Courtesy of April Koh
Last March, a team of Yale innovators launched Spring, an online medical tool connecting depression patients to care options, helping to incorporate medical technology into a field that often lacks entrepreneurship.
Depression is one of America’s most pervasive mental health problems and affects nearly 16 million American adults annually, according to the National Institute of Mental Health. Despite its prevalence, depression remains one of the most difficult disorders to treat. While antidepressants are the most prescribed medicine in the United States, only 11 to 30 percent of patients benefit from treatment, according to April Koh ’16, one of Spring’s cofounders.
“Treating depression is still a trial-and-error process,” Koh said. “Patients will take drug after drug. Taking a drug for weeks and weeks is long and expensive for the insurer and patient.”
Koh, along with Adam Chekroud GRD ’20 and Abhishek Chandra ’16, built the online platform, which uses data analysis of many previous studies to generate personalized antidepressant treatment recommendations. This eliminates the typical trial-and-error process of prescribing medication, Koh said.
According to Koh, by combining self-reported data with data from previous patients, Spring builds upon the precision medicine model — a concept whereby physicians tailor treatment methods to individual patients. She said that patients using Spring are screened for depression by means of a 10-minute questionnaire and are then recommended appropriate antidepressants, if necessary.
In January, Chekroud, a graduate student in the psychology department, published an article in The Lancet medical journal. Chekroud and his research team developed an algorithm that used existing clinical trial data to determine whether patients with depression would respond well to specific antidepressants. Koh said that this algorithm was more accurate in its prescription than existing genetic and biological tests, as well as those of a number of Yale-affiliated psychiatrists.
Koh, who took two years off from Yale after her junior year to work on another startup, said she ran into Chekroud’s article when she returned to school in fall 2015.
“When I read Adam’s paper, I knew at that moment the problem Adam was talking about, and found it compelling after having friends and family go through the trial-and-error process,” Koh said.
Chekroud and Koh said they wanted the algorithm research to be applied to real-life situations. Koh said that she and Chekroud worked together throughout early 2016 to design the questionnaire for depression.
Since Koh and Chekroud began to work together, Spring has won numerous awards, including venture capital support from the Yale Entrepreneurial Institute, Koh said. She added that YEI capital and mentorship has been very valuable to the Spring team.
According to Koh, one of the most important reasons for Spring’s success is the dynamic behind the platform’s team.
“Try to find the most talented people around you and work with them,” Koh said. “[Chandra] and [Chekroud] were new friends, but from the very beginning when we were putting the team together, we recognized in each other great talent. We really respect each other.”
Hundreds of patients around the country have used Spring since its launch. Koh said that a network of psychiatrists across the country is using Spring in their own practices, as well as 250 users who have completed their online test.
Girish Subramanyan, a San Francisco psychiatrist who specializes in mood and anxiety disorders, said that the addition of Spring to his practice was exciting.
“Most of my patients have found it interesting that answering a questionnaire might be able to predict treatment outcome,” Subramanyan said. “However, it’s a bit early on to tell how effective it is in our own clinical setting.”
Koh said that in the beginning, she was discouraged by the bureaucracies of health care and said the complexity of the US health care system represents the biggest problem for entrepreneurs in the field. She added that physicians and the health care industry are risk-averse, creating high barriers for the entry of new products into the market and making it difficult to quickly test or pilot a product like Spring. However, she acknowledged this reluctance to change comes from doctors’ good intentions to put their patients’ lives first.
Chekroud said that despite the usefulness of Spring, it cannot replace a clinician or physician, and that Spring therefore only serves a supplementary role.
“At the end of the day, an algorithm cannot supply a prescription, but rather provide support for the physician,” Chekroud said.