Returns on investment in public health: effect of public health expenditures on infant health, 1983-1990

In this paper a number of methodological issues relating to research on the relationship between the menopause, mood and hormone replacement therapy (HRT) are discussed. These issues relate to problems of design and statistical analyses, problems which have prevented the reaching of definite conclusions regarding the relationship between menopause, mood and hormones. These problems are discussed under three main headings, namely, the assessment of menopausal status, statistical modelling and the design and analyses of clinical trials. Problems relating to concepts and measurement of dependent variables are the subject matter of the papers that follow. Within the three main headings more specific issues are detailed. The paper concludes with a list of recommendations on how research in this important area might be further advanced.     

Quantitative synthesis in systematic reviews

The final common pathway for most systematic reviews is a statistical summary of the data, or meta-analysis. The complex methods used in meta-analyses should always be complemented by clinical acumen and common sense in designing the protocol of a systematic review, deciding which data can be combined, and determining whether data should be combined. Both continuous and binary data can be pooled. Most meta-analyses summarize data from randomized trials, but other applications, such as the evaluation of diagnostic test performance and observational studies, have also been developed. The statistical methods of meta-analysis aim at evaluating the diversity (heterogeneity) among the results of different studies, exploring and explaining observed heterogeneity, and estimating a common pooled effect with increased precision. Fixed-effects models assume that an intervention has a single true effect, whereas random-effects models assume that an effect may vary across studies. Meta-regression analyses, by using each study rather than each patient as a unit of observation, can help to evaluate the effect of individual variables on the magnitude of an observed effect and thus may sometimes explain why study results differ. It is also important to assess the robustness of conclusions through sensitivity analyses and a formal evaluation of potential sources of bias, including publication bias and the effect of the quality of the studies on the observed effect.     

Clinical trials with multiple outcomes: a statistical perspective on their design, analysis, and interpretation

This article tackles both practical and statistical issues in the handling of multiple outcomes in clinical trials, with relevance to trial design, analysis, and reporting. Specific topics illustrated by examples include: the advantage of prespecifying priorities amongst outcomes and analyses, corrections for multiple significance testing and their limited value, problems with adverse event data, the use of a single global test of significance for clinically related outcomes, the use of a combined outcome for clinical event data, and the value of exploring interrelationships amongst outcomes. The problems in handling multiple outcomes are enhanced by trials being too small, dichotomous attitudes (is the trial "positive" or not?), obsession with p-values, and the manipulative instincts of human nature. While predeclarations of priorities in analysis and reporting of multiple outcomes are important in suppressing distortive claims, it would be unfortunate if too inflexible an approach suppressed unpredictable findings from being seriously considered.     

Thrombolytic therapy for unstable angina

Thrombolytic therapy for unstable angina has not gained acceptance as a primary treatment for unstable angina (UA) despite the evidence showing a reduction in mortality when these agents are given for myocardial infarction. The purpose of this review is to examine the clinical value of thrombolytic therapy for UA. The multiple lines of evidence supporting intracoronary thrombus formation as a key mechanism in the pathogenesis of UA are reviewed. Studies examining the effect of thrombolytic therapy on angiographic endpoints have shown little effect on the extent of luminal narrowing, but do reveal a decrease in angiographically detected thrombus. Twelve randomized, controlled trials of thrombolytic agents in 611 UA patients with predefined clinical endpoints have been published. These trials varied widely in design and adjunctive therapy both in treated and control grops. Review of these trials show a tendency to fewer clinical events such as death, infarction, and need for revascularization in treated patients, with a corresponding increase in bleeding complications. Clinical efficacy of thrombolytic therapy cannot be excluded by the available data, perhaps in part because of insufficient numbers of patients treated. Determination of the net clinical value of thrombolytic therapy must await larger and more definitive trials.     

Hypertension and endstage renal disease

OBJECTIVE: To review current literature regarding the development of hypertensive renal disease, its epidemiology, and its pathophysiology. This review focuses on strategies to slow or halt the progression of endstage renal disease (ESRD) in hypertension, including the role of blood pressure control, different types of antihypertensive agents, early treatment, and dietary considerations. DATA SOURCES: Information was retrieved from searching the MEDLINE database for articles consisting of epidemiologic studies, clinical studies, and review articles pertaining to hypertension and ESRD. Information also was obtained from the US Renal Data System annual data reports. STUDY SELECTION: Emphasis was placed on clinical trials in the English language addressing issues in hypertension and ESRD. Clinical trials reporting relationships between blood pressure control and ESRD, as well as those comparing different antihypertensive agents, were evaluated. DATA EXTRACTION: The methodology and results from clinical trials were evaluated. Studies were assessed according to the measures of renal function used, baseline data collected, degree of blood pressure control, and antihypertensive therapy. DATA SYNTHESIS: Clinical trials including patients with essential hypertension, diabetes mellitus, and renal insufficiency of various etiologies were evaluated. The recommendations from these evaluations were based on study design and the types of populations used (i.e., blacks vs. whites, diabetics vs. nondiabetics). CONCLUSIONS: Blood pressure control is currently the most important strategy to slow or halt the progression of renal insufficiency in hypertensive individuals. Whether specific antihypertensives are renal protective is still controversial, but results from clinical trials are promising.     

