Problems extracting data from hospital maternity records

Publisher Website
Google Scholar
Abstract
Information extracted from hospital maternity records was first compared with information obtained from postal questionnaires to mothers. A check back to the notes when mothers reported events which had not been extracted revealed a number of omissions by the extractors and some ambiguities in the records. A second check was made by comparing data extracted by researchers with that done routinely by hospital staff for computer analysis. This revealed substantial discrepancies over a number of items. Further exploration indicated that extractors more often omitted events than identified them wrongly. Routine extractors tended to rely on summaries so events not included in summaries were liable to be overlooked. It is concluded that statistics based on extractions from records will tend to underestimate the frequency with which complications occur and procedures are carried out.

This publication has 0 references indexed in Scilit: