BINOCAR aims to provide data of the highest possible quality. Whilst each register is, to some extent, dependent upon the quality of the data supplied to it, there are a number of steps which can be taken to ensure published anomaly data is as complete and accurate as possible.
Each regional register, depending upon its individual circumstances, regularly assesses the extent to which it has captured all the anomaly cases within its area. This presents a number of difficulties, as the extent, and accuracy, of what is reported can vary with individual anomalies, clinical specialities, hospital units, geographical location, or reporting systems. Routine audits are undertaken to compare register data with, for example, hospital in-patient data, other registers' data, paediatric surgical databases, cyto- or clinical genetics databases, data obtained from research projects or as a result of clinical audits, to ensure that as many cases as possible are included.
Prevalence rates, usually expressed as the number of instances of a particular anomaly occurring per 10,000 live and still births per annum, is a useful method for comparison between registers, and is often used as a proxy measure for data completeness. Registers which collect similar data in similar ways (such as those belonging to the EUROCAT network) can more reliably be compared in this way.
Care should be taken when attempting to interpret data derived from different registers which have been collected in different ways. Confusion can arise over the number of cases (babies) or anomalies (each baby can have more than one), or between the inclusion (or not) of terminations and fetal losses, the inclusion (or not) of addtional anomalies recorded in cases of chromosomal anomaly (such as cardiac defects in children with Trisomy 21), the distinction between isolated and/or multiple anomalies and the standard definitions of stillbirth/fetal loss (gestational age or weight or a combination of both) and so on.
The EUROCAT network also publishes Data Quality Indices for each of its member registries, assessing both the quality of data, and the extent of 'missing' data, submitted over the most recent 5 year period.