The utility of the general practice research database to examine selected congenital heart defects: a validation study

Maternal sociodemographic characteristics, early pregnancy behaviours, and livebirth outcomes as congenital heart defects risk factors - Northern Ireland 2010-2014

BACKGROUND

Congenital Heart Defects (CHD) is the most commonly occurring congenital anomaly in Europe and a major paediatric health care concern. Investigations are needed to enable identification of CHD risk factors as studies have given conflicting results. This study aim was to identify maternal sociodemographic characteristics, behaviours, and birth outcomes as risk factors for CHD. This was a population based, data linkage cohort study using anonymised data from Northern Ireland (NI) covering the period 2010-2014. The study cohort composed of 94,067 live births with an outcome of 1162 cases of CHD using the International Statistical Classification of Diseases and Related Health Problems (ICD)-10 codes and based on the European Surveillance of Congenital Anomalies (EUROCAT) grouping system for CHD. CHD cases were obtained from the HeartSuite database (HSD) at the Royal Belfast Hospital for Sick Children (RBHSC), maternal data were extracted from the Northern Ireland Maternity System (NIMATS), and medication data were extracted from the Enhanced Prescribing Database (EPD). STATA version 14 was used for the statistical analysis in this study, Odds Ratio (OR), 95% Confident intervals (CI), P value, and logistic regression were used in the analysis. Ethical approval was granted from the National Health Service (NHS) Research Ethics Committee.

RESULT

In this study, a number of potential risk factors were assessed for statistically significant association with CHD, however only certain risk factors demonstrated a statistically significant association with CHD which included: gestational age at first booking (AOR = 1.21; 95% CI = 1.04-1.41; P < 0.05), family history of CHD or congenital abnormalities and syndromes (AOR = 4.14; 95% CI = 2.47-6.96; P < 0.05), woman's smoking in pregnancy (AOR = 1.22; 95% CI = 1.04-1.43; P < 0.05), preterm birth (AOR = 3.01; 95% CI = 2.44-3.01; P < 0.05), multiple births (AOR = 1.89; 95% CI = 1.58-2.60; P < 0.05), history of abortion (AOR = 1.12; 95% CI = 1.03-1.28; P < 0.05), small for gestational age (SGA) (AOR = 1.44; 95% CI = 1.22-1.78; P < 0.05), and low birth weight (LBW) (AOR = 3.10; 95% CI = 2.22-3.55; P < 0.05). Prescriptions and redemptions of antidiabetic (AOR = 2.68; 95% CI = 1.85-3.98; P < 0.05), antiepileptic (AOR = 1.77; 95% CI = 1.10-2.81; P < 0.05), and dihydrofolate reductase inhibitors (DHFRI) (AOR = 2.13; 95% CI = 1.17-5.85; P < 0.05) in early pregnancy also showed evidence of statistically significant association with CHD.

CONCLUSION

The results of this study suggested that there are certain maternal sociodemographic characteristics, behaviours and birth outcomes that are statistically significantly associated with higher risk of CHD. Appropriate prevention policy to target groups with higher risk for CHD may help to reduce CHD prevalence. These results are important for policy makers, obstetricians, cardiologists, paediatricians, midwives and the public.

The identification and validity of congenital malformation diagnoses in UK electronic health records: A systematic review

PURPOSE

To describe the methods used to identify and validate congenital malformation diagnoses recorded in UK electronic health records, and the results of validation studies.

METHODS

Medline and Embase were searched for publications between 1987 and 2019 that involved identifying congenital malformations from UK electronic health records using diagnostic codes. The methods and code-lists used to identify congenital malformations, and the methods and results of validations, were examined.

RESULTS

We retrieved 54 eligible studies; 36 identified congenital malformations from primary care data and 18 from secondary care data alone or in combination with birth and/or death records. Identification in secondary care data relied on codes from the 'Q' chapter for congenital malformations in ICD-10. In contrast, studies using primary care data frequently used additional codes outside of the 'P' chapter for congenital malformation diagnoses in Read, although the exact codes used were not always clear. Eight studies validated diagnoses identified in primary care data. The positive predictive value was highest (80%-100%) for congenital malformations overall, major malformations, and heart defects although the validity of the reference standard used was often uncertain. It was lowest for neural tube defects (71%) and developmental hip dysplasia (56%).

