SEARCH · Search PubMed

Results for “database”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

SourceResearch connectionFrom yearThrough year

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

At least 199 records · Page 11Linked to original sources

PubMed2004

Rice Proteome Database based on two-dimensional polyacrylamide gel electrophoresis: its status in 2003.

The Rice Proteome Database is the first detailed database to describe the proteome of rice. The current release contains 21 reference maps based on two-dimensional polyacrylamide gel electrophoresis (2D-PAGE) of proteins from rice tissues and subcellular compartments. These reference maps comprise 11 941 identified proteins showing tissue and subcellular localization, corresponding to 4180 separate protein entries in the database. The Rice Proteome Database contains the calculated properties of each protein such as molecular weight, isoelectric point and expression; experimentally determined properties such as amino acid sequences obtained using protein sequencers and mass spectrometry; and the results of database searches such as sequence homologies. The database is searchable by keyword, accession number, protein name, isoelectric point, molecular weight and amino acid sequence, or by selection of a spot on one of the 2D-PAGE reference maps. Cross-references are provided to tools for proteomics and to other 2D-PAGE databases, which in turn provide many links to other molecular databases. The information in the Rice Proteome Database is updated weekly, and is available on the World Wide Web at http://gene64.dna.affrc.go.jp/RPD/.

Computational Biology↗

PubMed2004

The EMBL Nucleotide Sequence Database.

The EMBL Nucleotide Sequence Database (http://www.ebi.ac.uk/embl/), maintained at the European Bioinformatics Institute (EBI), incorporates, organizes and distributes nucleotide sequences from public sources. The database is a part of an international collaboration with DDBJ (Japan) and GenBank (USA). Data are exchanged between the collaborating databases on a daily basis to achieve optimal synchrony. The web-based tool, Webin, is the preferred system for individual submission of nucleotide sequences, including Third Party Annotation (TPA) and alignment data. Automatic submission procedures are used for submission of data from large-scale genome sequencing centres and from the European Patent Office. Database releases are produced quarterly. The latest data collection can be accessed via FTP, email and WWW interfaces. The EBI's Sequence Retrieval System (SRS) integrates and links the main nucleotide and protein databases as well as many other specialist molecular biology databases. For sequence similarity searching, a variety of tools (e.g. FASTA and BLAST) are available that allow external users to compare their own sequences against the data in the EMBL Nucleotide Sequence Database, the complete genomic component subsection of the database, the WGS data sets and other databases. All available resources can be accessed via the EBI home page at http://www.ebi.ac.uk.

Animals↗

PubMed2001

A decade's experience with quality improvement in cardiac surgery using the Veterans Affairs and Society of Thoracic Surgeons national databases.

OBJECTIVE: To review the Department of Veteran Affairs (VA) and the Society of Thoracic Surgeons (STS) national databases over the past 10 years to evaluate their relative similarities and differences, to appraise their use as quality improvement tools, and to assess their potential to facilitate improvements in quality of cardiac surgical care. SUMMARY BACKGROUND DATA: The VA developed a mandatory risk-adjusted database in 1987 to monitor outcomes of cardiac surgery at all VA medical centers. In 1989 the STS developed a voluntary risk-adjusted database to help members assess quality and outcomes in their individual programs and to facilitate improvements in quality of care. METHODS: A short data form on every veteran operated on at each VA medical center is completed and transmitted electronically for analysis of unadjusted and risk-adjusted death and complications, as well as length of stay. Masked, confidential semiannual reports are then distributed to each program's clinical team and the associated administrator. These reports are also reviewed by a national quality oversight committee. Thus, VA data are used both locally for quality improvement and at the national level with quality surveillance. The STS dataset (217 core fields and 255 extended fields) is transmitted for each patient semiannually to the Duke Clinical Research Institute (DCRI) for warehousing, analysis, and distribution. Site-specific reports are produced with regional and national aggregate comparisons for unadjusted and adjusted surgical deaths and complications, as well as length of stay for coronary artery bypass grafting (CABG), valvular procedures, and valvular/CABG procedures. Both databases use the logistic regression modeling approach. Data for key processes of care are also captured in both databases. Research projects are frequently carried out using each database. RESULTS: More than 74,000 and 1.6 million cardiac surgical patients have been entered into the VA and STS databases, respectively. Risk factors that predict surgical death for CABG are very similar in the two databases, as are the odds ratios for most of the risk factors. One major difference is that the VA is 99% male, the STS 71% male. Both databases have shown a significant reduction in the risk-adjusted surgical death rate during the past decade despite the fact that patients have presented with an increased risk factor profile. The ratio of observed to expected deaths decreased from 1.05 to 0.9 for the VA and from 1.5 to 0.9 for the STS. CONCLUSION: It appears that the routine feedback of risk-adjusted data on local performance provided by these programs heightens awareness and leads to self-examination and self-assessment, which in turn improves quality and outcomes. This general quality improvement template should be considered for application in other settings beyond cardiac surgery.

