Scientists linked to CISM (Center for Research and Innovation in Mental Health) and 4 public universities created an online calculator that helps identify symptoms of depression and other disorders. Based on users’ responses to the 13 items that make up the SMFQ scale (Brief Mood and Feelings Questionnaire), the tool indicates diagnostic probabilities.
CISM is a CPA (Center for Applied Research) financed by Fapesp (Foundation for Research Support of the State of São Paulo) and created in partnership with the Department of Psychiatry of the Faculty of Medicine of USP (University of São Paulo).
The SMFQ Transdiagnostic Calculator was developed in a study that analyzed data from 1,905 young people, between 14 and 23 years old, participants in the BHRC (Brazilian High Risk Cohort), the Brazilian High Risk Cohort for Mental Conditions. The project is linked to CISM and investigates the genetic and environmental origins of mental disorders. This year, BHRC is completing 15 years of activity.
The SMFQ scale, on which the calculator is based, aims to identify how a person has been feeling or acting over the last two weeks. The answer options are “not true”, “sometimes” and “true” for questions related to emotional state, such as “I felt alone”, “I cried a lot” and “I did everything wrong”. The tool created by the researchers calculates a final score according to the weight of each answer provided.
The researchers calculated flexible cutoffs as a person increased their symptoms. The score is given in T-scores, a standardized measure that reflects how far a person is from the mean of a distribution, that is, a reference point for analyzes in relation to the population mean. From this, they developed the tool so that it could reflect, according to the calculated score, the probability of having a psychiatric diagnosis.
Based on the result of the calculated score, the tool indicates the percentage probability of falling into diagnoses such as depression, anxiety, panic/agoraphobia and post-traumatic stress. If the final score is below the cutoff points established for comorbidities, the platform indicates the absence or very low level of symptoms.
The study that led to the creation of the calculator, published in Journal of Psychiatric Research in August, it was led by Gabriele dos Santos Jobim, from UFSM (Federal University of Santa Maria), and involved researchers from USP, UFRGS (Federal University of Rio Grande do Sul) and Unifesp (Federal University of São Paulo), partner institutions of CISM.
The center’s coordinator, Euripedes Constantino Miguel, the vice-coordinator, Luis Augusto Rohde, and BHRC researchers Rodrigo Affonseca Bressan, Pedro Mario Pan and Giovanni Abrahão Salum also signed the work, which was supervised by Mauricio Scopel Hoffmann, professor at UFSM and member of CISM.
A differential of the research is that the calculator incorporates IRT (Item Response Theory), a statistical approach that considers that each item has a different weight – based on its difficulty and discrimination –, thus producing a factor score. In this way, according to Hoffmann, the IRT provides a more accurate assessment compared to the TCT (Classical Test Theory), which makes an estimate based on the gross sum of the points for each item.
The online calculator is free to use and can assist health professionals in clinical practice, especially in low- and middle-income countries such as Brazil, where the detection of disorders is still insufficient. The tool, which is available in Portuguese and English, also democratizes access to sophisticated statistical methods.
“By integrating a symptomatic calculator into clinical practice, healthcare professionals can achieve greater efficiency, as the tool automates data analysis, saving time and resources”highlights Hoffmann. “It can help healthcare professionals adjust their assessments based on the likelihood of specific diagnoses, that is, it helps estimate the probability of a diagnosis. This contributes to the personalization of mental health care”explains the teacher.
With information from .
