W E expect our doctors to be demi-gods—flawless, tireless, always right. But they are only human. Increasingly, they are stretched thin, working long hours, under immense pressure, and often with limited resources. Of course, better conditions would help, including more staff and improved systems. But even in the best-funded clinics with the most comprehensive specialists, standards can still lag; doctors, like the rest of us, are working with Stone Age minds. Despite years of training, human brains are not optimally equipped for the pace, pressure, and complexity of modern healthcare.
Given that patient care is the ultimate goal of medicine, the question is who or what is best equipped to deliver it? AI may still be suspect, but research is increasingly showing how it could help address some of the most persistent problems and overlooked failures – from misdiagnosis and error to unequal access to care.
As patients, each of us will experience at least one diagnostic error in our lifetime. In England, conservative estimates suggest that 5% of primary care visits result in a missed diagnosis, putting millions of patients at risk. In the US, diagnostic errors almost certainly result in death or permanent injury to 800,000 people a year. Misdiagnosis is a greater risk if you are one in ten people worldwide with a rare disease.
Modern medicine prides itself on being scientific, but doctors don’t always practice what the evidence suggests. Studies show that evidence-based treatments are offered only half the time to adults in the U.S. Doctors can also disagree with diagnoses. In a study of more than 12,000 radiology image reviewers offering second opinions, they disagreed with the initial assessment in about one in three cases—leading to a change in treatment nearly 20 percent of the time. As the workday is stretched, quality slips further: inappropriate antibiotic prescriptions are on the rise, while cancer screening rates are falling.
As troubling as it is, there are understandable reasons for these failures—and when viewed from another angle, it’s remarkable that doctors get it right as often as they do. Human reality—distractions, multitasking, even our body clocks—takes a toll. But burnout, depression, and cognitive aging don’t just wear doctors down; they raise the risk of clinical errors.
Medical knowledge is also moving faster than doctors can keep up. By the time they graduate, half of what medical students learn is already outdated. It takes an average of 17 years for research to reach clinical practice, and with a new biomedical article published every 39 seconds, even summarizing abstracts would take about 22 hours a day. There are more than 7,000 rare diseases, with 250 more identified each year.
AI, by contrast, learns medical data at lightning speed, 24/7, without sleep or bathroom breaks. Where doctors differ in undesirable ways, AI is consistent. And while these tools also make mistakes, it would be a stretch to deny how impressive the latest models are, with some studies showing that they significantly outperform human doctors in clinical judgment, including on complex medical conditions.
AI lielvara ir pamanāmi modeļi, kurus cilvēku pietrūkst, un šie rīki ir pārsteidzoši labi atpazīst retas slimības – bieži labāk nekā ārsti. Piemēram, iekšā Viens 2023. gada pētījums Pētnieki baroja 50 klīniskos gadījumus-ieskaitot 10 retus apstākļus-uz ChatGPT-4. Tika lūgts sniegt diagnozes ranžētu ieteikumu veidā. Tas atrisināja visus kopējos gadījumus ar otro ieteikumu un ieguva 90% no retajiem apstākļiem līdz astotajam – pārspējot cilvēku ārstus, kurus izmantoja kā salīdzinātājus. Pacienti un viņu ģimenes arvien vairāk atzīst šos ieguvumus. Viens bērns, Alekss, redzēja 17 ārstus Trīs gadi hroniskām sāpēm – Neviens nevarēja izskaidrot viņa simptomus. Izmisīgi, viņa māte pievērsās Chatgpt, kas ierosināja retu stāvokli, ko sauc par piesietu auklas sindromu. Ārsti apstiprināja diagnozi, un Alekss tagad saņem pienācīgu ārstēšanu.
Tad ir piekļuves problēma. Veselības aprūpe ir otrādi. Tie, kuriem visvairāk ir nepieciešami – visdrīzāk, visdrošākie un visvairāk atstumtākie sabiedrībā – ir tie, kas visdrīzāk tiks atstāti aiz muguras. Iesaiņoti grafiki un slikts sabiedriskais transports nozīmē miljoniem tikšanās. Vecāki un nepilna laika darbinieki, ieskaitot tos, kuriem ir koncerta ekonomikas darbi, bieži cīnās, lai apmeklētu pārbaudes. Amerikas laika lietošanas aptauja Dati rāda, ka pacienti upurē divas stundas par 20 minūšu ārsta apmeklējumu. Problēmas bieži ir sliktākas cilvēkiem ar invaliditāti, kuri ir par četras reizes Visticamāk, palaižot garām aprūpi Lielbritānijā, pateicoties transporta, izmaksu un garo gaidīšanas sarakstu dēļ. Salīdzinot ar vīriešiem bez invaliditātes, sievietēm ar invaliditāti vairāk nekā septiņas reizes biežāk ir neapmierinātas vajadzības aprūpes vai medikamentu izmaksu dēļ.
Un tomēr mēs reti apšaubām ideju gaidīt rindā ārsta kabinetā pilsētā, jo tas ir vienkārši veids, kā lietas vienmēr ir notikušas. AI to varētu mainīt. Iedomājieties ārstu jūsu kabatā, piedāvājot informāciju, kad un kur jums tas nepieciešams. Saskaņā ar Darba 10 gadu plānu veselības sekretārs Wes Streeting ir paziņojis, ka pacienti drīz varēs pārrunāt savas veselības problēmas ar AI, izmantojot NHS lietotni. Tas ir drosmīgs solis – un tas varētu sniegt ātrāku, praktisku klīnisku padomu miljoniem.
Tas darbosies tikai tiem, kas to var izmantot, protams. Piekļuve internetam uzlabojas visā pasaulē, taču joprojām pastāv nopietnas nepilnības: 2,5 miljardi cilvēku paliek bezsaistē. Lielbritānijā, 8,5 miljoni cilvēku Trūkst digitālo prasmju pamatprasmju, un 3,7 miljoni ģimeņu nokrītas zem “minimālā digitālā dzīves standarta”, kas nozīmē, ka tām ir slikta savienojamība, novecojušas ierīces un ierobežots atbalsts. Pārliecība ir arī barjera: 21% cilvēku Apvienotajā Karalistē saka, ka viņi jūtas atstāti no tehnoloģijas.
Currently, AI Healthcare Research focuses almost exclusively on its shortcomings. The crucial task is to test the feasibility of the technology for bias and error. But this orientation ignores the creaky and sometimes insecure systems we already rely on. Any fair assessment of AI must be compared to what we have now—a system that can all too often be frustrating, inaccessible, or simply flawed.
Charlotte Blease is a health researcher and author of Dr. Bot: Why Doctors Can Fail Us—and How AI Could Save Lives published by Yale on September 9th in the Local.
Further reading
Deep Medicine: How Artificial Intelligence Can Make Healthcare Human Again By: Eric Topol (Basic Books, £28)
Co-existence: Living and Working with AI by Ethan Mollick (Wh Allen, £16.99)
Artificial Intelligence: A Guide to Human Thinking by Melanie Mitchell (Pelicans, £10.99)
