[The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting of observational studies]. / Das Strengthening the Reporting of Observational Studies in Epidemiology (STROBE-) Statement.

Much of biomedical research is observational. The reporting of such research is often inadequate, which hampers the assessment of its strengths and weaknesses and of a study's generalizability. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Initiative developed recommendations on what should be included in an accurate and complete report of an observational study. We defined the scope of the recommendations to cover three main study designs: cohort, case-control, and cross-sectional studies. We convened a 2-day workshop in September 2004, with methodologists, researchers, and journal editors to draft a checklist of items. This list was subsequently revised during several meetings of the coordinating group and in e-mail discussions with the larger group of STROBE contributors, taking into account empirical evidence and methodological considerations. The workshop and the subsequent iterative process of consultation and revision resulted in a checklist of 22 items (the STROBE Statement) that relate to the title, abstract, introduction, methods, results, and discussion sections of articles. Eighteen items are common to all three study designs and four are specific for cohort, case-control, or cross-sectional studies. A detailed Explanation and Elaboration document is published separately and is freely available on the web sites of PLoS Medicine, Annals of Internal Medicine, and Epidemiology. We hope that the STROBE Statement will contribute to improving the quality of reporting of observational studies.

The cost-effectiveness of an early interventional strategy in non-ST-elevation acute coronary syndrome based on the RITA 3 trial.

BACKGROUND: Evidence suggests that an early interventional strategy for patients with non-ST-elevation acute coronary syndrome (NSTE-ACS) can improve health outcomes but also increase costs when compared with a conservative strategy. OBJECTIVE: The aim of this study was to assess the cost-effectiveness of an early interventional strategy in different risk groups from a UK health-service perspective. DESIGN: Decision-analytic model based on randomised clinical trial data. MAIN OUTCOME MEASURES: Costs in UK Sterling at 2003/2004 prices and quality-adjusted life years (QALYs) combined into an incremental cost-effectiveness ratio. METHODS: Data from the third Randomised Intervention Trial of unstable Angina (RITA 3) was employed to estimate rates of cardiovascular death and myocardial infarction, costs and health-related quality of life. Cost-effectiveness was estimated over patients' lifetimes within the decision-analytic model. RESULTS: The mean incremental cost per QALY gained for an early interventional strategy was approximately 55,000 pounds sterling, 22,000 pounds sterling and 12,000 pounds sterling for patients at low, intermediate and high risk, respectively. The early interventional strategy is approximately 1%, 35% and 95% likely to be cost-effective for patients at low, intermediate and high risk, respectively, at a threshold of 20,000 pounds sterling per QALY. The cost-effectiveness of early intervention in low-risk patients is sensitive to assumptions about the duration of the treatment effect. CONCLUSION: An early interventional strategy in patients presenting with NSTE-ACS is likely to be considered cost-effective for patients at high and intermediate risk, but this is less likely to be the case for patients at low risk.

Rosiglitazone RECORD study: glucose control outcomes at 18 months.

AIMS: To compare glucose control over 18 months between rosiglitazone oral combination therapy and combination metformin and sulphonylurea in people with Type 2 diabetes. METHODS: RECORD, a multicentre, parallel-group study of cardiovascular outcomes, enrolled people with an HbA(1c) of 7.1-9.0% on maximum doses of metformin or sulphonylurea. If on metformin they were randomized to add-on rosiglitazone or sulphonylurea (open label) and if on sulphonylurea to rosiglitazone or metformin. HbA(1c) was managed to < or = 7.0% by dose titration. A prospectively defined analysis of glycaemic control on the first 1122 participants is reported here, with a primary outcome assessed against a non-inferiority margin for HbA(1c) of 0.4%. RESULTS: At 18 months, HbA(1c) reduction on background metformin was similar with rosiglitazone and sulphonylurea [difference 0.07 (95% CI -0.09, 0.23)%], as was the change when rosiglitazone or metformin was added to sulphonylurea [0.06 (-0.09, 0.20)%]. At 6 months, the effect on HbA(1c) was greater with add-on sulphonylurea, but was similar whether sulphonylurea was added to rosiglitazone or metformin. Differences in fasting plasma glucose were not statistically significant at 18 months [rosiglitazone vs. sulphonylurea -0.36 (-0.74, 0.02) mmol/l, rosiglitazone vs. metformin -0.34 (-0.73, 0.05) mmol/l]. Increased homeostasis model assessment insulin sensitivity and reduced C-reactive protein were greater with rosiglitazone than metformin or sulphonylurea (all P < or = 0.001). Body weight was significantly increased with rosiglitazone compared with sulphonylurea [difference 1.2 (0.4, 2.0) kg, P = 0.003] and metformin [difference 4.3 (3.6, 5.1) kg, P < 0.001]. CONCLUSIONS: In people with diabetes, rosiglitazone in combination with metformin or sulphonylurea was demonstrated to be non-inferior to the standard combination of metformin + sulphonylurea in lowering HbA(1c) over 18 months, and produces greater improvements in C-reactive protein and basal insulin sensitivity but is also associated with greater weight gain.