CONCLUSIONS

Studies identifying congenital malformations from primary care data provided limited details about the methods used. The few validation studies were limited to diagnoses recorded in primary care. Further assessments of all measures of validity in both data sources and of other malformation subgroups are needed, using robust reference standards and adhering to reporting guidelines.

Toward the Development of Data Governance Standards for Using Clinical Free-Text Data in Health Research: Position Paper

BACKGROUND

Clinical free-text data (eg, outpatient letters or nursing notes) represent a vast, untapped source of rich information that, if more accessible for research, would clarify and supplement information coded in structured data fields. Data usually need to be deidentified or anonymized before they can be reused for research, but there is a lack of established guidelines to govern effective deidentification and use of free-text information and avoid damaging data utility as a by-product.

OBJECTIVE

This study aimed to develop recommendations for the creation of data governance standards to integrate with existing frameworks for personal data use, to enable free-text data to be used safely for research for patient and public benefit.

METHODS

We outlined data protection legislation and regulations relating to the United Kingdom for context and conducted a rapid literature review and UK-based case studies to explore data governance models used in working with free-text data. We also engaged with stakeholders, including text-mining researchers and the general public, to explore perceived barriers and solutions in working with clinical free-text.

RESULTS

We proposed a set of recommendations, including the need for authoritative guidance on data governance for the reuse of free-text data, to ensure public transparency in data flows and uses, to treat deidentified free-text data as potentially identifiable with use limited to accredited data safe havens, and to commit to a culture of continuous improvement to understand the relationships between the efficacy of deidentification and reidentification risks, so this can be communicated to all stakeholders.

CONCLUSIONS

By drawing together the findings of a combination of activities, we present a position paper to contribute to the development of data governance standards for the reuse of clinical free-text data for secondary purposes. While working in accordance with existing data governance frameworks, there is a need for further work to take forward the recommendations we have proposed, with commitment and investment, to assure and expand the safe reuse of clinical free-text data for public benefit.

Seasonal influenza vaccination during pregnancy and the risk of major congenital malformations in live-born infants: A 2010-2016 historical cohort study

Background

Available evidence indicates that seasonal inactivated influenza vaccination during pregnancy protects both the mother and her newborn and is safe. Nevertheless, ongoing safety assessments are important in sustaining vaccine uptake. Few studies have explored safety in relation to major congenital malformations (MCMs), particularly in the first trimester when most organogenesis occurs.

Methods

Anonymized UK primary care data (the Clinical Practice Research Datalink), including a recently developed Pregnancy Register, were used to identify live-born singletons delivered between 2010 and 2016. Maternal influenza vaccination was determined using primary care records and stratified by trimester. Ascertainment of MCMs from infant primary care records was maximized by linkage to hospitalization data and death certificates. The relationship between vaccination and MCMs recorded in the year after delivery and in early childhood was then assessed using multivariable Cox regression.

Results

A total of 78 150 live-birth pregnancies were identified: 6872 (8.8%) were vaccinated in the first trimester, 11 678 (14.9%) in the second, and 12 931 (16.5%) in the third. Overall, 5707 live births resulted in an infant with an MCM recorded in the year after delivery and the adjusted hazard ratio when comparing first-trimester vaccination to no vaccination was 1.06 (99% CI, .94–1.19; P = .2). Results were similar for second- and third-trimester vaccination and for analyses considering MCMs recorded beyond the first birthday.

Conclusions

In this large, population-based historical cohort study there was no evidence to suggest that seasonal influenza vaccine was associated with MCMs when given in the first trimester or subsequently in pregnancy.

Prevalence and Incidence Trends for Diagnosed Prescription Opioid Use Disorders in the United Kingdom

Introduction

The prevalence of prescription opioid use disorders in the US has increased markedly in parallel with increases in opioid prescribing. Whilst an increase in opioid prescribing has also occurred in the UK, it remains unknown if there have been concurrent increases in opioid use disorders. The aim of this study was to examine national trends in the prevalence and incidence of physician-diagnosed opioid use disorders in the UK.

