Misclassification and selection bias in case-control studies using an automated database

Automated databases are increasingly being used for pharmacoepidemiological research and fieldwork is often carried out to supplement and validate information held within them. In MEMO's case-control studies, patients are identified using computerized ICD9 diagnosis codes, the original medical records are retrieved and checked, and only patients fulfilling case inclusion criteria are used. The 20-30% of patients for whom medical records cannot be found are usually excluded. The aim is to eliminate misclassification bias. However, selection bias may be introduced if availability of medical records is associated with exposure. This investigation was therefore carried out to assess the relative importance of misclassification and selection bias. Data from four previous case-control studies, investigating the associations between NSAIDs and hospitalization for colitis, acute renal failure, appendicitis and colorectal cancer, were used. To assess misclassification, odds ratios (with 95% CI) for recent exposure to NSAIDs were compared in repeated studies that used all patients identified by ICD9 codes, with studies using validated cases only. Selection bias was assessed by comparing results in studies using patients for whom records could and could not be found. Results were plotted and the graphs indicated that misclassification bias was relatively unimportant, but that selection bias could be introduced into a study in this way.

Presence of bacteriuria caused by trimethoprim resistant bacteria in patients prescribed antibiotics: multilevel model with practice and individual patient data

OBJECTIVE

To look for evidence of a relation between antibiotic resistance and prescribing by general practitioners by analysis of prescribing at both practice and individual patient level.

DESIGN

Repeated cross-sectional study in 1995 and 1996.

SETTING

28 general practices in the Ninewells Hospital laboratory catchment area, Tayside, Scotland. SUBJECTS REVIEWED: 8833 patients registered with the 28 practices who submitted urine samples for analysis.

MAIN OUTCOME MEASURES

Resistance to trimethoprim in bacteria isolated from urine samples at practice and individual level simultaneously in a multilevel model.

RESULTS

Practices showed considerable variation in both the prevalence of trimethoprim resistance (26-50% of bacteria isolated) and trimethoprim prescribing (67-357 prescriptions per 100 practice patients). Although variation in prescribing showed no association with resistance at the practice level after adjustment for other factors (P = 0.101), in the multilevel model resistance to trimethoprim was significantly associated with age, sex, and individual-level exposure to trimethoprim (P < 0.001) or to other antibiotics (P = 0.002). The association with trimethoprim resistance was strongest for people recently exposed to trimethoprim, and there was no association for people with trimethoprim exposure more than six months before the date of the urine sample.

DISCUSSION

Analysis of practice level data obscured important associations between antibiotic prescribing and resistance. The results support efforts to reduce unnecessary prescribing of antibiotics in the community and show the added value of individual patient data for research on the outcomes of prescribing.

Cardiovascular prophylaxis with aspirin: costs of supply and management of upper gastrointestinal and renal toxicity

AIMS

To determine the cost to the NHS of prescribed low-dose aspirin.

METHODS

This was a population based observational cohort study. Patients from Tayside Scotland (17 244 new users of dispensed aspirin each with 10 matched comparators) were included. A pragmatic analysis totalled costs from the start to end of the study and compared these with a matched cohort of aspirin nonusers to estimate excess costs. Fastidious analyses were done of subjects with no prior history of upper gastrointestinal (UGI) or renal disease where the cost that occurred during aspirin exposure, the 30 days following aspirin exposure and subsequent nonexposure was calculated adjusting for risk factors in each period.

RESULTS

Subjects took aspirin for only 1.18 of the 2.53 years follow-up (47% compliance). Aspirin use cost an additional 49.86 UK pounds per year (pragmatic analysis) made up of 1.96 UK pounds for aspirin tablets (4%), 5.49 UK pounds for dispensing costs (11%), 24.60 UK pounds for UGI complications (49%) and 17.81 UK pounds for renal complications (36%). The costs for managing complications were substantially lower in the fastidious analysis (2.66 UK pounds for UGI complications and 2.92 UK pounds for renal complications). Assuming that the antiplatelet trial meta-analysis is an accurate assessment of the benefits of aspirin, the costs of preventing one vascular event lay between 62 500 UK pounds (primary prevention, pragmatic analysis) and 867 UK pounds (secondary prevention, fastidious analysis). These costs may be underestimates due to the low compliance observed.

