Analysis information tip sheet: Story concepts from Johns Hopkins Medication

Ad Blocker Detected

Our website is made possible by displaying online advertisements to our visitors. Please consider supporting us by disabling your ad blocker.

PICTURE: Illustration for tips on research history from the perspective of Johns Hopkins Medicine More

Photo credit: Johns Hopkins Medicine


Media contact: Michael E. Newman,

Diabetic retinopathy – a condition characterized by damage to the small blood vessels on the retina (light-focusing area) of the eye and a leading cause of vision loss worldwide – has increased in recent years with the number of children and adolescents in whom either type 1 or type 2 diabetes was diagnosed, increasing. Although the American Diabetes Association (ADA) recommends regular screening for pediatric diabetic retinopathy, it is estimated that less than half of all adolescents with diabetes follow the recommendation. Without early detection and treatment, these patients put themselves at risk of serious vision problems or blindness as they age.

In a recent study published online January 21, 2021 in Diabetes Care, researchers from Pediatric Endocrinology and Ophthalmology at Johns Hopkins Medicine and three other medical institutions in the United States showed that autonomous artificial intelligence (AI) can be used to treat pediatric diabetic retinopathy recognizable with a high degree of sensitivity, specificity and diagnostic capability (accuracy of recognition) – and without human interpretation. The technique has already been approved by the US Food and Drug Administration for adults with diabetes and is part of ADA’s guidelines for retinopathy screening for patients 21 years of age and older.

Because AI screening does not require eye dilatation, it takes less time to complete and is easier to perform for pediatric patients. Therefore, compliance with the regular retinopathy screenings defined by the ADA more than doubled in this study.

“The use of autonomous AI in adults has shown extremely high levels of sensitivity, specificity and accuracy in diagnosing reference levels [more than mild] diabetic retinopathy when the disease is most treatable, “says Dr. Risa Wolf, study director, pediatric endocrinologist at the Johns Hopkins Children’s Center, and assistant professor of pediatrics at the Johns Hopkins University School of Medicine.” So, with the incidence of pediatric disease diabetes increasing – especially type 2, which is associated with earlier onset of retinopathy – we felt it was important to consider whether AI can improve compliance with screening guidelines and early detection for younger patients. “

A total of 310 pediatric diabetes patients were enrolled in the study over a 12-month period. Participants had a mean age of 12 years, were 47% male, and represented a wide range of races (57% white, 32% black, 4% Hispanic, and 7% Asian or other). The majority of the patients had type 1 diabetes (82%) and an average age of 9 years when diabetes was first diagnosed, regardless of whether they were type 1 or type 2.

One hundred and fifty-two participants (49%) reported having had a diabetic eye exam with dilatation prior to the study, but only 17 (11.3%) had a recording of the screening test in their case files. Using a special statistical calculation, however, the researchers were able to measure the improvement in screening adherence for these patients and then estimate it for the entire group.

The study used digital fundus photography – which requires no expansion, takes only a few minutes, and provides high quality images for the detection of retinopathy by trained observers – in conjunction with a fully autonomous AI system built into the camera. This eliminated the need for human evaluation to obtain a diagnosis.

To review the diagnoses made by the AI ​​system, the same color photos were independently reviewed by two retinal specialists who were not informed of the AI ​​interpretations.

Of the 310 participants, AI gave an accurate interpretation of retinopathy or no retinopathy in 302 cases (97.5%). The eight uninterpreted image sets were due to the participant’s inability to keep their eyes open during the flash or focus as needed.

Overall, the sensitivity (85.7%), specificity (79.3%), and diagnostic ability (97%) of AI interpretations in children were high based on the reference standards for these traits defined by retinal specialists. This high level was observed regardless of race, ethnicity, age or gender.

After implementing the AI ​​screening system, the adherence rate improved from 49% to 95%, which is an increase of 111%.

“Our results show that autonomous AI – which has been shown to be a safe and effective tool for diagnosing diabetic retinopathy in adults – deserves a role in screening for this disease in younger patients,” says retinal specialist Dr. Roomasa Channa, senior study author and assistant professor of ophthalmology and visual sciences at the University of Wisconsin’s School of Medicine and Public Health.

Wolf is available for interviews.