Cardiac Surgical Procedures↗

PubMed1998

Research potentials and pitfalls in the use of an HIV clinical database: Chelsea and Westminster Hospital.

This article summarizes the various problems and pitfalls in using clinical databases for epidemiologic research, with particular reference to an HIV clinical database. The combined population of HIV-infected individuals attending the Chelsea and Westminster Hospital, the Charing Cross Hospital, and the Victoria Clinic in London is the largest cohort of HIV-positive individuals in the U.K. A computerized database was developed in the mid-1980s and was adapted into a clinically oriented observational database for approximately 6,653 HIV-1-positive registered patients from three hospital-based clinics within the Riverside Health Authority in London, U.K.: Chelsea and Westminster Hospital Clinic (n = 5,000); Charing Cross Hospital (n = 500); and the Victoria Clinic (n = 500). The majority (83%) of HIV-infected patients registered at these sites are homosexual or bisexual men. Of 2,078 patients seen within the last 6 months, 22% are asymptomatic and 33% have AIDS; 30% have a CD4 cell count of less than 100 cells/mm3 and 17% have a CD4 cell count of greater than 500 cells/mm3. Dates of seroconversion are known for approximately 285 patients. For each patient, information on demographic characteristics, clinical symptoms, and HIV-related diagnoses, outpatient pharmacy prescriptions, day care treatments and procedures, and enrollment into clinical trials is routinely collected at outpatient clinic visits and entered into the database. Inpatient diagnoses and treatments were integrated into the database in September 1995. Unused serum samples from routine AIDS antibody or antigen testing are stored in a local specimen repository. The main purpose of the HIV database is to provide a multipurpose resource for use by physicians, researchers, and managers for administration, clinical care, and research. The specific functions of the database are the following: to enhance patient management by providing access to a clinical summary sheet detailing up-to-date information; to serve as a research tool for clinical and epidemiologic research; to aid in the identification of patients eligible for planned or ongoing clinical trials; to provide a facility for local and regional AIDS surveillance and reporting; and to provide a facility for administration and resource management of HIV services. The major limitations of this database in the conduct of clinical research have been losses to follow-up and incomplete information about clinical outcomes, because physicians have failed to update the clinical information.

Acquired Immunodeficiency Syndrome↗

PubMed2000

A relational database for diagnosis of hematopoietic neoplasms using immunophenotyping by flow cytometry.

A relational database was developed to facilitate the diagnosis of hematopoietic neoplasms using results of immunophenotyping by flow cytometry. This database runs on personal computers and uses backward-chaining search to arrive at conclusions. Results of immunologic marker studies are processed by the database to obtain a set of differential diagnoses. The current version of this database includes diagnostic immunophenotyping pattern for 33 hematopoietic neoplasms. We tested this database using 92 clinical cases from 2 tertiary care medical centers. The database ranked the actual diagnosis as 1 of the top 5 differential diagnoses in 93% of the cases tested. The user can modify the database contents to suit individual needs. This database has been posted on the World Wide Web for direct access. We propose that this user-friendly database is a potential tool for computer-assisted diagnosis of hematopoietic neoplasms.

Antigens, CD↗

PubMed1997

Identifying the active general practice workforce in one division of general practice: the utility of public domain databases.

OBJECTIVE: To identify the non-specialist medical practitioner workforce engaged in active general practice in the region served by the Division of General Practice-Northern Tasmania and to determine the usefulness of public domain databases for enumeration of individual non-specialists providing general practice services. METHODS: A masterlist of the active general practice workforce was compiled by obtaining the names and addresses/postcodes of all non-specialist medical practitioners who were listed in at least one of nine public domain databases and who were confirmed by selected local medical practitioners to be in active general practice in the three months prior to 30 June 1994. This masterlist was used in calculating the sensitivity and positive predictive value (PPV) of each of the nine databases for enumerating non-specialist practitioners in active general practice. RESULTS: Combining the databases resulted in a list of 475 practitioners, which was refined to 139 practitioners who, by our criteria, were in active general practice. Databases had a range of sensitivities and PPVs, but those with high sensitivity tended to have low PPVs, and vice versa. The most useful database for enumerating these practitioners was the mailing list for Australian Family Physician (sensitivity, 94%; PPV, 0.79). CONCLUSIONS: When used alone, no single database had both high sensitivity and high positive predictive value for identifying the active general practice workforce. Combining multiple databases may improve precision. Developing methods to identify recent departures from local active practice has the potential to improve the PPV of existing highly sensitive databases.