CONCLUSIONS

Compliance with aspirin was poor. Serious adverse events were uncommon but despite this aspirin cost the NHS between 6 and 25 times the cost of aspirin tablets due to dispensing costs and the cost of managing adverse effects.

Channelling of patients taking NSAIDs or cyclooxygenase-2-specific inhibitors and its effect on interpretation of outcomes

When new drugs with improved safety or efficacy are introduced, they may be preferentially prescribed to specific populations of patients. Safety and efficacy may be underestimated if such channelling effects are not recognized. Meloxicam and cyclooxygenase (COX)-2-specific inhibitors were developed as safer alternatives to non-steroidal anti-inflammatory drugs (NSAIDs) for the treatment of osteoarthritis and rheumatoid arthritis. Studies of the use of meloxicam and COX-2-specific inhibitors demonstrate that both of these drugs are being prescribed to patients at increased risk of gastrointestinal adverse drug events. In the case of COX-2-specific inhibitors, this channelling appears to represent a prescribing pattern consistent with current recommendations. Subsequent analysis of the data, after adjusting for channelling bias, showed that the risk of gastrointestinal toxicity for meloxicam was similar to that for other NSAIDs, while COX-2-specific inhibitors reduced the risk of developing gastrointestinal adverse drug events by approximately 60%. These studies serve as examples of observed channelling bias and highlight the need for adjusting for channelling in order to provide a valid assessment of relevant outcomes for drugs likely to be preferentially prescribed to specific populations.

Channelling bias and the incidence of gastrointestinal haemorrhage in users of meloxicam, coxibs, and older, non-specific non-steroidal anti-inflammatory drugs

BACKGROUND

Although clinical trial results suggest that meloxicam has less gastrointestinal toxicity than most other non-steroidal anti-inflammatory drugs (NSAIDs), in practice it has been associated with a large number of yellow card reports of gastrointestinal complications.

AIMS

To estimate whether meloxicam and the coxibs, rofecoxib and celecoxib, have been channelled towards high risk patients, and to estimate the risk of hospitalisation for gastrointestinal haemorrhage associated with the use of these drugs, allowing for the effects of channelling.

PATIENTS

Using the UK General Practice Research Database, this study included 7.1 thousand patient years (tpy) exposure to meloxicam, 1.6 tpy exposure to coxibs, and 628 tpy exposure to older non-specific NSAIDs.

METHODS

Cohort study of patients who received a prescription for an NSAID between June 1987 and January 2001. Exposure to newer NSAIDs (meloxicam, rofecoxib, celecoxib) and to older non-specific NSAIDs was identified. Channelling was assessed on factors including: demographic variables; diagnosis of arthritis; history of NSAID use or gastrointestinal events, including gastrointestinal haemorrhage; and use of ulcer healing drugs.

RESULTS

Most risk factors for gastrointestinal haemorrhage were more prevalent among patients prescribed the newer NSAIDs. Adjusting for these risk factors reduced the relative risks of gastrointestinal haemorrhage on meloxicam and coxibs versus older non-specific NSAIDs to 0.84 (95% confidence interval 0.60, 1.17) and 0.36 (0.14, 0.97), respectively.

CONCLUSIONS

Channelling towards high risk gastrointestinal patients occurred in the prescribing of newer NSAIDs. After attempting to correct for channelling bias, coxib exposure, but not meloxicam exposure, was associated with a significantly lower risk of gastrointestinal haemorrhage than older non-specific NSAID exposure.

Why not prescribe the best drugs for hypertension now?

As Westernised societies have become more affluent, the attitudes of the population have become more risk-aware. People are now intolerant of small risks as well as the physical or mental discomforts from drug side effects. Safety and tolerability are now major forces driving the development of new medicines for the treatment of chronic illnesses and the prevention of increasingly rare events. For example, over the past decades, lower and lower treatment thresholds have been recommended in hypertension. Public perception of risk strongly influences the acceptability of lifetime treatment, especially for mild hypertension. This era has also witnessed great advances in the development of antihypertensive drugs that combine efficacy with unsurpassed tolerability. However, the philosophy of Scottish teachers of Materia medica still appears to be followed-'never be the first or the last to prescribe a new drug'. Angiotensin-converting enzyme (ACE) inhibitors and angiotensin II receptor antagonists are as safe and as efficacious as other antihypertensive medications and better tolerated. Large trials (HOT, HOPE, UKPDS and PROGRESS) point to the need for rigorous control of blood pressure particularly in high-risk individuals. Antihypertensive drugs that act on the renin-angiotensin system will probably impact significantly on achieving optimal blood pressure levels. Should it not now be accepted that high-risk patients should have ACE inhibitors and angiotensin II receptor antagonists prescribed as first-line agents? We review the evidence for the use of ACE inhibitors and angiotensin II receptor antagonists as antihypertensive agents.

