What do we know about the quality of routinely collected accident and
emergency department data in the UK?
Coding methods and classification systems are found wanting in this
West-midlands study presented by
Dr Amy Downing and Dr Richard Wilson.
abstract
Little has been published about the quality of routinely collected
accident and emergency (A&E) data in the UK. This study aims to assess the
quality of this data in terms of completeness and validity.
Attendance data from 12 A&E departments in the West Midlands for the
period 1 April 2002–31 March 2003 was analysed (750 970 records). Where
information was submitted, completeness levels were over 95% except for
the investigation, diagnosis and treatment fields. Validity levels were
above 90% for all except one field (second diagnosis). Information was not
always submitted by all departments, due to difficulties in coding the
information recorded in the department to the national standards.
There are many recommendations that can be made to improve the
usefulness and reliability of the A&E data on which service planning and
commissioning increasingly depends.
Br J Healthcare Comput Info Manage 2004; 21(9): 24–7.