Methods

In a retrospective electronic health care database analysis using data from the UK Clinical Practice Research Datalink (CPRD), we identified persons receiving a first opioid prescription between January 1, 2008 and December 31, 2012. Persons with an opioid use disorder were identified by Read codes assigned by patients’ physicians within 6 months following an opioid prescription. We calculated prevalence and incidence rates by dividing the analysis population by the total number of patients exposed (prevalence) or the total patient-years of exposure (incidence) using the ‘exact’ Clopper–Pearson Binomial method.

Results

Our analysis included 714,699 person-years of prescription opioid exposure. The 5-year period prevalence of opioid use disorders was 4.61 (95% CI 4.28–4.96) per 10,000 individuals, or 0.05%. The incidence rate of opioid use disorders was of 6.51 (95% CI 5.93–7.13) patients per 10,000 patient-years exposed. When examined by study year, there was no clear suggestion of a changing trend over time. When stratified by opioid drug, trends in the incidence rate during the study were either stable (i.e., codeine and tramadol), increasing (i.e., morphine) or decreasing (i.e., dihydrocodeine).

Conclusions

Our study demonstrates that despite the marked increase in overall opioid prescribing in the UK in the past decade, there has not been an increase in the incidence of physician-diagnosed opioid use disorders.

Improving case ascertainment of congenital anomalies: findings from a prospective birth cohort with detailed primary care record linkage

BACKGROUND

Congenital anomalies (CAs) are a common cause of infant death and disability. We linked children from a large birth cohort to a routine primary care database to detect CA diagnoses from birth to age 5 years. There could be evidence of underreporting by CA registries as they estimate that only 2% of CA registrations occur after age 1 year.

METHODS

CA cases were identified by linking children from a prospective birth cohort to primary care records. CAs were classified according to the European Surveillance of CA guidelines. We calculated rates of CAs by using a bodily system group for children aged 0 to <5 years, together with risk ratios (RRs) with 95% CIs for maternal risk factors.

RESULTS

Routinely collected primary care data increased the ascertainment of children with CAs from 432.9 per 10 000 live births under 1 year to 620.6 per 10 000 live births under 5 years. Consanguinity was a risk factor for Pakistani mothers (multivariable RR 1.87, 95% CI 1.46 to 2.83), and maternal age >34 years was a risk factor for mothers of other ethnicities (multivariable RR 2.19, 95% CI 1.36 to 3.54). Education was associated with a lower risk (multivariable RR 0.78, 95% CI 0.62 to 0.98).

CONCLUSION

98% of UK CA registrations relate to diagnoses made in the first year of life. Our data suggest that this leads to incomplete case ascertainment with a further 30% identified after age 1 year in our study. Risk factors for CAs identified up to age 1 year persist up to 5 years. National registries should consider using routine data linkage to provide more complete case ascertainment after infancy.

Risk of spontaneous abortion and other pregnancy outcomes in 15-25 year old women exposed to human papillomavirus-16/18 AS04-adjuvanted vaccine in the United Kingdom

BACKGROUND

We assessed the risk of spontaneous abortion (SA) after inadvertent exposure to HPV-16/18-vaccine during pregnancy using an observational cohort design.

METHODS

The study population included women aged 15-25 years registered with the Clinical Practice Research Datalink General Practice OnLine Database in the United Kingdom (UK), who received at least one HPV-16/18-vaccine dose between 1st September 2008 and 30th June 2011. Exposed women had the first day of gestation between 30 days before and 45 days (90 days for the extended exposure period) after any HPV-16/18-vaccine dose. Non-exposed women had the first day of gestation 120 days-18 months after the last dose. SA defined as foetal loss between weeks 1 and 23 of gestation (UK definition).

