Affordable access to medical assistance is the hallmark of a successful economy.
But with growing populations and other challenges, it may take more than mere money to increase the accessibility of healthcare to those in emerging markets. Artificial intelligence can do more than improve diagnostics and reduce costs — it can help us rethink how healthcare works, where it can be deployed and what it can achieve.
Access to healthcare is a human right. No one should die from something preventable purely because they’re impoverished. Good health does more than improve the quality of life (and lifespans) of individuals — it improves economic activity and productivity.
The corollary holds true, too: As the health of a country’s citizens improves, so does its economy. In a study on this phenomenon, the OECD Observer found that a 10% increase in life expectancy results in economic growth of roughly 0.4% per year.
The quality of a country’s healthcare is also often seen as a measure of its economic stability, which can affect foreign investment.
Accessibility, affordability and quality
Good healthcare rests on three pillars: accessibility, affordability and quality. And even the best among them are set to be tested in coming years, especially in developed economies where populations are ageing and birth rates are declining.
The World Economic Forum projects that by 2050, one in six people on Earth will be over 65 years old. This shift in global demographics will necessitate increased spending on the sector, even as new technologies seek to alleviate some of the challenges.
One of these new technologies is artificial intelligence (AI), but before AI can realise widespread adoption, it needs to build widespread trust.
Each country has a budget for healthcare, and some (like the UK) have systems of public healthcare, while others have hybrid solutions that combine public and private services. More than 70% of South Africans depend on public healthcare services, but these lag behind those available to the less than a third of the population who can afford private healthcare, which is primarily funded by their contributions to medical insurance schemes.
According to the UN, African countries spend between $8 and $129 per capita on healthcare per annum, while high-income nations often spend more than $4,000.
Despite increased prosperity for many African countries, closing that gap remains a gargantuan task and relies on the proportion of funding being allocated to healthcare growing in line with increased resources, which it may not do, depending on the priorities (or promises) of those in power — among myriad other variables.
Another massive obstacle facing developing markets is a shortage of qualified workers. While European nations often boast 10 health workers per 1,000 people, in sub-Saharan Africa the average ratio is 1 per 1,000 (and in Zambia, it’s as bad as 1 per 6,000).
In some emerging markets, counterfeit drugs are a growing problem, as are shifting drug requirements. Where malaria and HIV/Aids were once the most prevalent threats to life expectancy, growing middle classes now face more of the healthcare challenges prevalent in developed markets: cardiovascular disease, diabetes and various cancers. Increases in sedentary lifestyles, tobacco and alcohol use, and fast-food consumption, only exacerbate these issues.
AI has the potential to transform healthcare, both in developed economies and emerging ones. AI can improve access and affordability while also freeing healthcare professionals to focus on the sorts of things they excel at, but which AI is less adept at.
AI diagnostics
One of the areas AI is especially well-suited to is diagnostics based on large sample sizes. For instance, AI can parse thousands of images of glaucoma, and better recognise it than even the most experienced ophthalmologist, because it’s able to simultaneously consider far larger sample sizes and more precisely recognise abnormalities, even those that are extremely subtle.
AI-powered diagnostic services, for instance, have massive advantages when compared with conventional solutions. For starters, they’re cumulative. Improvements from one set of tools can immediately be shared with all others of the same category, meaning improvements are exponential, not linear.
Hidden Health Impacts of Industrial Livestock Production Systems in Africa
Similarly, progress isn’t limited by geography. If one region identifies something unique in its datasets, it can share the findings with all other users of the same tools nearly instantly.
Progress in AI can be cumulative and shared, which also enables progress at a speed previously impossible. This is why it’s estimated that global medical spending on AI will reach $7-billion by 2022, but that AI-powered healthcare solutions could save $150-billion by 2026, according to Frontiers in Medical Health. This is why, by the end of the decade, AI medical systems could generate more than $900-billion in value, based on current projections.
AI can serve many roles in the healthcare industry. It can aid training and research, improve palliative treatment and end-of-life care, make diagnoses and early protection more precise, and aid decision-making.
It can also reduce false positives — according to the American Cancer Society, so many mammograms result in false positives that one in two healthy women are told they have cancer. AI is enabling the review of mammograms 30 times faster and with 99% accuracy, which is massively reducing the number of unnecessary biopsies performed.
The Kenya National Union of Medical Laboratory Officers, for instance, has said it estimates that three out of 10 patients in Kenya get the wrong diagnosis or treatment, in part because of insufficient systems, and partly because of incomplete medical histories.
