Publications
scientifiques

BACKGROUND: The roles of depression and antidepressants in triggering reflux symptoms remain unclear. AIM: To compare the incidence of gastro-oesophageal reflux disease (GERD) in individuals with and without a depression diagnosis and to evaluate risk factors for a GERD diagnosis. The relationship between antidepressant treatment and GERD was also assessed. METHODS: The Health Improvement Network UK primary care database was used to identify patients with incident depression and an age- and sex-matched control cohort with no depression diagnosis. Incident GERD diagnoses were identified during a mean follow-up of 3.3 years. Furthermore, we performed nested case-control analyses where odds ratios (OR) with 95% confidence intervals (CI) were estimated by unconditional logistic regression in multivariable models. RESULTS: The incidence of GERD was 14.2 per 1000 person-years in the depression cohort and 8.3 per 1000 person-years in the control cohort. The hazard ratio of GERD in patients with depression compared with controls was 1.72 (95% CI: 1.60-1.85). Among patients with depression, tricyclic antidepressant use was associated with an increased risk of GERD (OR: 1.71; 95% CI: 1.34-2.20), while selective serotonin reuptake inhibitors were not associated with GERD. CONCLUSIONS: A depression diagnosis is associated with an increased risk of a subsequent GERD diagnosis, particularly in individuals using tricyclic antidepressants.

OBJECTIVE: Patients with COPD are at risk for cardiovascular events. This is attributed to increased systemic inflammation. The course of COPD is punctuated by exacerbations, which further increase systemic inflammation, but the risk of vascular events in the postexacerbation period has never been defined. METHODS: We analyzed data from 25,857 patients with COPD entered in The Health Improvement Network database over a 2-year period. Exacerbations were defined using a health-care use definition of prescription of oral corticosteroids > 20 mg/d and/or selected oral antibiotics. The risk of myocardial infarction (MI) and stroke in the postexacerbation period was calculated relative to the patient's baseline risk using the self-controlled case series approach. RESULTS: We identified 524 MIs in 426 patients and 633 ischemic strokes in 482 patients. The incidence rates of MI and stroke were 1.1 and 1.4 per 100 patient-years, respectively. There was a 2.27-fold (95% CI, 1.1-4.7; P = .03) increased risk of MI 1 to 5 days after exacerbation (defined by prescription of both steroids and antibiotics). This relative risk diminished progressively with time and was not significantly different from the baseline MI risk at any other postexacerbation time interval. One in 2,513 exacerbations was associated with MI within 1 to 5 days. There was a 1.26-fold (95% CI, 1.0-1.6; P = .05) increased risk of stroke 1 to 49 days after exacerbation. CONCLUSION: The results suggest that exacerbations of COPD increase the risk of MI and stroke. This may have implications for therapy in both stable and exacerbated COPD.

BACKGROUND: There are currently limited data regarding the epidemiology of anaphylaxis. OBJECTIVE: To estimate the incidence of anaphylaxis from all causes, to explore the variety of diagnoses that may predispose to an anaphylactic episode, and to estimate the rate of recurrence of anaphylaxis in patients with no asthma, nonsevere asthma, and severe asthma. METHODS: The Health Improvement Network database provided data on individuals 10 to 79 years old who had been enrolled for at least 1 year with a general practitioner in the United Kingdom and had at least 1 health contact in the year before entering the study. RESULTS: Anaphylaxis incidence rates (per 100,000 person-years) were 21.28 (95% CI, 17.64-25.44) and 50.45 (95% CI, 44.67-56.76) in the no asthma and overall asthma cohorts, respectively. Risk of anaphylaxis was greater in the nonsevere asthma (relative risk, 2.07; 95% CI, 1.65-2.60) and severe asthma (relative risk, 3.29; 95% CI, 2.47-3.47) subgroups compared with the no asthma cohort. The incidence rate of anaphylaxis was higher in women than men (22.65 vs 19.56 per 100,000 person-years). Within the overall asthma population, patients at significantly increased risk of anaphylaxis included those with allergic rhinitis or atopic dermatitis, and current users of antihistamines, oral steroids, or antibiotics (compared with nonusers). Drug and food allergies were the most common known causes of anaphylaxis. CONCLUSION: Patients with asthma have a greater risk of anaphylaxis than those without asthma, and the risk is greater in severe than nonsevere asthma. Women are at higher risk of anaphylaxis than men, especially those with severe asthma.

