Publications
scientifiques

PURPOSE: To determine trends and predictors of clopidogrel and proton pump inhibitor (PPI) coprescription with low-dose acetylsalicylic acid (ASA) prescribed for secondary cardiovascular or cerebrovascular disease (CVD) prevention in UK primary care. METHODS: Patients aged 50-84 years who received a first prescription for low-dose ASA for secondary CVD prevention in 2000-2001 (n = 10,330) or 2006-2007 (n = 8154) were identified in The Health Improvement Network UK primary care database. Clopidogrel or PPI coprescriptions received within 15 days after the first low-dose ASA prescription were ascertained. RESULTS: Clopidogrel coprescription with low-dose ASA increased from 1.6% to 25.2% between the two study periods; PPI coprescription increased from 11.6% to 28.3%. Low-dose ASA indications of myocardial infarction [odds ratio (OR) 11.7, 95% confidence interval (CI) 10.2 to 13.4] and unstable angina (OR 1.73, 95%CI 1.09 to 2.75) were positive predictors of clopidogrel coprescription in 2006-2007, relative to chronic ischaemic heart disease. Patients at high risk of upper gastrointestinal bleeding were more likely to receive a PPI than those at lower risk in 2006-2007 (OR 4.36, 95%CI 3.93 to 4.84). In this period, 65.5% of patients who required a clopidogrel coprescription according to guideline recommendations received one, and 44.3% of patients at high risk of upper gastrointestinal bleeding received a PPI. CONCLUSION: Clopidogrel and PPI coprescription with low-dose ASA increased markedly between 2000-2001 and 2006-2007; however, many patients on low-dose ASA did not receive the recommended coprescriptions at the end of the study period.

PURPOSE: To describe the use of 3 prostaglandin/timolol fixed combinations (FCs) in UK primary care, to summarize characteristics of recipients, and to assess 12-month persistence. METHODS: This retrospective cohort study included first-time recipients of latanoprost/timolol FC, bimatoprost/timolol FC, or travoprost/timolol FC treated between April 1, 2007, and November 30, 2008, identified in The Health Improvement Network database, a large database of anonymized longitudinal electronic medical records of patients treated in UK primary care. Eligible patients were = 18 years old at the index date (date of first prescription). Persistence, defined as a gap =60 days between consecutive prescriptions, was assessed through 12 months post-index for each cohort (Cox proportional hazards models). RESULTS: A total of 2,015 patients were included: latanoprost/timolol FC, n = 898 (44.6%); bimatoprost/timolol FC, n = 733 (36.4%); travoprost/timolol FC, n = 384 (19.1%). The mean age was approximately 72 years across cohorts (p = 0.792). Glaucoma was the diagnosis for >90% of patients in each cohort. Twelve-month persistence was similar across treatments: latanoprost/timolol FC: 38.2%; bimatoprost/timolol FC: 38.6%; travoprost/timolol FC: 38.3% (p = 0.985). Mean time to therapy change for nonpersistent patients was also similar: 143.3 ± 89.8, 151.0 ± 87.9, and 151.8 ± 87.7 days, respectively (p = 0.095). Among persistent patients, additional therapy was prescribed for 36.2%, 41.7%, and 41.5% of patients, respectively. Among nonpersistent patients, 64.0%, 70.4%, and 69.2%, respectively, restarted the index therapy. CONCLUSIONS: The largest proportion of first-time recipients of prostaglandin/beta-blocker FC products treated in UK primary care was prescribed latanoprost/timolol FC. Twelve-month persistence was similar (<40%) across the 3 FCs evaluated.