Australia↗

PubMed1999

Comparison of literature searches on quality and costs for health technology assessment using the MEDLINE and EMBASE databases.

Biomedical databases are an important source of information for health technology assessment. However, there is considerable variation in the costs of accessing commercial databases. We sought to measure the quality, amount of overlap, and costs of information retrieved from two of the main database sources--MEDLINE and EMBASE. Librarians at two health technology assessment agencies ran a total of eight literature searches on various medical technologies, using both databases. All search results were independently reviewed by two researchers. The researchers were asked to identify relevant references and to rank each of these according to a level of evidence scale. The results were tabulated to show the number of references identified by each database, the number of relevant references ranked by level of evidence, and the number of these references that were unique to one or the other database. The cost of retrieving references from each source was also calculated. Each database contained relevant references not available in the other. Because of the longer time lag for indexing in MEDLINE, many of the references that originally appeared to be unique to EMBASE were subsequently available in MEDLINE as well. Since our study was conducted, MEDLINE has been made available worldwide, free of charge, via the Internet. Hence, the cost difference between the databases is now even greater. However, notwithstanding the costs, it appears that literature searches that rely on only one or the other database will inevitably miss pertinent information.

Abstracting and Indexing↗

PubMed2000

[The ZEBET database on alternative methods to animal experiments in the Internet--a concrete contribution to the protection of animals].

Up from February of the year 2000 ZEBET (German Centre for the Documentation and Validation of Alternative Methods) at the Federal Institute for Consumer Health Protection and Veterinary Medicine (BgVV) put the ZEBET-database on alternative methods to animal experiments on the Internet in English via DIMDI, the German Institute for Medical Documentation and Information (http://gripsdb.dimdi.de/engl/guieng.html). The access is free, moreover DIMDI's complete service is available to visitors of the ZEBET-database. The ZEBET database contains documents on alternatives to testing in animals, which have been carefully evaluated by ZEBET's staff according to the "3Rs"-concept established by Russel and Burch in 1959. Therefore, methods documented in the ZEBET database must meet at least one of the following criteria: "replacement" of an animal experiment by a non-animal method, "reduction" of the number of animals used, "refinement" of an experiment by minimising pain and suffering of animals. In addition, the ZEBET-database provides information on the current stage of development and validation of a method and on the acceptance for either scientific or regulatory purposes. Each document is characterised by the following criteria: the title of a method, keywords, assessment, summary and bibliographic references. To search DIMDI<<s-database and host-system the grips software has to be used. Examples are given for searching in the ZEBET-database. Currently 125 alternative methods are meeting the criteria of the ZEBET database. 50 of them are available online on the internet via DIMDI, the remaining 75 documents will be available by the end of the year 2000. International fellow organisations, e.g. FRAME (Fund for the Replacement of Animals in Medical Experiments) in the UK and CAAT (Johns Hopkins Center for Alternatives to Animal Testing) in the USA, have established links on their web sites to provide visitors free access to the ZEBET-database.

Animal Testing Alternatives↗

PubMed2001

Answering family physicians' clinical questions using electronic medical databases.

OBJECTIVE: We studied the ability of electronic medical databases to provide adequate answers to the clinical questions of family physicians. STUDY DESIGN: Two family physicians attempted to answer 20 questions with each of the databases evaluated. The adequacy of the answers was determined by the 2 physician searchers, and an arbitration panel of 3 family physicians was used if there was disagreement. DATA SOURCE: We identified 38 databases through nominations from national groups of family physicians, medical informaticians, and medical librarians; 14 met predetermined eligibility criteria. OUTCOMES MEASURED: The primary outcome was the proportion of questions adequately answered by each database and by combinations of databases. We also measured mean and median times to obtain adequate answers for individual databases. RESULTS: The agreement between family physician searchers regarding the adequacy of answers was excellent (k=0.94). Five individual databases (STAT!Ref, MDConsult, DynaMed, MAXX, and MDChoice.com) answered at least half of the clinical questions. Some combinations of databases answered 75% or more. The average time to obtain an adequate answer ranged from 2.4 to 6.5 minutes. CONCLUSION: Several current electronic medical databases could answer most of a group of 20 clinical questions derived from family physicians during office practice. However, point-of-care searching is not yet fast enough to address most clinical questions identified during routine clinical practice.