OBJECTIVE: To evaluate the performance of the QRISK2 score for predicting 10-year cardiovascular disease in an independent UK cohort of patients from general practice records and to compare it with the NICE version of the Framingham equation and QRISK1. DESIGN: Prospective cohort study to validate a cardiovascular risk score. SETTING: 365 practices from United Kingdom contributing to The Health Improvement Network (THIN) database. PARTICIPANTS: 1.58 million patients registered with a general practice between 1 January 1993 and 20 June 2008, aged 35-74 years (9.4 million person years) with 71 465 cardiovascular events. MAIN OUTCOME MEASURES: First diagnosis of cardiovascular disease (myocardial infarction, angina, coronary heart disease, stroke, and transient ischaemic stroke) recorded in general practice records. RESULTS: QRISK2 offered improved prediction of a patient's 10-year risk of cardiovascular disease over the NICE version of the Framingham equation. Discrimination and calibration statistics were better with QRISK2. QRISK2 explained 33% of the variation in men and 40% for women, compared with 29% and 34% respectively for the NICE Framingham and 32% and 38% respectively for QRISK1. The incidence rate of cardiovascular events (per 1000 person years) among men in the high risk group was 27.8 (95% CI 27.4 to 28.2) with QRISK2, 21.9 (21.6 to 22.2) with NICE Framingham, and 24.8 (22.8 to 26.9) with QRISK1. Similarly, the incidence rate of cardiovascular events (per 1000 person years) among women in the high risk group was 24.3 (23.8 to 24.9) with QRISK2, 20.6 (20.1 to 21.0) with NICE Framingham, and 21.8 (18.9 to 24.6) with QRISK1. CONCLUSIONS: QRISK2 is more accurate in identifying a high risk population for cardiovascular disease in the United Kingdom than the NICE version of the Framingham equation. Differences in performance between QRISK2 and QRISK1 were marginal.

OBJECTIVES: Varenicline is a licensed smoking cessation medication in the EU, USA and many other countries worldwide. This study was designed to assess its effectiveness in a UK general practice setting. METHODS: The main outcome measure was the rate of smoking cessation, defined as the seven-day point prevalence after six months from starting varenicline. Varenicline users were identified from records in The Health Improvement Network (THIN) database. A questionnaire on smoking cessation was sent to patients who commenced treatment close to the selection date (six months prior to the date of questionnaire dispatch). RESULTS: The response rate was 26.4%: 193 responses were received. Ninety percent had previously attempted to stop smoking and 87.4% had used nicotine replacement therapy during the previous attempt to stop smoking. The overall smoking cessation rate was 49.5%. There was a strong association between the duration of varenicline treatment and smoking cessation. Patients who reported using varenicline for 9-12 weeks were 11 times more likely to stop smoking than those who completed less than two weeks of treatment. There was some evidence that patients with a longer history of smoking were less likely to stop. No association was observed between smoking cessation and: previous number of cigarettes smoked per day; number of previous attempts to stop smoking; or motivations for stopping. CONCLUSIONS: Varenicline appeared to be a useful pharmacological aid to smoking cessation in a general practice setting. The observed effectiveness was similar to the efficacy estimates from previously reported clinical trials. However, the response rate was lower than expected and responders tended to be older, more likely to suffer from chronic obstructive pulmonary disease and to live in more affluent areas than non-responders. Responses were self-reported and not clinically validated therefore recall bias may be an issue.

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.

AIM: To determine current treatment patterns for infants with cow milk allergy (CMA) and the associated resource implications and budget impact, from the perspective of the UK's National Health Service (NHS). METHODS: A computer-based model was constructed depicting current management of newly-diagnosed infants with CMA derived from patients suffering from this allergy in The Health Improvement Network (THIN) Database. The model spanned a period of 12 months following initial presentation to a general practitioner (GP) and was used to estimate the 12-monthly healthcare cost (at 2006/07 prices) of treating an annual cohort of 18,350 infants from when they initially present to their GP. RESULTS: Patients presenting with a combination of gastrointestinal and atopic symptoms accounted for 59% of all patients. From the initial GP visit for CMA it took a mean 2.2 months to be put on diet, although treatment varied according to presenting symptoms. A total of 60% of all infants were initially treated with soy, 18% with an extensively hydrolysed formula and 3% with an amino acid formula. A mean 9% of patients remained symptomatic on soy and 29% on an extensively hydrolysed formula. The total cost of managing CMA over the first 12 months following initial presentation to a GP was estimated to be £1,381 per patient and £25.6 million for an annual cohort of 18,350 infants. LIMITATIONS: Patients were not randomised to treatment and resource use was not collected prospectively. Nevertheless, 1,000 eligible patients have been included in the analysis, which should be a sufficiently large sample to accurately assess treatment patterns and healthcare resource use in actual clinical practice. The diagnosis of CMA may not be secure in all cases. Nevertheless, patients were diagnosed as having CMA by a clinician and have been managed by their GP as if they had CMA. CONCLUSION: CMA imposes a substantial burden on the NHS. Any strategy that improves healthcare delivery and thereby shortens time to treatment, time to diagnosis and time to symptom resolution should potentially decrease the burden this allergy imposes on the health service and release resources for alternative use.