OBJECTIVE: The incidence and the risk of suicidal behaviors and severe neuropsychiatric disorders in people treated with systemic glucocorticoids are poorly known. The authors assessed the incidence rates of depression, mania, delirium, panic disorder, and suicidal behaviors in patients treated with glucocorticoids in primary care settings and the risk factors for developing these outcomes. METHOD: Data were obtained for all adult patients registered between 1990 and 2008 at U.K. general practices contributing to The Health Improvement Network (THIN) primary care database. The incidence rates for the outcomes of interest were assessed in patients who received prescriptions for oral glucocorticoids and compared with those in patients who did not receive such prescriptions. The predictors of these outcomes in exposed patients were ascertained using Cox proportional hazards models. RESULTS: Overall, 786,868 courses of oral glucocorticoids were prescribed for 372,696 patients. The authors identified 109 incident cases of suicide or suicide attempt and 10,220 incident cases of severe neuropsychiatric disorders in these patients. The incidence of any of these outcomes was 22.2 per 100 person-years at risk for first-course treatments. Compared to people with the same underlying medical disease who were not treated with glucocorticoids, the hazard ratio for suicide or suicide attempt in exposed patients was 6.89 (95% CI=4.52–10.50); for depression, 1.83 (95% CI=1.72–1.94); for mania, 4.35 (95% CI=3.67–5.16); for delirium, confusion, or disorientation, 5.14 (95% CI=4.54–5.82); and for panic disorder, 1.45 (95% CI=1.15–1.85). Older men were at higher risk of delirium/confusion/disorientation and mania, while younger patients were at higher risk of suicide or suicide attempt. Patients with a previous history of neuropsychiatric disorders and those treated with higher dosages of glucocorticoids were at greater risk of neuropsychiatric outcomes. CONCLUSIONS: Glucocorticoids increase the risk of suicidal behavior and neuropsychiatric disorders. Educating patients and their families about these adverse events and increasing primary care physicians' awareness about their occurrence should facilitate early monitoring.

BACKGROUND: Care home residents are vulnerable to the adverse effects of prescribing but there is limited monitoring in the UK. AIM: To compare prescribing quality in care homes in England and Wales with the community and with US nursing homes. DESIGN AND SETTING: Cross-sectional analysis of a UK primary care database and comparison with the US National Nursing Home Survey including 326 general practices in 2008-2009 in England and Wales, with 10 387 care home and 403 259 community residents aged 65 to 104 years. METHOD: Comparison of age- and sex-standardised use of 'concern' and common drug groups in the last 90 days and potentially inappropriate prescribing based on a consensus list of medications best avoided in older people (Beers criteria). RESULTS: Compared to the community, care home residents were more likely to receive 'concern' drugs, including benzodiazepines (relative risk (RR) = 2.05, 95% confidence interval (CI) = 1.90 to 2.22), anticholinergic antihistamines (RR = 2.78, 95% CI = 2.38 to 3.23), loop diuretics (RR = 1.47, 95% CI = 1.41 to 1.53), and antipsychotics (RR = 22.7, 95% CI = 20.6 to 24.9). Use of several common drug groups, including laxatives, antidepressants, and antibiotics, was higher, but use of cardiovascular medication was lower. Thirty-three per cent (95% CI = 31.7% to 34.3%) of care home residents in England and Wales received potentially inappropriate medication, compared to 21.4% (95% CI = 20.9% to 21.8%) in the community. The potentially inappropriate prescribing rate in US nursing homes was similar to England and Wales. CONCLUSION: Care home prescribing has the potential for improvement. High use of anticholinergic and psychotropic medication may contribute to functional and cognitive decline. The targeting and effectiveness of medication reviews in care homes needs to be improved.

