Adverse event categorisation across NHS Scotland.

OBJECTIVE This paper describes a scoping of adverse event categories and definitions in use across NHS Scotland and makes recommendations about the categories used. METHODS The design involved a network analysis making use of categories and subcategories supplied by 17 different Health Boards via their risk management/clinical governance functions. A total of 128 discrete categories and 1348 subcategories are used across 17 Health Boards (mean number of subcategories used: 164; SD 61; range 56-288). RESULTS There is some redundancy with around half of all categories (65; 51%) used by one single board and a further 10 (8%) by just two boards. Many health boards use similar subcategories within different categories. This shows conceptual overlap between confounded categories. This overlap has implications for reliability and has been quantified via the network analysis. CONCLUSION The eight most frequently used categories across the Health Boards have an average of 14.9 overlaps with other categories, suggesting some degree of variability in how adverse events are captured. This has implications for nationally collated data; however the formal links between categories identified in this work offer the opportunity of reliable data extraction at national level without substantial recoding by individual Health Boards.