Cross-validation of bioanalytical methods between laboratories

Increased reliance on pharmacokinetic studies in regulatory submissions emphasizes the need for cross-validating bioanalytical methods between different laboratories to allow comparison of data. Globalization of pharmaceutical development results in a greater need to define cross-validation standards. A strategy for performing cross-validation experiments using prepared biological samples of known concentration and "real" samples from clinical trials is presented. The statistical techniques used to compare data sets and establish acceptability of the assays are illustrated by practical examples.     

Research fundamentals: III. elements of a research protocol for clinical trials

A clinical trial is a powerful technique for evaluating the effectiveness of an experimental intervention. The initial stages of planning a clinical trial involve choosing and refining a research question, selecting a study design, and deciding on appropriate statistical tests and sample sizes. The success of the study depends upon how well these issues are thought out in advance, and how they can be put into practice. The protocol is the written document that allows the investigator to communicate details of how the research question will be answered. In the following article, the basic components of the research protocol are described. Issues related to quality control, data entry, and pilot testing are discussed. This is the third in a series of research fundamental concept papers, written by members of the SAEM Research Committee.     

Data monitoring committees and problems of lower-than-expected accrual or events rates

Data monitoring committees for randomized clinical trials must frequently decide what action, if any, is required for trials whose accrual has been slower than expected, or whose event rates have been less than expected. We discuss in this article some of the practical issues concerning modifying or closing such trials, including what data and analyses could be helpful to the data monitoring committee in their deliberations.     

Statistical approaches to trial durations in episodic affective illness

In light of the high variability in illness characteristics and patterns among patients with bipolar illness, parallel group designs present severe methodologic difficulties. Crossover, off-on-off-on (B-A-B-A), and other individualized designs may be a useful substitute, but no consensus exists about how to estimate the individual trial durations required in these instances. Several methods for determining optimum trial lengths in crossover designs are presented, illustrated, and discussed. These include: chi-square (chi2) for the expected versus observed number of either episodes or days well; exceeding two standard deviations for average duration of episodes or euthymic intervals; or the Sequential Probability Ratio Test (SPRT), which detects when mean values differ from prior statistical expectations. Each method was applied to three demonstration cases using data from actual clinical trials of three patients with different patterns of recurrent affective illness. Each method detected changes in illness severity, although different tests appeared to be sensitive to differing cycle patterns in the patients illustrated. We suggest that these types of analyses and others can be used as indicator statistics to augment global impressions and clinical judgment, and to assist in determining individualized trial durations, both in formal clinical trials and in clinical treatment settings. Once individual responsivity is confirmed with an appropriate interplay of trial design and statistical analysis, the percentage response in a given population can then be compared to other agents or in other populations. Moreover, meta-analytic techniques based on addition of z scores from individuals' effect sizes can then be used to assess overall significance of a drug effect in a given population or subpopulation. The need for further development of appropriate and alternate study designs and analysis methods for bipolar illness is highlighted. Approaches to estimating required trial durations in individuals with different cycle frequencies in crossover and B-A-B-A designs constitute one element of that exploration.     

Introducing Critical Appraisal to studies of animal models investigating novel therapies in sepsis

OBJECTIVES: To discuss theoretical and practical aspects relating to the design of animal studies investigating the efficacy of novel therapeutic agents for the treatment of sepsis, and to make explicit the process whereby these studies can be evaluated for the purpose of designing clinical trials in humans. DATA SOURCES: Relevant articles from the pertinent literature were reviewed. STUDY SELECTION: Studies relevant to an evidence-based assessment of clinical studies on therapeutic efficacy, and studies relevant to the design of animal models of sepsis were selected. DATA EXTRACTION: Concepts relevant to an evidence-based assessment of the animal literature were extracted. DATA SYNTHESIS: Articles were reviewed and an evidence-based framework for the assessment of animal studies was developed. In this process, we discuss the steps that are necessary to assess the internal validity of an individual study and review topics relevant to the application of animal data to the design of clinical trials. CONCLUSIONS: The success of clinical trials of sepsis therapies is predicated on the generation and interpretation of sound preclinical data. In this review, we have attempted to outline an evidence-based approach to the assessment of preclinical animal studies evaluating novel therapeutic interventions in sepsis.     

