Breaking News
October 17, 2018 - Nine cases of polio-like illness suspected in children in illinois
October 17, 2018 - Eisai enters into agreement with Eurofarma for development and sales of lorcaserin in 17 countries
October 17, 2018 - Patients once thought incurable can benefit from high-dose radiation therapy
October 17, 2018 - Researchers awarded grant to advance testing of experimental heroin vaccine
October 17, 2018 - Researchers examine SSRI use during pregnancy and major gestational malformations
October 17, 2018 - Study reveals link between childhood abuse and higher arthritis risk in adulthood
October 17, 2018 - FDA Approves Liletta (levonorgestrel-releasing intrauterine system) 52 mg to Prevent Pregnancy for up to Five Years
October 17, 2018 - Weight gain after smoking cessation linked to increased short-term diabetes risk
October 17, 2018 - Researchers find opportunity to control salt-sensitive hypertension without exercising
October 17, 2018 - Metrohm offers robust handheld Raman analyzer for Defense and Security
October 17, 2018 - Modeling Non-Numerical Data in Systems Biology
October 17, 2018 - Research aims to address health disparities in African-American men
October 17, 2018 - Human and cattle decoys trap outdoor-biting mosquitoes in malaria endemic regions
October 17, 2018 - High Circulating Prolactin Level Inversely Linked to T2DM Risk
October 17, 2018 - Study finds gene variant predisposes people to both Type 2 diabetes and low body weight
October 17, 2018 - Metrohm software products make it easy to comply with ALOCA and ALCOA+ guidelines
October 17, 2018 - Network of doctors identify the cause of 31 new conditions
October 17, 2018 - Notable improvement in brain cancer survival among younger patients but not much for elderly
October 17, 2018 - Scientists shed light on roles of transcription factors, TP63 and SOX2, in squamous cell carcinoma
October 17, 2018 - Costs of Medicare Diabetes Prevention Program may be higher than expected reimbursement
October 17, 2018 - Misuse of prescription opioids or benzodiazepines associated with suicidal thoughts
October 17, 2018 - New research seeks to address sex disparities in women’s health
October 17, 2018 - C-Section Rates Have Nearly Doubled Since 2000: Study
October 17, 2018 - Talking to Your Kids About STDs
October 17, 2018 - New classification of periodontal and peri-implant diseases and conditions
October 17, 2018 - Herbert D. Kleber, Pioneer in Addiction Treatment, Dies at 84
October 17, 2018 - Health effects of smoke-filled atmosphere
October 17, 2018 - Down syndrome may hold important clues to onset of Alzheimer’s disease
October 17, 2018 - A special report on US’ aging societies
October 17, 2018 - Birth mode may have acute effects on neurodevelopment, study suggests
October 17, 2018 - Global health innovation system fails to deliver affordable treatments to patients, says report
October 17, 2018 - Simple, inexpensive test quickly detects antibiotic-resistant ‘superbugs’
October 17, 2018 - New drugs could reduce risk of heart disease when added to statins
October 17, 2018 - Visible and valued: Stanford Medicine’s first-ever LGBTQ+ Forum
October 17, 2018 - HVP vaccination not linked with rise in teen risky sex
October 17, 2018 - Potential ‘early warning markers’ for sepsis discovered
October 17, 2018 - Who knew? Life begins (again) at 65
October 17, 2018 - Application of blood pressure guidelines ups treatment
October 17, 2018 - Stanford researchers find that small molecule may help treat enzyme deficiency
October 17, 2018 - Speed Cameras Save Money and Lives in New York City
October 17, 2018 - Men who conform to ‘the man box’ more likely to consider suicide and violence
October 17, 2018 - Researchers aim to create more authentic organoids for drug testing, transplantation
October 16, 2018 - New blood test for pediatric brain tumor patients offers safer approach than surgical biopsies
October 16, 2018 - Age-related estrogen increase may be the culprit behind inguinal hernias in men
October 16, 2018 - Skills-Based Intervention Did Not Cut Systolic BP After Stroke, TIA
October 16, 2018 - Researchers uncover new role of TIP60 protein in controlling tumour formation
October 16, 2018 - Behind the scenes of a lifesaving heart surgery
October 16, 2018 - ‘To See the Suffering’
October 16, 2018 - Drinking concentrated rosemary extract can boost memory by up to 15%, shows research
October 16, 2018 - Medicare Advantage riding high as new insurers flock to sell to seniors
October 16, 2018 - NHS tackles prescription fraud to save millions
October 16, 2018 - New molecular switch may help develop sophisticated photomedications
October 16, 2018 - Improving access to behavioral health screenings for pregnant and postpartum women
October 16, 2018 - Health Highlights: Oct. 12, 2018
October 16, 2018 - Study holds promise for new pediatric brain tumor treatment
October 16, 2018 - Patient advocate uses MRI scans to create art and spark conversations about life with illness
October 16, 2018 - Fish oil based diets may suppress growth and spread of breast cancer cells
October 16, 2018 - Number of VHA facilities offering acupuncture has increased rapidly
October 16, 2018 - Influential Leapfrog Group jumps in to rate 5,600 surgery centers
October 16, 2018 - HIV-infected infants more likely to acquire congenital cytomegalovirus infection
October 16, 2018 - Study pinpoints new marker that can predict Crohn’s disease subtype
October 16, 2018 - Simple procedure could be efficacious intervention for failed back surgery
October 16, 2018 - New research identifies modifiable dementia risk factor in elderly people
October 16, 2018 - Zebrafish study uncovers molecular ‘brake’ that helps control eye lens development
October 16, 2018 - Overlapping copy number variations underlie autism and schizophrenia in Japanese patients
October 16, 2018 - Early menopause and diabetes may reduce life expectancy
October 16, 2018 - Majority of Americans’ ancestry can be traced through existing DNA databases
October 16, 2018 - Patients coerced into mental health care less likely to perceive treatment as effective
October 16, 2018 - Healthy elders can consume walnuts without having negative impact on weight gain, finds study
October 16, 2018 - Interactive robot helps older people exercise and detects underlying health problems
October 16, 2018 - What you need to know about autism spectrum disorder
October 16, 2018 - Antidepressants can be used to treat Alzheimer’s disease
October 16, 2018 - Study uncovers important role of PRMT1 in dilated cardiomyopathy
October 16, 2018 - Nutritional quality of breakfast linked to cardiovascular and metabolic risk factors in children
October 16, 2018 - Study uses novel approach to investigate genetic origins of mental illnesses
October 16, 2018 - Scientists develop dual anthrax-plague vaccine
October 16, 2018 - Poor Outcomes for Hispanic Infants With Congenital Heart Dz
October 16, 2018 - Global study finds youngest in class more likely to be diagnosed with ADHD
October 16, 2018 - Researchers sequence two selfish genes in the fungus Neurospora intermedia
October 16, 2018 - Survey results highlight the need for better communication between patients and HCPs about bacterial vaginosis
How should an algorithm generate recommendations for patient care?