Effect of ibuprofen on cardioprotective effect of aspirin

Treatment with ibuprofen might limit the cardioprotective effects of aspirin. We aimed to assess whether patients with known cardiovascular disease who take low-dose aspirin and ibuprofen have increased risk of cardiovascular mortality. We studied 7107 patients who were discharged after first admission for cardiovascular disease between April, 1989, and April, 1997, and who were prescribed low-dose aspirin (<325 mg/day) and survived for at least 1 month. Compared with those who used aspirin alone, patients taking aspirin plus ibuprofen had an increased risk of all-cause mortality (adjusted hazard ratio 1.93, 95% CI 1.30-2.87, p=0.0011) and cardiovascular mortality (1.73, 1.05-2.84, p=0.0305). Our finding lends support to the hypothesis that ibuprofen may interact with the cardioprotective effects of aspirin, at least in patients with established cardiovascular disease.

A pharmacoeconomic comparison of misoprostol/diclofenac with diclofenac

OBJECTIVES

To estimate the economic impact of misoprostol/diclofenac in a fixed combination tablet compared with diclofenac.

DESIGN

Cohort study with a prospectively constructed, population-based, record-linkage database containing details of exposure to all community dispensed NSAIDs and all admissions to hospital for upper gastrointestinal (GI) diagnoses. Costs associated with each study drug exposure were analysed using generalized linear models.

SETTING

The population of Tayside, Scotland.

SUBJECTS

Subjects aged 20 years and over who received misoprostol/diclofenac or any other NSAID between January 1989 and 31 December 1995.

MAIN OUTCOME MEASURES

Total costs for the number of days of exposure to diclofenac and misoprostol/diclofenac, plus costs of concomitant ulcer healing drug therapy plus endoscopy procedures plus costs of admissions to hospital for upper GI diagnoses.

RESULTS

The rate of hospitalization with gastrointestinal events was 30% higher among patients receiving diclofenac than that for patients receiving misoprostol/diclofenac. Among patients who received diclofenac and an ulcer-healing drug (UHD), the event rate was more than twice that for patients receiving misoprostol/diclofenac. In patients with a prior GI history, switching from diclofenac to misoprostol/diclofenac would reduce the costs of hospitalization. The resulting savings would more than offset the extra prescription costs. In patients without a prior GI history, the greatest potential saving would arise due to reduced use of UHDs and net savings would occur in subjects aged between 60 and 70 years of age or more.

CONCLUSION

Use of misoprostol/diclofenac instead of diclofenac can produce cost savings due to reduced hospitalization rates and decreased use of UHDs in subjects with a prior history of GI disease and in older subjects without prior GI disease. These findings have implications for the management of patients who require treatment with NSAIDs.

Adherence to statin treatment and readmission of patients after myocardial infarction: a six year follow up study

OBJECTIVE

To investigate patients' adherence to statin treatment prescribed following their first myocardial infarction (MI) and to estimate the effect of adherence to statins on recurrence of MI and all cause mortality.

DESIGN

Cohort study using a record linkage database.

SETTING

Tayside, Scotland, UK.

PATIENTS

Patients who experienced their first MI between January 1990 and November 1995.

MAIN OUTCOME MEASURES

Percentage of statin use and adherence to statins by patients after an MI and the relative risk of hospitalisation for recurrent MI. The effect of adherence on all cause mortality was also examined. The covariates used were age, sex, socioeconomic deprivation, serum cholesterol concentration, diabetes mellitus, cardiovascular drug use, and other hospitalisations.

