Suchergebnisse

ABSTRACT
ObjectivesThis study shows how to improve PRL when using a limited number of personal identifiable information (PIIs) available in HES such as NHS number, DOB, Hospital Number, Gender and Postcode. The focus is on new approaches in categorisation of m-u probabilities for fields related to Postcode. In addition, a brief outline of our procedure to generate an appropriate training dataset will be presented as well as reasons for splitting HES dataset into single and multi-admissions, and proposed techniques to avoid computational cost.
ApproachPIIs such as DOB and Postcode can be categorised in order to increase the number of potential comparison vectors which is necessary in order to deal with the multi-million size of the HES dataset. We propose a multinomial approach for comparison between postcodes both for agreement and disagreement case scenarios, based on geographical information and a similarity score resulting from a Probit model. In the absence of a 'gold-standard' dataset, generating an appropriate training set is possible by combining two-way Deterministic Linkage (DL) and simple random sampling. Splitting HES into single and multi-admissions is recommended for reasons related to computational performance and demographic differences of the underlying population. Selection of the lower and upper thresholds regarding linkage weights can be done in a practical way by using conditional Normal mixture modelling, or more formally by using sampling techniques based on Fellegi-Sunter decision rule. We show how the latter can be further improved avoiding the computational cost by introducing appropriate mapping functions.
ResultsWe show the advantages of PRL over DL between Surveillance Laboratory Flu Data and HES. Preliminary results indicate an additional 10% of matches for < 1% expected false positives. EM convergence problems when treating missing values as separate outcome category will be discussed comparing results with the suggested re-normalisation approach in the bibliography.
ConclusionPRL between Laboratory Surveillance Data and Hospital Episode Statistics (HES) is a common practice and performed in a number of epidemiological studies conducted at Public Health England (PHE). Therefore, improving linkage will subsequently benefit those studies. Furthermore, this study proposes new approaches for Postcode comparison, and the utilisation of mapping functions to avoid the computational cost of the Fellegi-Sunter decision rule. These approaches could be used in a wide range of applications. Finally, we suggest a structural way of handling the HES dataset in order to achieve high efficiency and robustness of linkage.

ABSTRACTObjectivesThis study aims to describe the burden of respiratory syncytial virus (RSV) in children <5 years of age on secondary care in England using linked laboratory surveillance and hospital admissions databases. We will compare our results with previous estimates of the burden of RSV in secondary care based on statistical modelling techniques.
ApproachNational laboratory surveillance data was probabilistically linked to the Hospital Episode Statistics (HES) database for children aged <5 years in England from 01/08/2010 to 31/07/2012. Linkage was based on NHS and hospital number, date of birth, sex and post code. Only records where the laboratory test was within one week of hospital admission were included. Using the linked data, we estimated the probability of a respiratory hospital admission being related to RSV based on patient characteristics, calendar week and diagnostic coding. We used these probabilities to estimate the national incidence of RSV-associated hospital admissions. Admission rates were calculated by age group (<1 year, 1-4 years) using Office for National Statistics (ONS) mid-year population estimates.
ResultsThere was an average of 11,289 RSV-positive laboratory records per year during the study period. Preliminary results indicate that 75% of RSV-positive laboratory records linked to a hospital admission. 92% of linked hospital admissions occurred between November and March each year. 76% of the linked hospital admissions were in children aged <1 year. There were significant differences in coding of RSV-associated hospital admissions by age, and over two-thirds of RSV-associated hospital admissions did not have a primary diagnosis indicating RSV as the cause of disease. We will present results on the number of linked hospital admissions, unlinked hospital admissions potentially caused by RSV and the total estimated number of RSV-associated hospital admissions, stratified by age, calendar week and primary diagnosis.
ConclusionThis study is the first to link laboratory and hospital data for RSV in England, and therefore the first to describe laboratory-confirmed RSV-associated hospital admissions on a national scale. Our approach to estimate the national incidence of admissions based on developing probability weights from a smaller, linked sample has applications not just for RSV but for other seasonal infections. In addition, with a World Health Organization (WHO) consultation last year estimating that an RSV vaccine will be available commercially within 5-10 years, the results of this study can be used in mathematical models of potential vaccine strategy, contributing to the identification of optimal target groups for a potential licensed vaccine.