RESULTS

The frequency of SA was 11.6% (among 207 exposed) and 9.0% (632 non-exposed), women: hazard ratio (HR) adjusted for age at first day of gestation 1.30 (95% confidence interval: 0.79-2.12). Sensitivity analysis per number of doses administered (-30 to +45-day risk period) showed a HR for SA of 1.11 (0.64-1.91) for 18/178 women with one dose during the risk period versus 2.55 (1.09-5.93) in 6/29 women with two doses within a 4-5 weeks period. The proportion of pre-term/full-term/postterm deliveries, small/large for gestational age infants, and birth defects was not significantly different between exposed and non-exposed women. Results were consistent using a (United States) SA definition of foetal loss between weeks 1-19 and/or the extended risk period.

CONCLUSION

There was no evidence of an increased risk of SA and other adverse pregnancy outcomes in young women inadvertently HPV-16/18-vaccinated around gestation. Nevertheless, women who are pregnant or trying to become pregnant are advised to postpone vaccination until completion of pregnancy.

Nicotine replacement therapy in pregnancy and major congenital anomalies in offspring

BACKGROUND AND OBJECTIVES

Nicotine replacement therapy (NRT) is now being used as a smoking cessation aid during pregnancy, although little is known about fetal safety. We assessed the relationship between early pregnancy exposure to NRT or smoking with major congenital anomalies (MCA) in offspring.

METHODS

We studied 192,498 children born in the United Kingdom between 2001 and 2012 with linked mother-child primary care records. The absolute risks of MCAs in the NRT group (women prescribed NRT during the first trimester or 1 month before conception [and therefore likely consumed during the first trimester]) and odds ratios (ORs) and 99% confidence intervals (CIs) were compared with those of women who smoked during pregnancy and with a control group (women who neither smoked nor were prescribed NRT); logistic regression models adjusted for maternal morbidities that increase MCA risk were used for analysis.

RESULTS

MCA prevalence was 288 per 10,000 live births (5535 children with ≥ 1 MCA). Maternal morbidities were most common in the NRT group (35%) followed by smokers (27%) and the control group (20%). Compared with the control group, adjusted ORs for MCAs in the NRT group and smokers were 1.12 (99% CI: 0.84-1.48) and 1.05 (99% CI: 0.89-1.23), respectively. The OR comparing the NRT group directly with smokers was 1.07 (99% CI: 0.78-1.47). There were no statistically significant associations between maternal NRT and system-specific anomalies except for respiratory anomalies (OR: 4.65 [99% CI: 1.76-12.25]; absolute risk difference: 3 per 1000 births), which was based on 10 exposed cases.

CONCLUSIONS

For most system-specific MCAs, we found no statistically significant increased risks associated with maternal NRT prescribed during pregnancy, except for respiratory anomalies. Although this study is the largest published to date, NRT use in pregnancy remains rare; thus, the statistical power was limited. Higher morbidities in those women prescribed NRT may also be an explanatory factor. Nevertheless, absolute MCA risks were similar between women who smoked and those prescribed NRT during pregnancy.

First trimester exposure to anxiolytic and hypnotic drugs and the risks of major congenital anomalies: a United Kingdom population-based cohort study

Background

Despite their widespread use the effects of taking benzodiazepines and non-benzodiazepine hypnotics during pregnancy on the risk of major congenital anomaly (MCA) are uncertain. The objectives were to estimate absolute and relative risks of MCAs in children exposed to specific anxiolytic and hypnotic drugs taken in the first trimester of pregnancy, compared with children of mothers with depression and/or anxiety but not treated with medication and children of mothers without diagnosed mental illness during pregnancy.

Methods

We identified singleton children born to women aged 15–45 years between 1990 and 2010 from a large United Kingdom primary care database. We calculated absolute risks of MCAs for children with first trimester exposures of different anxiolytic and hypnotic drugs and used logistic regression with a generalised estimating equation to compare risks adjusted for year of childbirth, maternal age, smoking, body mass index, and socioeconomic status.