Prevalence and prognostic implications of electrocardiographic left ventricular hypertrophy in heart failure: evidence from the CHARM programme.

BACKGROUND: Electrocardiographic left ventricular hypertrophy (ECG LVH) is a powerful independent predictor of cardiovascular morbidity and mortality in hypertension. OBJECTIVE: To determine the contemporary prevalence and prognostic implications of ECG LVH in a broad spectrum of patients with heart failure with and without reduced left ventricular ejection fraction (LVEF). METHODS AND OUTCOME: The Candesartan in Heart failure: Assessment of Reduction in Mortality and morbidity (CHARM) programme randomised 7599 patients with symptomatic heart failure to receive candesartan or placebo. The primary outcome comprised cardiovascular death or hospital admission for worsening heart failure. The relative risk (RR) conveyed by ECG LVH compared with a normal ECG was examined in a Cox model, adjusting for as many as 31 covariates of prognostic importance. RESULTS: The prevalence of ECG LVH was similar in all three CHARM trials (Alternative, 15.4%; Added, 17.1%; Preserved, 14.7%; Overall, 15.7%) despite a more frequent history of hypertension in CHARM-Preserved. ECG LVH was an independent predictor of worse prognosis in CHARM-Overall. RR for the primary outcome was 1.27 (95% confidence interval (CI) 1.04 to 1.55, p = 0.018). The risk of secondary end points was also increased: cardiovascular death, 1.50 (95% CI 1.13 to 1.99, p = 0.005); hospitalisation due to heart failure, 1.19 (95% CI 0.94 to 1.50, p = 0.148); and composite major cardiovascular events, 1.35 (95% CI 1.12 to 1.62, p = 0.002). CONCLUSION: ECG LVH is similarly prevalent in patients with symptomatic heart failure regardless of LVEF. The simple clinical finding of ECG LVH was an independent predictor of a worse clinical outcome in a broad spectrum of patients with heart failure receiving extensive contemporary treatment. Candesartan had similar benefits in patients with and without ECG LVH.

Interventional versus conservative treatment in acute non-ST elevation coronary syndrome: time course of patient management and disease events over one year in the RITA 3 trial.

OBJECTIVE: To determine whether, in acute non-ST elevation coronary syndrome, the benefit from early invasive coronary intervention compared with a conservative strategy of later symptom-guided intervention varies over time. METHODS: In RITA 3 (Randomised Intervention Trial of unstable Angina 3) patients were randomly assigned to coronary angiography (median 2 days after randomisation) and appropriate intervention (n = 895) or to a symptom-guided conservative strategy (n = 915). RESULTS: In the first week patients in both groups were at highest risk of death, myocardial infarction (MI) or refractory angina (incidence rate 40 times higher than in months 5-12 of follow up). There were 22 MIs and 6 deaths in the intervention group (largely due to procedure-related events, 14 MIs and 3 deaths) versus 17 MIs and 3 deaths in the conservative group. In the rest of the year there were an additional 12 versus 27 MIs, respectively (treatment-time interaction p = 0.021). Over one year in the intervention group there was a 43% reduction in refractory angina; 22% of patients underwent coronary artery bypass surgery and 35% underwent percutaneous coronary intervention only, which reduced refractory angina but provoked some early MIs; and 43% were still treated medically, mostly because of a favourable initial angiogram. CONCLUSION: Any intervention policy needs to recognise the high risk of events in the first week and the substantial minority of patients not needing intervention. Intervention may be best targeted at higher risk patients, as the early hazards of the procedure are then offset by reduced subsequent events.

