What the data tells you?
The Patient Episode Database for Wales (PEDW) records all episodes of inpatient and daycase activity in NHS Wales hospitals. This includes planned and emergency admissions, minor and major operations, and hospital stays for giving birth. Hospital activity for Welsh residents treated in hospitals in England is also included.
How are the data collected?
The data are collected and coded at each hospital. The records are then electronically transferred to the NHS Wales Informatics Service (NWIS)*, where they are validated and merged into the main database.
* previously known as Health Solutions Wales (HSW)
Diagnoses are coded using the International Classification of Diseases (ICD). Further details on ICD can be accessed: http://www.who.int/classifications/icd/en/index.html
Procedures are coded using the OPCS classification of surgical operations.
How accurate or complete will the data be?
PEDW should be a complete record of all inpatient and daycase activity in NHS Wales hospitals.
However, NWIS acknowledge that “While the PEDW team liaise closely with the NHS in order to maintain data quality and consistency, it is inevitable in such a complex undertaking that a few errors will occur.” (extract from NWIS's PEDW notes and definitions) Such errors could include shortfalls in the number of records submitted, or missing/invalid clinical information. Adjustments are not made for these errors, but considering that PEDW comprises around one million episodes of hospital care per year, only a small proportion of the data is affected.
Also, where there is no definitive diagnosis within a hospital episode, the main symptom, abnormal finding or problem is recorded as the primary diagnosis.
Analyses produced using PEDW should be interpreted with a degree of caution, considering the potential for minor data errors, uncertain diagnoses and the broader issues listed below.
PEDW records are coded within the hospitals across Wales by clinical coders. If it is not obvious in a patient's notes, the coder may have to make a judgement on what they consider to be the primary diagnosis being cared for in each finished consultant episode for that patient.
Are there any problems associated with the data?
The data held in PEDW is of interest to public health services since it can provide information regarding both health service utilisation and also the incidence and prevalence of disease. However, since PEDW was created to track hospital activity from the point of view of payments for services, rather than epidemiological analysis, the use of PEDW for public health work is not straightforward. For example:
Coding practices vary. In particular, coding practices for recording secondary diagnoses is likely to vary for different hospitals. This makes regional variations more difficult to interpret. The validation process led by the Corporate Health Improvement Programme and implemented by NWIS is aiming to address some of these inconsistencies.
Due to the complexity and pitfalls of PEDW it is recommended that any PEDW requests for public health purposes are discussed with a member of the Observatory. In turn the Obseravtory will seek advice from NWIS if required.
Where can you get hold of further information?
NWIS publish annual PEDW data tables here.
Further notes and definitions surrounding PEDW data can be found here.