SMR01 and Associated Data Quality Assurance 2004 - 2006

Detail
The SMR01 and Associated Data Quality Assurance 2004 - 2006 project included several elements and the following areas were agreed with the relevant stakeholders. 

Hospital Assessment Details

SMR01 Random Sample
In line with the last SMR01 assessment, all hospitals with 3000 or more discharges per annum were assessed.   (An assessment of the hospitals with fewer than 3000 records is felt to require a separate project). The sample was approximately 5000 records across Scotland. This equated to 1.75% of quarterly returns for the larger hospitals.   A minimum of 75 records were assessed for smaller hospitals. 

The project assessed the accuracy of the following data items:

Consultant/Health Care Professional responsible for care
Waiting List Date
Main Condition
Other Conditions
Main Operation/Procedure
Date of Main Operation/Procedure
Clinician responsible for Main Operation/Procedure
Other Operations/Procedures

A total of 5430 records across 38 hospitals have now been assessed by the Data Quality Assurance team and the results were published in the cumulative Scotland report in September 2007.

Mental and Behavioural Disorders
At current around one third of recorded drug misuse episodes are coded to the non-specific F19 code denoting 'multiple/other' drug use.  DQA assessed the use of F19, with the objective of gathering information which could be used to improve specificity, and therefore the value, of drug misuse coding.  All records from 2004/2005 containing F19 code were sampled.  DQA assessed these records in any hospital with a minimum of 10 records up to a maximum of 100 to ascertain whether or not a more specific code from F10 to F18 could have been used instead of the non-specific F19.

Ethnic Group
As part of the pre-assessment process, DQA investigated the recording of Ethnic Group.  The hospital contact was asked if this information will be recorded, and if so, where it is found.

Clinical Coding Resources
DQA seeked to measure any correlation between the quality of coding, staff complement, length of service and provision of training.  During the pre-assessment DQA gathered data relating to the clinical coding resources in each hospital.

Revision Arthroplasty
A 100% sample of episodes (approximately 1000 records) with codes for revision arthroplasty and a discharge date in the year 2004/2005 was taken from the National SMR01 file by the Scottish Arthroplasty Project (SAP).  Using this sample, DQA assessed the accuracy of revision arthroplasty clinical coding, including the application of the April 2004 guidance on the coding of two-stage revisions, and also assessed the accuracy of recording of Consultant/HCP responsible for care, Clinician responsible for main operation and Date of main operation.  The finding of this data collection strand will not be included in the Scotland report.

Psychological Distress
Psychological distress is apparently much less recorded on SMR01 than might be expected given current research about the prevalence of such conditions in the population.  For each record in the SMR01 random sample, DQA when examining the documents pertaining to the particular episode being assessed, recorded the appearance of any of a set of clinical terms describing psychological distress. The findings of this data collection strand will be analysed at national level and will not be included in the Scotland report.

Community Health Index
For each record in the main SMR01 sample, DQA recorded the appearance (or not) of the CHI number on the patient case notes/identification labels, GP referral letters, discharge and clinic letters.  The finding of this data collection strand will not be included in the Scotland report.