Results

Overall MCA prevalence was 2.7% in 1,159 children of mothers prescribed diazepam, 2.9% in 379 children with temazepam, 2.5% in 406 children with zopiclone, and 2.7% in 19,193 children whose mothers had diagnosed depression and/or anxiety but no first trimester drug exposures. When compared with 2.7% in 351,785 children with no diagnosed depression/anxiety nor medication use, the adjusted odds ratios were 1.02 (99% confidence interval 0.63–1.64) for diazepam, 1.07 (0.49–2.37) for temazepam, 0.96 (0.42–2.20) for zopiclone and 1.27 (0.43–3.75) for other anxiolytic/hypnotic drugs and 1.01 (0.90–1.14) for un-medicated depression/anxiety. Risks of system-specific MCAs were generally similar in children exposed and not exposed to such medications.

Conclusions

We found no evidence for an increase in MCAs in children exposed to benzodiazepines and non-benzodiazepine hypnotics in the first trimester of pregnancy. These findings suggest that prescription of these drugs during early pregnancy may be safe in terms of MCA risk, but findings from other studies are required before safety can be confirmed.

The Clinical Practice Research Datalink for drug safety in pregnancy research: an overview

Medicine use during pregnancy is common; however the safety of medicine use during pregnancy is largely unknown when a medicine comes to market. Electronic healthcare databases, including the Clinical Practice Research Datalink (CPRD), are increasingly being used for post-marketing surveillance in this field. The CPRD contains anonymised, longitudinal medical records routinely collected in primary care. Using CPRD data it is possible to identify medical records indicative of pregnancy, including pregnancy losses. Data on prescriptions issued can be used to determine maternal exposure and for about 80% of pregnancies it is possible to link the mother's medical record to the medical record of the child. Data in the medical records of the mother and child can then be used to identify adverse pregnancy outcomes, including congenital malformations. This paper describes some of the complexities involved in using CPRD data for pregnancy related research and discusses some of its strengths and limitations.

Sex prevalence of major congenital anomalies in the United Kingdom: a national population-based study and international comparison meta-analysis

Background

The aim of this study was to assess sex differences in major congenital anomaly (CA) diagnoses within a national population sample; to examine the influence of sociodemographic and maternal factors on these risks; and to conduct a meta-analysis using estimates from other population-based studies.

Methods

We conducted a population-based study in a United Kingdom research database of prospectively collected primary care data (The Health Improvement Network) including children born 1990 to 2009 (n = 794,169) and identified major CA diagnoses using EUROCAT (European Surveillance of Congenital Anomalies) classification. Prevalence ratios (PR) were used to estimate the risk of CA in males compared with females for any CA, system-specific subgroups and specific CA diagnoses. In a subpopulation of children whose medical records were linked to their mothers', we assessed the effect of adjusting for sociodemographic and maternal factors on sex odds ratios. PRs were pooled with measures from previously published studies.

Results

The prevalence of any CA was 307/10,000 in males (95% CI, 302–313) and 243/10,000 in females (95% CI, 238–248). Overall the risk of any CA was 26% greater in males (PR (male: female) 1.26, 95% CI, 1.23–1.30) however there was considerable variation across specific diagnoses. The magnitude and direction of risk did not change for any specific CA upon adjustment for sociodemographic and maternal factors. Our PRs were highly consistent with those from previous studies.

Conclusion

The overall risk of CA is greater in males than females, although this masked substantial variation by specific diagnoses. Sociodemographic and maternal factors do not appear to affect these risks. Birth Defects Research (Part A) 100:79–91, 2014. © 2014 Wiley Periodicals, Inc.

Determining the predictive value of Read codes to identify congenital cardiac malformations in the UK Clinical Practice Research Datalink

PURPOSES

The purposes of this study were to determine (i) the positive predictive value (PPV) of multiple Read codes used to identify congenital cardiac malformation (CCM) cases in the UK Clinical Practice Research Datalink (CPRD); (ii) the accuracy of the diagnosis date; and (iii) the source of information that the general practitioners (GPs) use for validating the diagnosis suggested by the code.

METHODS

Eight hundred eighty-eight records with Read diagnostic and procedures codes for CCM, between January 1996 and November 2010, were identified from CPRD. Questionnaires were sent to GPs to verify the diagnoses and date of the code-identified events.