BACKGROUND: A substantial fraction of UK general practitioners' salaries is now intended to reflect the quality of care provided. This performance-related pay system has probably improved aspects of primary health care but, using the observational data available, disentangling the impacts of different types of targets set within this unique payment system is challenging. DISCUSSION: Financial incentives undoubtedly influence GPs' activities, however, those aimed at encouraging GPs' delivery of health promotion counselling may not always have the effects intended. There is strong, observational evidence that targets and incentives intended to increase smoking cessation counselling by GPs have merely increased their propensity to record this activity in patients' medical records. The limitations of using financial incentives to stimulate the delivery of counselling in primary care are discussed and a re-appraisal of their use within UK GPs' performance-related pay system is argued for. SUMMARY: The utility of targets employed by the system for UK General Practitioners' performance related pay may be inappropriate for encouraging the delivery of health promotion counselling interventions. An evaluation of these targets is essential before they are further developed or added to.

OBJECTIVE: To estimate the prevalence and incidence of depression; investigate its association with risk factors including comorbidities and drug and health care use; and describe treatment patterns of depression in primary care using The Health Improvement Network database. METHOD: In this cohort study, subjects with a first recorded diagnosis of depression (Read code) between January 1, 2002, and December 31, 2004 (n=47,170) were identified from a source population of 1,287,829 subjects aged 10-79 years. A comparison group was sampled from the same population and frequency matched to the depression cohort by age, sex, and calendar year (n=50,000). Depression diagnoses were validated using physician-completed questionnaires. Odds ratios and 95% CIs for the relationship of depression with a range of factors were estimated using unconditional logistic regression in a nested case-control analysis. RESULTS: The prevalence of depression was 11.23% (95% CI, 11.18-11.28). This prevalence decreased with increasing age and was higher in women than in men. The incidence was 13.89 per 1,000 person-years (95% CI, 13.82-14.08). Depression was associated with frequent use of health services, smoking, pregnancy in the previous year, anxiety, stress, sleep disorders, digestive and respiratory disorders, and pain. In the trimester following diagnosis, 82% of cases were treated-98% with antidepressants and 81.5% with selective serotonin reuptake inhibitors (SSRIs). CONCLUSIONS: We found a high prevalence and incidence of depression diagnoses in primary care in the United Kingdom. Following diagnosis, the majority of individuals were prescribed SSRIs. A diagnosis of depression is associated with a number of prior comorbidities, which could mask the depression. This fact should be taken into account when screening individuals in primary care.

BACKGROUND: Anxiety disorders are common and can cause substantial quality of life impairment. OBJECTIVE: The aim of this study was to investigate the frequency of anxiety in UK primary care. Treatment patterns and factors associated with an anxiety diagnosis were also assessed. METHODS: The Health Improvement Network was used to identify all patients aged 10-79 years with a new diagnosis of anxiety in 2002-04 (n = 40 873) and age-, sex- and calendar-year-matched controls (n = 50 000). A nested case-control analysis was used to quantify potential risk factors for anxiety by multivariate logistic regression. RESULTS: The prevalence of anxiety was 7.2% and the incidence was 9.7 per 1000 person-years. Incidence and prevalence were highest in women and young adults (20-29 years). Anxiety was associated with heavy alcohol use, smoking and addiction problems as well as stress, sleep and depression disorders. Anxiety patients used health care services more frequently than controls. Among patients diagnosed with anxiety, 63% were treated pharmacologically. Antidepressants accounted for almost 80% of prescriptions. CONCLUSIONS: The prevalence and incidence of anxiety are high in UK primary care and are almost twice as high in women than in men. Anxiety is associated with other psychiatric morbidity as well as frequent health care use. Antidepressants are the most commonly used pharmacological treatment.