BACKGROUND: Several studies have demonstrated that the use of single tablet regimens (STRs) in hypertension is associated with improved outcomes and reduced healthcare costs compared with individual component regimens. The objective was to carry out a retrospective analysis of a UK general practice population to test these conclusions in a UK context. METHOD: A retrospective cohort study was carried out using a primary care database (The Health Improvement Network; THIN), comparing 9929 hypertensive patients on STRs with 18,665 patients on individual component therapy. Data were collected for prescriptions, significant cardiovascular events, and out-patient referrals over a minimum follow-up period of 5 years after initiation of therapy. Current NHS costings were applied to the data, to arrive at an estimate of comparative resource use. RESULTS: There were significantly more cardiovascular events in the individual component group than those treated with a single tablet regimen. Five year event rates: 8.3% vs 13.6%; Absolute Risk Reduction (ARR) =5.3%; Number needed to treat (NNT) =18.9. After correction for potential confounders, the hazard ratio was 0.74 (95%CI=0.70-0.77), p<0.0001. Hospital admission costs were lower in the STR group, but drug costs were higher. Overall, the mean annual management cost per patient was similar in the two groups (£191.49 vs £189.35). KEY LIMITATIONS: The study was based on a retrospective cohort and the result may therefore be influenced by unidentified confounders. It was not possible to identify the reasons for individual prescriptions, some of which may have been issued for reasons other than hypertension. Costings for some components of the outcome could not be assessed from the dataset and are therefore omitted from the analysis. Finally, no attempt was made to distinguish outcomes associated with individual classes of anti-hypertensives. CONCLUSIONS: This study confirms the association observed by other authors that patients treated with STRs are less likely to experience serious cardiovascular events than those on individual component therapy. In a UK context this analysis has shown that potential hospital savings broadly offset the additional drug acquisition costs associated with STRs. These agents can therefore be considered cost neutral.

OBJECTIVE: To compare the relative hazard of muscle toxicity, renal dysfunction, and hepatic dysfunction associated with the drug interaction between statins and concomitant medications that inhibit the CYP3A4 isoenzyme. BACKGROUND: Although statins provide important clinical benefits related to mitigating the risk of cardiovascular events, this class of medications also has the potential for severe adverse reactions. The risk for adverse events may be potentiated by concomitant use of medications that interfere with statin metabolism. METHODS: Data from The Health Improvement Network (THIN) from 1990 to 2008 were used to conduct a retrospective cohort study. Cohorts were created to evaluate each outcome (muscle toxicity, renal dysfunction, and hepatic dysfunction) independently. Each cohort included new statin initiators and compared the relative hazard of the outcome. The interaction ratio (I*R) was the primary contrast of interest. The I*R represents the relative effect of each statin type (statin 3A4 substrate vs. statin non-3A4 substrate) with a CYP3A4 inhibitor, independent of the effect of the statin type without a CYP3A4 inhibitor. We adjusted for confounding variables using the multinomial propensity score. RESULTS: The median follow-up time per cohort was 1.5 years. There were 7889 muscle toxicity events among 362,809 patients and 792,665 person-years. The adjusted muscle toxicity I*R was 1.22 (95% confidence interval [CI] = 0.90-1.66). There were 1449 renal dysfunction events among 272,099 patients and 574,584 person-years. The adjusted renal dysfunction I*R was 0.91 (95%CI = 0.58-1.44). There were 1434 hepatic dysfunction events among 367,612 patients and 815,945 person-years. The adjusted hepatic dysfunction I*R was 0.78 (95%CI = 0.45-1.31). CONCLUSIONS: Overall, this study found no difference in the relative hazard of muscle toxicity, renal dysfunction, or hepatic dysfunction for patients prescribed a statin 3A4 substrate versus a statin non-3A4 substrate with CYP3A4 inhibitor concomitancy.