RESULTS

A total of 719 questionnaires were returned (81% response rate). The PPV of the CCM codes was 93% (670/719). Thirty-one percent of cases had a different event date than the one recorded in the electronic medical record (EMR); 10% of these differing dates were within 30 days of the code-identified CCM date. GPs used a variety of data sources to confirm CCM diagnoses. Although the EMR was the most frequently used data source (70%), 66% reported using consultation letters, 9% reported using clinical notes or paper charts, and 35% of GPs reported using the hospital record to confirm the CCM diagnosis.

CONCLUSIONS

Clinical Practice Research Datalink Read codes for CCMs have 93% PPV and most likely point to true cases. However, the accuracy of diagnosis dates and the age at diagnosis may not be as reliable. The findings of this study indicate that GPs use information beyond what is available for researchers in the EMR to confirm clinical diagnoses when responding to validation questionnaires. Published 2013. This article is a U.S. Government work and is in the public domain in the USA.

Challenges of using primary care electronic medical records in the UK to study medications in pregnancy

PURPOSE

This paper aimed to assess the prescription of medications during pregnancy by primary care physicians in the UK.

METHODS

We identified both completed pregnancies and pregnancies losses (ectopic pregnancies, miscarriages, terminations, and stillbirths) in women aged 13-49 years enrolled in The Health Improvement Network (THIN) between 1996 and 2010 following an algorithm with three sequential cycles that searched for Read Code groups in hierarchical order: (1) indicators of conception; (2) delivery or pregnancy loss; and (3) other codes suggestive of pregnancy. Completed pregnancies were linked to liveborn infants by means of the family identification number and date of birth. Prescription of specific drugs during the first trimester and time trends during the last decade were calculated.

RESULTS

A total of 191 000 pregnancies were identified, including 148 544 completed pregnancies and 42 456 (22.2%) pregnancies losses (ectopic pregnancies, miscarriages, terminations, and stillbirths). Of the completed pregnancies, 131670 (88.6%) were successfully linked with the offspring. The most commonly prescribed drugs were antibiotics, antimycotics, asthma/allergy medications, and analgesics. From 1996 to 2010, the proportion of completed pregnancies with at least one prescription during the first trimester increased for antidepressants (1.8% to 4.2%), thyroid hormones (0.7% to 1.6%), and opiods (1.5% to 2.6%), among other drugs.

CONCLUSION

In conclusion, the prescription of several medications by primary care physicians during the first trimester of pregnancy has risen in the UK during the last decade. We discuss how, when important challenges are considered, THIN may be a promising resource to study specific therapies during pregnancy.

Using free text information to explore how and when GPs code a diagnosis of ovarian cancer: an observational study using primary care records of patients with ovarian cancer

BACKGROUND

Primary care databases provide a unique resource for healthcare research, but most researchers currently use only the Read codes for their studies, ignoring information in the free text, which is much harder to access.

OBJECTIVES

To investigate how much information on ovarian cancer diagnosis is 'hidden' in the free text and the time lag between a diagnosis being described in the text or in a hospital letter and the patient being given a Read code for that diagnosis.

DESIGN

Anonymised free text records from the General Practice Research Database of 344 women with a Read code indicating ovarian cancer between 1 June 2002 and 31 May 2007 were used to compare the date at which the diagnosis was first coded with the date at which the diagnosis was recorded in the free text. Free text relating to a diagnosis was identified (a) from the date of coded diagnosis and (b) by searching for words relating to the ovary.

RESULTS

90% of cases had information relating to their ovary in the free text. 45% had text indicating a definite diagnosis of ovarian cancer. 22% had text confirming a diagnosis before the coded date; 10% over 4 weeks previously. Four patients did not have ovarian cancer and 10% had only ambiguous or suspected diagnoses associated with the ovarian cancer code.

CONCLUSIONS

There was a vast amount of extra information relating to diagnoses in the free text. Although in most cases text confirmed the coded diagnosis, it also showed that in some cases GPs do not code a definite diagnosis on the date that it is confirmed. For diseases which rely on hospital consultants for diagnosis, free text (particularly letters) is invaluable for accurate dating of diagnosis and referrals and also for identifying misclassified cases.