BACKGROUND: It is unknown whether macrogol 3350 (Movicol) affords the UK's National Health Service (NHS) a cost-effective addition to the current range of laxatives. AIM: To estimate the cost-effectiveness of macrogol 3350 compared with lactulose in the treatment of chronic constipation, from the perspective of the UK's NHS. METHODS: A decision model depicting the management of chronic constipation was constructed using clinical outcomes and resource use values derived from patients suffering from chronic constipation in The Health Independent Network (THIN) database. The model was used to estimate the cost-effectiveness of a GP prescribing macrogol 3350 instead of lactulose to treat adults > or =18 years of age suffering from chronic constipation. RESULTS: Sixty-eight percent of patients given macrogol 3350 were successfully treated within 6 months after starting treatment compared to 60% of patients given lactulose.Patients' health status at 6 months was estimated to be 0.458 and 0.454 quality-adjusted life years (QALYs) in the macrogol 3350 and lactulose groups respectively. The total 6-monthly NHS cost of initially treating patients with macrogol 3350 or lactulose was estimated to be pound420 (US $688) and pound419 (US $686) respectively. Hence, the cost per QALY gained with macrogol 3350 was estimated to be pound250 (US $410). CONCLUSION: Macrogol 3350 affords the NHS a cost-effective addition to the range of laxatives available for this potentially resource-intensive condition.

McCandless, Gustafson and Austin (2009) describe a Bayesian approach to regression adjustment for the propensity score to reduce confounding. A unique property of the method is that the treatment and outcome models are combined via Bayes theorem. However, this estimation procedure can be problematic if the outcome model is misspecified. We observe feedback that can bias propensity score estimates. Building on new innovation in Bayesian computation, we propose a technique for cutting feedback in a Bayesian propensity analysis. We use the posterior distribution of the propensity scores as an input in the regression model for the outcome. The method is approximately Bayesian in the sense that it does not use the full likelihood for estimation. Nonetheless, it severs feedback between the treatment and outcome giving propensity score estimates that are free from bias but modeled with uncertainty. We illustrate the method in a matched cohort study investigating the effect of statins on primary stroke prevention.

BACKGROUND: There is no specific treatment against acute pancreatitis (AP). A protective effect by angiotensin II receptor blockers (ARB) on AP has been suggested experimentally, but clinical evidence is scarce. METHODS: We conducted a population-based case-control study using The Health Improvement Network in the United Kingdom, comprising about 167,000 hypertensive patients in the study period 1996-2005. In multivariate logistic regression analysis, odds ratios were calculated with 95% confidence intervals (CI). Adjustments included sex, age, calendar year, body mass index, tobacco smoking, alcohol, general practitioner visits per year, and various antihypertensive medications with regard to exposure to ARB, and risk of AP. RESULTS: Among 633,281 person-years at risk, 265 new cases of AP were identified. Current users of ARB had a 37% statistically non-significant reduced risk of developing AP as compared to non-users (OR 0.63, 95% CI 0.38-1.02). No clear association was found between use of other antihypertensive drugs and risk of AP. CONCLUSION: Our study adds some support to previous experimental findings. Use of ARB might be associated with a reduced risk of AP. More research is needed to elucidate the potential role of ARB in the development of AP in the clinical setting.

BACKGROUND: There is a paucity of data describing how general practitioners (GPs) label or record depression. AIMS: To determine incidence and sociodemographic variation in GP-recorded depression diagnoses and depressive symptoms. METHOD: Annual incidence rates calculated using data from 298 UK general practices between 1996 and 2006, adjusted for year of diagnosis, gender, age and deprivation. RESULTS: Incidence of diagnosed depression fell from 22.5 to 14.0 per 1000 person-years at risk (PYAR) from 1996 to 2006. The incidence of depressive symptoms rose threefold from 5.1 to 15.5 per 1000 PYAR. Combined incidence of diagnoses and symptoms remained stable. Diagnosed depression and symptoms were more common in women and in more deprived areas. CONCLUSIONS: Depression recorded by general practitioners has lower incidence rates than depression recorded in epidemiological studies, although there are similar associations with gender and deprivation. General practitioners increasingly use symptoms rather than diagnostic labels to categorize people's illnesses. Studies using standardised diagnostic instruments may not be easily comparable with clinical practice.