Clinical Competence↗

PubMed2002

Spanish personal name variations in national and international biomedical databases: implications for information retrieval and bibliometric studies.

OBJECTIVES: The study sought to investigate how Spanish names are handled by national and international databases and to identify mistakes that can undermine the usefulness of these databases for locating and retrieving works by Spanish authors. METHODS: The authors sampled 172 articles published by authors from the University of Granada Medical School between 1987 and 1996 and analyzed the variations in how each of their names was indexed in Science Citation Index (SCI), MEDLINE, and Indice Medico Español (IME). The number and types of variants that appeared for each author's name were recorded and compared across databases to identify inconsistencies in indexing practices. We analyzed the relationship between variability (number of variants of an author's name) and productivity (number of items the name was associated with as an author), the consequences for retrieval of information, and the most frequent indexing structures used for Spanish names. RESULTS: The proportion of authors who appeared under more then one name was 48.1% in SCI, 50.7% in MEDLINE, and 69.0% in IME. Productivity correlated directly with variability: more than 50% of the authors listed on five to ten items appeared under more than one name in any given database, and close to 100% of the authors listed on more than ten items appeared under two or more variants. Productivity correlated inversely with retrievability: as the number of variants for a name increased, the number of items retrieved under each variant decreased. For the most highly productive authors, the number of items retrieved under each variant tended toward one. The most frequent indexing methods varied between databases. In MEDLINE and IME, names were indexed correctly as "first surname second surname, first name initial middle name initial" (if present) in 41.7% and 49.5% of the records, respectively. However, in SCI, the most frequent method was "first surname, first name initial second name initial" (48.0% of the records) and first surname and second surname run together, first name initial (18.3%). CONCLUSIONS: Retrievability on the basis of author's name was poor in all three databases. Each database uses accurate indexing methods, but these methods fail to result in consistency or coherence for specific entries. The likely causes of inconsistency are: (1) use by authors of variants of their names during their publication careers, (2) lack of authority control in all three databases, (3) the use of an inappropriate indexing method for Spanish names in SCI, (4) authors' inconsistent behaviors, and (5) possible editorial interventions by some journals. We offer some suggestions as to how to avert the proliferation of author name variants in the databases.

Databases, Bibliographic↗

PubMed2002

Draft secure medical database standard.

Medical database security is a particularly important issue for all Healthcare establishments. Medical information systems are intended to support a wide range of pertinent health issues today, for example: assure the quality of care, support effective management of the health services institutions, monitor and contain the cost of care, implement technology into care without violating social values, ensure the equity and availability of care, preserve humanity despite the proliferation of technology etc.. In this context, medical database security aims primarily to support: high availability, accuracy and consistency of the stored data, the medical professional secrecy and confidentiality, and the protection of the privacy of the patient. These properties, though of technical nature, basically require that the system is actually helpful for medical care and not harmful to patients. These later properties require in turn not only that fundamental ethical principles are not violated by employing database systems, but instead, are effectively enforced by technical means. This document reviews the existing and emerging work on the security of medical database systems. It presents in detail the related problems and requirements related to medical database security. It addresses the problems of medical database security policies, secure design methodologies and implementation techniques. It also describes the current legal framework and regulatory requirements for medical database security. The issue of medical database security guidelines is also examined in detailed. The current national and international efforts in the area are studied. It also gives an overview of the research work in the area. The document also presents in detail the most complete to our knowledge set of security guidelines for the development and operation of medical database systems.

Computer Security↗

PubMed1994

Performance of online biomedical databases in rheumatology.