5-year outcome of an interventional strategy in non-ST-elevation acute coronary syndrome: the British Heart Foundation RITA 3 randomised trial.

BACKGROUND: The long-term outcome of an interventional strategy in patients with non-ST-elevation acute coronary syndrome is unknown. We tested whether an interventional strategy (routine angiography followed by revascularisation) was better than a conservative strategy (ischaemia-driven or symptom-driven angiography) over 5 years' follow-up. METHODS: In a multicentre randomised trial, 1810 patients (from 45 hospitals in England and Scotland, UK) with non-ST-elevation acute coronary syndrome were randomly assigned to receive an early intervention (n=895) or a conservative strategy (n=915) within 48 h of the index episode of cardiac pain. In each group, the aim was to provide the best medical treatment, and also to undertake coronary arteriography within 72 h in the interventional strategy with subsequent management guided by the angiographic findings. Analysis was by intention to treat and the primary outcome (composite of death or non-fatal myocardial infarction) had masked independent adjudication. RITA 3 has been assigned the International Standard Randomised Control Trial Number ISRCTN07752711. FINDINGS: At 1-year follow-up, rates of death or non-fatal myocardial infarction were similar. However, at a median of 5 years' follow-up (IQR 4.6-5.0), 142 (16.6%) patients with intervention treatment and 178 (20.0%) with conservative treatment died or had non-fatal myocardial infarction (odds ratio 0.78, 95% CI 0.61-0.99, p=0.044), with a similar benefit for cardiovascular death or myocardial infarction (0.74, 0.56-0.97, p=0.030). 234 (102 [12%] intervention, 132 [15%] conservative) patients died during follow-up (0.76, 0.58-1.00, p=0.054). The benefits of an intervention strategy were mainly seen in patients at high risk of death or myocardial infarction (p=0.004), and for the highest risk group, the odds ratio of death or non-fatal myocardial infarction was 0.44 (0.25-0.76). INTERPRETATION: In patients with non-ST-elevation acute coronary syndrome, a routine invasive strategy leads to long-term reduction in risk of death or non-fatal myocardial infarction, and this benefit is mainly in high-risk patients. The findings provide support for national and international guidelines in the need for more robust risk stratification in acute coronary syndrome.

Rosiglitazone Evaluated for Cardiac Outcomes and Regulation of Glycaemia in Diabetes (RECORD): study design and protocol.

AIMS/HYPOTHESIS: Studies suggest that in addition to blood glucose concentrations, thiazolidinediones such as rosiglitazone improve some cardiovascular (CV) risk factors and surrogate markers, that are abnormal in type 2 diabetes. However, fluid retention might lead to cardiac failure in a minority of people. The aim of the Rosiglitazone Evaluated for Cardiac Outcomes and Regulation of Glycaemia in Diabetes (RECORD) study is to evaluate the long-term impact of these effects on CV outcomes, as well as on long-term glycaemic control, in people with type 2 diabetes. MATERIALS AND METHODS: RECORD is a 6-year, randomised, open-label study in type 2 diabetic patients with inadequate blood glucose control (HbA1c 7.1-9.0%) on metformin or sulphonylurea alone. The study is being performed in 327 centres in Europe and Australasia. After a 4-week run-in, participants were randomised by current treatment stratum to add-on rosiglitazone, metformin or sulphonylurea, with dose titration to a target HbA1c of < or = 7.0%. If confirmed HbA1c rises to > or = 8.5%, either a third glucose-lowering drug is added (rosiglitazone-treated group) or insulin is started (non-rosiglitazone group). The same criterion for failure of triple oral drug therapy in the rosiglitazone-treated group is used for starting insulin in this group. The primary endpoint is the time to first CV hospitalisation or death, blindly adjudicated by a central endpoints committee. The study aim is to evaluate non-inferiority of the rosiglitazone group vs the non-rosiglitazone group with respect to CV outcomes. Safety, tolerability and study conduct are monitored by an independent board. All CV endpoint and safety data are held and analysed by a clinical trials organisation, and are not available to the study investigators while data collection is open. RESULTS: Over a 2-year period a total of 7,428 people were screened in 25 countries. Of these, 4,458 were randomised; 2,228 on background metformin, 2,230 on background sulphonylurea. Approximately half of the participants are male (52%) and almost all are Caucasian (99%). CONCLUSIONS/INTERPRETATION: The RECORD study should provide robust data on the extent to which rosiglitazone, in combination with metformin or sulphonylurea therapy, affects CV outcomes and progression of diabetes in the long term.