Medical databases in studies of drug teratogenicity: methodological issues

More than half of all pregnant women take prescription medications, raising concerns about fetal safety. Medical databases routinely collecting data from large populations are potentially valuable resources for cohort studies addressing teratogenicity of drugs. These include electronic medical records, administrative databases, population health registries, and teratogenicity information services. Medical databases allow estimation of prevalences of birth defects with enhanced precision, but systematic error remains a potentially serious problem. In this review, we first provide a brief description of types of North American and European medical databases suitable for studying teratogenicity of drugs and then discuss manifestation of systematic errors in teratogenicity studies based on such databases. Selection bias stems primarily from the inability to ascertain all reproductive outcomes. Information bias (misclassification) may be caused by paucity of recorded clinical details or incomplete documentation of medication use. Confounding, particularly confounding by indication, can rarely be ruled out. Bias that either masks teratogenicity or creates false appearance thereof, may have adverse consequences for the health of the child and the mother. Biases should be quantified and their potential impact on the study results should be assessed. Both theory and software are available for such estimation. Provided that methodological problems are understood and effectively handled, computerized medical databases are a valuable source of data for studies of teratogenicity of drugs.

Validity of diagnostic coding within the General Practice Research Database: a systematic review

BACKGROUND

The UK-based General Practice Research Database (GPRD) is a valuable source of longitudinal primary care records and is increasingly used for epidemiological research.

AIM

To conduct a systematic review of the literature on accuracy and completeness of diagnostic coding in the GPRD.

DESIGN OF STUDY

Systematic review.

METHOD

Six electronic databases were searched using search terms relating to the GPRD, in association with terms synonymous with validity, accuracy, concordance, and recording. A positive predictive value was calculated for each diagnosis that considered a comparison with a gold standard. Studies were also considered that compared the GPRD with other databases and national statistics.

RESULTS

A total of 49 papers are included in this review. Forty papers conducted validation of a clinical diagnosis in the GPRD. When assessed against a gold standard (validation using GP questionnaire, primary care medical records, or hospital correspondence), most of the diagnoses were accurately recorded in the patient electronic record. Acute conditions were not as well recorded, with positive predictive values lower than 50%. Twelve papers compared prevalence or consultation rates in the GPRD against other primary care databases or national statistics. Generally, there was good agreement between disease prevalence and consultation rates between the GPRD and other datasets; however, rates of diabetes and musculoskeletal conditions were underestimated in the GPRD.

CONCLUSION

Most of the diagnoses coded in the GPRD are well recorded. Researchers using the GPRD may want to consider how well the disease of interest is recorded before planning research, and consider how to optimise the identification of clinical events.

Validation and validity of diagnoses in the General Practice Research Database: a systematic review

AIMS

To investigate the range of methods used to validate diagnoses in the General Practice Research Database (GPRD), to summarize findings and to assess the quality of these validations.

METHODS

A systematic literature review was performed by searching PubMed and Embase for publications using GPRD data published between 1987 and April 2008. Additional publications were identified from conference proceedings, back issues of relevant journals, bibliographies of retrieved publications and relevant websites. Publications that reported attempts to validate disease diagnoses recorded in the GPRD were included.

RESULTS

We identified 212 publications, often validating more than one diagnosis. In total, 357 validations investigating 183 different diagnoses met our inclusion criteria. Of these, 303 (85%) utilized data from outside the GPRD to validate diagnoses. The remainder utilized only data recorded in the database. The median proportion of cases with a confirmed diagnosis was 89% (range 24–100%). Details of validation methods and results were often incomplete.

CONCLUSIONS

A number of methods have been used to assess validity. Overall, estimates of validity were high. However, the quality of reporting of the validations was often inadequate to permit a clear interpretation. Not all methods provided a quantitative estimate of validity and most methods considered only the positive predictive value of a set of diagnostic codes in a highly selected group of cases. We make recommendations for methodology and reporting to strengthen further the use of the GPRD in research.

