Suchergebnisse

ABSTRACTObjectivesData linkage algorithms are used to link together multiple episodes of care belonging to the same patient. For example, the HESID algorithm is used to generate Hospital Episode Statistics (HES) in England. HESID is a deterministic algorithm, requiring identifiers to agree or disagree at each step. Data linkage errors occur when episodes belonging to two patients are incorrectly linked (a false match) or when episodes belonging to the same patient are not linked (a missed match). This typically occurs because patient identifiers (e.g. NHS number, postcode) contain errors or have missing data. We previously showed that HESID has a low false match rate (0.2%) but a high missed match rate (4.1%) when applied to paediatric intensive care data. This biased the true readmission rate, particularly for some patient groups including ethnic minorities. The aim of our study was to evaluate whether an additional step involving probabilistic matching would lower the missed match rate in HES without increasing the false matched rate.
ApproachWe simulated three datasets having the same characteristics as HES, for three age groups expected to have different levels of postcode stability (at age 0/1, 5/6 and 18/19). We compared the deterministic algorithm to a probabilistic algorithm, and then to a deterministic algorithm with an additional probabilistic step. In sensitivity analyses, we evaluated the algorithms under different data quality scenarios.
ResultsResults show that deterministic followed by probabilistic matching is the best solution for reducing missed matches, particularly in scenarios where errors in patient identifiers are more common.
ConclusionData linkage algorithms need to be evaluated against good quality reference standard data sets. For hospital data in England, the Personal Demographics Service (PDS) could be used to evaluate our approach, because it contains many of the same patient identifiers used in HES. Reducing data linkage error will improve monitoring of hospital activity in England.

Record linkage of administrative and survey data is increasingly used to generate evidence to inform policy and services. Although a powerful and efficient way of generating new information from existing data sets, errors related to data processing before, during and after linkage can bias results. However, researchers and users of linked data rarely have access to information that can be used to assess these biases or take them into account in analyses. As linked administrative data are increasingly used to provide evidence to guide policy and services, linkage error, which disproportionately affects disadvantaged groups, can undermine evidence for public health. We convened a group of researchers and experts from government data providers to develop guidance about the information that needs to be made available about the data linkage process, by data providers, data linkers, analysts and the researchers who write reports. The guidance goes beyond recommendations for information to be included in research reports. Our aim is to raise awareness of information that may be required at each step of the linkage pathway to improve the transparency, reproducibility, and accuracy of linkage processes, and the validity of analyses and interpretation of results.

Record linkage of administrative and survey data is increasingly used to generate evidence to inform policy and services. Although a powerful and efficient way of generating new information from existing data sets, errors related to data processing before, during and after linkage can bias results. However, researchers and users of linked data rarely have access to information that can be used to assess these biases or take them into account in analyses. As linked administrative data are increasingly used to provide evidence to guide policy and services, linkage error, which disproportionately affects disadvantaged groups, can undermine evidence for public health. We convened a group of researchers and experts from government data providers to develop guidance about the information that needs to be made available about the data linkage process, by data providers, data linkers, analysts and the researchers who write reports. The guidance goes beyond recommendations for information to be included in research reports. Our aim is to raise awareness of information that may be required at each step of the linkage pathway to improve the transparency, reproducibility, and accuracy of linkage processes, and the validity of analyses and interpretation of results.

Statistical data and evidence-based claims are increasingly central to our everyday lives. Critically examining 'Big Data', this book charts the recent explosion in sources of data, including those precipitated by global developments and technological change. It sets out changes and controversies related to data harvesting and construction, dissemination and data analytics by a range of private, governmental and social organisations in multiple settings. Analysing the power of data to shape political debate, the presentation of ideas to us by the media, and issues surrounding data ownership and access, the authors suggest how data can be used to uncover injustices and to advance social progress