Publications
scientifiques

PURPOSE: Pharmacoepidemiology researchers often utilize data from two UK electronic medical record databases, the Clinical Practice Research Datalink (CPRD) and The Health Improvement Network (THIN), and may choose to combine the two in an effort to increase sample size. To minimize duplication of data, previous studies examined the practice-level overlap between these databases. However, the proportion of overlapping patients remains unknown. We developed a method using demographic and pharmacy variables to identify patients included in both CPRD and THIN, and applied this method to measure the proportion of overlapping patients who initiated the oral anti-diabetic drug saxagliptin. METHODS: We conducted a cross-sectional study among patients initiating saxagliptin in CPRD and THIN between October 2009 and September 2012. Within both databases, we identified patients: (i) ≥18 years, (ii) newly prescribed saxagliptin, and (iii) with ≥180 days enrollment prior to saxagliptin initiation. Demographic data (birth year, sex, patient registration date, family number, and marital status) and prescriptions (including dates) for the first two oral anti-diabetic drugs prescribed within the study period were used to identify matching patients. RESULTS: Among 4202 CPRD and 3641 THIN patients initiating saxagliptin, 2574 overlapping patients (61% of CPRD saxagliptin initiators; 71% of THIN saxagliptin initiators) were identified. Among these patients, 2474 patients (96%) perfectly matched on all demographic and prescription data. CONCLUSIONS: Within each database, over 60% of patients initiating saxagliptin were included within both CPRD and THIN. Combined demographic and prescription data can be used to identify patients included in both CPRD and THIN.

BACKGROUND: We aimed to assess the potential usefulness of primary care data in the UK for estimating smoking prevalence in pregnancy by comparing the primary care data estimates with those obtained from other data sources. METHODS: In The Health Improvement Network (THIN) primary care database, we identified pregnant smokers using smoking information recorded during pregnancy. Where this information was missing, we used smoking information recorded prior to pregnancy. We compared annual smoking prevalence from 2000 to 2012 in THIN with measures from the Infant Feeding Survey (IFS), Smoking At Time of Delivery (SATOD), Child Health Systems Programme (CHSP) and Scottish Morbidity Record (SMR). RESULTS: Smoking estimates from THIN data converged with estimates from other sources after 2004, though still do not agree completely. For example, in 2012 smoking prevalence at booking was 11.6% in THIN using data recorded only during pregnancy, compared with 19.6% in SMR data. However, the use of smoking data recorded up to 27 months before conception increased the THIN prevalence to 20.3%, improving the comparability. CONCLUSIONS: Under-recording of smoking status during pregnancy results in unreliable prevalence estimates from primary care data and needs improvement. However, in the absence of gestational smoking data, the inclusion of pre-conception smoking records may increase the utility of primary care data. One strategy to improve gestational smoking status recording in primary care could be the inclusion of pregnancy in the Quality and Outcome's Framework as a condition for which smoking status and smoking cessation advice must be recorded electronically in patient records.

IMPORTANCE: Reports on the association between statins and memory impairment are inconsistent. OBJECTIVE: To assess whether statin users show acute decline in memory compared with nonusers and with users of nonstatin lipid-lowering drugs (LLDs). DESIGN, SETTING, AND PARTICIPANTS: Using The Health Improvement Network database during January 13, 1987, through December 16, 2013, a retrospective cohort study compared 482,543 statin users with 2 control groups: 482,543 matched nonusers of any LLDs and all 26,484 users of nonstatin LLDs. A case-crossover study of 68,028 patients with incident acute memory loss evaluated exposure to statins during the period immediately before the outcome vs 3 earlier periods. Analysis was conducted from July 7, 2013, through January 15, 2015. RESULTS: When compared with matched nonusers of any LLDs (using odds ratio [95% CI]), a strong association was present between first exposure to statins and incident acute memory loss diagnosed within 30 days immediately following exposure (fully adjusted, 4.40; 3.01-6.41). This association was not reproduced in the comparison of statins vs nonstatin LLDs (fully adjusted, 1.03; 0.63-1.66) but was also present when comparing nonstatin LLDs with matched nonuser controls (adjusted, 3.60; 1.34-9.70). The case-crossover analysis showed little association. CONCLUSIONS AND RELEVANCE: Both statin and nonstatin LLDs were strongly associated with acute memory loss in the first 30 days following exposure in users compared with nonusers but not when compared with each other. Thus, either all LLDs cause acute memory loss regardless of drug class or the association is the result of detection bias rather than a causal association.

