Consulting a dermatologist over the Internet may have serious drawbacks, a new study suggests.
Direct-to-consumer telemedicine, in which a patient receives an examination and treatment over the Internet, is rapidly expanding, and in 2015 attracted an estimated 1.25 million people seeking advice in the United States. And while telemedicine has substantial benefits, the researchers say — it is convenient and typically saves money – many of the websites they looked at were not providing adequate care.
For their study, published in JAMA Dermatology, the researchers submitted six simulated cases involving skin symptoms to 16 direct-to-consumer telemedicine sites and apps for diagnosis and advice. The study excluded sites that failed to respond or responded intermittently, or that delivered incomplete or inconsistent responses. In all, 62 responses were included in the analysis.
Most sites allowed patients to submit photographs, usually one to three images of the skin condition in question. In four cases, the site provided a diagnosis based only on the patient’s description.
Two thirds of the sites failed to collect a complete medical history, and only 52 percent asked female patients about pregnancy or lactation, even though such information can be essential in the diagnosis and treatment of some dermatologic disorders.
Diagnoses were sometimes missed. In one case, a 28-year-old woman presented with acne, but not a single website asked about excessive hair growth or irregular menstrual periods, symptoms that would have led to the correct diagnosis of polycystic ovarian syndrome. In another case, a woman with eczema complicated by a potentially fatal herpes infection was given a diagnosis of having an ordinary eczema flare-up in seven of nine encounters.
The lead author, Dr. Jack S. Resneck Jr., a professor of dermatology at the University of California, San Francisco, acknowledged that such misdiagnoses could occur even in a face-to-face meeting with a doctor. But, he said, none of the telemedicine sites engaged in the kind of patient-doctor discussion and “give and take” that would happen in an office and provide the information that leads to successful diagnosis.
“You could imagine a telemedicine conversation in which the patient discusses the diagnosis with the clinician,” Dr. Resneck said, “but none of that was going on. There were no instances in which the clinician discussed the problem, asked about other symptoms and so on.”
Of patients who got a diagnosis, 65 percent were given a prescription, but the risks and side effects of the medicines were mentioned only 32 percent of the time. In addition, “when people go to an office in person, they’re aware of who they’re seeing — a clinician, a doctor, a nurse practitioner,” Dr. Resneck said. “But only a minority of sites identified the person or gave information about licensure.” Only six of the 62 responses offered to send a report to a patient’s primary care doctor.
Dr. Resneck said he would like to see telemedicine succeed. “We’re inclined to support this kind of innovation,” he said, “but it needs to be done right. Our results were disappointing.”