SEARCH · PubMed Health

Results for “record linkage”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 37 records · Page 2Linked to original sources

PubMed1997

The diabetes audit and research in Tayside Scotland (DARTS) study: electronic record linkage to create a diabetes register. DARTS/MEMO Collaboration.

OBJECTIVES: To identify all patients with diabetes in a community using electronic record linkage of multiple data sources and to compare this method of case ascertainment with registers of diabetic patients derived from primary care. DESIGN: Electronic capture-recapture linkage of records included data on all patients attending hospital diabetes clinics, all encashed prescriptions for diabetes related drugs and monitoring equipment, all patients discharged from hospital, patients attending a mobile unit for eye screening, and results for glycated haemoglobin and plasma glucose concentrations from the regional biochemistry database. Diabetes registers from primary care were from a random sample of eight Tayside general practices. A detailed manual study of relevant records for the 35,144 patients registered with these eight general practices allowed for validation of the case ascertainment. SETTING: Tayside region of Scotland, population 391,274 on 1 January 1996. MAIN OUTCOME MEASURES: Prevalence of diabetes; population of patients identified by different data sources; sensitivity and positive predictive value of ascertainment methods. RESULTS: Electronic record linkage identified 7596 diabetic patients, giving a prevalence of known diabetes of 1.94% (0.21% insulin dependent diabetes, 1.73% non-insulin dependent): 63% of patients had attended hospital diabetes clinics, 68% had encashed diabetes related prescriptions, 72% had attended the mobile eye screening unit, and 48% had biochemical results diagnostic of diabetes. A further 701 patients had isolated hyperglycaemia (plasma glucose > 11.1 mmol/l) but were not considered diabetic by general practitioners. Validation against the eight general practices (636 diabetic patients) showed electronic linkage to have a sensitivity of 0.96 and a positive predictive value of 0.95 for ascertainment of known diabetes. General practice lists had a sensitivity of 0.91 and a positive predictive value of 0.98. CONCLUSIONS: Electronic record linkage was more sensitive than general practice registers in identifying diabetic subjects and identified an additional 0.18% of the population with a history of hyperglycaemia who might warrant screening for undiagnosed diabetes.

Diabetes Mellitus↗

PubMed2002

An empirical comparison of record linkage procedures.

We consider the problem of record linkage in the situation where we have only non-unique identifiers, like names, sex, race etc., as common identifiers in databases to be linked. For such situations much work on probabilistic methods of record linkage can be found in the statistical literature. However, although many groups undoubtedly still use deterministic procedures, not much literature is available on deterministic strategies. Furthermore, there appears to exist almost no documentation on the comparison of results for the two strategies. In this work we compare a stepwise deterministic linkage strategy with a probabilistic strategy, as implemented in AUTOMATCH, for a situation in which the truth is known. The comparison was carried out on a linkage between medical records from the Regional Perinatal Intensive Care Centers database and educational records from the Florida Department of Education. Social security numbers, available in both databases, were used to decide the true status of each record pair after matching. Match rates and error rates for the two strategies are compared and a discussion of their similarities and differences, strengths and weaknesses is presented.

Child↗

PubMed1990

Development of a record linkage protocol for use in the Dutch Cancer Registry for Epidemiological Research.

A method has been developed to determine the optimal linkage key for record linkage between the cancer registry and a large-scale prospective cohort study in the Netherlands. The proposed linkage procedure is a two-stage process in which the initial computerized linkage using a particular linkage key is followed by visual inspection with additional information to separate the computer matches into true and false positives. In the determination of the optimal key, both informativeness and susceptibility to error of personal identifiers were taken into account. The performance of the various keys in the linkage was expressed in terms of sensitivity and predictive value of a reported computer match. The key, consisting of date of birth, first four characters of the family name and gender was the optimal choice, with a sensitivity of 98% and an initial predictive value of a computer match of 98%. When additional information on migration, place of birth and first initial was collected in the second stage, it was possible to eliminate the false positives from the reported computer matches without loss of true positives. Thus, the sensitivity remained constant whereas the secondary predictive value of accepted matches was maximized.

Databases, Factual↗

PubMed2003

[Use of medical record linkage in epidemiologic studies].

The article presents two methods of records linkage, deterministic and probabilistic, applied in epidemiological studies and medical registers. Sources of selection bias linkage was considered. The modified deterministic linkage by random selection of one from non-unique records was elaborated. The quality of applied method of record linkage in survival analysis was empirically evaluated. Usefulness of record linkage in epidemiology was talked over.

Adolescent↗

PubMed2001

Electronic health summaries--building on the foundation of Scottish Record Linkage system.