In England, hospital admissions for severe staphylococcal boils and abscesses trebled between 1989 and 2004. We investigated this trend using routine data from primary and secondary care. We used The Health Improvement Network (THIN), a large primary-care database and national data on hospital admissions from Hospital Episode Statistics (HES). Time trends in the incidence of primary-care consultations for boils and abscesses were estimated for 1995-2010. HES data were used to calculate age-standardized hospital admission rates for boils, abscesses and cellulitis. The incidence of boil or abscess was 450 [95% confidence interval (CI) 447-452] per 100 000 person-years and increased slightly over the study period (incidence rate ratio 1·005, 95% CI 1·004-1·007). The rate of repeat consultation for a boil or abscess increased from 66 (95% CI 59-73) per 100 000 person-years in 1995 to peak at 97 (95% CI 94-101) per 100 000 person-years in 2006, remaining stable thereafter. Hospital admissions for abscesses, carbuncles, furuncles and cellulitis almost doubled, from 123 admissions per 100 000 in 1998/1999 to 236 admissions per 100 000 in 2010/2011. Rising hospitalization and recurrence rates set against a background of stable community incidence suggests increased disease severity. Patients may be experiencing more severe and recurrent staphylococcal skin disease with limited treatment options.

BACKGROUND: Primary care databases from the UK have been widely used to produce evidence on the epidemiology and health service usage of a wide range of conditions. To date there have been few evaluations of the comparability of estimates between different sources of these data. AIM: To estimate the comparability of two widely used primary care databases, the Health Improvement Network Database (THIN) and the General Practice Research Database (GPRD) using venous leg ulceration as an exemplar condition. DESIGN OF STUDY: Cross prospective cohort comparison. SETTING: GPRD and the THIN databases using data from 1998 to 2006. METHOD: A data set was extracted from both databases containing all cases of persons aged 20 years or greater with a database diagnosis of venous leg ulceration recorded in the databases for the period 1998-2006. Annual rates of incidence and prevalence of venous leg ulceration were calculated within each database and standardized to the European standard population and compared using standardized rate ratios. RESULTS: Comparable estimates of venous leg ulcer incidence from the GPRD and THIN databases could be obtained using data from 2000 to 2006 and of prevalence using data from 2001 to 2006. CONCLUSIONS: Recent data collected by these two databases are more likely to produce comparable results of the burden venous leg ulceration. These results require confirmation in other disease areas to enable researchers to have confidence in the comparability of findings from these two widely used primary care research resources.

OBJECTIVE: To characterize the relationship between body mass index (BMI) and all-cause mortality among adults with hypertension. METHODS: We conducted a retrospective cohort study of adults with hypertension from The Health Improvement Network primary care research database in the United Kingdom. Mortality rates within each BMI category were compared after adjustment for age, sex, pre existing medical conditions, smoking, alcohol use, and socioeconomic status. RESULTS: During a median follow-up period of 8.0 years, 34,354 deaths occurred (8.8%) among the 388,724 study patients. In multivariable analysis that included interaction terms between smoking, heavy alcohol use, preexisting cardiovascular disease and chronic obstructive pulmonary disease (COPD), and BMI category to account for these conditions' association with increased mortality at lower BMI levels, a U-shaped relationship was observed between BMI and mortality, with no difference in mortality in overweight (25.0-29.9 kg/m(2) ) versus normal (18.5-24.9 kg/m(2) ) BMI (risk ratio 1.00, 95% confidence interval 0.95 to 1.04, P = 0.80). In multivariable analysis that used narrower BMI categories, lowest mortality was observed between BMI of 23.0 and 26.9 kg/m(2) . CONCLUSIONS: Patients with hypertension who have no history of smoking, heavy alcohol use, cardiovascular disease, or COPD have lowest mortality between BMI of 23.0 and 26.9 kg/m(2) .

