Smartphone app screens for rare liver disease in newborns

A smartphone app can accurately interpret and record the colour of a newborn's poop as a possible early symptom of a rare life-threatening liver disease, scientists have found.
Biliary atresia (BA) is a rare disorder that accounts for nearly half of pediatric end-stage liver disease in the US.
For the vast majority of parents using the app, named PoopMD, the results can provide reassurance that their newborn's stool colour is normal.
For the one in 14,000 newborns with BA - about 400 babies each year in the US - parents using the app can rely on it to help detect the symptomatic pale yellow to chalky grey stools that mean urgent medical assessment is needed.

"Days matter in diagnosing BA," said lead author Douglas Mogul, a pediatric gastroenterologist at the Johns Hopkins Children's Center.
That is because babies with BA treated within the first two months of life have the best outcomes and are far less likely to need a liver transplant later.

However, the 60-day window is all too often missed, with the average time to diagnosis in the US standing at 70 days.
"PoopMD does what it says it will do," said Mogul, who worked with HCB Health to create the app, first released in 2014.
For the study of the app, which builds on an earlier "colour card" that is distributed to new parents, the team first gathered the medical opinions of seven pediatricians who looked at 34 photographs of pale-coloured stool.

Twenty-seven of the pictures were determined to be of normal stool, and seven were deemed acholic, or bile deficient, signalling high risk for BA.
Next, one expert and three laypeople were asked to use the app to look at and analyse the same pictures under a variety of lighting conditions and using a variety of smartphone models.
"These individuals were essentially asked to take a picture of the stool photograph and determine if the app identifies the photo as normal or pale," Mogul said, adding that in normal use, a parent just takes pictures of the contents of a diaper.

Even with the picture of the picture, the researchers said, the app correctly identified all of the acholic stool samples and correctly identified 24 of the 27 normal stools, while three normal stools were mislabelled "indeterminate."
Once downloaded on a smartphone, parents use the app by taking a picture of the baby's stool and identifying the part of the picture that has a stool colour of concern.
The app then immediately identifies whether the stool colour matches those associated with gastrointestinal illnesses or problems with the liver, including BA.
The app can store results for future and comparative reference, and parents can email a photo to a pediatrician directly from the app.

The study was published in the journal PLOS One.