Background: Hospital admissions for preventable reasons [ambulatory care sensitive (ACS) conditions] can indicate gaps in access to or quality of primary care. This paper seeks to document the numbers and causes of these admissions in England for people with intellectual disabilities (ID) compared with those without. 
Methods: Observational cohort study of number and duration of emergency admitted patient episodes for ACS conditions, overall and by cause, using the Clinical Practice Research Datalink GOLD primary care database and the linked Hospital Episode Statistics Admitted Patient Care dataset. 
Results: The study covered 5.2% of the population of England from April 2010 to March 2014 giving a total population base of 59 280 person-years for people with ID and 11 103 910 for people without identified ID. The rate of emergency admissions for ACS conditions for people with ID was 77.5 per 1000 person-years. As a crude comparison, this was 3.0 times the rate for those without ID, but standardising for the distinct demography of this group, the number of episodes was 4.8 times that expected if they had the same age-specific and sex-specific rates. Stay durations for these episodes were longer for both young-age and working-age people with ID. Overall people with ID used 399.8 bed-days per 1000 person-years. As a crude comparison, this is 2.8 times the figure for people without ID. Standardising for their age and sex profile, it is 5.4 times the number expected if they had the same age-specific and sex-specific rates. For patients with ID, 16.6% (one in six) of all admitted patient episodes and 24.3% (one in four) of in-patient care days for people with ID were for ACS conditions. Corresponding figures for those without ID were 8.3% (one in 12) and 14.4% (one in seven). The difference in rates between those with and without ID was most marked in people of working age. The three most common causes of emergency episodes for ACS conditions in people with ID were convulsions and epilepsy, influenza pneumonia and aspiration pneumonitis. Influenza pneumonia was also a common cause for people without ID. Episodes for convulsions and epilepsy and aspiration pneumonitis were specifically associated with people with ID. 
Conclusions: Rates of hospital admissions for ACS conditions provide an important indicator of health literacy, basic self-care (or support by carers) and the accessibility of primary care. High rates are seen for some conditions specifically associated with premature death in people with ID. Local monitoring of these figures could be used to indicate the effectiveness of local primary health services in providing support to people with ID.