BACKGROUND AND OBJECTIVE: Recent evidence has linked childhood antibiotic use and microbiome disturbance to autoimmune conditions. This study tested the hypothesis that antibiotic exposure was associated with newly diagnosed juvenile idiopathic arthritis (JIA). METHODS: We performed a nested case-control study in a population-representative medical records database from the United Kingdom. Children with newly diagnosed JIA were compared with age- and gender-matched control subjects randomly selected from general practices containing at least 1 case, excluding those with inflammatory bowel disease, immunodeficiency, or other systemic rheumatic diseases. Conditional logistic regression was used to examine the association between antibacterial antibiotics (including number of antibiotic courses and timing) and JIA after adjusting for significant confounders. RESULTS: Any antibiotic exposure was associated with an increased rate of developing JIA (adjusted odds ratio: 2.1 [95% confidence interval: 1.2-3.5]). This relationship was dose dependent (adjusted odds ratio over 5 antibiotic courses: 3.0 [95% confidence interval: 1.6-5.6]), strongest for exposures within 1 year of diagnosis, and did not substantively change when adjusting for number or type of infections. In contrast, nonbacterial antimicrobial agents (eg, antifungal, antiviral) were not associated with JIA. In addition, antibiotic-treated upper respiratory tract infections were more strongly associated with JIA than untreated upper respiratory tract infections. CONCLUSIONS: Antibiotics were associated with newly diagnosed JIA in a dose- and time-dependent fashion in a large pediatric population. Antibiotic exposure may play a role in JIA pathogenesis, perhaps mediated through alterations in the microbiome.

Big longitudinal observational medical data potentially hold a wealth of information and have been recognised as potential sources for gaining new drug safety knowledge. Unfortunately there are many complexities and underlying issues when analysing longitudinal observational data. Due to these complexities, existing methods for large-scale detection of negative side effects using observational data all tend to have issues distinguishing between association and causality. New methods that can better discriminate causal and non-causal relationships need to be developed to fully utilise the data. In this paper we propose using a set of causality considerations developed by the epidemiologist Bradford Hill as a basis for engineering features that enable the application of supervised learning for the problem of detecting negative side effects. The Bradford Hill considerations look at various perspectives of a drug and outcome relationship to determine whether it shows causal traits. We taught a classifier to find patterns within these perspectives and it learned to discriminate between association and causality. The novelty of this research is the combination of supervised learning and Bradford Hill's causality considerations to automate the Bradford Hill's causality assessment. We evaluated the framework on a drug safety gold standard known as the observational medical outcomes partnership's non-specified association reference set. The methodology obtained excellent discrimination ability with area under the curves ranging between 0.792 and 0.940 (existing method optimal: 0.73) and a mean average precision of 0.640 (existing method optimal: 0.141). The proposed features can be calculated efficiently and be readily updated, making the framework suitable for big observational data.

AIMS: We aimed to identify the incidence and risk factors for first ever atrial fibrillation among patients with newly diagnosed heart failure following initial heart failure diagnosis. METHODS: A heart failure inception cohort of patients aged 20-89 years without atrial fibrillation or cancer (N = 14 457) from 2000 to 2005 was identified from The Health Improvement Network primary care database in the United Kingdom and followed for a mean of 2.67 years. First ever cases of atrial fibrillation were identified and controls (N = 3000) were frequency matched to cases by age and sex. Odds ratios (ORs) with 95% confidence intervals (CIs) were estimated using unconditional logistic regression. RESULTS: One thousand four hundred and eighty-nine patients (10.3%) developed a first episode of atrial fibrillation: incidence rate 27.3/1000 person-years. A three-fold increased risk of atrial fibrillation was seen in the first 6 months after heart failure diagnosis, OR 3.62 (95% CI: 2.97-4.42) with the risk decreasing thereafter. Other risk factors were excessive alcohol consumption (OR 2.91, 1.60-5.30) and valvular heart disease (OR 1.98, 1.63-2.40) and use of oral steroids (OR 1.76, 95% CI: 1.40-2.22). Reduced risks of atrial fibrillation were found with use of statins (OR 0.65, 95% CI: 0.56-0.76) and β-blockers (OR 0.78, 95% CI: 0.67-0.91). CONCLUSIONS: The incidence of first ever atrial fibrillation among newly diagnosed heart failure patients is high, especially in the first 6 months after diagnosis. This time relationship, together with the identified risk factors for atrial fibrillation, warrants consideration in the medical care of patients with heart failure.