The epidemiology of osteonecrosis: findings from the GPRD and THIN databases in the UK

We conducted a case-control study to examine osteonecrosis (ON) incidence, patient characteristics, and selected potential risk factors using two health record databases in the UK. Statistically significant risk factors for ON included systemic corticosteroid use, hospitalization, referral or specialist visit, bone fracture, any cancer, osteoporosis, connective tissue disease, and osteoarthritis.

INTRODUCTION

The purpose of this case-control study was to examine the incidence of osteonecrosis (ON), patient characteristics, and selected potential risk factors for ON using two health record databases in the UK: the General Practice Research Database and The Health Improvement Network.

METHODS

ON cases (n = 792) were identified from 1989 to 2003 and individually matched (age, sex, and medical practice) up to six controls (n = 4,660) with no record of ON. Possible risk factors were considered for inclusion based on a review of published literature. Annual incidence rates were computed, and a multivariable logistic regression model was derived to evaluate selected risk factors.

RESULTS

ON of the hip represented the majority of cases (75.9%). Statistically significant risk factors for ON were systemic corticosteroid use in the previous 2 years, hospitalization, referral or specialist visit, bone fracture, any cancer, osteoporosis, connective tissue disease, and osteoarthritis within the past 5 years. Only 4.4% of ON cases were exposed to bisphosphonates within the previous 2 years.

CONCLUSIONS

This study provides further perspective on the descriptive epidemiology of ON. Studies utilizing more recent data may further elucidate the understanding of ON key predictors.

Phynx: an open source software solution supporting data management and web-based patient-level data review for drug safety studies in the general practice research database and other health care databases

PURPOSE

To develop a software solution that supports management and clinical review of patient data from electronic medical records databases or claims databases for pharmacoepidemiological drug safety studies.

METHODS

We used open source software to build a data management system and an internet application with a Flex client on a Java application server with a MySQL database backend. The application is hosted on Amazon Elastic Compute Cloud. This solution named Phynx supports data management, Web-based display of electronic patient information, and interactive review of patient-level information in the individual clinical context. This system was applied to a dataset from the UK General Practice Research Database (GPRD).

RESULTS

Our solution can be setup and customized with limited programming resources, and there is almost no extra cost for software. Access times are short, the displayed information is structured in chronological order and visually attractive, and selected information such as drug exposure can be blinded. External experts can review patient profiles and save evaluations and comments via a common Web browser.

CONCLUSIONS

Phynx provides a flexible and economical solution for patient-level review of electronic medical information from databases considering the individual clinical context. It can therefore make an important contribution to an efficient validation of outcome assessment in drug safety database studies.

Determining the predictive value of Read/OXMIS codes to identify incident acute myocardial infarction in the General Practice Research Database

PURPOSE

To determine (1) the positive predictive value (PPV) of multiple Read/OXMIS codes to identify incident acute myocardial infarction (AMI) cases in General Practice Research Database (GPRD); (2) the ability to capture the correct timing of the clinical event.

METHODS

A random sample of 238 records (from 155 general practitioner (GP) practices) with AMI codes, between 1 January 1997 and 31 December 2004, was selected from GPRD. Questionnaires were sent to the GPs to verify the diagnosis and timing of code-identified incident AMI events and collect supporting information. We calculated the PPV of the AMI codes as the proportion of code-identified AMIs that the GPs confirmed as AMI cases. Two physicians from Food and Drug administration (FDA), blinded to the GP response, reviewed the supporting hospital records returned by GPs for 98 AMI cases.

RESULTS

A total of 217 questionnaires were completed (91% response rate). The PPV of the AMI codes was 93% (201/217). Thirty one (15%) cases had a different event date than the one recorded in the electronic medical records (EMR); 28 (90%) of the dates were within 15 days. One GP indicated that a patient had a previous AMI, 110 days before the codes that captured the AMI diagnosis. A total of 159 (79%) AMI cases were hospitalized; hospital records were provided for 98 patients. Physician review of the hospital records found that 96% of these records had enough information for classification, but not independent diagnosis, of AMI.

CONCLUSIONS

Information in GPRD is sufficient to identify incident AMI cases and determine the event date with reasonable accuracy.