A national cohort study of serial missed appointments in general practice- health outcomes, utilisation and social vulnerability.

Talk Code: 
Andrea E Williamson or David A Ellis
David A Ellis, Ross McQueenie, Philip Wilson, Alex McConnachie
Author institutions: 
University of Glasgow, University of Lancaster, University of Aberdeen


Low engagement in health care requires to be better understood if the health services’ contribution to tackling health inequalities is to improve. Low engagement includes having a pattern of missing health care appointments. This study tests the hypothesis that serially missing appointments are a risk marker for social vulnerability and poorer health outcomes.


Based on demographic information from the literature about single missed appointments, 155 urban, rural and high deprivation (Deep End) General Practices across Scotland were recruited. A sample of 909,073 patients who had scheduled a GP practice appointment in the preceding 3 years was categorised based on our pilot study into zero, low, medium or high missed appointment patterns. Analysis of extracted Read code categories and limited text data in the national Safehaven is underway to establish: 1. the factors and interactions that predict patients who serially miss appointments2. whether serial missed appointments may be a predictor of future patient outcomes This presentation will focus on clinical diagnoses associated with serial missed appointments, general practice utilisation and social vulnerability. The wider study also includes linked data from outpatient, A&E attendances, hospital admissions and school attendance and attainment. Analyses will initially be descriptive, summarising missed appointment rates in relation to these factors recorded at the point of entry to the study. Regression models (Poisson or negative binomial) will help us understand how these categories are associated. Appropriate regression models, according to the outcome, will be used to assess whether any associations with serial missed appointment rates are independent of other factors.


The clinical diagnoses data are being derived from multi-morbidity counts, additional diagnoses based on priority 1 Read codes and levels of psychotropic medicines prescriptions.The health utilisation data focuses on screening and immunisation, use of primary care and other services, and evidence about low or non- engagement in care such as for long term conditions monitoring.Social vulnerability maps GP data for the first time onto Adverse Childhood Experiences, Severe and Multiple Disadvantage and other iteratively derived categories based on the literature about marginalisation and vulnerability.


Using this large data set to explore our hypothesis will provide general practice derived data on whether seeking to reduce patterns of missed appointments are worthy of targeted attention.

Submitted by: 
Andrea E Williamson
Funding acknowledgement: 
Scottish Government Chief Scientist Office research grant (CZH/4/1118) with Safehaven and data linkage costs supported in lieu by the DSLS at Scottish Government.