Issues in data monitoring and interim analysis of trials.

OBJECTIVES: To address issues about data monitoring committees (DMCs) for randomised controlled trials (RCTs). DATA SOURCES: Electronic databases. Handsearching of selected books. Personal contacts with experts in the field. REVIEW METHODS: Systematic literature reviews of DMCs and small group processes in decision-making; sample surveys of: reports of RCTs, recently completed and ongoing RCTs and policies of major organisations involved in RCTs; case studies of four DMCs; and interviews with experienced DMC members. All focused on 23 prestated questions. RESULTS: Although still a minority, RCTs increasingly have DMCs. There is wide agreement that nearly all trials need some form of data monitoring. Central to the role of the DMC is monitoring accumulating evidence related to benefit and toxicity; variation in emphasis has been reflected in the plethora of names. DMCs for trials performed for regulatory purposes should be aware of any special requirements and regulatory consequences. Advantages were identified for both larger and smaller DMCs. There is general agreement that a DMC should be independent and multidisciplinary. Consumer and ethicist membership is controversial. The chair is recognised as being particularly influential, and likely to be most effective if he or she is experienced, understands both statistical and clinical issues, and is facilitating in style and impartial. There is no evidence available to judge suggested approaches to training. The review suggested that costs should be covered, but other rewards must be so minimal as to not affect decision-making. It is usual to have a minimum frequency of DMC meetings, with evidence that face-to-face meetings are preferable. It is common to have open sessions and a closed session. A report to a DMC should cover benefits and risks in a balanced way, summarised in an accessible style, avoiding excessive detail, and be as current as possible. Disadvantages of blinded analyses seem to outweigh advantages. Information about comparable studies should be included, although interaction with the DMCs of similar ongoing trials is controversial. A range of formal statistical approaches can be used, although this is only one of a number of considerations. DMCs usually reach decisions by consensus, but other approaches are sometimes used. The general, but not unanimous, view is that DMCs should be advisory rather than executive on the basis that it is the trial organisers who are ultimately responsible for the conduct of the trial. CONCLUSIONS: Some form of data monitoring should be considered for all RCTs, with reasons given where there is no DMC or when any member is not independent. An early DMC meeting is helpful, determining roles and responsibilities; planned operations can be agreed with investigators and sponsors/funders. A template for a DMC charter is suggested. Competing interests should be declared. DMC size (commonly three to eight people) is chosen to optimise performance. Members are usually independent and drawn from appropriate backgrounds, and some, particularly the chair, are experienced. A minimum frequency of meetings is usually agreed, with flexibility for more if needed. The DMC should understand and agree the statistical approach (and guidelines) chosen, with both the DMC statistician and analysis statistician competent to apply the method. A DMC's primary purpose is to ensure that continuing a trial according to its protocol is ethical, taking account of both individual and collective ethics. A broader remit in respect of wider ethical issues is controversial; arguably, these are primarily the responsibility of research ethics committees, trial steering committees and investigators. The DMC should know the range of recommendations or decisions open to it, in advance. A record should be kept describing the key issues discussed and the rationale for decisions taken. Errors are likely to be reduced if a DMC makes a thorough review of the evidence and has a clear understanding of how it should function, there is active participation by all members, differences are resolved through discussion and there is systematic consideration of the various decision options. DMCs should be encouraged to comment on draft final trial reports. These should include information about the data monitoring process and detail the DMC membership. It is recommended that groups responsible for data monitoring be given the standard name 'Data Monitoring Committee' (DMC). Areas for further research include: widening DMC membership beyond clinicians, trialists and statisticians; initiatives to train DMC members; methods of DMC decision-making; 'open' data monitoring; DMCs covering a portfolio of trials rather than single trials; DMC size and membership, incorporating issues of group dynamics; empirical study of the workings of DMCs and their decision-making, and which trials should or should not have a DMC.

