Distributions of Recorded Pain in Mental Health Records: A Natural Language Processing Based Study

Abstract

AbstractObjectiveThe objective of this study is to determine demographic and diagnostic distributions of physical pain recorded in the clinical notes of a mental health electronic health records database by utilising natural language processing and to examine the level of overlap in recorded physical pain between primary and secondary care.Design, Setting and ParticipantsThe data were extracted from an anonymised version of the electronic health records from a large mental community and secondary healthcare provider serving a catchment of 1.3M residents in south London. These included patients under active referral and aged 18+ at the index date of July 1, 2018, and had at least one clinical document (>=30 characters) associated with their record between July 1, 2017 and July 1, 2019. This cohort was compared to linked primary care records from one of the four catchment boroughs.OutcomeThe primary outcome of interest was the presence or absence of recorded physical pain within the clinical notes of the patients. This does not include mental, psychological or metaphorical pain.ResultsA total of 27,211 patients were retrieved based on the extraction criteria. Of these, 52% (14,202) had narrative text containing relevant mentions of physical pain. Patients who were older (OR 1.17, 95% CI 1.15-1.19), female (OR 1.42, 95% CI 1.35-1.49), of Asian (OR 1.30, 95% CI 1.16-1.45) or Black (OR 1.49, 95% CI 1.40-1.59) ethnicities, and living in deprived neighbourhoods (OR 1.64, 95% CI 1.55-1.73) showed higher odds of recorded pain. Patients with an SMI diagnosis were found to be less likely to report pain (OR 0.43, 95% CI 0.41-0.46, p<0.001). When comparing the overlap between primary and secondary care, 17% of the CRIS cohort also had records within LDN, and 31% of these had recorded pain in both records.ConclusionThe findings of this study show the sociodemographic and diagnostic differences in recorded pain, and have significant implications for the assessment and management of physical pain in patients with mental health disorders.Strengths and Limitations of this studyThis study utilises natural language processing on clinical notes to access a large sample with information about pain.This is the first cross-sectional study to summarise and describe the distribution of recorded pain within the clinical notes of mental health records.The recorded mentions of pain within clinical notes clearly depend on the patient sharing and the clinician recording their experiences.The findings are not generalisable to the general population since this study only looks at patients receiving mental healthcare within a specific geographic catchment.