Developing electronic patient records is currently a priority for Health Services, so that the clinician treating the patient has all the relevant information available on the desktop. However, as with paper records, there is the danger of information overload, with key information being missed. It is now recognized that electronic health summaries of patient events will be needed. This paper describes the Scottish Record Linkage system, which has been routinely used over the last 10 years to build event histories for patients in the NHS in Scotland. Probability matching has been used to identify records belonging to an individual since a unique patient identification number has not been in general use. We present examples of use of this rich information resource, either for clinical decision making, evidence based medicine, or population needs assessment, and investigate how the system may be developed in a data-warehousing environment to populate electronic health summaries for the five million residents of Scotland.

Decision Making↗

PubMed1980

Use of automated record linkage to measure patient fertility after family planning service.

Record linkage was used to identify births in 1974 to patients who had received contraceptive services from public family planning clinics in metropolitan Atlanta. The record linkage methodology is described. Results indicated that patient educational level, age, contraceptive choice, and race were the factors most associated with subsequent patient fertility. Black teenage patients experienced the highest fertility. Patients age 20 years and over with some education beyond high school experienced approximately one-third the fertility of less-educated patients. The program effectiveness of the IUD and the Pill was substantially better than that of foam and condoms. Excluding teenagers, the program effectiveness of the IUD was better than the Pill. Changes in family planning services for high-risk patients may help them achieve their fertility control objectives. Linking family planning clinic records to official birth records as described in this study can help determine clinic success in meeting patient objectives. Other applications of record linkage are discussed briefly.

Adolescent↗

PubMed1988

An experience in psychiatric record linkage.

This paper describes a personal experience in setting up a psychiatric record linkage system in an Eastern Ontario city. It discusses the rationale, background and methodology of the Kingston Psychiatric Record Linkage System and includes a detailed description of the practical issues encountered in its establishment and operation. The issues include funding, cooperation of local facilities, system start-up, operating costs, and data collection, linkage, entry, processing, storage and analysis. The limitations and uses of this system are discussed. Potential problems such as financial support and reliance on others have been resolved and the system has been functioning remarkably well since 1984. Published reports from other record linkage systems do not include such mundane details which would have been useful to know before planning this project. It is hoped that this paper will be beneficial to others who are interested in record linkage.

Costs and Cost Analysis↗

PubMed1995

Feasibility study for identifying adverse events attributable to vaccination by record linkage.

To investigate the feasibility of using a record linkage method for identifying vaccine attributable adverse events, computerized hospital admissions and vaccination records from South East Kent district were linked and checked for accuracy. Records for 90% of children under 2 years of age admitted to hospital over a 2-year period were matched with vaccination records using a computer algorithm based on name, date of birth, sex, and post-code supplemented by visual inspection. Relative to this gold standard, matching on date of birth, sex and postcode alone had a sensitivity of 60% and an incorrect match rate of 0.2% after matches to more than one vaccine recipient were excluded. Manual checking of a sample of admissions showed that only 4% had been assigned incorrect International Classification of Disease (ICD) codes. Routine record linkage of ICD admission codes to vaccination records therefore yields data of good quality which may be used for surveillance purposes.

Adverse Drug Reaction Reporting Systems↗

PubMed2002

Probabilistic record linkage and a method to calculate the positive predictive value.

BACKGROUND: Computerized record linkage is commonly used in cohort studies to ascertain the study outcome, and as such its accuracy classifying the outcome can be described using the standard epidemiological terms of sensitivity and positive predictive value (PPV). METHOD: We describe a 'duplicate method' to calculate the PPV of record linkage when each record can only be involved in one match (e.g. linking population files to death files). The method does not require a validation subset of records from both files with detailed personal information (e.g. name and address), and is therefore ideal for linkage projects using anonymous data. The duplicate method assumes that the number of records from one file with zero, one, two, etc., links from the other file is distributed in a manner predicted by combinatorial probabilities. Having made this assumption, the number of false positive links, and hence the PPV, are estimable. We demonstrate this duplicate method using output from anonymous and probabilistic record linkage of census and mortality records in New Zealand. RESULTS: The PPV estimates conform to the pattern expected based on the underlying theory of probabilistic record linkage, and were robust to sensitivity analyses. We encourage other researchers to further assess the accuracy of this method.

Epidemiologic Methods↗

PubMed2005

The EpiLink record linkage software: presentation and results of linkage test on cancer registry files.

OBJECTIVES: Record linkage, the process of bringing together separately compiled but related records from different databases, is essential in many areas of biomedical research. We developed a record linkage program (EpiLink), which employs a simple mathematical approach. We describe the program and present results obtained testing it in a linkage task. METHODS: EpiLink was designed to be flexible with user-friendly settings to tailor linkage and operating parameters to specific linkage tasks, and employ deterministic, probabilistic or sequential deterministic-probabilistic linkage strategies as required. The user can also standardize data format, examine linkage results and accept or discard them. We used EpiLink to link a subset of cases of the Lombardy Cancer Registry (20,724 records) with the Social Security file of the population (1,021,846 records) covered by the registry. The linkage strategy was deterministic, followed by several probabilistic linkage steps. RESULTS: Manual inspection of the results showed that EpiLink achieved 98.8% specificity and 96.5% sensitivity. CONCLUSIONS: EpiLink is a practical and accurate means of linking records from different databases that can be used by non-statisticians and is efficient in terms of human and financial resources.

