JMIR Publications

ABSTRACT

Background:

Personality disorders (PDs) are associated with higher service utilisation, however individual patterns of engagement within complex needs are poorly understood. The study aims to identify subgroups of individuals based on patterns of service receipt in secondary mental health services, and examine how routinely collected information is associated with these subgroups.

Objective:

Describe and predict patterns of engagement in personality disorder treatments using routine data.

Methods:

A sample of 3,941 patients with a diagnosis of personality disorder in contact with secondary services in South London was assembled from healthcare records, with an exposure period of 11 years (2007 – 2018). Basic demographic information, service use and treatment data were included in the analysis. Service use measures included number of contacts from clinical teams and Did Not Attends (DNAs). Latent profile analysis (LPA) identified two statistically distinct groups of patients as a function of service contact measures. Cross-validation modelling tested out-of-sample predictive performance.

Results:

LPA identified two subgroups, characterised by high and low service receipt. Variables associated with profiles included diagnostic information and number of contacts from clinical teams. Model evaluation considered classification accuracy and predictive relevance of routine measures in determining likely subgroup. Nursing contacts were shown to differentially impact profiles by fitted rate of DNAs.

Conclusions:

Results suggest that routinely collected data may be used to classify likely engagement archetypes in complex needs. Employed algorithm identified factors associated with service utilisation, and may shape clinical decision making in complex needs treatments.