PURPOSE: Electronic healthcare databases are commonly used in comparative effectiveness and safety research of therapeutics. Many databases now include additional confounder information in a subset of the study population through data linkage or data collection. We described and compared existing methods for analyzing such datasets. METHODS: Using data from The Health Improvement Network and the relation between non-steroidal anti-inflammatory drugs and upper gastrointestinal bleeding as an example, we employed several methods to handle partially missing confounder information. RESULTS: The crude odds ratio (OR) of upper gastrointestinal bleeding was 1.50 (95% confidence interval: 0.98, 2.28) among selective cyclo-oxygenase-2 inhibitor initiators (n = 43 569) compared with traditional non-steroidal anti-inflammatory drug initiators (n = 411 616). The OR dropped to 0.81 (0.52, 1.27) upon adjustment for confounders recorded for all patients. When further considering three additional variables missing in 22% of the study population (smoking, alcohol consumption, body mass index), the OR was between 0.80 and 0.83 for the missing-category approach, the missing-indicator approach, single imputation by the most common category, multiple imputation by chained equations, and propensity score calibration. The OR was 0.65 (0.39, 1.09) and 0.67 (0.38, 1.16) for the unweighted and the inverse probability weighted complete-case analysis, respectively. CONCLUSIONS: Existing methods for handling partially missing confounder data require different assumptions and may produce different results. The unweighted complete-case analysis, the missing-category/indicator approach, and single imputation require often unrealistic assumptions and should be avoided. In this study, differences across methods were not substantial, likely due to relatively low proportion of missingness and weak confounding effect by the three additional variables upon adjustment for other variables.

OBJECTIVE: To investigate a possible increased risk observed in tiotropium clinical trials of stroke and other adverse events. DESIGN: New users of long-acting anticholinergic therapy (tiotropium HandiHaler®) were compared with new users of long-acting β-agonist (LABA) monotherapy, and propensity scores were used to control confounding. SETTING: UK healthcare system general practitioner electronic medical record database. PARTICIPANTS: 10 840 patients newly prescribed tiotropium (n=4767) or LABA (n=6073), at least 40 years old, and not having asthma as their only respiratory illness. PRIMARY AND SECONDARY OUTCOME MEASURES: Incidence rates of total stroke, myocardial infarction, angina and other adverse events. RESULTS: Tiotropium was associated with increased rates of stroke (HR=1.49, 95% CI 0.91 to 2.45), angina (HR=1.38, 95% CI 0.88 to 2.16) and myocardial infarction (HR=1.26, 95% CI 0.72 to 2.21). Groups had similar rates of chronic obstructive pulmonary disease exacerbation (HR=0.95, 95% CI 0.80 to 1.12) and pneumonia (HR=0.96, 95% CI 0.58 to 1.58). Tiotropium was associated with a lower rate of total mortality (HR=0.70, 95% CI 0.56 to 0.89) and asthma exacerbations (HR=0.46, 95% CI 0.36 to 0.57) than users of LABA. CONCLUSION: Small increased risks of serious ischaemic cardiovascular events have been reported with inhaled anticholinergic medication from randomised and nonrandomized studies of ipratropium, tiotropium HandiHaler® and tiotropium Respimat®. Additional research is needed to understand the full extent of cardiovascular effects of inhaled anticholinergic medications and the patients who may be most susceptible.

OBJECTIVE: To describe the validity of recorded diabetic retinopathy (DR) and diabetic maculopathy (DMP) diagnoses, including edema (DMO) in The Health Improvement Network (THIN) database. RESEARCH DESIGN AND METHODS: In two independent computer searches, we detected 20,838 patients with diabetes aged 1-84 years with a first DR computer Read entry in 2000-2008 and 4,064 with a first DMP entry. A two-step strategy was used to validate both outcomes as follows: 1) review of patient profiles including free-text comments from primary care practitioners (PCPs) (containing referral information and test results) of a random sample of 500 DR and all DMP computer-detected patients. We classified them in probable, possible, and noncase according to the diagnosis plausibility based on the manual review of the computerized information; and 2) review of questionnaires sent by PCPs and medical records in a random sample (N = 200 for each outcome including 36 diabetic macular edema [DMO]). Gold standard was PCPs' confirmation. RESULTS: After profiles review, we categorized 418 as probable/possible DR. In addition, 3,676 DMP were categorized as probable/possible (including 711 DMO). After review of information sent by PCPs, confirmation rates were 87.3 and 87.2%, respectively (90.3% for DMO). When we applied them to the whole sample of computer-detected patients, the weighted confirmation rate was 78.0% for DR and 78.8% for DMP (86.2% for DMO). CONCLUSIONS: Read codes for DR, DM, and DMO are moderately accurate in identifying incident case subjects of these ophthalmologic complications. The validity improved when incorporating PCPs' text comments to the patient's profile. THIN database proved to be a valuable resource to study ophthalmological diabetes complications.

