Accuracy of Public Health Data Linkages

Qayad, Mohamed; Zhang, Hui

doi:10.1007/s10995-008-0377-6

Cited by 11 publications

(11 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, it is not always possible to trace all individuals; consequently the active follow up is also prone to errors 19,24,25 . Another strategy is to manually inspect a random sample of links designated as matches and non-matches by two independent reviewers with the human judgment being considered the gold standard 26 . This procedure might be timeconsuming and because of its subjective nature, it is also subject to error.…”

Section: Discussionmentioning

confidence: 99%

Accuracy of a probabilistic record linkage strategy applied to identify deaths among cases reported to the Brazilian AIDS surveillance database

et al. 2010

View full text Add to dashboard Cite

Since record linkage errors can bias measures of disease occurrence and association, it is important to assess their accuracy. The aim of this study is to assess the accuracy of a multiple pass probabilistic record linkage strategy to identify deaths among persons reported to the Brazilian AIDS surveillance database. An HIV/AIDS national surveillance database (N = 559,442) was linked to a total of 6,444,822 deaths registered (all causes) in the Brazilian mortality database. To estimate standard measures of accuracy, we selected all AIDS cases with a date of death registered in the surveillance database from 2002 to 2005 (N = 19,750) and 38,675 cases known to be alive in 2006. The linkage strategy presented a sensitivity of 87.6% (95%CI: 87.1-88.2), a specificity of 99.6% (95%CI: 99.6-99.7), and a positive predictive value of 99.2% (95%CI: 99.1-99.3). We observed a small variation in the validity measures according to some putative predictors of mortality. Our findings suggest that even large and heterogeneous databases can be linked with a satisfactory accuracy.

show abstract

Section: Discussionmentioning

confidence: 99%

Accuracy of a probabilistic record linkage strategy applied to identify deaths among cases reported to the Brazilian AIDS surveillance database

et al. 2010

View full text Add to dashboard Cite

show abstract

“…Qayad et al compared the PRAMS data set to birth certificate data and reported 90% agreement between the PRAMS survey and the computer-matched files. 6 The information in the PRAMS database is gathered from surveys that are sent to women who deliver a live-born infant within New Jersey. The New Jersey PRAMS questionnaire is built around a mandatory core developed by CDC.…”

Section: Methodsmentioning

confidence: 99%

Obstetric Outcomes in Normal Weight and Obese Women in Relation to Gestational Weight Gain: Comparison between Institute of Medicine Guidelines and Cedergren Criteria

et al. 2009

View full text Add to dashboard Cite

We compared obstetric outcomes based on gestational weight gain in normal-weight and obese women using traditional Institute of Medicine (IOM) guidelines and newly recommended Cedergren criteria. Using the New Jersey Pregnancy Risk Assessment Monitoring System (PRAMS) database and electronic birth records, perinatal outcomes were analyzed to estimate the independent effects of prepregnancy body mass index (BMI) and gestational weight gain by IOM versus Cedergren criteria. Of 9125 subjects in PRAMS database from 2002 to 2006, 53.7% had normal BMI, 12.3% were overweight, 18.2% were obese, and the rest were underweight. Among normal-weight mothers, when compared with the IOM guidelines, macrosomia (6.45% versus 4.27%) and cesarean delivery rates (30.42% versus 29.83%) were lower using Cedergren criteria but the rates of preterm delivery (5.06% versus 9.44%), low birth weight (0.38% versus 2.42%), and neonatal intensive care unit (NICU) admissions (7.02% versus 10.86%) were higher with the Cedergren criteria. Similarly, among obese patients, when compared with IOM guidelines, macrosomia (10.79% versus 5.47%) and cesarean delivery rates (43.95% versus 40.71%) were lower using Cedergren criteria but the rates of preterm delivery (6.83% versus 8.32%), low birth weight (0.87% versus 1.88%), and NICU admissions (8.92% versus 13.78%) were higher with the Cedergren criteria. Based on our results, ideal gestational weight gain is presumably somewhere between the IOM and Cedergren's guidelines.

show abstract

“…184,198,200,201 In probabilistic linkage, a pair of records from two datasets are classified as belonging to the same individual, or not, based on the statistical probability that common identifiers in the datasets belong to the same entity, using weights or scores to gauge the agreement and disagreement between the records. 184,[193][194][195]202,203 The statistical probability, and not the evidence of an exact match, as needed in deterministic approaches, is then used to decide whether a pair of records refers to the same entity. Probabilistic linkage is therefore more suitable in exercises where unique identifiers are not available, or where there is evidence of variation in reporting, coding or transcription of identifiers.…”