PURPOSE: To examine patterns of outpatient and community antibiotic use among adults in five European countries. METHODS: We used healthcare data of 28.8 million adults from six population-based ARITMO project databases to ascertain information on systemic antibiotic use in Denmark (2000-2008), the Netherlands (1999-2010), Italy (2000-2010), the UK (1996-2009), and Germany (2004-2008). We estimated overall, and age-group and sex specific antibiotic use as defined daily doses (DDD) per 1000 inhabitants per day. We computed annual age- and sex-standardized population prevalence of antibiotic use per 1000 persons-years (p-y) and the mean duration (in days) of antibiotic use. RESULTS: The overall antibiotic use varied from 8.7 DDD per 1000 inhabitants per day in the UK to 18.1 DDD in Denmark, representing a 2.1-fold geographical variation. In all countries, prescribing was relatively high among individuals aged 15-19 years; lower in those aged 20-50 years; and then increased steadily reaching 41.8 DDD per 1000 inhabitants per day in individuals ≥ 85 years in Denmark. After age- and sex-standardization, prevalence of antibiotic use varied threefold from 160.2/1000 p-y in the UK to 421.1/1000 p-y in Italy. The ratio of broad- to narrow-spectrum penicillin, cephalosporin, and macrolide use varied from 0.6 in Denmark to 120.2 in Italy. Women used more antibiotics than men did in all countries. Across countries, the mean duration of antibiotic use varied 1.3 to 21.1-fold for different antibiotics. CONCLUSIONS: Antibiotic use is high in women and the elderly. Prescribing patterns vary substantially across European countries, both according to overall consumption, user prevalence, duration, and narrow- versus broad-spectrum antibiotics.

PURPOSE: The study aims to analyse overall as well as subgroup-specific outpatient paediatric macrolide use in five European countries, including time trends of macrolide prescription rates, and to provide potential targets for future interventions aiming to promote judicious macrolide use. METHODS: Macrolide prescription rates per 1000 person years to paediatric outpatients (≤18 years) were calculated using healthcare databases from Denmark, Germany, Italy, The Netherlands and the UK. Poisson regression analysis was used to estimate the influence of increasing calendar year on total macrolide and subgroup-specific prescription rates based on monthly data, adjusted for seasonal variations. Time periods for which data were available varied between 4 (Italy 2007-10, Germany 2005-8) and 10 years (UK 2000-9). RESULTS: Paediatric macrolide use in 2008 varied between 199 (Italy) and 47 (Netherlands) prescriptions per 1000 person years. Prescription rates of short-acting macrolides declined significantly in all countries but the UK. The use of intermediate-acting macrolides significantly rose with increasing calendar year in Denmark (rate ratio (RR) = 1.12) and the UK (RR = 1.06), but decreased in Germany (RR = 0.84) and The Netherlands (RR = 0.97). Prescription rates of long-acting agents increased in Denmark (RR = 1.05), The Netherlands (RR = 1.05) and the UK (RR = 1.11) (all trends p < 0.05). The greatest seasonal variations of macrolide use between summer and winter months were observed in Italy and Germany. CONCLUSIONS: The observed trend toward increased prescribing of intermediate- and/or long-acting agents might further increase resistance pressure on bacterial pathogens due to their prolonged plasma half-life and broader antibacterial activity. Marked seasonality of prescription rates in the high-utilising countries, Italy and Germany, suggests frequent prescription of macrolides to treat respiratory infections which may be of viral origin.