OBJECTIVE: To compare the performance of MEDLINE, EMBASE, and BIOSIS in selected rheumatology topics. METHODS: Online literature searches were conducted with regard to the epidemiology of rheumatoid arthritis (RA), systemic lupus erythematosus (SLE), and ankylosing spondylitis (AS), as well as for 3 specific questions representing clinical, clinical/laboratory, and therapeutic topics in rheumatology. Total number of citations retrieved, type and language of publication, percentage of contribution from rheumatology journals, and degree of overlap among the databases were recorded. Publications retrieved for the 3 specific questions were also graded for relevance. RESULTS: For 1991, each online biomedical database (OBD) retrieved more than 1,100 citations for RA, over 600 for SLE, and over 110 for AS. For the epidemiology subtopic, fewer than 25% of the citations were retrieved by more than one of the databases. About 3/4 of the citations obtained for the specific search questions were retrieved by a single database. No major differences were observed among databases in relation to number of relevance of citations retrieved. Over 60% of the papers assessed had low relevance in relation to the topic of the search. Efficiency was estimated as the percentage of all relevant citations retrieved by each OBD. Results varied according to the topic, but in most cases each database retrieved at least 50% of the relevant citations. About 45% of the citations retrieved for the 3 search questions were published in nonrheumatology journals. CONCLUSION: No database was superior in all respects. The majority of the citations were retrieved by a single database. A high percentage of the articles retrieved were not relevant, implying low specificity. If a comprehensive online search in rheumatology is required, 2 or more databases should be utilized.

Arthritis, Rheumatoid↗

PubMed1996

Maintaining patient confidentiality in the public domain Internet Autopsy Database (IAD).

The Internet provides the opportunity of permitting public access to large databases containing patient information that can be shared and utilized by epidemiologists, health planners, and medical researchers. Until now, large databases containing patient information have been held in strict confidence, with database access available only to approved researchers or to researchers with access limited to only specific portions of the database. The Internet Autopsy Database (IAD) consists of demographic and pathologic data from over 49,000 autopsies contributed by over a dozen academic medical institutions. Each autopsy record in the public database consists of a uniform set of demographics and SNOMED-compatible terms. To make the database publicly available, a strategy had to be devised that assured the privacy of every person included in the database. A key step involved translating the autopsy facesheets into a listing of SNOMED-compatible terms that effectively eliminated identifying terminology, replacing free text with a generic nomenclature that preserves diagnostic information. The entire database is available on the Internet at: http:@www.med.jhu.edu/pathology/iad.html

Autopsy↗

PubMed1996

Use of large databases for resolving critical care problems.

Large databases allow for rapid access to large volumes of data. To convert raw data to information, large numbers of data points must be correlated into a descriptive pattern that can be interpreted by the user. Databases must be constructed so as to allow reliable extraction of the raw data into a format that supports analysis of events in a meaningful, objective, and reproducible manner. Databases must be responsive to a variety of users. They must not demand unrealistic amounts of effort on those responsible for data entry. Standard protocols in various stages of development will make databases easier to use and more reliable. Database management tools such as the Internet and the National Library of Medicine will become more integrated into the practice of critical care medicine at all levels, including administration, clinical care, and research. This article provides an overview of the capabilities and difficulties associated with large databases. The major areas of use of large databases in the hospital setting are administration, bibliographic, patient care, research, and education. Each of these areas has different requirements and is supported by different types of databases. The advantages and disadvantages of linear, relational, and object-oriented databases are discussed. Issues relating to methods of data entry and the accuracy and reliability of data are discussed. The challenges involving integration of various sources of data and the interfacing of devices are reviewed.

Critical Care↗

PubMed2003

AraCyc: a biochemical pathway database for Arabidopsis.

AraCyc is a database containing biochemical pathways of Arabidopsis, developed at The Arabidopsis Information Resource (http://www.arabidopsis.org). The aim of AraCyc is to represent Arabidopsis metabolism as completely as possible with a user-friendly Web-based interface. It presently features more than 170 pathways that include information on compounds, intermediates, cofactors, reactions, genes, proteins, and protein subcellular locations. The database uses Pathway Tools software, which allows the users to visualize a bird's eye view of all pathways in the database down to the individual chemical structures of the compounds. The database was built using Pathway Tools' Pathologic module with MetaCyc, a collection of pathways from more than 150 species, as a reference database. This initial build was manually refined and annotated. More than 20 plant-specific pathways, including carotenoid, brassinosteroid, and gibberellin biosyntheses have been added from the literature. A list of more than 40 plant pathways will be added in the coming months. The quality of the initial, automatic build of the database was compared with the manually improved version, and with EcoCyc, an Escherichia coli database using the same software system that has been manually annotated for many years. In addition, a Perl interface, PerlCyc, was developed that allows programmers to access Pathway Tools databases from the popular Perl language. AraCyc is available at the tools section of The Arabidopsis Information Resource Web site (http://www.arabidopsis.org/tools/aracyc).