How should an algorithm generate recommendations for patient care?

image_pdfDownload PDFimage_print

Say you’re a doctor. You’d like guidance on how to treat a particular patient, and you have the opportunity to query a group of physicians about what they’d do next.

Who do you include in that group? All doctors who have seen similar patients? A smaller number of doctors who are considered the best?

These questions are at the root of recent Stanford research.

The study, published in the Journal of Biomedical Informatics, examines a key aspect of medical artificial intelligence: If machines are to provide advice for patient care, who should those machines be learning from?

Jonathan Chen, MD, PhD, tackled this question as the latest step in his quest to build OrderRex, a tool that will mine data from electronic health records to inform medical decisions. In a prototype, an algorithm that works like Amazon’s recommendation feature supplied information to doctors on how their peers managed similar patient cases — i.e., it tells users that other doctors in this situation ordered this medication or this test.

Now, Chen and his colleagues are working through logistics of how the computer decides what to recommend. Research assistant Jason K. Wang is the study’s lead author. Chen, the senior author, told me,

The distinguishing question was, if you had to learn medical practices, should you just try to get everybody’s data and learn from everybody? Or should you try to find the ‘good’ doctors? Then that evokes some really deep follow-up questions, such as, what is a preferred doctor or a good expert? What does that mean? How would we define that?

He asked multiple doctors a simplified version of these questions — if you had to grade each doctor by a number, how would you come by that number? — and tumbled into the usual thicket surrounding physician evaluations. Which outcomes should be considered? What if you see sicker patients? What about hospital readmission, length of stay, and patient satisfaction?

Ultimately, Chen and his colleagues settled on the measure of 30-day patient mortality rate: actual number of deaths compared to expected deaths, as reliably calculated by a computer using three years of electronic health record data.

“We tried a couple of things, but found they were correlated enough that it basically didn’t matter. The doctors who rose to the top of the list versus the bottom — it was a pretty stable list,” Chen said. “Plus, you can’t really game mortality: it’s the most patient-centered outcome there is.”

With “expert” doctors identified as those with low patient mortality rates, the researchers turned to defining “optimal” care to compare with recommendations that the algorithm would produce. They decided to try two different options. One was care that followed standards from clinical practice guidelines based on literature available through such venues as the National Guideline Clearinghouse. The second was care reflecting patterns that typically led to better-than-expected patient outcomes — calculated by the computer from data in the electronic patient records.

With this established, they tested the algorithm to see how its recommendations for care for a condition such as pneumonia compared to both standard clinical practice guidelines and to computer-generated above-average patient care. First, they fed the algorithm information from patients seen by expert doctors. Then, separately, the algorithm looked at patients seen by all doctors.

The result? Not much difference between recommendations derived from care in the expert group versus all of the doctors.

Essentially, Chen said, outlier behavior from specific doctors was neutralized. And without significant differences in the results, the findings argue for building machine-learning models around data from more physicians rather than a curated bunch. The logic is similar to calculating an average: with more numbers in the equation, each number carries less weight and the average is less influenced by individual numbers — including anomalies.

Chen summed it up:

As a human, it’s not feasible to interview thousands of people to help learn medical practice, so I’m going to seek out a few key mentors and experts. What large datasets and computational algorithms enable us to do is to essentially learn from ‘everybody,’ rather than being constrained to learn from a small number of “experts.”

With this question resolved, Chen and his colleagues have now built a prototype user interface incorporating the suggestion system, and they plan to bring in doctors to test it with simulated patient cases.

“The next big challenge to closing the loop on a learning health care system,” he said, “is… where we don’t just learn interesting things from clinical data, but we design, study, and evaluate how to deliver that information back to clinicians and patients.”

Photo by Markus Spiske

Tagged with:

About author

Related Articles