Humans↗

PubMed1997

Effects of record linkage errors on registry-based follow-up studies.

The importance of reliable record linkage for high quality-population-based disease registration is widely recognized. Systematic methodologic work is lacking, however, on the effects of record linkage errors on the use of disease registries for epidemiologic purposes. The present paper provides algebraic models describing the effects of record linkage errors on monitoring survival of registered patients, which is commonly performed by matching registry records against a database of death certificates, and on registry-based incidence follow-up of external cohorts. Homonym errors, that is, erroneous linkage of records that pertain to distinct individuals, lead to underestimation of survival of registered patients and to overestimation of disease incidence among external cohorts linked to the registry, while synonym errors, that is, failure to link notifications on the same individual, have opposite effects. The magnitude of effects of record linkage errors are illustrated with numerical examples.

Bias↗

PubMed1988

Record linkage studies for postmarketing drug surveillance: data quality and validity considerations.

Large automated databases are the source of information for many record linkage studies, including postmarketing drug surveillance. Despite this reliance on prerecorded data, there have been few attempts to assess data quality and validity. This article presents some of the basic data quality and validity issues in applying record linkage methods to postmarketing surveillance. Studies based on prerecorded data, as in most record linkage studies, have all the inherent problems of the data from which they are derived. Sources of threats to the validity of record linkage studies include the completeness of data, the ability to accurately identify and follow the records of individuals through time and place, and the validity of data. This article also describes techniques for evaluating data quality and validity. Postmarketing surveillance could benefit from more attention to identifying and solving the problems associated with record linkage studies.

Database Management Systems↗

PubMedDate not supplied

[The importance of the common cancer registry for the identification of cancer cases in the EPIC Potsdam-study -- results of the first record linkage].

AIM: The follow-up of the European Prospective Investigation into Cancer and Nutrition (EPIC)-Potsdam study has been identifying cancer diagnoses by an active follow-up through participant's self-report which is then confirmed by pathology registries or medical doctors. To utilise another source of information, a record linkage was performed with the Common Cancer Registry (CCR) of the states of Berlin, Brandenburg, Mecklenburg-Western Pomerania, Saxony-Anhalt and of the free states Saxony and Thuringia in November 2001, using the EPIC-Potsdam database of compliant study participants at that point in time (n = 27 087). This methodological study compares data of identified cancer cases in the EPIC-Potsdam study with those generated through the record linkage. METHODS: The GKR identified 1468 study participants of the EPIC-Potsdam study. Cases (n = 108) with non-melanoma skin cancer were excluded. A separate analysis was done for prevalent (date of cancer diagnosis before baseline examination) and incident cancers (date of diagnosis after baseline examination). We examined how many subjects with at least one cancer were known to EPIC-Potsdam and the CCR, how many had been known only to the CCR and how many only to the EPIC-Potsdam study. Of incident cancer cases each diagnosis was counted. It was analyzed whether the diagnoses as they had been collected by EPIC-Potsdam were identified by the record linkage. Focus was on the diagnoses that had been unknown in EPIC-Potsdam before. The concordance of information from both sources was also considered. RESULTS: According to the record linkage, 1298 subjects with at least one prevalent cancer were identified in the EPIC-Potsdam cohort, 593 (46 %) of whom were identified in both data sources, 241 (19 %) were newly identified, and 464 (36 %) participants self-reported at least one prevalent tumour which was not identified through the record linkage. Seven tumours were diagnosed before 1961 (the CCR's computerised data are only available since 1961). Fifty-seven percent (n = 187) of all tumours which could not be identified in the cancer registry dated from 1990 - 1995. After completion of the record linkage a total of 582 medically verified incident cancer cases were identified. These comprised 397 medically verified cases that were identified in the EPIC-cohort study prior to the record linkage, 18 self-reported cases where the medical verification was pending, and 167 newly identified cases. Eighty-five of the latter cases have not been identified by other sources in the two following years after record linkage indicating that 15 % of the medically verified cases have been found solely through the linkage. CONCLUSION: The EPIC-Potsdam study had a significant benefit of the record linkage due to both the additional identification and verification of prevalent and incident cancer cases, as well as a timely advanced identification of incident diagnoses. The record linkage with the cancer registry improved the quality of the endpoint data of the EPIC-Potsdam study.

Adult↗

PubMed1998

How to ensure data security of an epidemiological follow-up: quality assessment of an anonymous record linkage procedure.