Arabidopsis↗

PubMedDate not supplied

Exploring performance issues for a clinical database organized using an entity-attribute-value representation.

BACKGROUND: The entity-attribute-value representation with classes and relationships (EAV/CR) provides a flexible and simple database schema to store heterogeneous biomedical data. In certain circumstances, however, the EAV/CR model is known to retrieve data less efficiently than conventionally based database schemas. OBJECTIVE: To perform a pilot study that systematically quantifies performance differences for database queries directed at real-world microbiology data modeled with EAV/CR and conventional representations, and to explore the relative merits of different EAV/CR query implementation strategies. METHODS: Clinical microbiology data obtained over a ten-year period were stored using both database models. Query execution times were compared for four clinically oriented attribute-centered and entity-centered queries operating under varying conditions of database size and system memory. The performance characteristics of three different EAV/CR query strategies were also examined. RESULTS: Performance was similar for entity-centered queries in the two database models. Performance in the EAV/CR model was approximately three to five times less efficient than its conventional counterpart for attribute-centered queries. The differences in query efficiency became slightly greater as database size increased, although they were reduced with the addition of system memory. The authors found that EAV/CR queries formulated using multiple, simple SQL statements executed in batch were more efficient than single, large SQL statements. CONCLUSION: This paper describes a pilot project to explore issues in and compare query performance for EAV/CR and conventional database representations. Although attribute-centered queries were less efficient in the EAV/CR model, these inefficiencies may be addressable, at least in part, by the use of more powerful hardware or more memory, or both.

Database Management Systems↗

PubMed1989

Maintenance of a nutrient database for clinical trials.

Maintenance of a nutrient database for use in dietary analysis for clinical trials and other medical research studies is described. The database, maintained at the University of Minnesota's Nutrition Coordinating Center (NCC), has been used to calculate dietary intake data for a wide range of diet-disease related investigations including studies on cardiovascular disease, hypertension, cancer, gastroenterology, and osteoporosis. Potential sources of error associated with nutrient databases are identified. Criteria are provided for the selection of a nutrient database to meet study objectives and to minimize the potential for errors and inconsistencies. NCC database maintenance procedures, designed to provide updated and verified nutrient calculations for clinical research, involve adherence to standardized procedures for all aspects of database maintenance including data selection, imputations, quality control, recipe calculations, and documentation. By maintaining multiple versions of the database, the NCC is able to update and expand a working version of the database while providing database stability for individual research studies.

Clinical Trials as Topic↗

PubMed1995

An assessment of data quality in the Vermont-Oxford Trials Network database.

The Vermont-Oxford Trials Network is a voluntary collaborative research group of neonatologists that maintains a database for very low birthweight infants (501-1500 g). The database (1) provides core data for randomized trials, (2) serves as a resource for outcomes research in neonatology, and (3) generates quality management reports for participating sites. To assess the reliability of this database and to determine the sources of error, we reviewed 635 medical records chosen at random from among the 4341 eligible infants born at 40 participating data generating sites during an 18-month period beginning January 1, 1990. The estimated frequencies of disagreement between the medical record and database for each of the 10 data items studied and the standard errors of the estimates (in parentheses) were: date of birth 1.3% (0.4), date of admission 2.5% (0.6), date of discharge 8.8% (1.0), birthweight (difference > 50 g) 2.9% (0.6), location of birth (inborn or outborn) 2.1% (0.5), multiple birth 2.2% (0.5), cesarean section 2.5% (0.6), gender 2.1% (0.5), status 28 days after birth 3.4% (0.6), final status 2.9% (0.6). The overall proportions and mean values for items in the database were close to the estimated values based on the random sample of records. There were a total of 247 disagreements between the database and the medical records in the sample. Twenty-three were due to data keying errors. Two hundred twenty-four were due to errors in transcription or interpretation. The rate of data keying errors decreased from over 50 errors per 10,000 fields to less than 15 errors per 10,000 fields when specific quality control procedures, including visual inspection, were instituted. Data keying errors accounted for 13.7% of all disagreements between the database and medical record before improved data entry methods were introduced, and only 3.7% of all errors after they were introduced. We concluded that the Vermont-Oxford Trials Network Database is reliable. Data keying errors have been reduced by the introduction of additional quality control measures. Further reductions in database errors will require measures aimed at minimizing transcription or interpretation errors by individuals completing the data forms.

Computer Communication Networks↗