RESULTS

Of 5590 patients who experienced an incident MI, 717 (12.8%) experienced at least one further MI. Only 7.7% of patients used statins after an MI during the study period. Compared with those not taking statins, those who had 80% or better adherence to statin treatment had an adjusted relative risk of recurrent MI of 0.19 (95% confidence interval (CI) 0.08 to 0.47) and all cause mortality of 0.47 (95% CI 0.22 to 0.99). There was no significant reduction in either end point for those who were less than 80% adherent to statins.

CONCLUSIONS

Despite the infrequent use of statin during the study period, good adherence to statin treatment was associated with lower risk of recurrent MI.

Reversible hypertension following coeliac disease treatment: the role of moderate hyperhomocysteinaemia and vascular endothelial dysfunction

The vascular endothelium maintains a relatively vasodilated state via the release of nitric oxide (NO), a process that could be disrupted by hyperhomocysteinaemia. Since endothelial dysfunction is associated with increased systemic vascular resistance that is the hallmark of sustained arterial hypertension, we hypothesised that in patients with both hypertension and coeliac disease with hyperhomocysteinaemia (via malabsorption of essential cofactors), treatment of the latter disease could improve blood pressure (BP) control. A single patient with proven sustained hypertension and newly-diagnosed coeliac disease had baseline and post-treatment BP and endothelial function assessed by ambulatory BP monitoring (ABPM) and brachial artery forearm occlusion plethysmography respectively. This 49 year-old woman had uncomplicated sustained hypertension proven on repeated ABPM carried out 6 weeks apart (daytime mean 151/92 mm Hg and 155/95 mm Hg), and sub-clinical coeliac disease (gluten-sensitive enteropathy). Initial assessments revealed raised homocysteine levels with low normal vitamin B(12) level. It was likely that she had impaired absorption of essential cofactors for normal homocysteine metabolism. She adhered to a gluten-free diet and was give oral iron, folate and B(6) supplementations as well as B(12) injections for 3 months. Her BP had improved by 6 months and normalised by 15 months (daytime ABPM mean 128/80 mm Hg). There was parallel restoration of normal endothelial function with normalisation of her homocysteine levels. These observations suggest that sub-clinical coeliac disease related hyperhomocysteinaemia might cause endothelial dysfunction, potentially giving rise to a reversible form of hypertension. In addition, this case study supports the notion that irrespective of aetiology, endothelial dysfunction may be the precursor of hypertension. This highlights the need to resolve co-existing vascular risk factors in patients with hypertension.

Prognosis following first acute myocardial infarction in Type 2 diabetes: a comparative population study

AIMS

To estimate the incidence of death and macrovascular complications after a first myocardial infarction for patients with Type 2 diabetes.

RESEARCH DESIGN

In a retrospective, incidence cohort study in the Tayside Region of Scotland we studied all patients hospitalized with a diagnosis of first acute myocardial infarction from 1 April 1993 to 31 December 1994. The primary endpoint was time to death. Secondary endpoints were 2-year incidence of hospital admission for angina, myocardial infarction, stroke, heart failure, coronary angiography, coronary artery bypass graft (CABG) and percutaneous transluminal coronary angioplasty (PTCA).

RESULTS

The 147 patients with Type 2 diabetes had significantly worse survival with an increase in relative hazard of 67% compared with non-diabetic patients. After adjustment for age, sex, smoking status, prior heart failure, prior angina, delay to hospitalization, site of infarction, drug therapy with aspirin, beta-blockers, streptokinase and hyperlipidaemia and treated hypertension, Type 2 diabetes was still associated with a 40% higher death rate compared with people without diabetes (P < 0.05) There was no significant difference in death rates in those aged over 70 years, but an indication of a trend in younger individuals with a four-fold increase in death rate in those with diabetes aged < 60 years, compared with a rate ratio of 2.6 in those with diabetes aged 61-70 years.

CONCLUSIONS

Among hospitalized patients with first acute myocardial infarction, Type 2 diabetes mellitus is consistently associated with increased mortality and increased hospital admission for heart failure. The estimated 4-year survival rate is only 50%. Our results indicate that younger subjects with Type 2 diabetes and acute myocardial infarction are a high-risk group deserving of special study, and support the argument for aggressive targeting of coronary risk factors among patients with Type 2 diabetes.

