Connor Iapoce is an assistant editor for HCPLive and joined the MJH Life Sciences team in April 2021. He graduated from The College of New Jersey with a degree in Journalism and Professional Writing. He enjoys listening to records, going to concerts, and playing with his cat Squish. You can reach him at email@example.com.
A new study from APA 2021 finds algorithm use can reduce clinical misdiagnosis of bipolar disorder as other mood disorders, including major depressive disorder.
New data show that evidence-based algorithms may accurately diagnose mood disorders in a clinically relevant group of participants.
The findings were presented this weekend at the American Psychiatric Association Annual Meeting.
Investigators, led by Tony Olmert of the Cambridge Center for Neuropsychiatric Research at the University of Cambridge, conducted a diagnostic trial, labeled the Delta Trial, to develop algorithms to improve the speed and accuracy of mood disorder diagnosis, a condition affecting millions of people worldwide.
The algorithms of the Delta Trial combine data from digital questionnaires and proteomic analysis of dried blood spots for an accurate diagnosis of individuals suffering from low mood with bipolar disorder (BD), major depressive disorder (MDD), or now meeting criteria for a mood disorder (low mood).
Olmert and team’s primary objectives of the study were to reduce the misdiagnosis of BD as MDD, while the secondary objective was achieving a more accurate diagnosis of MDD.
The investigators follow-up objective included an assessment of quality of life of trial participants after participating in the study and receiving a personalized mental well-being results report.
Participants in the trial were recruited through the internet and social media advertising, then screened for eligibility.
A total of 26,200 participants were screened for the study, with participants required to complete enrollment (n = 5422), digital questionnaires (n = 3232), and a returned DBS sample (n = 1377).
A total of 924 participants completed the World Health Organization World Mental Health Composite International Diagnostic Interview (CIDI) as the final step in the study subset.
Investigators primary analysis for the primary objective took participants who self-reported a clinical diagnosis of MDD in comparison with those who were diagnosed with BD during the trial CIDI interview.
A further analysis compared participants who did not self-report an existing clinical diagnosis of MDD, but were diagnosed with BD, MDD, or low mood in the CIDI. Investigators then validated their models with a positive control of participants who had prior diagnosis of BD.
The team found the mean duration of diagnosis (2.7, -1.6 years) and the percentage of general practitioner diagnosis (81.2%), as well as those diagnosed by a psychiatrist (18.5%).
Data show the positive control group had 78% of participants initially diagnosed with BD, with an average duration of correct diagnosis of 5.5, -5.9 years.
Investigators found their model shows a strong ability (AUC 0.80) to determine participants with new onset MDD from those with low mood, without meeting the criteria for depression.
In the follow-up, a majority (62.69%) of the 2064 participants who completed at least 1 follow-up questionnaire sought care from a mental health professional, with 275 participants seeking mental health care for the first time.
Investigators concluded the Delta Trial was a successful proof of concept that evidence-based algorithms can accurately diagnosis mood disorders in a relevant group of participants.
“After participating in the trial, participant mental well-being, as measured by the Warwick-Edinburgh Mental Wellbeing Scale, improved significantly at 6 and 12 months after participation,” investigators wrote.
The team also mentioned that further work would be necessary to integrate the capability into the clinical workflow.
The study, “A Combined Digital and Biomarker Diagnostic Aid for Mood Disorders: The Delta Trial,” was presented online at APA 2021.