Decision Support for Tailored Biopsychosocial Rehabilitation: In Non-specific Low Back Pain

Linda Nieminen

Research output: Book/ReportDoctoral thesisCollection of Articles

Abstract

Low back pain is globally the most burdensome symptom causing disability. It is most commonly defined as non-specific, which means no pathoanatomical cause can be demonstrated as the cause. Different biopsychosocial factors are widely related to the experience and prolongation of pain and disability. Some of these factors can be affected by targeting timely interventions and decreasing the risk for pain chronicity. Pain related biopsychosocial factors and their connections can be understood more profoundly with the help of the International Classification of Functioning, Disability, and Health (ICF) framework developed by the World Health Organization (WHO), which describes disability from a wide biopsychosocial perspective.

The main aim of this dissertation was to develop methods to support the decision-making in the tailored biopsychosocial rehabilitation of patients with non- specific LBP. The secondary aims were to produce a topical summary of the known biopsychosocial risk factors for low back pain chronicity, and to find methods to recognize those factors as well as support the assessment and execution of tailored interventions targeted to the individually recognized factors.

A systematic literature review was compiled from the results of 25 different studies on the risk factors associated with low back pain chronicity. The studies had to evaluate the possible risk factor before the chronic phase of pain (3 months) in order to be regarded as a preceding factor for pain. To help the recognition of biopsychosocial factors at the individual level, an artificial intelligence algorithm application was developed that identifies disability information from electronic health records in accordance with the ICF framework. The results of the application were compared to the findings of a domain expert. The processes of patients with low back pain in primary and occupational health care were developed to more comprehensively assess possible risk factors and better tailor interventions to the individuals. A multidisciplinary team was formed from primary, occupational, and special health care professionals for the process design. For the purposes of developing new methods, a patient population of 93 patients with chronic low back pain were gathered. The data comprised free text from electronic health records and quantitative information from medical history forms.

According to the systematic review, 45 different factors were identified as being associated with low back pain chronification. The factors were divided into demographical and medical history related factors, biomechanical factors, symptom related factors, psychological and psychosocial factors, and lifestyle factors. The factors were interrelated with the description of disability in the ICF framework, with the exception of the demographic and medical history related factors. The applied artificial intelligence algorithm was able to recognize disability information from the electronic health records with a sensitivity of 83.1% and specificity of 99.84% compared to the results of the domain expert. The rehabilitation process design was presented in a logic model that guides the needed professionals into the process according to the patients’ needs, clearly states the activities of the professionals, and comprehensively exploits a multidisciplinary community over sector boundaries.

The findings of this dissertation open new research possibilities in the areas of low back pain and the exploitation of disability information. The results of the systematic review will help clinicians to better understand the biopsychosocial entity of low back pain more competently and researchers to extend their intervention study designs. In future, a feasibility study on the rehabilitation process should be executed before a larger intervention. The benefits of the artificial intelligence algorithm application are planned to be expanded to other patient groups and languages.
Original languageEnglish
Place of PublicationTampere
ISBN (Electronic) 978-952-03-2851-1
Publication statusPublished - 2023
Publication typeG5 Doctoral dissertation (articles)

Publication series

NameTampere University Dissertations - Tampereen yliopiston väitöskirjat
Volume780
ISSN (Print)2489-9860
ISSN (Electronic)2490-0028

Fingerprint

Dive into the research topics of 'Decision Support for Tailored Biopsychosocial Rehabilitation: In Non-specific Low Back Pain'. Together they form a unique fingerprint.

Cite this