Does practice analysis agree with the ambulatory care sensitive conditions list of avoidable unplanned admissions: cross-sectional study in the East of England.

Robert Fleetcroft, Antonia Hardcastle, Nicholas Steel, Gillian Price, Sarah Purdy, Alistair Lipp, Phyo K. Myint, Amanda Howe

Research output: Contribution to journalArticlepeer-review

8 Citations (Scopus)
26 Downloads (Pure)


Objectives: To use Significant Event Audits (SEA) in primary care to determine which of a sample of emergency (unplanned) admissions were potentially avoidable; and compare to the NHS list of ambulatory care sensitive conditions (ACSCs).
Design: Analysis of unplanned medical admissions randomly identified in secondary care.
Setting: Primary Care in the East of England.
Participants: 20 general practice teams trained to use SEA on unplanned admissions to identify potentially preventable factors.
Interventions: SEA of admissions.
Main outcome measures: Level of agreement between those admissions identified as potentially preventable by SEA and the NHS ACSC list.
Results: 132 (26%) of randomly selected patients with unplanned admissions gave consent and an SEA was performed by their primary practice team. 130 SEA reports had sufficient data for our analysis. Practices concluded that 17 (13%) of admissions were potentially preventable. The NHS ACSC list identified 36 admissions (28%) as potentially preventable. There was a low level of agreement between the practices and the NHS list as to which admissions were preventable (Kappa = 0.253). The ACSC list consisted mainly of respiratory admissions whereas the practice list identified a wider range of cases and identified context-specific factors as important.
Conclusions: There was disagreement between the NHS list and practice conclusions of potentially avoidable admissions. The SEAs suggest that the pathway into unplanned admission may be less dependent on the condition than on context-specific factors, and the assumption that unplanned admissions for ACSC conditions are reasonable indicators of performance for primary care may not be valid.
Original languageEnglish
Article numbere020756
JournalBMJ Open
Issue number4
Publication statusPublished - 28 Apr 2018

Cite this