Hepatitis C virus (HCV) infection is associated with systemic inflammation and metabolic complications that might predispose patients to atherosclerosis. However, it remains unclear if HCV infection increases the risk of acute myocardial infarction (MI). To determine whether HCV infection is an independent risk factor for acute MI among adults followed in general practices in the United Kingdom (UK), a retrospective cohort study was conducted in The Health Improvement Network, from 1996 through 2008. Patients ≥18 years of age with at least 6 months of follow-up and without a prior history of MI were eligible for study inclusion. HCV-infected individuals, identified with previously validated HCV diagnostic codes (n = 4809), were matched on age, sex and practice with up to 15 randomly selected patients without HCV (n = 71 668). Rates of incident MI among patients with and without a diagnosis of HCV infection were calculated. Adjusted hazard ratios were estimated using Cox proportional hazards regression, controlling for established cardiovascular risk factors. During a median follow-up of 3.2 years, there was no difference in the incidence rates of MI between HCV-infected and -uninfected patients (1.02 vs 0.92 events per 1000 person-years; P = 0.7). HCV infection was not associated with an increased risk of incident MI (adjusted HR, 1.10; 95% confidence interval [CI], 0.67-1.83). Sensitivity analyses including the exploration of a composite outcome of acute MI and coronary interventions yielded similar results (adjusted HR, 1.16; 95% CI, 0.77-1.74). In conclusion, HCV infection was not associated with an increased risk of incident MI.

BACKGROUND: Chronic kidney disease is a major health concern that, if left untreated, may progress to end-stage kidney failure (ESKF). Identifying individuals at an increased risk of kidney disease and who might benefit from a therapeutic or preventive intervention is an important challenge. AIM: To evaluate the performance of the QKidney® scores for predicting 5-year risk of developing moderate-severe kidney disease and ESKF in an independent UK cohort of patients from general practice records. DESIGN AND SETTING: Prospective cohort study to evaluate the performance of two risk scores for kidney disease in 364 practices from the UK, contributing to The Health Improvement Network (THIN) database. METHOD: Data were obtained from 1.6 million patients registered with a general practice surgery between 1 January 2002 and 1 July 2008, aged 35-74 years, with 43,186 incident cases of moderate-severe kidney disease and 2663 incident cases of ESKF. This is the first recorded evidence of moderate-severe chronic kidney and ESKF as recorded in general practice records. RESULTS: The results from this independent and external validation of QKidney scores indicate that both scores showed good performance data for both moderate-severe kidney disease and ESKF, on a large cohort of general practice patients. Discrimination and calibration statistics were better for models including serum creatinine; however, there were considerable amounts of missing data for serum creatinine. QKidney scores both with and without serum creatinine were well calibrated. CONCLUSION: QKidney scores have been shown to be useful tools to predict the 5-year risk of moderate-severe kidney disease and ESKF in the UK.

OBJECTIVES: This study aims to assess whether digoxin has a different effect on mortality risk for women than it does for men in patients with heart failure (HF). DESIGN: This study uses the UK-based The Health Information Network population database in a cohort study of the impact of digoxin exposure on mortality for men and women who carry the diagnosis of HF. Digoxin exposure was assessed based on prescribing data. Multivariable Cox proportional hazards models were used to assess whether there was an interaction between sex and digoxin affecting mortality hazard. SETTING: The setting was primary care outpatient practices. PARTICIPANTS: The study cohort consisted of 17 707 men and 19 227 women with the diagnosis of HF who contributed only time without digoxin exposure and 9487 men and 10 808 women with the diagnosis of HF who contributed time with digoxin exposure. MAIN OUTCOME MEASURES: The main outcome measure was all-cause mortality. RESULTS: The primary outcome of this study was the absence of a large interaction between digoxin use and sex affecting mortality. For men, digoxin use was associated with a HR for mortality of 1.00, while for women, the HR was also 1.00 (p value for interaction 0.65). The results of sensitivity analyses were consistent with those of the primary analysis. CONCLUSION: Observational data do not support the concern that there is a substantial increased risk of mortality due to the use of digoxin in women. This finding is consistent with previous observational studies but discordant with results from a post hoc analysis of a randomised controlled trial of digoxin versus placebo.