Section: Record Linkage Approachesmentioning

confidence: 99%

“…203 In particular, advances in the affordability and capacity of computing equipment were an enabling factor. 175 Linkage activities moreover increased as a result of researchers and research institutions departing from traditional methods that are prohibitively resource-consuming-in terms of financial costs, and human-resource and infrastructural investments-and likely subject to complicating factors referred to before.…”

Section: Increased Linkage: Reasons and Enabling Factorsmentioning

confidence: 99%

“…For such studies, different source materials and approaches have been used to identify mis-attribution, errors, or inconsistencies on individual death certificates, and to form an opinion of the accuracy of cause-of-death information at the population level. Source materials for comparing, evaluating or validating certified causes of death include medical records and clinical notes assessed by clinical review committees or review panels comprising disease specialists or experts; 231,[240][241][242][243][244][245][246][247] autopsy reports and pathology results, [248][249][250] police, coroner and medical examiner reports; 251,252 interviews with the certifying physician 253,254 or the relatives of the deceased; 242,255,256 health insurance claims; 184,203,257 verbal autopsy questionnaires; 258 entries into disease-specific registers; 259,260 participant records in follow-up studies; 183,223 or combinations of these. 183,203,242,244,256,257,[261][262][263] …”

Section: Validation Studies Against a Reference Standardmentioning

confidence: 99%

See 1 more Smart Citation

Availability and quality of vital registration mortality data in South Africa: a national and surveillance site perspective

Joubert¹

View full text Add to dashboard Cite

The importance of valid and reliable mortality data as a key input for public and population health planning and research, and development programmes, is globally acknowledged.However, few low-and middle-income countries have such data. Even fewer have assessed the quality of their mortality data. Despite the particular utility of periodic, independent evaluations of national mortality statistics by countries themselves, such an evaluation has not yet been conducted for South Africa. Neither has a review been conducted of the sources of mortality data in South Africa since 1995. Regarding causes of death, a few small-scale studies assessed data quality in urban settings, but no study yet in civil registration mortality data in a rural setting. In addition, notwithstanding district health managers' critical need to know the health care needs of their district population current district mortality profiles are accompanied by strong warnings of invalid data.These gaps in essential public-and population-health intelligence present important reasons for research on the availability and quality of mortality data in South Africa.The primary aim of this body of research was to assess the quality of mortality data in the national CRVS system of South Africa. As alternative mortality data sources can complement or evaluate the quality of CRVS data, a secondary aim was to identify and review alternative mortality data sources, and assess the availability, strengths, and limitations of their data. A further aim was to conduct a field study to link and compare CRVS and Agincourt HDSS mortality data for the same individuals with the purpose of quantifying the level of completeness of death registration into the CRVS system, and assess the level of agreement between CRVS and verbal autopsy cause-of-death data. Seven research questions have been addressed towards achieving these aims. Different approaches were followed, including: (a) a mortality data source review and appraisal of the availability, strengths and limitations of different sources; (b) a comprehensive evaluation of national CRVS mortality data based on demographic and epidemiological principles; (c) a detailed record-linkage study applying both deterministic and probabilistic approaches; and (d) an application of adjustment factors from the linkage study to improve district-level cause-of-death data quality.iii The data source review found a rich and varied list of mortality data sources. Recent transformation in South Africa has enhanced mortality data availability from CRVS. Detailed cause-of-death data are available from civil registration and demographic surveillance.Additionally, a range of mortality data items are available in three population censuses, three demographic surveillance systems, and various national surveys, mortality audits and disease-notification programmes. Employing nine criteria to evaluate the quality of national CRVS mortality data, the data were rated satisfactory for coverage and completeness of death registration, temporal c...

show abstract

Accuracy of Public Health Data Linkages

Abstract: The computer-matched records evaluated were sufficiently accurate. This provides a unique source to public health practitioners and researchers to perform epidemiological studies that could not be possible with single sources data.

Cited by 11 publications

References 10 publications

Accuracy of a probabilistic record linkage strategy applied to identify deaths among cases reported to the Brazilian AIDS surveillance database

Accuracy of a probabilistic record linkage strategy applied to identify deaths among cases reported to the Brazilian AIDS surveillance database

Obstetric Outcomes in Normal Weight and Obese Women in Relation to Gestational Weight Gain: Comparison between Institute of Medicine Guidelines and Cedergren Criteria

Availability and quality of vital registration mortality data in South Africa: a national and surveillance site perspective

Contact Info

Product

Resources

About