|front |1 |2 |3 |4 |5 |6 |7 |8 |9 |10 |11 |12 |13 |14 |15 |16 |17 |18 |19 |20 |21 |22 |23 |24 |25 |review|
The purpose of the MCHP research report “Defining and Validating Chronic Diseases: An Administrative Data Approach”, was to investigate the validity of administrative data for chronic disease surveillance.
In order to use administrative data for chronic disease surveillance, a disease case definition needs to be constructed. A case definition is the set of rules to distinguish one disease from another. Essentially, applying a case definition to a databases filters the “hits” from the “misses”. These “hits” are the identified disease cases.
Some elements of a case definition include: the type of data source, the number of years to include, diagnostic/treatment code(s), and the number of ‘contacts’ an individual has with a specific diagnostic/treatment code. Studies have suggested that the ability of a case definition to identify a disease from administrative data is sensitive to how elements of these case definitions are manipulated (Robinson et al., 1997; Powell, 2003; Rector et al., 2004). One solution is to apply multiple case definitions to administrative data to produce a probable range of estimates.
Three sources of population-based administrative data were used to construct the case definitions – 1) The physician database, which captures billing information from all fee-for-service physicians and the majority of salaried physicians in the province of Manitoba, Canada; 2) The hospital database, which records all separations from acute care facilities in Manitoba, Canada; and 3) The Drug Programs Information Network (DPIN), which is a centralized database that captures pharmaceutical dispensations from all retail pharmacy outlets in Manitoba, Canada. Diagnoses in the physician and hospital data are recorded using ICD-9-CM. Prescription drugs can be linked to Anatomic, Therapeutic, Chemical (ATC) codes.