Do men benefit more than women from an interventional strategy in patients with unstable angina or non-ST-elevation myocardial infarction? The impact of gender in the RITA 3 trial.

AIMS: The RITA 3 trial randomized patients with non-ST-elevation myocardial infarction or unstable angina to strategies of early intervention (angiography followed by revascularization) or conservative care (ischaemia or symptom driven angiography). The aim of this analysis was to investigate the impact of gender on the effect of these two strategies. METHODS AND RESULTS: In total, 1810 patients (682 women and 1128 men) were randomized. The risk factor profile of women at presentation was markedly different to men. There was evidence that men benefited more from an early intervention strategy for death or non-fatal myocardial infarction at 1 year (adjusted odds ratios 0.63, 95% confidence interval 0.41-0.98 for men and 1.79, 95% confidence interval 0.95-3.35 for women; interaction p-value=0.007). Men who underwent the assigned angiogram were more likely to be put forward for coronary artery bypass surgery, even after allowing for differences in disease severity. CONCLUSION: An early intervention strategy resulted in a beneficial effect in men which was not seen in women although caution is needed in interpretation. Further research is needed to evaluate why women do not appear to benefit from early intervention and to identify treatments that improve the prognosis of women.

A comparison of two methods for the estimation of precision with incomplete longitudinal data, jointly modelled with a time-to-event outcome.

Several methods for the estimation and comparison of rates of change in longitudinal studies with staggered entry and informative drop-outs have been recently proposed. For multivariate normal linear models, REML estimation is used. There are various approaches to maximizing the corresponding log-likelihood; in this paper we use a restricted iterative generalized least squares method (RIGLS) combined with a nested EM algorithm. An important statistical problem in such approaches is the estimation of the standard errors adjusted for the missing data (observed data information matrix). Louis has provided a general technique for computing the observed data information in terms of completed data quantities within the EM framework. The multiple imputation (MI) method for obtaining variances can be regarded as an alternative to this. The aim of this paper is to develop, apply and compare the Louis and a modified MI method in the setting of longitudinal studies where the source of missing data is either death or disease progression (informative) or end of the study (assumed non-informative). Longitudinal data are simultaneously modelled with the missingness process. The methods are illustrated by modelling CD4 count data from an HIV-1 clinical trial and evaluated through simulation studies. Both methods, Louis and MI, are used with Monte Carlo simulations of the missing data using the appropriate conditional distributions, the former with 100 simulations, the latter with 5 and 10. It is seen that naive SEs based on the completed data likelihood can be seriously biased. This bias was largely corrected by Louis and modified MI methods, which gave broadly similar estimates. Given the relative simplicity of the modified MI method, it may be preferable.

Neonatal head circumference, neonatal weight, and risk of hayfever, asthma and eczema in a large cohort of adolescents from Sheffield, England.

BACKGROUND: Neonatal physical characteristics, including head circumference and birth weight, have been hypothesized to be markers of in utero thymic development. Greater head circumference and lower birth weight have been linked in previous studies to subsequent development of asthma, and greater birth weight has been associated with subsequent development of eczema. OBJECTIVE: To investigate potential associations between neonatal head circumference and weight and hayfever, asthma and eczema in a cohort of adolescents from Sheffield, England. METHODS: Responses to a questionnaire inquiring about physician-diagnosed hayfever, asthma and eczema among adolescents in Sheffield, England, were linked to previously recorded measurements of weight at birth and at 1 month and head circumference at 1 month. Logistic regression methods were used to relate diagnoses to neonatal measurements and potential confounders. RESULTS: The cohort consisted of 10,809 adolescents, of whom 16.5% reported hayfever, 18.0% asthma, and 16.2% eczema. After adjusting for sex, age at the time of the questionnaire, maternal age and gestational age at birth, number of older and younger siblings, time since birth of next older sibling, neonatal sickness, type of neonatal feeding, and maternal and paternal educational backgrounds, hayfever was the only disease associated with neonatal measurements. Comparing the highest with the lowest fifths of distributions, lifetime prevalence of hayfever was positively associated with neonatal head circumference (adjusted odds ratio 1.23, 95% CI 1.03 to 1.47) and with birth weight (1.17, 0.99 to 1.39). Hayfever was inversely related to the ratio of head circumference to birth weight (0.89, 0.75 to 1.05) and to gestational age. The associations with head circumference and birth weight were not substantially altered by further adjustment for gestational age. CONCLUSION: Greater neonatal head circumference may be associated with an increased risk of hayfever, but the inverse relationship between hayfever prevalence and the ratio of head circumference to birth weight challenges the prior hypothesis that greater head circumference relative to body mass reflects abnormal thymic development in utero, increasing the likelihood of allergic sensitization.

