, drop us a line and we’ll be happy to help you with your idea.

19 October 2020