Implications of Digitizing the Medical Landscape
The evolution of the digital age presents new, unprecedented challenges in regards to the implications new technologies can have on public health and safety. Our world, as we know it, is encapsulated by technology; we rely on technology to ease access to information, save time, communicate more efficiently, and be more cost efficient. Particularly catalyzed by the COVID-19 pandemic, digital health tools, such as machine learning, patient portals, health trackers, remote monitoring devices, and mobile health applications, are becoming a critical component of healthcare. However, while designed with the intent of improving health outcomes and experiences, the adoption of new technology into healthcare presents challenges in perpetuating biases and exacerbating existing inequalities.
For one, the healthcare industry is adapting to the implementation of machine learning as a diagnostic, analytic, and predictive tool for patients. Machine learning can be a valuable asset to the healthcare industry. Algorithms can predict medical events before they arise, such as whether patients will be hospitalized, how long they will stay in the hospital, and whether their health is worsening (Bresnick). Recent developments in Google’s machine learning can identify skin cancers, where artificial intelligence can match the diagnostic performance of practicing clinicians (Bresnick). While the implementation of machine learning and other forms of artificial intelligence in the healthcare industry does provide many benefits, the nature of the development of the algorithm and its violation of the code of ethics presents barriers to patients. For example, an algorithm that predicts health cost reflects and reinforces society’s underlying racial bias (Benjamin). The development of the predictive tool did not take into consideration that twice as many Black patients need to be identified for intervention on the basis of severity of active chronic disease, so the system underestimates the level of attention Black patients may need (Benjamin). In general, data used to train artificial intelligence for healthcare comes from segregated hospital facilities, racist medical curricula, and unequal insurance structures (Benjamin). This design exploits sampling bias, deepens existing racial inequities in the medical industry, and presents life-threatening barriers to patients. The utilization of racist data to power and train new technologies negatively impacts the patient and ultimately puts a strain on the trustworthiness of the medical system, which could prevent people from getting proper care, or getting any form of care in the first place.
Additionally, public health systems and resources, such as telehealth, are designed to improve healthcare outcomes, but the benefits of these tools are limited when there are challenges in affordability and access. Transitioning to remote living was a necessary step in preventing the further spread of the COVID-19 pandemic, but also highlighted the true extent of the digital divide and its healthcare consequences. According to a Brookings Institution report, 12-24% of Americans lack access to broadband connection to the Internet (Sieck, et al.). This inequity hampered access to telehealth, which was a critical component of healthcare during and post pandemic in New Mexico, Montana, Vermont, Iowa, West Virginia, Oklahoma, Indiana, Missouri, Tennessee, and South Carolina (Chakravorti). Rural areas, tribal lands, and low income communities were particularly disproportionately affected by this inequity in lack of access to adequate connection to the internet (Saeed & Masters). Outside of the context of just the COVID-19 Pandemic, digital health technologies, such as telehealth, removes the transportation barrier to healthcare that some patients may face, and can be beneficial for patients with large time constraints (Saeed & Masters). However, this also requires stable internet access, sufficient devices, a private space for conversations with clinicians, and at least a baseline level of health and digital literacy. So, while designed with the intent of improving healthcare and the possibility of improving health outcomes, the need for access to these resources in order to reap their benefits has a detrimental impact on low-income, underserved communities. As healthcare continues to digitize, it is important to focus efforts on minimizing the divide in access to digital resources to prevent disproportionate health outcomes, biases, and inequalities.
Efforts that have currently been implemented to help mend the challenges presented in regards to digitizing the medical field include training for future practitioners to ensure the proper utilization and implementation of health information technology, distributing educational tools to aid patients in understanding how to access certain technologies and improve their health literacy, and improving electronic health literacy and intervention research to gain a better understanding of the implications of these technologies on patients of various backgrounds (Saeed & Masters). Additional possible efforts include ensuring a sufficient foundation of health technology at an organizational level with on-call IT staff and clear patient-physician communication to prevent possible technical issues that could arise during care, as well as implementing smart phone applications to provide widespread accessible healthcare tools (Saeed & Masters). While these are steps in the right direction, we still need to be aware of the pressing challenge of lack of internet access, a large determinant in improving research and care. As technology continues to infiltrate the medical landscape, we must consider the groups of people that are most negatively affected by the implementation of technology in healthcare and work to ensure that these changes are equitable for all populations.
Chakravorti, Bhaskar. “How to Close the Digital Divide in the U.S.” Harvard Business Review,
20 July 2021, https://hbr.org/2021/07/how-to-close-the-digital-divide-in-the-u-s.
Benjamin, Ruha. Assessing Risk, Automating Racism | Science. 25 Oct. 2019,
Bresnick, Jennifer. “Google Using FHIR, Deep Learning for Healthcare Predictive Analytics.”
HealthITAnalytics, 19 May 2017, https://healthitanalytics.com/news/google-using-fhir-deep-learning-for-healthcare-predictive-analytics.
Saeed, Sy Atezaz, and Ross MacRae Masters. “Disparities in Health Care and the Digital
Divide.” Current Psychiatry Reports, U.S. National Library of Medicine, 23 July 2021, https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8300069/#:~:text=These%20disparities%20persist%20in%20spite,health%20outcomes%20despite%20technological%20improvements.
Sieck, Cynthia J., et al. “Digital Inclusion as a Social Determinant of Health.” Nature News,
Nature Publishing Group, 17 Mar. 2021, https://www.nature.com/articles/s41746-021-00413-8.