Adherence to prescribed oral hypoglycaemic medication in a population of patients with Type 2 diabetes: a retrospective cohort study

AIMS

To evaluate the patterns and predictors of adherence in all patients with Type 2 diabetes in the community receiving treatment with a single oral hypoglycaemic drug. In particular, to test the hypothesis that one tablet per day is associated with better adherence than more than one.

METHODS

The study design was a retrospective cohort study set in the Tayside region of Scotland (population approx. 400 000). Participants were residents of Tayside from 1 January 1993 until 31 December 1995 with at least 12 months of prescriptions of oral hypoglycaemic drugs (OHDs). The main outcome measures were adherence indices for sulphonylureas and metformin separately, adjusting for prescribing while hospitalized.

RESULTS

Of the total 2920 subjects identified, adequate adherence (> or = 90%) was found in 31% of those prescribed sulphonylureas alone (n = 1329, median adherence = 300 days per year), and in 34% of those prescribed metformin alone (n = 528, median = 302 days per year). There were significant linear trends of poorer adherence with each increase in the daily number of tablets taken (p = 0.001) and increase in co-medication (p = 0.0001) for sulphonylureas alone after adjustment for other factors.

CONCLUSIONS

In the community only one in three with Type 2 diabetes had adequate adherence to OHDs. One tablet per day administration was associated with greater adherence than multiple tablets. Poor adherence is a major obstacle to the benefit of complex drug regimens in the treatment of Type 2 diabetes.

Is aldosterone the missing link in refractory hypertension?: aldosterone-to-renin ratio as a marker of inappropriate aldosterone activity

Use of the random aldosterone-to-renin ratio (ARR) as a reliable marker of inappropriate aldosterone activity has led to primary aldosteronism (PA) being increasingly diagnosed in hypertensive patients. At least 10% of hypertensives have been found to have PA, the majority of whom presumably have bilateral adrenal hyperplasia or idiopathic hyperaldosteronism as an aetiology for PA. Whilst these patients clearly have excess aldosterone activity, they have in common many features that are found in hypertensive patients in general, amongst which include heightened angiotensin II adrenal sensitivity. Whether these individuals belong within the spectrum of 'essential hypertension' is being debated, but is probably irrelevant clinically since they appear to respond favourably to spironolactone treatment. In addition, there is recent evidence suggesting that these patients overexpress a key enzyme involved in aldosterone production, the aldosterone synthase, the activity of which appears to relate to its genotypic variation.

Effect of allopurinol on mortality and hospitalisations in chronic heart failure: a retrospective cohort study

OBJECTIVE

To examine whether allopurinol is associated with any alteration in mortality and hospitalisations in patients with chronic heart failure (CHF). This hypothesis is based on previous data that a high urate concentration is independently associated with mortality with a risk ratio of 4.23 in CHF.

DESIGN

Retrospective cohort study.

SETTING

Medicines Monitoring Unit, Ninewells Hospital, Dundee, UK.

PATIENTS

1760 CHF patients divided into four groups: those on no allopurinol, those on long term low dose allopurinol, those on short term low dose allopurinol, and those on long term high dose allopurinol.

MAIN OUTCOME MEASURES

Total mortality, cardiovascular mortality, cardiovascular hospitalisations, cardiovascular mortality or hospitalisations.

RESULTS

Long term low dose allopurinol was associated with a significant worsening in mortality over those who never received allopurinol (relative risk 2.04, 95% confidence interval (CI) 1.48 to 2.81). This may be because low dose allopurinol is insufficient to negate the adverse effect of a high urate concentration. However, long term high dose (> or = 300 mg/day) allopurinol was associated with a significantly better mortality than longstanding low dose allopurinol (relative risk 0.59, 95% CI 0.37 to 0.95). This may mean that high dose allopurinol can fully negate the adverse effect of urate and return the mortality to normal.

CONCLUSIONS

Long term high dose allopurinol may be associated with a better mortality than long term low dose allopurinol in patients with CHF because of a dose related beneficial effect of allopurinol against the well described adverse effect of urate. Further work is required to substantiate or refute this finding.

Epidemiology and economic burden of viral hepatitis: an observational population based study

OBJECTIVE

To describe the epidemiology and estimate the health resource use of patients with viral hepatitis in Tayside, Scotland, using record linkage techniques.

