Medical education and clinical practice have been based on memory, and in the case of dermatology—pattern recognition—for over a century, the standard has not changed.
As a medical student and resident, I find it frustrating that we need to consume and process vast amounts of clinical data. Physicians and advanced practice providers are expected to work on memorizing nearly limitless amounts of data, and this expectation begins in school. However, as the healthcare industry becomes more and more digital, information is becoming more accessible to us.
We must distill information to meet the fast pace of healthcare decision-making. As an emergency physician, it’s not uncommon to manage more than 10 patients at a time (sometimes as many as 40 at a time) and answer questions from residents, paramedics, and others. Trying to remember each patient’s diagnosis, medical history, drug allergies, and other details is impractical.
The challenge is to find clinically relevant puzzles from all the available data. There is only so much information a person can keep, and this impossible expectation can become overwhelming, eventually leading to burnout.
Burnout not only hurts doctors, it also leads to medical errors. 2018 Coverage report The most common root cause of medical professional liability claims was found to be related to diagnosis. These diagnostic errors can harm patients and erode their trust in providers, putting doctors and institutions at financial risk.
we have to provide data context Therefore, decision makers (providers, patients, or both) feel the information is relevant to them when making decisions. We can achieve this through the use of technologies such as artificial intelligence in and outside the exam room.
The goal is not to replace human intelligence, but to increase knowledge
Artificial intelligence (AI) refers to the use of software in programming software to simulate human intelligence. Thanks to new algorithms developed over the past five years or so, software can now detect patterns in images to identify features of specific diagnoses.
For example, some software can now be used to scan radiological images to detect changes over time or to point to a diagnosis.In ophthalmology, an FDA-approved software application can detect diabetic retinopathy More sensitive than an ophthalmologist.
In the field of dermatology, machine learning has been in the spotlight: training software to detect patterns in images. For example, with machine learning, software can detect patterns in images and indicate the likelihood of lesions becoming cancerous. For example, the software could also be used more broadly to analyze images of lesions and generate a short list of potential diagnoses for doctors to consider.
When thinking about AI in healthcare, the idea is not to replace clinicians, but to support them.
Software can start seeing patterns humans can’t, reducing bias and improving outcomes
Humans are biased. We all have these internal shortcuts that we use in our lives, whether we are aware of them or not.Cognitive biases such as close prematurely, Anchoring Biasand representational biasas well as gender and racial biases, all contribute to diagnostic errors and suboptimal care.
What we use in medical education resources and reference materials and AI technology in exam rooms has to be fair – and this is true across all professions.Consider evidence showing women’s experiences More diagnostic errors Around myocardial infarction (MI), this is a bias that arises from the false belief that women do not experience MI at the same rate as men.
We also know that people with darker skin face worse health care outcomes than people with lighter skin. One reason for this discrepancy is the underrepresentation of darker skin in our educational materials: a recent study found that, across more than 15,000 medical images Only 19.5% included dark skin. This lack of representation can impact diagnosis and delay treatment for people of color.
AI can help fill the knowledge gaps created by these biases. Over time and with properly developed and vetted software, we can move from memory-based systems to a new world where we can augment our thinking based on population datasets. To do this, we need to ensure that we use trusted solutions built by clinicians, developed on diverse datasets, and free from cognitive biases based on factors such as race, gender, socioeconomic status, age, and more.
The sheer volume of clinical data that emerges every day is unsustainable for memory. Professionals need access to comprehensive and highly reliable knowledge systems as they work. Information technology can ultimately improve care and provide a better experience for patients in exam rooms and at home—not by replacing clinicians, but by empowering them with augmented intelligence.
We need to improve all our medical decisions; to do this, we need to employ fair, data-driven evidence and tools. By equipping clinicians with the right technology, we can foster partnerships with patients, improve experiences and outcomes, and even work on addressing systemic issues such as unconscious bias and racial disparities in care.
Image credit: ipopba, Getty Images