A multitude of startups is trying to make up for this shortfall in quality data, including some that use patient-powered tracking and diaries or smartphones to help keep track of key metrics. As the sensors and other capabilities of smartphones and wearable devices improve, self-monitoring could prove increasingly valuable and accurate, especially when combined with ever-more nuanced anomaly detection algorithms.
Freenome, for instance, is an early cancer detection service that uses AI to access early screenings, diagnostic tests and blood work to detect cancer — or signs of it. By harnessing AI during general screenings and beyond, Freenome tries to detect cancer in its earliest stages. It’s also being used to develop new treatments.
Monitoring
Meanwhile, the Institute for Intelligent Systems at the University of Johannesburg has developed a service to automatically and remotely monitor patients’ vital signs in real time, and collect, process and analyse the resultant data. The device is called “e-mutakalo” (Tshivenda for “health”) and alerts medical professionals in the case of an emergency.
Collecting and collating quality patient diagnostic data can be extremely costly and this is where the increasing availability and functionality of smartphones can help — especially if patients are willing to share their data with healthcare professionals — and those professionals can demonstrate that the manner in which they are using it is ultimately valuable to the end-user.
AI and QR codes can also help medical practitioners keep track of patients while they’re in a hospital — from tracking every treatment or medication they get, to their physical location and the details of any surgeries.
Combine this with AI or machine learning and you can also make hospitals more efficient, freeing up beds and freeing up doctors (who may well be overworked already) to focus on the things they’re best at — while also reducing the odds of medication errors.
Similarly, AI can help medical staff to deliver precision doses, based on diagnosis, specifics of the patient (age, weight, and other metrics) and the nature of the ailment being treated. This reduces the odds of erroneous treatments, while also cutting costs and allowing for custom treatments based on patients’ unique circumstances.
Surgery
AI can also assist with making surgery available to more people and ensuring superfluous surgeries are avoided. In Africa, only 5% of required surgeries are performed, and the continent needs 30 times the surgical staff it currently has.
In October 2021, Groote Schuur Hospital in Cape Town became the first public hospital in Africa to offer robot-assisted surgery.
The surgical system is called the Da Vinci XI and includes a three-dimensional viewfinder for the surgeon that allows for precise depth perception, while finger callipers enable the surgeon to make minute movements with instruments.
“This translates into minimised invasiveness, ease of doing surgery and, therefore, limited trauma, which means better outcomes for the patient,” Groote Schuur’s robotic surgery coordinator, Dr Samkele Salukazana, explains.
“Greater precision means less blood loss, reduced pain, less inflammation and, ultimately, a reduced hospital stay.”
Robot-assisted surgery is an industry already valued at over $6-billion, and is expected to be worth more than $22-billion annually by 2028. It can not only save lives, but can also save on malpractice litigation.
Record-keeping
AI also helps ensure accurate record-keeping. Lost data and poor record-keeping costs the sector around $100-billion a year — a staggering amount of money that could be spent on patients, staff or equipment.
Furthermore, inadequate record-keeping can slow the development of new solutions, from medicines to diagnostic tools… and it can result in superfluous delays for patients, or in them receiving the incorrect treatment.
It can also help in life-threatening situations. A tool called SMART (Scalable Medical Alert Response Technology) monitors unattended, but mobile patients using wireless sensors. If something goes wrong, medical professionals can be alerted immediately.
AI can also help medical professionals overcome language barriers thanks to its growing aptitude for languages and translation. In the UK, a video consultation platform with real-time translation makes it easier for doctors to communicate with patients who aren’t first-language English speakers.
Various companies are also making home assistant robots for the elderly, like Mabu from Catalia Health, and ElliQ from Intuition Robotics. Some help people connect with friends and family; others are able to provide patient monitoring.
Africa
These solutions aren’t for developed or wealthy markets alone, but can also be applied to Africa, especially as they become more prevalent and the cost of implementing them falls.
Whether it’s reducing the cost of drug discovery and development, improving the accuracy of diagnostics or ensuring more accurate record-keeping, AI undoubtedly has a role to play in the future of medicine.
Moreover, it can help emerging markets or other resource-strapped economies to provide their citizens with the sort of care they deserve, but may not be able to afford.
AI need not be limited to those who can afford to develop it alone.
As it becomes more reliable, it’s also likely to become more affordable, and that means it’ll be possible to roll it out to a wider selection of beneficiaries, all of whom deserve to enjoy its benefits, regardless of their socioeconomic standing or geographic location.