DESIGN

A retrospective observational study.

SETTING

Liver disease database, Tayside, Scotland.

PATIENTS

All subjects resident in Tayside in the study period 1989-1999 and registered on the Epidemiology of Liver Disease in Tayside (ELDIT) database.

MAIN OUTCOME MEASURES

Incidence and prevalence of known viral hepatitis in Tayside, survival of subjects diagnosed with viral hepatitis, and the health resource use with respect to hospital admissions compared with the general population.

RESULTS

There were 4992 patients identified with viral hepatitis in the study period 1989-1999; 86 were IgM positive anti-hepatitis A, 187 patients were hepatitis B surface antigen (HBsAg) positive, and 469 were anti-hepatitis C (HCV) positive. HCV and HBsAg seropositive patients were more likely to be hospitalised and stay in hospital longer, less likely to survive after six years, and used more drugs of potential abuse than the general population. There was an increase in cost per admission and per patient as a consequence of liver disease.

CONCLUSIONS

A record linkage population based study of viral hepatitis allows outcomes to be identified and costed. Those at risk of viral hepatitis infection in the Tayside population should be informed about the future implication to their health and costs to society. The health service should investigate the cost effectiveness of vaccination and opportunity costs to the health service of viral hepatitis taking into consideration the increasing incidence and prevalence of disease.

Hypersensitivity reactions associated with exposure to naproxen and ibuprofen: a cohort study

The aim of the study was to evaluate the risks of hospitalisation and death due to hypersensitivity reactions associated with the NSAIDs naproxen and ibuprofen, using a record-linkage database for Tayside, Scotland (population 400,000). Cohorts of patients exposed to naproxen (n=54,038) and ibuprofen (n=79,513) were assembled. There were no deaths due to hypersensitivity. There was an increased risk of unvalidated hypersensitivity reactions during periods on-drug versus off-drug in patients exposed to naproxen and ibuprofen. However, after checking medical records, none of the three valid cases of hypersensitivity in the naproxen cohort and neither of the two in the ibuprofen cohort were judged to be due to NSAID exposure. A "worst-case" scenario gave an adjusted rate-ratio of on-drug with naproxen versus on-drug with ibuprofen of 1.63 (0.50, 5.29). The study shows that hypersensitivity reactions associated with NSAID use are rare, and provides no evidence that the risks of hypersensitivity reactions differ between naproxen and ibuprofen.

Pharmacoepidemiological research at the Medicines Monitoring Unit, Scotland: data protection and confidentiality

The Medicines Monitoring Unit (MEMO) is a University-based organization that uses record-linkage techniques to construct an observational database for the population of Tayside, Scotland (approximately 400,000 people). This contains healthcare data indexed by a unique identifier, including data on all prescriptions dispensed, which facilitates pharmacoepidemiological (and other) research. It has hitherto been possible to carry out drug safety studies in the entire population, with access to original medical records of patients where necessary, that have satisfied ethical concerns and confidentiality legislation. However, the recent UK Data Protection Act 1998 (which enforces the 1995 European Directive on Data Protection) has important implications for MEMO's research. The Act has necessitated changes to the way in which research studies are carried out, with MEMO's objective being to ensure that research can continue while protecting the rights and privacy of individual patients. This involves anonymization of data, seeking specific ethical approval for research studies and obtaining relevant permissions from 'Caldicott Guardians', as described in this article.

Seasonal changes in temperature and light drive acclimation of photosynthetic physiology and macromolecular content in Lobaria pulmonaria