A computerised record hash coding and linkage procedure is proposed to allow the chaining of medical information within the framework of epidemiological follow-up. Before their extraction, files are rendered anonymous using a one-way hash coding based on the standard hash algorithm (SHA) function, in order to respect the legislation on data privacy and security. To avoid dictionary attacks. two keys have been added to SHA coding. Once rendered anonymous, the linkage of patient information can be accomplished by means of a statistical model, taking into account several identification variables. Quality assessment of this anonymous record linkage procedure shows a specificity of 100% and a sensitivity of 95%.

Algorithms↗

PubMed1984

Postmarketing drug surveillance by record linkage in Tayside.

The feasibility of conducting postmarketing drug surveillance by record linkage in Tayside was assessed. The key feature of the method is that all hospital discharge data are already computerised by the area health board and may be accessed through the unique community health number (CHNo) which has been allocated to all Tayside residents. The 12 861 prescriptions for cimetidine dispensed in Tayside over a nine month period were collected and the CHNo identified for 76%. These corresponded to 3802 individuals and their discharge data, together with those for an equal number of controls matched by age, sex, and general practitioner were retrieved from the computer. The expected excess of those diseases for which cimetidine is prescribed (peptic ulcer and oesophagitis) was observed. Other drug-disease associations were observed but may have been due to confounding and emphasise the inadequacy of community based controls. The major advantages of record linkage are, firstly, the low cost of the method, the present study costing just over pounds 12 000, and, secondly, the duration of patient follow up which may may be extended for as many years as required simply by rerunning the computer programs. To assemble large patient cohorts the study would have to be extended to other area health boards that are currently developing similar computer systems. Record linkage may provide a cost effective method for the follow up of patients to identify serious adverse drug reactions, particularly those that take several years to develop.

Cimetidine↗

PubMed2006

Completeness of notification of tuberculosis in The Netherlands: how reliable is record-linkage and capture-recapture analysis?

The aim of this study was to describe a systematic process of record-linkage, cross-validation, case-ascertainment and capture-recapture analysis to assess the quality of tuberculosis registers and to estimate the completeness of notification of incident tuberculosis cases in The Netherlands in 1998. After record-linkage and cross-validation 1499 tuberculosis patients were identified, of whom 1298 were notified, resulting in an observed under-notification of 13.4%. After adjustment for possible imperfect record-linkage and remaining false-positive hospital cases observed under-notification was 7.3%. Log-linear capture-recapture analysis initially estimated a total number of 2053 (95% CI 1871-2443) tuberculosis cases, resulting in an estimated under-notification of 36.8%. After adjustment for possible imperfect record-linkage and remaining false-positive hospital cases various capture-recapture models estimated under-notification at 13.6%. One of the reasons for the higher than expected estimated under-notification in a country with a well-organized system of tuberculosis control might be that some tuberculosis cases, e.g. extrapulmonary tuberculosis, are managed by clinicians less familiar with notification of infectious diseases. This study demonstrates the possible impact of violation of assumptions underlying capture-recapture analysis, especially the perfect record-linkage, perfect positive predictive value and absent three-way interaction assumptions.

Disease Notification↗

PubMed2001

Record linkage of healthcare insurance claims.

Limitations of current record linkage techniques include difficulty in handling large and heterogeneous data sets, low sensitivity in deterministic matching, the necessity to provide a priori weights for probabilistic matching, low computational efficiency, and complex software. This paper provides a detailed description of a method developed for purposes of linking records of individuals across time and geography. The procedure for record-linkage consists of three major components: data standardization, weight estimation, and matching. The proposed method was designed to incorporate a combination of exact and probabilistic matching techniques. The procedure was validated using convergent, divergent, and criterion- validity measures. The authors feel that the procedure outlined in this paper is a first step in addressing the current trend toward larger and more complex databases.

Health Services Research↗

PubMed2004

Probabilistic record linkage and an automated procedure to minimize the undecided-matched pair problem.

Probabilistic record linkage allows the assembling of information from different data sources. We present a procedure when a one-to-one relationship between records in different files is expected but not found. Data were births and infant deaths, 1998-birth cohort, city of São Paulo, Brazil. Pairs for which a one-to-one relationship was obtained and a best-link was found with the highest weight were taken as unequivocally matched pairs and provided information to decide on the remaining pairs. For these, an expected relationship between differences in dates of death and birth registration was found; and places of birth and death registration for neonatal deaths were likely to be the same. Such evidence was used to solve for the remaining pairs. We reduced the number of non-uniquely matched records and of uncertain matches, and increased the number of uniquely matched pairs from 2,249 to 2,827. Future research using record linkage should use strategies from first record linkage runs before a full clerical review (the standard procedure under uncertainty) to efficiently retrieve matches.

Birth Rate↗