CONTEXT: Endogenous glucocorticoids are known to increase blood pressure, but very little is known about the early effects of synthetic glucocorticoids (eg, prednisone, dexamethasone) on blood pressure. OBJECTIVE: To assess longitudinal variations of blood pressure before and after initiation of systemic glucocorticoid therapy. DESIGN, SETTING, AND PATIENTS: Adult patients prescribed synthetic glucocorticoid therapy for at least 3 months and registered between 2004 and 2012 in The Health Improvement Network (THIN) primary care database. MAIN OUTCOME MEASURE: Systolic and diastolic blood pressure as measured in primary care. RESULTS: Among the 16,351 patients prescribed antihypertensive drug within the year before glucocorticoid initiation (women, 57.1%; age [mean ± SD], 72.0 ± 11.4 y), the mean (SD) blood pressure within the year before glucocorticoid initiation was 140 (18)/78 (9) mm Hg. It was 139 (19)/77 (11), 139 (19)/77 (10), and 139 (19)/77 (11) mm Hg during the first, second, and third months of exposure, respectively. We did not find any evidence of the effects of glucocorticoids on blood pressure in these patients. Among the 6914 patients not prescribed antihypertensive drug within the year before glucocorticoid initiation (women, 64.9%; age, 60.0 ± 17.8 y), the figures were 133 (17)/78 (10), 135 (20)/79 (11), 133 (19)/78 (11), and 133 (19)/78 (11) mm Hg before exposure and during the first, second, and third months of exposure, respectively. In this group of patients, glucocorticoid exposure was associated with a slight (<1 mm Hg; P = .03) increase of systolic but not diastolic (P = .52) blood pressure. However, in 4% of people or fewer, extreme increases or decreases (ie, ≥30 mm Hg for systolic blood pressure) in blood pressure were observed after glucocorticoid initiation. Prednisone/prednisolone use was associated with a higher risk of extreme increase in systolic blood pressure than the other synthetic glucocorticoids (odds ratio, 4.9 [95% confidence interval, 1.9-12.6]; P = .001). CONCLUSION: Contrary to what is usually thought, the increase of blood pressure during the first months of exposure to synthetic glucocorticoids seems clinically nonsignificant.

BACKGROUND: For patients with asthma, COPD, or asthma-COPD overlap syndrome (ACOS), inter-country comparisons of seasonal changes in drug prescriptions are scarce or missing. Hence, we aimed to compare seasonal changes in prescription rates of long-acting beta-2-agonist (LABA) in four European countries. METHODS: A common study protocol was applied to six health care databases (Germany, Spain, the Netherlands (2), and the UK (2)) to calculate age- and sex-standardized point prevalence rates (PPRs) of LABA-containing prescriptions by the 1st of March, June, September, and December of each year during the study period 2002-2009. Seasonal variation of PPRs was quantified using seasonal indexes (SIs; based on the ratio-to-moving-average-method) and SIs averaged over the study period (aSI) stratified by sex, age, and indication (asthma, COPD, or ACOS). RESULTS: There was a moderate seasonal change in LABA-containing prescriptions which was more pronounced in asthma or COPD patients compared to ACOS patients. For asthma and ACOS patients, highest seasonal variation was found for patients living in Spain (aSI: 87.3-110.7, aSI: 93.2-103.1) whereas for COPD highest seasonal variation was revealed for the NPCRD database (the Netherlands) (aSI: 92.2-105.6). Regarding age and sex, highest seasonal variation was found in Spanish boys under 10 years of age having a diagnosis of asthma. CONCLUSIONS: By applying a common analysis in six databases, we could observe moderate overall seasonal changes in LABA-containing prescription rates in patients with asthma, COPD, or ACOS.