Basic principles of cost-benefit analysis in long-term treatment of risk factors

Health care decision makers are increasingly forced to identify and implement the options for potential spendings and savings. Historically, preventive measures for cardiovascular diseases have been scrutinised a great deal. The main reason for this was that substantial financial investments would have to be undertaken long before the clinical benefits were apparent. In the past, most economic evaluations of lipid lowering therapy have been based on models combining logistic regression risk functions from epidemiological cohort studies, such as the Framingham study, with the extent of cholesterol reduction. Recently, however, data from the large controlled outcome studies (4S, CARE, LIPID, WOSCOPS) have been reported which allow a direct estimate of the economic benefits of cholesterol reduction in primary and secondary prevention. The economic evaluation of such therapies can be performed in several ways, from relatively easy cost-consequence analyses to more complex cost-effectiveness analysis. Initial economic analyses of the trials are already available. There are, however, still considerable practical and methodological issues which have to be taken into account in assessing lipid lowering drugs. Available pharmaco-economic data on lipid lowering therapies suggest that statins are a cost-effective option in primary and secondary coronary prevention.     

Evaluation of coronary artery disease in women

Sepsis syndrome and septic shock remain significant causes of morbidity and mortality. To date, clinical trials of novel agents to treat sepsis have failed to demonstrate clinical efficacy despite considerable animal data to suggest a positive therapeutic benefit. This article reviews the recent major clinical trials on sepsis and discusses the hypotheses on which these therapies are based and the critical issues associated with clinical sepsis. Recommendations for future clinical trials on sepsis are made.     

Summary measures and statistics for comparison of quality of life in a clinical trial of cancer therapy

Assessment of health related quality of life (QOL) has become an important endpoint in many clinical trials of cancer therapy. Most of these studies entail multiple QOL scales that are assessed repeatedly over time. As a result, the problem of multiple comparisons is a primary analytic challenge with these trials. The use of summary measures and statistics both reduces the number of hypotheses tested and facilitates the interpretation of trial results where the primary question is 'Does the overall QOL differ between treatment arms?' I present two classes of summary measures that are sensitive to consistent trends in the same direction across multiple assessment times or multiple QOL scales. Missing data strongly influences the choice between the two classes, where one class handles missing data on an individual basis, while the other class uses model-based strategies. I present the results from a clinical trial of adjuvant therapy for breast cancer that use summary measures with a focus on the practical issues that affect these analysis strategies, such as missing data and integration of QOL with efficacy endpoints such as survival.     

Missing data in quality of life research in Eastern Cooperative Oncology Group (ECOG) clinical trials: problems and solutions

Incorporation of quality of life (QOL) investigation into Eastern Cooperative Oncology Group (ECOG) multi-centre clinical trials has led to innovative strategies for protocol design and high quality data collection. A scientific advisory committee reviews protocol design components, measurement selection, timing of assessments and compliance issues. Extensive educational programmes provide information about the scientific and clinical relevance of QOL protocols, as well as practical strategies for data collection and management. Compliance with QOL data collection standards is prospectively monitored and evaluated. Preliminary results from eight ECOG-run protocols found overall compliance to be approximately 85 per cent (94 per cent at baseline and 73 per cent during treatment). Selected patient and institutional factors were evaluated for their association with compliance.     

Kids count--a children''s rights project

Offers a protocol of antibacterial therapy of abdominal sepsis in surgical patients. Analyzes the principal etiological agents of abdominal septic complications. Presents optimal schemes of antibacterial therapy of abdominal sepsis. Pays special attention to choice of antibiotics in different clinical situations.     

Interim analyses and early termination of clinical trials

The group sequential tests (GST) are appropriate for performing interim analyses in clinical trials. Various GST are reviewed and compared in this paper in terms of the expected sample size, the maximum sample size, and other practical aspects. Also discussed are the p-values of the significant differences for GST. Common problems and difficulties of using GST in practice are examined. One problem is difficulties associated with the delayed data accumulated after a trial is terminated at an interim test. The GST with O'Brien-Fleming type of boundaries are found to be safe in dealing with delayed observations. Possible approaches for performing futility analyses are illustrated with examples. It is recommended that futility analysis with GST be built into the design of large clinical trials.     