Lobaria pulmonaria (L.) Hoffm. is an epiphytic lichen common to temperate deciduous forests where it copes with large changes in temperature and light levels through repeated annual cycles. Samples of L. pulmonaria were taken from a deciduous forest in southeastern Canada at 35-day intervals from February 1999 to February 2000 and also from a rare population in an evergreen forest in March and August 1999. At field-ambient temperatures and light levels, the realised photosystem II (PSII) electron transport was low both in the summer and winter, with transient peaks in the spring and autumn. In contrast, the seasonal pattern of potential electron transport measured at a fixed 20 degrees C peaked in winter, showing the importance of temperature in driving photosynthesis to low levels in the winter despite an acclimation of electron-transport potential to exploit the high ambient light. Realised gross CO2 uptake was correlated with PSII electron transport at mechanistically plausible rates at all sampling sites in the summer but not in the winter, indicating electron diversion away from CO2 fixation in the winter. Chlorophyll content was highest in the dark summer months. The amount of ribulose-1,5-bisphosphate carboxylase-oxygenase (RuBisCO) large subunit (LSU) was highest in spring. Changes in the level of this hyperabundant protein and in the activity of PSII maintained a relatively constant rate of maximum CO2 uptake per RuBisCO LSU from April through November, despite great changes in the seasonal light and temperature. L. pulmonaria acclimates between light and temperature stress in the winter months to light-limitation in the dark summer months. Transition intervals in the spring and autumn, with warm, bright and wet conditions, are likely the most amenable times for growth.

Blood pressure determinants of left ventricular wall thickness and mass index in hypertension: comparing office, ambulatory and exercise blood pressures

Left ventricular (LV) mass relates positively and continuously to cardiac mortality and thus its regression is a rational therapeutic aim. Whilst the office blood pressure (BP) relates poorly to LV mass, it was unclear whether the 24-h ambulatory BP or the exercise systolic BP (ExSBP) was the stronger correlate of LV structural indices. We studied 49 hypertensive patients with a mean age of 45 (s.d. 12) years with a mean body mass index of 27.1(3.9) kg/m(2). The mean (s.d.) of office BP, ambulatory BP and ExSBP measured at the end of the first three stages of Bruce protocol treadmill exercise I, II and III were 161(20)/99(10), 140(13)/89(10), 190(30), 198(30) and 201(33) mm Hg respectively. The LV indices measured echocardiographically were LV septal thickness (IVSd) (1.1(0.2) cm), LV posterior wall thickness (LVPWd) (1.0(0.1) cm) and LV mass indexed to body surface area (LVMI) (123(30) g/m(2)). Age and gender (male) had the highest correlations with the LV indices. Of the BP measures, the stage II ExSBP's correlation with the LV indices was consistently higher than all other ExSBP, office systolic BP and 24-h systolic ambulatory BP. In a stepwise multiple regression analysis on IVSd, after adjusting for age and gender, the stage II ExSBP was independently associated with IVSd (beta= 0.018 (s.e. 0.008), P = 0.024). When only BP measures were considered as explanatory variables only stage II ExSBP was a significant predictor (P = 0.0001) of IVSd as was the case with LVPWd (P = 0.006) and LVMI (P = 0.0008). Submaximal exercise BP measured at a workload comparable to physical activity encountered in daily life correlated more closely with the left ventricular wall thickness and mass. The exercise BP should perhaps be normalised in hypertension management to optimise regression of LV hypertrophy.

Record-linkage methodology for prescribing research

Record-linkage refers to the linking together of data relating to the same individual from separate source files. In this paper, we discuss ways in which the technique can enhance observational prescribing research in large populations. We draw upon the work of the Medicines Monitoring Unit (MEMO), University of Dundee, to illustrate its contribution to prescribing research.

Nebivolol reverses endothelial dysfunction in essential hypertension: a randomized, double-blind, crossover study

BACKGROUND

Vascular endothelial dysfunction may predict future atherosclerosis. Hence, an antihypertensive agent that reverses endothelial dysfunction and lowers blood pressure might improve the prognosis of patients with hypertension. We hypothesized that nebivolol, a vasodilating beta-blocker, could improve endothelial dysfunction. We tested this hypothesis by comparing the effects of nebivolol and atenolol on endothelial function.

METHODS AND RESULTS

Twelve hypertensive patients with a mean ambulatory blood pressure of 154+/-7/97+/-10 mm Hg were randomized after a 2-week placebo run-in period (baseline) in a double-blind, crossover fashion to 8-week treatment periods with either 5 mg of nebivolol with 2.5 mg of bendrofluazide or 50 mg of atenolol with 2.5 mg of bendrofluazide. Forearm venous occlusion plethysmography and intra-arterial infusions of acetylcholine and N(G)-monomethyl-L-arginine (L-NMMA) were used to assess stimulated and basal endothelium-dependent nitric oxide release, respectively. Sodium nitroprusside was used as an endothelium-independent control. Nebivolol/bendrofluazide and atenolol/bendrofluazide each lowered the clinic blood pressure to the same extent (132+/-7/82+/-6 and 132+/-9/83+/-8 mm Hg, respectively; P<0.001 from baseline). The vasodilatory response to acetylcholine was significantly increased with nebivolol/bendrofluazide (maximum percentage change in forearm blood flow [mean+/-SEM], 435+/-27%, P<0.001) but not with atenolol/bendrofluazide. Similarly, the endothelium-dependent vasoconstrictive response to L-NMMA was significantly improved only with nebivolol treatment (percentage change in forearm blood flow, -54+/-5%; P<0.001). The response to sodium nitroprusside was not different between treatments, suggesting that the endothelium-independent pathway was unaffected.

