How to be a Digital Pioneer in Cancer Diagnostics

In my last blog, I called for pathologists and hospital administrators to start working together, right now, to create a vision to integrate digital pathology (DP) and artificial intelligence (AI) into pathology services. Why the urgency? Firstly, as outlined in the Future of Pathology report, cancer numbers and requests for pathology testing are both on the rise.^1,2 Numbers of pathologists, meanwhile, are falling.³ The introduction by the UK’s Department of Health of a new faster diagnostic standard* requires efficient testing turnaround times.⁴ The second reason is that we are starting to garner enough experience of DP to understand the potential of AI. We need to take advantage of what we know to decide what our pathology services will look like in the future. This is all very well, you may say, but how do we go about this? We can see the bigger picture, but that may not be enough to support healthcare decision-makers who need more detail as part of a shorter-term plan.

The response has to be to build a strong case for change. The people best poised to do this are highly passionate advocates. These may be supporters of DP and AI in particular, or experts in efficiency and effectiveness in general. Advocates can play a key role in creating consensus among influential colleagues by raising their awareness of the need for change. Raising awareness of the problem is every bit as important, indeed essential, as is making the case for change itself. The first step, therefore, is to find advocates. Talk to clinical and academic colleagues, and to hospital management to seek out enthusiasts, their ideas, and what they can contribute. Once you have found people who are willing to lend their voice and their energy, the next step is to equip them with evidence that makes the case.

One key element of any case for change is financial, but shaping our pathology services through the introduction of DP and AI is not only a question of cost savings. Through the change, we will need services that can ensure continued accuracy of diagnoses, while optimizing patient care and protecting the reputations of our hospitals. The contributions that DP and AI may bring include workforce optimisation, improved efficiency and quality in our services, and, importantly, enhanced patient safety. It would be difficult to show gains in all of these parameters at one time, across all elements of the pathology department, so I recommend first choosing one element or ‘use case’. For example, one ‘use case’ could be how DP might allow increased insourcing or outsourcing of diagnostic work; another could be using DP instead of a microscope in primary diagnosis from all pathological specimens from a particular part of the body.

Let’s take an example of how to make a case for the latter. After outlining the entire current workflow for the diagnosis of prostate cancer, for example, you can identify the reasons for investment in DP. Perhaps you want to reach out on a regular basis to a specialist pathologist in another center. By doing so, you may predict improved efficiency, as the need for off-site glass slide storage and transport should diminish, which in turn may free staff time to cover another overburdened area. Projected costs of the initiative need to be detailed, as well as whether this is a service that might be used beyond your own hospital, regionally, nationally or even internationally. Economically, would this change make more sense than doing nothing? In addition to economics, consider sustainability, affordability, quality, benefits, and risks. Might there be costs saved in terms of streamlining workflows and quicker report turnaround times?

You can read in more detail about how I and my colleagues at Leeds think we can make such cases, including potential areas of focus, in two publications: Future-proofing pathology: the case for clinical adoption of digital pathology – part one and part two . Also, see what other institutions are learning about benefits DP and AI can bring in terms of efficiency and effectiveness. Projected cost savings of a fully digital workflow in a routine histopathology context in a large healthcare organisation in the US were recently estimated at $12.4 million for an institution with 219,000 annual lab accessions over a 5-year period – however, these were only projected savings.⁵ Another center which was an early adopter of DP has shown how whole slide imaging has resulted in a >90% reduction in requests for glass slides leading to cost and efficiency savings in filing, retrieval and transport of glass slides.⁶ I anticipate we will hear of many more creative use cases for adoption of DP in the next few years.

I hope this information and these examples empower and inspire you to make a case for the adoption of DP and AI in your workplace. The future is in our hands – let’s transform cancer diagnostics together.

*The new Faster Diagnosis Standard will ensure that all patients who are referred for the investigation of suspected cancer find out, within 28 days, if they do or do not have a cancer diagnosis – further information on the policy is available on the NHS England website at www.england.nhs.uk/cancer/early-diagnosis

Follow #TheFutureOfPathology on Twitter, Facebook or LinkedIn.