As vaccination rates increase across the world, we find ourselves gradually transitioning from pandemic Covid-19 to endemic. The familiar challenges of the pandemic will hopefully continue to subside, but there seems to be universal agreement that as they do, we will arrive not to pre-Covid conditions, but to a new normal. What is not clear is what this new normal will look like. It’s time to begin dissecting population health data to understand what the next chapter of healthcare delivery entails. What will in fact remain stable? Which diseases will become more prevalent? What changes should we make in healthcare systems and finance? Understanding these questions will be crucial for society and the medical community, and we are certain that the road toward that understanding will be paved by data science and artificial intelligence applications.
The Rise of Hypercoagulable Disease
From the beginning of the pandemic, researchers have published data supporting the association between Covid-19 and thrombotic events. A study conducted in France found that, in 2020, the total number patients hospitalized with pulmonary embolism (PE) increased by an estimated 16 percent compared to 2019. They noted, “The final PE odds ratio for Covid-19 hospitalized patients was four compared with other hospitalized patients in 2020” with PE mortality rates rising to 10.3 percent in that same year-to-year period.
The thrombotic-Covid correlation also extends into stroke. A group of researchers in China first published a report in JAMA demonstrating that 36.4% of patients with coronavirus displayed neurologic symptoms. The American Journal of Neuroradiology published data showing that Covid-19 is an independent risk factor for the development of an acute ischemic stroke. Moreover, the Global Covid-19 Stroke Registry has proven that patients with Covid-19 that develop ischemic strokes are more likely to suffer from more severe symptomatology and have a higher mortality rate than case match control ischemic stroke patients.
If this association is as robust as it seems presently, and if Covid will truly become an endemic problem, then it would seem imperative that the medical community carefully considers the future impact Covid-19 will have on stroke and PE incidence and care. We predict that institutions leveraging impactful and well-proven technologic solutions which address these thrombotic diseases will be better equipped to deal with the post-pandemic new normal.
“The overall incidence of pulmonary embolisms has increased worldwide during the Covid-19 pandemic. Even patients with mild Covid-19 symptoms treated primarily for other pathologies, such as acute injuries, are at an increased risk for pulmonary embolisms. The practical use of AI can support the detection of incidental pulmonary emboli in trauma and cancer patients scheduled for contrast enhanced CT of thorax and abdomen,” says Prof. Mutze- Director of the radiology and neuroradiology institute at Unfallkrankenhaus Berlin.
Where Can AI Add Real Value?
By integrating AI-based radiologic screening tools into the clinical workflow, hospitals have seen increased detection rates rate of a variety of pathologies. At the forefront of innovation is an improved specificity and sensitivity in the detection of ischemic stroke and pulmonary embolism. Beyond improved detection, rapid communication tools resting on the backbone of an AI-based infrastructure empower clinical teams to deliver the right treatment at the right time to the right patient. Whether it be administering catheter directed thrombolysis for an acute pulmonary embolism, or transferring an ischemic stroke patient from a rural hospital to a stroke center, advances in AI have proven to enhance clinical workflows that result in improved patient outcomes.
Yet, as with other areas of innovation in healthcare, we eventually arrive at the question of how to finance the application of AI in medicine? A 2019 report that surveyed 284 healthcare and digital health professionals revealed that 42% of respondents felt they were likely to partner with an AI company in 2020. However, 26% felt there were unique challenges to doing so as it relates to pricing and reimbursement strategy. Nearly 60% of respondents believed that “strongly entrenched business and reimbursement models make it difficult to bring digital health products to market,” demonstrating that reimbursement is a key challenge for the adoption of advanced technological solutions in the healthcare setting.
Funding Healthcare for Covid: Germany and UK examples
Due to their increased incidence in Covid-positive patients, improved detection and treatment of thromboembolic conditions are critical to the care of this population. Some counties have already begun their response, acknowledging through funding the importance of Covid-related illnesses. Health agencies in both Germany and the United Kingdom have committed to increase funding towards treating patients with Covid-related disease states. In the UK, the NHS has allocated upwards of £33.8 billion to this end. Germany has also allocated specific funds towards the management of Covid related medical management; for example, hospitals may be reimbursed for the use of AI on the federal state level when diagnosing and treating patients that are suffering from medical conditions attributable to Covid-19. Other options are public funding initiatives for healthcare digitalization in general or AI specifically, like the “Krankenhauszukunftsgesetz” (KHZG) in Germany or the NHS AI Lab program in the UK.
The Cost of AI in Managing Covid-19: Now and Into the Future
As patients enter emergency rooms across Europe, physicians are now accustomed to ask about or test for their Covid status. It is now one of many independent predisposing factors to be considered if there is a suspicion of a thromboembolic disorder. As such, our prediction is that workups for thromboembolic disease – including imaging, where indicated – will increase in frequency in the post-pandemic world. To that end specifically, AI-based screening tools for PE and ischemic stroke have proven to be valuable. By improving the ability of radiologists to detect these medical conditions in real-time, as well as offering emergency staff and interventionists a platform for efficient communication, AI is at the forefront of expediting efficient care in critical thromboembolic diseases. Furthermore, the added efficiency observed in departments where such tools are deployed is predicted to result in positive downstream effects for the ER and beyond.
Although the prospect of funding these initiatives may seem daunting, the coming trend of state funding for Covid-related disease, which will likely spread beyond the UK and Germany, may represent a unique opportunity for institutions to step out of the pandemic, and into the new normal.
Co-written by Ayden Jacob, Director of Health Economics and Outcomes Research in Business Development at Aidoc