BACKGROUND: Antiplatelet (AP) therapy is well established for the secondary prevention of acute coronary events. However, patients may discontinue treatment, often owing to gastrointestinal (GI) complications, leaving them at elevated risk of recurrent cardiovascular events. OBJECTIVES: This descriptive retrospective study assessed trends in prescription of AP agents and coprescription of gastroprotective therapy, after an acute coronary event. Discontinuation of AP therapy within 2 years of an event and factors predicting discontinuation were investigated. METHODS: The study was conducted in a UK primary care setting from 2000 to 2008; a total of 27, 351 patients aged 50 to 84 years were included in the analysis. Main outcome measures were exposures to low-dose acetylsalicylic acid (ASA), clopidogrel, and proton pump inhibitors (PPIs). RESULTS: At 90 days after an acute coronary event, 85.9% of patients had been prescribed some form of AP therapy and 33.6% of patients who were issued at least 1 ASA prescription in this period were also issued a PPI prescription. The use of dual antiplatelet therapy (DAT) 90 days after an event increased from 2% in 2000 to over 50% in 2008. An estimated 15.1% of patients on ASA monotherapy and 37.5% on DAT discontinued treatment within 1 year. A bleeding event during follow-up, including upper GI bleeding or hemorrhagic stroke, was the strongest predictor of discontinuation. CONCLUSION: Although most patients were prescribed AP therapy in the 90 days following an acute coronary event, a substantial proportion discontinued DAT or ASA monotherapy within 1 year. It is essential that physicians consider strategies to reduce the risk of discontinuation of AP therapy.

AIMS: The aim of the analysis was to investigate whether insulin intensification, based on the use of intensive insulin regimens as recommended by the current standard of care in routine clinical practice, would be cost-effective for patients with type 2 diabetes in the UK. METHODS: Clinical data were derived from a retrospective analysis of 3185 patients with type 2 diabetes on basal insulin in The Health Improvement Network (THIN) general practice database. In total, 48% (614 patients) intensified insulin therapy, defined by adding bolus or premix insulin to a basal regimen, which was associated with a reduction in HbA1c and an increase in body mass index. Projections of clinical outcomes and costs (2011 GBP) over patients' lifetimes were made using a recently validated type 2 diabetes model. RESULTS: Immediate insulin intensification was associated with improvements in life expectancy, quality-adjusted life expectancy and time to onset of complications versus no intensification or delaying intensification by 2, 4, 6, or 8 years. Direct costs were higher with the insulin intensification strategy (due to the acquisition costs of insulin). Incremental cost-effectiveness ratios for insulin intensification were GBP 32,560, GBP 35,187, GBP 40,006, GBP 48,187 and GBP 55,431 per QALY gained versus delaying intensification 2, 4, 6 and 8 years, and no intensification, respectively. CONCLUSIONS: Although associated with improved clinical outcomes, insulin intensification as practiced in the UK has a relatively high cost per QALY and may not lead to cost-effective outcomes for patients with type 2 diabetes as currently defined by UK cost-effectiveness thresholds.

BACKGROUND: the objective of the study was to estimate and compare the incidence rates of ischaemic and haemorrhagic stroke and seizure among cohorts with and without Alzheimer's disease (AD) dementia. METHODS: we conducted a retrospective cohort study using electronic medical records (EMRs) from primary care practices that participated in The Health Improvement Network (THIN) in the United Kingdom from 1 January 1990 to 31 July 2009. For each AD-dementia patient, we selected one general population control patient without AD-dementia matched to one AD-dementia patient on year of birth, sex and physician practice. FINDINGS: the AD-dementia cohorts were 68% female and averaged 80 years of age at the start of follow-up. Populations for analysis included 19,902 AD-dementia and matched non-AD-dementia patients with no history of stroke at baseline in which 790 incident cases of stroke occurred, and similarly, 22,084 AD-dementia and matched patients with no history of seizure at baseline in which 286 cases of seizure occurred. After adjusting for risk factors for each outcome, hazard ratios comparing AD-dementia with non-AD-dementia patients indicated higher rates among AD-dementia patients for stroke (HR = 1.29, 95% CI 1.11, 1.50) and seizure (HR = 5.31, 95% CI 3.97, 7.10). For stroke and seizure, the incidence rate ratios comparing AD-dementia patients with non-AD-dementia controls were greatest for the younger age groups. AD-dementia was observed to be a risk factor for both haemorrhagic stroke and seizures. Increasing age was associated with a decrease in relative risk and an increase in absolute risk.

