Accuracy of cancer registration

Area of Science:

• Public Health
• Cancer Epidemiology
• Health Informatics

Background:

• Cancer registration is crucial for epidemiological studies and health planning.
• Hospital Activity Analysis (HAA) is a primary method for cancer registration in South Wales.
• Accurate cancer data is essential for monitoring disease trends and evaluating interventions.

Purpose of the Study:

• To assess the accuracy of cancer registration data in South Wales.
• To identify specific areas of error and omission in the registration process.
• To evaluate the quality of data used for cancer surveillance by the Office of Population Censuses and Surveys (OPCS).

Main Methods:

• A study of 1460 hospital cancer records in South Wales.
• Comparison of 're-registrations' from HAA with original registrations at the Welsh Hospital Board's cancer bureau.
• Item-by-item analysis of differences in diagnostic summaries, dates, and demographic information.

Main Results:

• Significant errors were found in 234 diagnostic summaries, 164 dates of birth, and 198 dates of registration.
• Some errors in diagnostic codes could lead to misclassification under the International Classification of Diseases (ICD).
• High error and omission rates were noted for National Health Service (NHS) number, occupation, place of birth, and histology.

Conclusions:

• The quality of cancer registration data in South Wales requires improvement.
• Errors in key data fields can impact the accuracy of cancer statistics and epidemiological research.
• Further efforts are needed to enhance data collection and validation processes for cancer registries.