CONCLUSIONS

Nebivolol/bendrofluazide increased both stimulated and basal endothelial nitric oxide release, whereas for the same degree of blood pressure control, atenolol/bendrofluazide had no effect on nitric oxide bioactivity. Thus, nebivolol may offer additional vascular protection in treating hypertension.

Prior trimethoprim use and trimethoprim-resistant urinary tract infection: a nested case-control study with multivariate analysis for other risk factors

Trimethoprim resistance is increasingly prevalent in community-acquired urinary infections. The objective of this study was to evaluate the association between exposure to community-prescribed trimethoprim and other risk factors in subjects and subsequent trimethoprim-resistant urinary tract infection. The design was a nested case-control study using a record-linkage database. Study subjects submitted a urine sample to the Ninewells Hospital Laboratory between July 1993 and December 1995. Antibiotic exposure in subjects with trimethoprim-resistant isolates (cases) was compared with antibiotic exposure in subjects with trimethoprim-susceptible isolates (controls). Study subjects were drawn from the catchment area of a large teaching hospital in Tayside, Scotland. There were 13765 males and females aged 1-106 years who submitted their first urine sample for culture during the study period. After adjustment for significant risk factors and confounding variables, logistic regression analysis showed exposure to trimethoprim [odds ratio (OR) 4.35] or any antibiotic other than trimethoprim (OR 1.32) to be predictive of resistance. The growth of Proteus spp. (OR 115.14) and bacterial growth other than Escherichia coli and Proteus spp. (OR 2.83) were also predictor variables. Hospitalization in the previous 6 months was not independently associated with trimethoprim resistance. In conclusion, trimethoprim resistance was independently associated with exposure to trimethoprim and to antibiotics other than trimethoprim. Reduction in trimethoprim prescribing alone may not reduce the prevalence of trimethoprim resistance.

A review of the medical treatment of primary aldosteronism

PURPOSE

Use of the aldosterone-to-renin ratio (ARR) has suggested that at least one in 10 hypertensive subjects have primary aldosteronism (PA). There is thus a timely need to review the literature for effective drug therapies and to speculate on other therapeutic options by taking into account recent advances in understanding of the PA disease pathophysiological process.

DATA SOURCE

A MEDLINE and EMBASE search of all articles published from the start of the databases until July 1999 and reviews of the bibliographies of textbooks.

STUDY SELECTION

Primary research articles on the medical treatment of PA with emphasis on diagnosis, treatment option, drug dosage, therapeutic response and adverse drug effect.

DATA EXTRACTION

Study design and quality were assessed. Relevant data on diagnostic methodology, drug usage and response were analysed and compared.

DATA SYNTHESIS

A select number of subjects with aldosterone-producing adenoma (APA) can be expected to respond well to surgical treatment For the majority of PA cases especially subjects with idiopathic hyperaldosteronism (IHA), long-term medical treatment is now safe and feasible although no randomized controlled trials have been carried out to date. The best therapeutic response is obtained by directly antagonizing aldosterone at the receptor level using medium to low dose spironolactone and this response can be predicted by a raised ARR. The response to other potassium-sparing diuretics and calcium channel blockers are modest. IHA responds better than angiotensin II-unresponsive APA to angiotensin converting enzyme inhibitors and this may also be true with angiotensin II receptor blockers. The discovery of the aldosterone synthase gene opens up the possibility for gene therapy.

CONCLUSION

The diagnosis of PA allows appropriate management with resultant blood pressure control in many hypertensive subjects who otherwise have resistant hypertension despite multiple drug therapy.