BACKGROUND: Antenatal antiepileptic drug (AED) use has been found to be associated with increased major congenital anomaly (CA) risks. However whether such AED-associated risks were different according to periconceptional high dose (5mg daily) folic acid supplementation is still unclear. METHODS: We included 258,591 singleton live-born children of mothers aged 15-44 years in 1990-2013 from The Health Improvement Network, a large UK primary care database. We identified all major CAs according to the European Surveillance of Congenital Anomalies classification. Absolute risks and adjusted odds ratios (aOR) were calculated comparing children of mothers prescribed AEDs to those without such prescriptions, stratified by folic acid prescriptions around the time of conception (one month before conception to two months post-conception). RESULTS: CA risk was 476/10,000 in children of mothers with first trimester AEDs compared with 269/10,000 in those without AEDs equating to an aOR of 1.82, 95% confidence interval 1.30-2.56. The highest system-specific risks were for heart anomalies (198/10,000 and 79/10,000 respectively, aOR 2.49,1.47-4.21). Sodium valproate and lamotrigine were both associated with increased risks of any CA (aOR 2.63,1.46-4.74 and aOR 2.01,1.12-3.59 respectively) and system-specific risks. Stratification by folic acid supplementation did not show marked reductions in AED-associated risks (e.g. for CAs overall aOR 1.75, 1.01-3.03 in the high dose folic acid group and 1.94, 95%CI 1.21-3.13 in the low dose or no folic acid group); however, the majority of mothers taking AEDs only initiated high dose folic acid from the second month of pregnancy. CONCLUSIONS: Children of mothers with AEDs in the first trimester of pregnancy have a 2-fold increased risk of major CA compared to those unexposed. We found no evidence that prescribed high dose folic acid supplementation reduced such AED-associated risks. Although statistical power was limited, prescribing of folic acid too late for it to be effective during the organogenic period or selective prescribing to those with more severe morbidity may explain these findings.

OBJECTIVE: To assess clinical outcomes and cost-effectiveness of using a two-layer cohesive compression bandage (TLCCB; Coban 2) compared with a two-layer compression system (TLCS; Ktwo) and a four-layer compression system (FLCS; Profore) in treating venous leg ulcers (VLUs) in clinical practice in the UK, from the perspective of the National Health Service (NHS). METHOD: This was a retrospective analysis of the case records of VLU patients, randomly extracted from The Health Improvement Network (THIN) database (a nationally representative database of clinical practice among patients registered with general practitioners in the UK), who were treated with either TLCCB (n=250), TLCS (n=250) or FLCS (n=175). Clinical outcomes and health-care resource use (and costs) over six months after starting treatment with each compression system were estimated. Differences in outcomes and resource use between treatments were adjusted for differences in baseline covariates. RESULTS: Patients' mean age was 75 years old and 57% were female. The mean time with a VLU was 6-7 months and the mean initial wound size was 77-85 cm2. The overall VLU healing rate, irrespective of bandage type, was 44% over the six months' study period. In the TLCCB group, 51% of wounds had healed by six months compared with 40% (p=0.03) and 28% (p=0.001) in the TLCS and FLCS groups, respectively. The mean time to healing was 2.5 months. Patients in the TLCCB group experienced better health-related quality of life (HRQoL) over six months (0.374 quality-adjusted life years (QALYs) per patient), compared with the TLCS (0.368 QALYs per patient) and FLCS (0.353 QALYs per patient). The mean six-monthly NHS management cost was £2,413, £2,707 and £2,648 per patient in the TLCCB, TLCS and FLCS groups, respectively. CONCLUSION: Despite the systems studied reporting similar compression levels when tested in controlled studies, real-world evidence demonstrates that initiating treatment with TLCCB, compared with the other two compression systems, affords a more cost-effective use of NHS-funded resources in clinical practice, since it resulted in an increased healing rate, better HRQoL and a reduction in NHS management cost. The evidence also highlighted the lack of continuity between clinicians managing a wound, the inconsistent nature of the administered treatments and the lack of specialist involvement, all of which may impact on healing. DECLARATION OF INTEREST: This study was supported by an unrestricted research grant from 3M Health Care, UK. 3M Health Care had no influence on the study design, the collection, analysis, and interpretation of data, or on the writing of, and decision to submit for publication, the manuscript.