Interventional versus conservative treatment for patients with unstable angina or non-ST-elevation myocardial infarction: the British Heart Foundation RITA 3 randomised trial. Randomized Intervention Trial of unstable Angina.

BACKGROUND: Current guidelines suggest that, for patients at moderate risk of death from unstable coronary-artery disease, either an interventional strategy (angiography followed by revascularisation) or a conservative strategy (ischaemia-driven or symptom-driven angiography) is appropriate. We aimed to test the hypothesis that an interventional strategy is better than a conservative strategy in such patients. METHODS: We did a randomised multicentre trial of 1810 patients with non-ST-elevation acute coronary syndromes (mean age 62 years, 38% women). Patients were assigned an early intervention or conservative strategy. The antithrombin agent in both groups was enoxaparin. The co-primary endpoints were a combined rate of death, non-fatal myocardial infarction, or refractory angina at 4 months; and a combined rate of death or non-fatal myocardial infarction at 1 year. Analysis was by intention to treat. FINDINGS: At 4 months, 86 (9.6%) of 895 patients in the intervention group had died or had a myocardial infarction or refractory angina, compared with 133 (14.5%) of 915 patients in the conservative group (risk ratio 0.66, 95% CI 0.51-0.85, p=0.001). This difference was mainly due to a halving of refractory angina in the intervention group. Death or myocardial infarction was similar in both treatment groups at 1 year (68 [7.6%] vs 76 [8.3%], respectively; risk ratio 0.91, 95% CI 0.67-1.25, p=0.58). Symptoms of angina were improved and use of antianginal medications significantly reduced with the interventional strategy (p<0.0001). INTERPRETATION: In patients presenting with unstable coronary-artery disease, an interventional strategy is preferable to a conservative strategy, mainly because of the halving of refractory or severe angina, and with no increased risk of death or myocardial infarction.

A score for predicting risk of death from cardiovascular disease in adults with raised blood pressure, based on individual patient data from randomised controlled trials.

OBJECTIVE: To create a risk score for death from cardiovascular disease that can be easily used. DESIGN: Data from eight randomised controlled trials of antihypertensive treatment. SETTING: Europe and North America. PARTICIPANTS: 47 088 men and women from trials that had differing age ranges and differing eligibility criteria for blood pressure. MAIN OTUCOME MEASURE: 1639 deaths from cardiovascular causes during a mean 5.2 years of follow up. RESULTS: Baseline factors were related to risk of death from cardiovascular disease using a multivariate Cox model, adjusting for trial and treatment group (active versus control). A risk score was developed from 11 factors: age, sex, systolic blood pressure, serum total cholesterol concentration, height, serum creatinine concentration, cigarette smoking, diabetes, left ventricular hypertrophy, history of stroke, and history of myocardial infarction. The risk score is an integer, with points added for each factor according to its association with risk. Smoking contributed more in women and in younger age groups. In women total cholesterol concentration mattered less than in men, whereas diabetes had more of an effect. Antihypertensive treatment reduced the score. The five year risk of death from cardiovascular disease for scores of 10, 20, 30, 40, 50, and 60 was 0.1%, 0.3%, 0.8%, 2.3%, 6.1%, and 15.6%, respectively. Age and sex distributions of the score from the two UK trials enabled individual risk assessment to be age and sex specific. Risk prediction models are also presented for fatal coronary heart disease, fatal stroke, and all cause mortality. CONCLUSION: The risk score is an objective aid to assessing an individual's risk of cardiovascular disease, including stroke and coronary heart disease. It is useful for physicians when determining an individual's need for antihypertensive treatment and other management strategies for cardiovascular risk.

