To explore how well the primary care system in Scotland works for adults with intellectual disabilities (ID), using the rate of unplanned hospital admissions for ambulatory care sensitive conditions (ACSC) as a proxy indicator. As part of this, to investigate those rates and rate ratios among adults with ID and without ID, adjusting for the prevalence of a given ACSC in each population. The secondary aim was to explore deaths due to ACSC among the ID and no-ID populations.

A population-based retrospective cohort data linkage study of adult respondents to Scotland’s 2011 Census. Self-reported or proxy-reported ID status from the Census was linked to hospital admissions data and deaths data. The cohort was followed until the end of 2019. The prevalence of ACSCs in each population was calculated from aggregate-level data published by the National Health Service, as it was not possible to use the linked dataset for this purpose.

Whole population of Scotland.

People aged 18+ on census day (27 March 2011), including all adults with ID (n=16 840) and a 15% randomly selected comparator sample of adults without ID (n=566 074).

Crude and age-sex standardised incidence rates and ratios; cumulative incidence; prevalence ratios. The exposure was ID status, and the outcomes were (1) unplanned ACSC hospital admission, (2) death with an ACSC condition listed as the main cause on the death certificate and (3) death with an ACSC condition listed as one of the causes on the death certificate.

Adults with ID under the age of 55 had only a slightly higher risk of an unplanned ACSC hospitalisation than their general population counterparts (standardised incidence ratio 1.11; 95% CI 1.03 to 1.20). After adjusting for different ACSC prevalence in ID and non-ID cohorts, this difference in risk disappeared. These findings contrast with existing evidence from England, where a much higher unadjusted risk of unplanned ACSC hospitalisations was found among people with ID. Adults with ID had a higher risk of dying due to ACSC than adults without ID (standardised mortality ratio 2.54; 95% CI 2.19 to 2.95).

Our findings on unplanned ACSC hospitalisations suggest that the primary care system in Scotland appears to be similarly effective for adults with ID than for adults without ID. However, the higher risk of dying from ACSC among people with ID suggests that this system is less effective for people with ID. Future research should investigate this tension and aim to understand why the operation of the primary healthcare system seems to be worse with regards to ACSC mortality than with regards to unplanned ACSC hospitalisations.