Experience with a cross-study endpoint review committee for AIDS clinical trials. Terry Beirn Community Programs for Clinical Research on AIDS

OBJECTIVES: To describe the methods and results of a standardized system for clinical endpoint determination for defining and reviewing endpoints in clinical trials for HIV-infected individuals. DESIGN: A system was developed utilizing standard definitions for the 24 diagnoses or clinical events that serve as trial endpoints and together define the combined endpoint 'progression of HIV disease. A common set of case report forms were used for all trials. Thus, an event of Pneumocystis carinii pneumonia (PCP), for example, for a subject co-enrolled in an antiretroviral trial and a PCP prophylaxis trial was only reported once. METHODS: A central committee was established to define clinical events and review endpoints across all studies. Events were classified according to established criteria for confirmed, probable and possible levels of certainty. RESULTS: This report describes the methods used to ascertain and review endpoints, and summarized 2299 clinical events for 8097 subjects enrolled in one or more of nine clinical trials. Data on the diagnostic certainty of events and agreement between site clinicians and the endpoint committee are presented. CONCLUSIONS: Uniform classification of endpoints across AIDS clinical trials can be accomplished by multicenter, multitrial organizations with standardized definitions and review of endpoint documentation. Our experience suggests that nurse coordinators reviewing all submitted endpoints for every trial are warranted and the need for external review by a clinical events committee may depend on the type of trial conducted.     

The design and implementation of trials of host modulation agents

Over the last two decades it has become more evident that although oral microorganisms are essential agents of periodontal pathogenesis, interpatient variability in the host response is a major determinant of the expression of periodontal disease extent and severity. Data from animal models and human studies have identified many of the components of the host inflammatory response which serve as critical mediators of clinical inflammation, attachment loss, and bone resorption. Studies suggest that certain pharmacologic agents, which act at a molecular level to block specific inflammatory mediators, appear to attenuate disease progression. These promising findings herald a new era in periodontal medicine. Anti-infective therapies may soon be supplemented with anti-inflammatory pharmacological agents. However, there are many unanswered issues regarding formulation design, clinical application, potential indication claims, and clinical study design. Furthermore, current considerations of fundamental mechanisms of pathogenesis, as well as new data from epidemiologic studies emphasizing the multifactorial nature of disease, are changing the underlying assumptions which have served to guide our design of anti-infective drug trials over the last two decades. There are new questions regarding appropriate outcome measurements which are to be reconsidered. For example, the measurement of a change in periodontal disease status, either during progression or in response to therapy, is fundamentally unidimensional and may be only mildly informative when one considers that the disease is multifactorial by nature. Using an example from intensive care medicine, pathophysiologic studies of septic shock have demonstrated that the microbial dose and the host inflammatory mediator response are far better predictors of patient morbidity and mortality than any combination of clinical signs associated with clinical shock. Clinical trials of anti-cytokine and anti-inflammatory drugs to treat shock are now designed and conducted taking strategic advantage of this knowledge by including measurements of microbial dose and host response. It appears prudent that the design and implementation of clinical trials of host modulation agents also benefit from our current insights into pathogenesis and not represent a template-driven adaptation of historical, anti-infective clinical trial protocols.     

Pharmacokinetic considerations in the design of optimal chemotherapeutic regimens for the treatment of breast carcinoma: a comceptual approach

Pharmacological data are currently available for a number of antineoplastic agents which have shown clinical activity in advanced breast carcincoma. Preclinical data reveal a relationship between therapeutic response and certain pharmacokinetic parameters such as time of effective cytotoxic exposure (Teff) and the product of concentration with time (Cxt). We have attempted to apply human pharmacologic data to get estimates of these parameters for 6 active agents in breast cancer, to relate them to response rates, and to suggest a method for estimating the role of individual drugs in a multidrug combination. The response rates for 6 single agents were obtained from literature review and related to estimates of Teff and Cxt. The Cxt-response relations for single drugs were linear for cyclophosphamide, 5-fluorouracil, and thiotepa; exponential for vincristine; and relatively flat for methotrexate and cytoxine arabinoside. Most Teff values for the active single agents clustered about 15 hr/dose. From the graphs of response rate vs Cxt, the individual contribution of each agent in a combination study was estimated to arrive at a predicted response rate. The predicted response rates for the combination studies correlated with the actual response rates determined in the clinical study, for 6 of 6 nonrandomized studies and for 12 of 14 randomized studies analyzed. In 2 studies, deviations from the predicted response rate were attributed to differences in study design or analysis. There was no correlation between Teff and predicted response rate. Analyses of pharmacokinetic data may be useful to simulate combination chemotherapy studies to predict the effectiveness of clinical trials in breast cancer. Since the pharmacologic data were not obtained for any of the agents in the actual clinical trials done, we can only speculate on the usefulness of this method. We would encourage the prospective collection of this data in future clinical trails.     

General issues of study design and analysis in the use of biomarkers in cancer epidemiology

Other contributions to this volume have discussed sources of variation (see Vineis) and measurement error (see White). In this article, we focus on statistical issues involved in the design and analysis of epidemiological studies that use biomarkers. We do not consider statistical issues of laboratory analyses.