Impact of missing data due to drop-outs on estimators for rates of change in longitudinal studies: a simulation study.

Many cohort studies and clinical trials are designed to compare rates of change over time in one or more disease markers in several groups. One major problem in such longitudinal studies is missing data due to patient drop-out. The bias and efficiency of six different methods to estimate rates of changes in longitudinal studies with incomplete observations were compared: generalized estimating equation estimates (GEE) proposed by Liang and Zeger (1986); unweighted average of ordinary least squares (OLSE) of individual rates of change (UWLS); weighted average of OLSE (WLS); conditional linear model estimates (CLE), a covariate type estimates proposed by Wu and Bailey (1989); random effect (RE), and joint multivariate RE (JMRE) estimates. The latter method combines a linear RE model for the underlying pattern of the marker with a log-normal survival model for informative drop-out process. The performance of these methods in the presence of missing data completely at random (MCAR), at random (MAR) and non-ignorable (NIM) were compared in simulation studies. Data for the disease marker were generated under the linear random effects model with parameter values derived from realistic examples in HIV infection. Rates of drop-out, assumed to increase over time, were allowed to be independent of marker values or to depend either only on previous marker values or on both previous and current marker values. Under MACR all six methods yielded unbiased estimates of both group mean rates and between-group difference. However, the cross-sectional view of the data in the GEE method resulted in seriously biased estimates under MAR and NIM drop-out process. The bias in the estimates ranged from 30 per cent to 50 per cent. The degree of bias in the GEE estimates increases with the severity of non-randomness and with the proportion of MAR data. Under MCAR and MAR all the other five methods performed relatively well. RE and JMRE estimates were more efficient(that is, had smaller variance) than UWLS, WLS and CL estimates. Under NIM, WLS and particularly RE estimates tended to underestimate the average rate of marker change (bias approximately 10 per cent). Under NIM, UWLS, CL and JMRE performed better in terms of bias (3-5 per cent) with the JMRE giving the most efficient estimates. Given that markers are key variables related to disease progression, missing marker data are likely to be at least MAR. Thus, the GEE method may not be appropriate for analysing such longitudinal marker data. The potential biases due to incomplete data require greater recognition in reports of longitudinal studies. Sensitivity analyses to assess the effect of drop-outs on inferences about the target parameters are important.

Wheeze associated with prenatal tobacco smoke exposure: a prospective, longitudinal study. ALSPAC Study Team.

AIMS: To determine whether maternal smoking during pregnancy is a risk factor for reported wheeze in early childhood that is independent of postnatal environmental tobacco smoke (ETS) exposure and other known risk factors. METHODS: A total of 8561 mothers and infants completed questions about smoking during pregnancy, ETS exposure, and the mother's recall of wheeze during early childhood. RESULTS: A total of 1869 (21.8%) children had reported wheeze between 18 and 30 months of age, and 3496 (40.8%) had reported wheeze in one or more of the three study periods (birth to 6 months, 6-18 months, 18-30 months). The risk of wheeze between 18 and 30 months of age was higher if the mother smoked during pregnancy. This relation did not show a dose-response effect and became less obvious after adjustment for the effects of other factors. Average daily duration of ETS exposure reported at 6 months of age showed a dose-response effect and conferred a similar risk of reported wheeze. Factors associated with early childhood wheeze had the following adjusted odds ratios: maternal history of asthma 2.03 (1.74 to 2. 37); preterm delivery 1.66 (1.30 to 2.13); male sex 1.42 (1.28 to 1. 59); rented accommodation 1.29 (1.11 to 1.51); and each additional child in household 1.13 (1.04 to 1.24). CONCLUSIONS: Maternal smoking during pregnancy may be a risk factor for reported wheeze during early childhood that is independent of postnatal ETS exposure. For wheeze between 18 and 30 months of age, light smoking during the third trimester of pregnancy appears to confer the same risk as heavier smoking.

Subgroup analysis and other (mis)uses of baseline data in clinical trials.

