A new tool to measure risk for pain development

With the help of artificial intelligence and using 99 factors, a research team has created a scale to predict the likelihood, intensity and extent of chronic pain. A simplified, six-question version is also in the works. Published in Nature Medicine, the findings are drawn from U.K. biobank data of nearly 500,000 people and were later validated in two other cohorts in Finland and Canada. They showed key risk factors as being insomnia, feelings of distress, fatigue, stress and a body mass index above 30. Read: Understanding pain and what healthcare professionals can do Chronic pain is a disease in its own right, not just a symptom, according to the World Health Organization's International Classification of Diseases. The pain is chronic if it persists for more than three months. It can be primary, without determinable cause, or secondary, when it is clearly associated with diagnosed diseases. More than 7.6 million Canadians live in this situation, including one in three people aged 65 and over, according to Statistics Canada's Canadian Community Health Survey. Although the numerical scale from zero to 10 remains the absolute reference, not all pain assessment systems are suitable for complex situations, according to study co-author Étienne Vachon-Presseau, assistant professor at McGill University. He gives the example of arthritis in the knee for which X-ray imaging and MRI are used to quantify seriousness. “These are not very good predictors. There are people who will have a knee that looks really damaged but who will not necessarily report very high pain.” Read: Looking after survivors Given that pain is not just the consequence of a signal sent to the brain but influenced by many cultural and sociological phenomena, the research team relied on the biopsychosocial model of pain. It is a model "well anchored in the literature," says Vachon-Presseau, but often fragmented. “There are people who study sleep and pain, sleep and genetics, trauma.” Here, the team incorporated as many variables as possible in order to create a new, more “comprehensive” model, explains Christophe Tanguay-Sabourin, a doctoral student in biomedical sciences at the University of Montreal and study co-author. "By combining a lot of factors, we manage to obtain a lot of information that can help to better understand the patient's perspective, his trajectory." This multivariate model includes elements such as the experience of stress, calculated by the loss of a partner or the arrival of retirement which can lead to increased financial instability. "We noticed that people who had little pain, but a very high score, were those who were most likely to develop it, while those who had a very low score but a lot of pain, tended to see their pain lessen,” says Tanguay-Sabourin. Particular attention has been paid to people living with a chronic overlapping pain condition (COPC), meaning they experience intense pain in several parts of the body. Machine learning algorithms were able to determine the number of sites where people would experience pain, based on data available nine years prior to this release. "The model is not infallible," admits Vachon-Presseau, but the results were particularly accurate for people who developed pain at four or more sites. The concept of overlapping pain is well-known in the literature, points, he said, noting that 44% of the population in the UK biobank is affected. "People who have pain on several sites have more impact on their activities, disabilities, disorders, anxiety, depression." According to Tanguay-Sabourin, healthcare professionals should not fear these situations where intense pain overlaps and adds up. "Populations that move away from traditional diagnoses are a challenge for doctors, who are very reluctant to take them on." Based on the biopsychosocial model, he recommends personalizing pain treatments by taking into account patients as a whole, including their mental health. The research team's six-question assessment tool, giving a score ranging from zero to six, has not yet been validated for clinical application, but will be tested in the field. “We are in talks with the doctors at the McGill pain clinic to see if it can be useful," says Vachon-Presseau. This article was translated from French and originally ran in our sister publication Profession Santé.

A new tool to measure risk for pain development

Register for free access or login.

Related Topics