BACKGROUND: Injuries in childhood are largely preventable yet an estimated 2,400 children die every day because of injury and violence. Despite this, the factors that contribute to injury occurrence have not been quantified at the population scale using primary care data. We used The Health Improvement Network (THIN) database to identify risk factors for thermal injury, fractures and poisoning in pre-school children in order to inform the optimal delivery of preventative strategies. METHODS: We used a matched, nested case-control study design. Cases were children under 5 with a first medically recorded injury, comprising 3,649 thermal injury cases, 4,050 fracture cases and 2,193 poisoning cases, matched on general practice to 94,620 control children. RESULTS: Younger maternal age and higher birth order increased the odds of all injuries. Children's age of highest injury risk varied by injury type; compared with children under 1 year, thermal injuries were highest in those age 1-2 (OR = 2.43, 95%CI 2.23-2.65), poisonings in those age 2-3 (OR = 7.32, 95%CI 6.26-8.58) and fractures in those age 3-5 (OR = N 3.80, 95%CI 3.42-4.23). Increasing deprivation was an important modifiable risk factor for poisonings and thermal injuries (tests for trend p ≤ 0.001) as were hazardous/harmful alcohol consumption by a household adult (OR = 1.73, 95%CI 1.26-2.38 and OR = 1.39, 95%CI 1.07-1.81 respectively) and maternal diagnosis of depression (OR = 1.45, 95%CI 1.24-1.70 and OR = 1.16, 95%CI 1.02-1.32 respectively). Fracture was not associated with these factors, however, not living in single-adult household reduced the odds of fracture (OR = 0.88, 95%CI 0.82-0.95). CONCLUSIONS: Maternal depression, hazardous/harmful adult alcohol consumption and socioeconomic deprivation represent important modifiable risk factors for thermal injury and poisoning but not fractures in preschool children. Since these risk factors can be ascertained from routine primary care records, pre-school children's frequent visits to primary care present an opportunity to reduce injury risk by implementing effective preventative interventions from existing national guidelines.

BACKGROUND: The aetiology of meningiomas is largely unknown although hormones have been suggested to play a role. METHODS: A cohort study was performed to evaluate hormone-related factors associated with meningioma. Patients (12-89 years) with a first diagnosis of meningioma (January 1996-June 2008) were identified from The Health Improvement Network UK primary care database and age- and sex-matched to controls (n=10000) from the same cohort. Odds ratios (ORs) were calculated following a nested case control analysis using unconditional logistic regression. RESULTS: In total, 745 patients with meningioma were identified from a study population of 2171287. No significantly increased risk of meningioma was found among female users of oral contraceptives (OR: 1.15; CI: 0.67-1.98), hormone replacement therapy (OR: 0.99; CI: 0.73-1.35) or low-dose cyproterone acetate (CPA; OR: 1.51; CI: 0.33-6.86) compared with non-users. There was a significantly increased risk of meningioma among male users of androgen analogues (OR: 19.09; CI: 2.81-129.74) and among users of high-dose CPA (OR: 6.30; CI: 1.37-28.94) compared with non-users, however there were only three cases currently using these drugs. No significant association was found between meningioma and prostate, breast, or genital cancers. CONCLUSIONS: Our results do not support a role for exogenous hormone use by females in meningioma development. The risk in males was only observed with high-dose, short-term (<1 year) therapy. IMPACT: While hormonal cancers and therapies are not associated with meningioma in females, the risk in males requires further investigation.