BACKGROUND: Baseline data collected on each patient at randomisation in controlled clinical trials can be used to describe the population of patients, to assess comparability of treatment groups, to achieve balanced randomisation, to adjust treatment comparisons for prognostic factors, and to undertake subgroup analyses. We assessed the extent and quality of such practices in major clinical trial reports. METHODS: A sample of 50 consecutive clinical-trial reports was obtained from four major medical journals during July to September, 1997. We tabulated the detailed information on uses of baseline data by use of a standard form. FINDINGS: Most trials presented baseline comparability in a table. These tables were often unduly large, and about half the trials inappropriately used significance tests for baseline comparison. Methods of randomisation, including possible stratification, were often poorly described. There was little consistency over whether to use covariate adjustment and the criteria for selecting baseline factors for which to adjust were often unclear. Most trials emphasised the simple unadjusted results and covariate adjustment usually made negligible difference. Two-thirds of the reports presented subgroup findings, but mostly without appropriate statistical tests for interaction. Many reports put too much emphasis on subgroup analyses that commonly lacked statistical power. INTERPRETATION: Clinical trials need a predefined statistical analysis plan for uses of baseline data, especially covariate-adjusted analyses and subgroup analyses. Investigators and journals need to adopt improved standards of statistical reporting, and exercise caution when drawing conclusions from subgroup findings.

Quality of life after coronary angioplasty or continued medical treatment for angina: three-year follow-up in the RITA-2 trial. Randomized Intervention Treatment of Angina.

OBJECTIVES: We sought to evaluate the impact of percutaneous transluminal coronary angioplasty (PTCA) and medical treatment on self-perceived quality of life among patients with angina. BACKGROUND: The second Randomized Intervention Treatment of Angina trial (RITA-2) implemented initial policies of PTCA or continued medical treatment in patients with angina, allowing assessment of long-term health consequences. METHODS: A total of 1,018 patients were randomly assigned (504 to PTCA and 514 to medical treatment). The short form 36 (SF-36) self-administered quality-of-life questionnaire was completed at randomization and three months, one year and three years later. To date, 98% of patients reached one year and 67% reached three years. RESULTS: The PTCA group had significantly greater improvements in physical functioning, vitality and general health at both three months and one year, but not at three years. These quality-of-life scores were strongly related to breathlessness, angina grade and treadmill exercise time both at baseline and at one year. The treatment differences in quality of life are explained by the PTCA group's improvements in breathlessness, angina and exercise time. The attenuation of treatment difference at three years is partly attributed to 27% of medically treated patients receiving nonrandomized interventions in the interim. For both groups, there were also improvements in ratings of physical role functioning, emotional role functioning, social functioning, pain and mental health, but for these the superiority of PTCA over medical treatment was less pronounced. After one year, 33% and 22% of the PTCA and medical groups, respectively, rated their health much better. CONCLUSIONS: Coronary angioplasty substantially improves patient-perceived quality of life, especially physical functioning and vitality, as compared with continued medical treatment. These differences are attributed to alleviation of cardiac symptoms (specifically, breathlessness and angina), but must be balanced against the small procedure-related risks of PTCA.

Estimation and comparison of rates of change in longitudinal studies with informative drop-outs.

Many cohort studies and clinical trials have designs which involve repeated measurements of disease markers. One problem in such longitudinal studies, when the primary interest is to estimate and to compare the evolution of a disease marker, is that planned data are not collected because of missing data due to missing visits and/or withdrawal or attrition (for example, death). Several methods to analyse such data are available, provided that the data are missing at random. However, serious biases can occur when missingness is informative. In such cases, one needs to apply methods that simultaneously model the observed data and the missingness process. In this paper we consider the problem of estimation of the rate of change of a disease marker in longitudinal studies, in which some subjects drop out prematurely (informatively) due to attrition, while others experience a non-informative drop-out process (end of study, withdrawal). We propose a method which combines a linear random effects model for the underlying pattern of the marker with a log-normal survival model for the informative drop-out process. Joint estimates are obtained through the restricted iterative generalized least squares method which are equivalent to restricted maximum likelihood estimates. A nested EM algorithm is applied to deal with censored survival data. The advantages of this method are: it provides a unified approach to estimate all the model parameters; it can effectively deal with irregular data (that is, measured at irregular time points), a complicated covariance structure and a complex underlying profile of the response variable; it does not entail such complex computation as would be required to maximize the joint likelihood. The method is illustrated by modelling CD4 count data in a clinical trial in patients with advanced HIV infection while its performance is tested by simulation studies.