Three Essential Questions for QA Leaders to Ask about Automated Healthcare Software Testing
The future of digital healthcare innovation is tightly coupled with the evolution of software testing.
The various systems through which patient data flows – electronic medical records (EMR) systems, lab software, patient portals, and connected medical devices – are complex and unique, creating a complex mesh of possible test scenarios, often far too many for even an army of manual testers to cover.
In the meantime, quality assurance (QA) leaders are grappling with the rising demand for at-home care, interoperability, and generative AI adoption in healthcare. To shed light on the path ahead, we recently collaborated with HIMSS on a survey aimed at helping healthcare software testing professionals prepare for the future.
Two-thirds of respondents revealed uncertainty in scaling current testing practices to meet future requirements. Consequently, 75% of them are inclined to invest in or increase their investment in automation to navigate the “perfect storm” of technological, regulatory, and cultural changes.
QA professionals can play a crucial role in helping their IT organizations move forward by asking the right questions. In this article, we present three essential questions that you should consider posing to your leaders. If you're interested in learning more, join our "Vital Signs: State of Healthcare Software Testing" webinar on August 17 at 1 pm CT. You'll hear from industry experts from Ernst & Young, Keysight, and HIMSS as they delve deeper into this important issue.
How will test automation impact patient safety in the digital age?
The adoption of test automation in healthcare requires a thorough understanding of how this technology will impact patient care. The U.S. healthcare industry has made significant strides in digitalization, with more than 90% of U.S. healthcare providers adopting EMR systems by 2021. These systems have become the cornerstone of modern health systems, but each EMR update can introduce bugs or vulnerabilities that affect interconnected systems. Manually identifying these changes with each update is both costly and inefficient. Alarmingly, our research suggests reveals that only 6% healthcare executive leaders feel confident that current software test coverage is sufficient to minimize patient risks.
Consequently, the adoption of automated software testing is rapidly gaining momentum in the healthcare industry. Over the next five years, the percentage of U.S. providers planning to automate software testing is projected to increase from 25% to 75%.
Figure 1. A five-year projection on the percentage of U.S. providers using automated software testing from 2023 to 2028
Healthcare IT leaders must seize immediate opportunities to close existing EMR testing gaps. Failure to do so may result in patient risks and falling behind competitors who are shortening release cycles with automated processes. In addition, IT leaders must consider the ever-expanding digital healthcare landscape. How can the QA team scale to support emerging IoT applications and meet the regulatory requirements? How to move beyond functional testing to improve digital experience?
Is our testing strategy keeping up with the growing impact of AI?
Assessing the new frontiers opened by AI-driven software testing will undoubtedly ignite a desire for capitalizing on its value. However, AI raises concerns related to patient data privacy and security. Testing managers must strike a balance between the level of autonomy given to AI in executing tests and their control over test selection.
Several testing tools have been harnessing the power of AI. Broadly speaking, they divide into four categories:
AI in test case selection
Tools in this category leverage AI to analyze the existing repository of manual and automated test cases, determining which ones to include for a specific release. While this use of AI enhances the workflow efficiency, it still requires significant human input.
AI in test case execution
Tools use AI to survey data and determine which automated scripts to execute. If any scripts fail due to changes in the application under test or within the script itself, the AI updates and repairs them. In this case, the efficacy of AI heavily relies on test data quality.
AI in test case selection and execution
These tools combine the first two approaches, allowing AI to select and execute test cases, reducing the need for human intervention. However, there is an increased risk of low-quality data or incorrect fixes.
AI in automation intelligence
The fourth type of tool, known as Automation Intelligence, leverages AI in a more comprehensive manner. It incorporates inputs from various algorithms to determine the areas of the application under test. These inputs can include factors such as past bug history, application changes, or coverage history. AI then builds, executes, and reports on the test cases.
The bottom line is that AI-driven software testing must always be subject to the effective oversight of those using it. Support for the effort can come from industry regulatory frameworks and it will be important for healthcare QA leaders to keep abreast of these.
Do we have the necessary capabilities?
To keep pace with the rapid digital healthcare transformation, providers may need to review their organizational capabilities on two fronts.
Technology
A modern software testing stack is critical for ensuring patient safety in the era of rapid digital transformation. Currently, the healthcare software ecosystem, including EMR systems, lab software, patient portals, and connected medical devices, generates approximately 30% of the world's data, according to IDC. With the wider adoption of connected medical devices, this data will continue to grow exponentially. Health systems that have used homegrown automated testing solutions will face a laborious process to maintain multiple test automation frameworks. Our survey results found only 38% of respondents were content with their organization's scalability to meet emerging requirements - a satisfaction rate 10 percentage points lower than that of organizations using manual testing. Does the management team have a clear understanding of the QA resources and tools required? Do they have a strategy in place for acquiring them?
Talent
The introduction of test automation, like any major change, necessitates a reassessment of the organization's talent. According to a recent Forrest survey, nearly 50% of enterprises lack in-house expertise to implement or maintain an automated testing platform.
The specific skills required will vary depending on the use case. For relatively straightforward test cases with automation platforms featuring low-code capabilities, a nurse who is familiar with the clinical process can automate tests with support from a QA engineer. However, for more complex test cases or homegrown tools requiring custom programming, the company may need to hire experts in test automation frameworks.
Conclusion
Software testing plays a crucial role in ensuring positive patient outcomes, particularly in the rapidly evolving landscape of digital healthcare. It is the responsibility of QA professionals to effectively educate management teams on the need to modernize testing approaches to keep the organization at the forefront of this latest technological development while minimizing risks to care outcomes.
The questions presented in this article are not exhaustive, and as technology progresses, new questions will arise. However, they serve as an excellent starting point for discussions and considerations. For a deep diver, we invite you to register for the webinar titled "Vital Signs: State of Healthcare Software Testing" on August 17 at 1 pm CT. During this webinar, esteemed speakers, including Aloha McBride, Global Health Leader at Ernst & Young, Dave Hester, Tech Consultant at Keysight, and Patty Enrado, Director of Content at HIMSS, will provide further insights into this critical topic.