Media contact:
Vanessa McMains,

The longer someone stays awake, the more likely they are to get tired because their brain needs sleep. But how the brain perceives this need for sleep has not always been clear. Researchers at Johns Hopkins Medicine have now shown in fruit flies that certain groups of brain cells, called astrocytes, detect electrical activity in different regions of the brain and use these signals to make it easier for people to fall asleep. The more activity they detect, the stronger the sleep need signals get until they trigger a trigger mechanism that promotes sleep.

In their findings, published January 11, 2021 in the journal Current Biology, the researchers say understanding how we get sleepy can help us understand, and ultimately treat, the types of sleep disorders in people who do never feel rested no matter how much sleep they have received.

“If you nod off during a boring lecture in class and still hear the professor calling your name, it’s because only part of your brain is sleeping,” says Dr. Mark Wu, professor of neurology at the University of Johns Hopkins Medical School. “We believe that different groups of these astrocyte cells monitor different parts of the brain to initiate the drive to sleep in those specific regions.”

The researchers showed in their study that prolonged vigilance leads to an accumulation of calcium ions in the astrocytes, which ultimately leads to a whole cascade of genes being switched on. In this case, the astrocytes release chemical molecules that induce sleep by acting on a central sleep drive circuit (an electrochemical network) in the brain.

Two recent papers by researchers at other institutions showed similar results in mice to those published in the Johns Hopkins Medicine Paper. Taken together, these studies suggest that these processes are persistent throughout the animal kingdom and are likely to be applicable to humans as well.

Wu is available for interviews.


Media contact:
Michel Morris,

Researchers at Johns Hopkins Medicine recently found that although primary care physicians should discuss hypoglycemia or low blood sugar issues with patients with diabetes every time they visit, and should take high-risk drugs like insulin, the topic was only discussed in three months of those visits.

Hypoglycemia is the most common serious side effect caused by treatment for diabetes. Severe hypoglycemic episodes can lead to negative consequences, including falls and visits to the emergency room, and increase the risk of stroke and death. In a 2018 survey of 20,188 adults with diabetes, 12% said they had severe hypoglycemia in the past year.

“In order for patients to receive safe diabetes treatment, there must be open communication between them and their healthcare provider about drug side effects, particularly hypoglycemia,” says Dr. Scott Pilla, MHS, Assistant Professor of Medicine at the Johns Hopkins University School of Medicine. “For example, in our study we found that doctors almost never advised against driving a car if a patient thought their blood sugar was low or might get low. This is an important discussion as low blood sugar makes a person Could cause you to think unclearly and have an accident. ”

The results from Pilla and his research team were published in the Journal of General Internal Medicine on January 21, 2021.

Most of the US outpatient diabetes care is provided in primary care. Therefore, visits to the doctor for patients with diabetes are an important opportunity to advance the prevention of hypoglycemia. To find ways to improve hypoglycemia communication during a doctor visit, Pilla and his team attempted to define the frequency and content of assessments and counseling in primary care related to hypoglycemia.

To that end, the researchers looked at 83 visits to primary care an urban health practice representing eight clinicians who saw 33 patients with diabetes using insulin or sulfonylureas such as glipizide and glyburide. Audio during visits was recorded as part of the “Achieving Blood Pressure Control Together” study, a randomized study of behavioral interventions for high blood pressure.

Communication between the clinician and the patient about hypoglycemia occurred in 24% of the visits, while communication about prevention of hypoglycemia occurred in 21%. Despite the patients’ fear of hypoglycemia, doctors rarely rated the frequency of hypoglycemia, its severity, or the potential impact on the patient’s quality of life.

While office visits are sometimes complicated and often focus on a variety of topics, according to Pilla, the study results should encourage primary care physicians to make counseling about assessing hypoglycemia a priority for patients taking high-risk diabetes medications. He says there is currently a lack of a system for routinely assessing hypoglycemia during GP visits, and he believes his team’s research shows the need for such a system.

Pilla also suggests that patients talk about low blood sugar when they see a doctor. “Family doctors should work with patients to find out how best to prevent episodes of low blood sugar and choose the safest treatment for diabetes,” he says.

Pilla says he ultimately hopes to study hypoglycemia communications on a larger scale. With more data, he explains, researchers can better understand how such discussions can be made more effective and productive, which could lead to improved safety for primary care diabetes treatment.

Pilla is available for interviews.