■ Are the results such that they would alter patient management in the presence of associated factors reported? Harm Some reports focus on adverse effects or events related to a given exposure or intervention A specific example might be a report addressing the question, “What is the risk of congenital cardiac disease in offspring of mothers taking angiotensin receptor blocker medications during pregnancy?” In this case, the most efficient design is often a case-control study, in which a group of pregnant mothers were enrolled who were taking the medication and were matched on key characteristics to a group of mothers who were taking no medications The critical appraisal of studies of harm proceeds as follows: Was the study conducted in manner that minimized bias? ■ Was a comparison group without the exposure or factor of interest clearly identified and similar regarding other important factors that might influence outcome? ■ Were assessments of both the exposure or factor and the relevant outcomes conducted and interpreted in an identical manner irrespective of the exposure status of the subject? ■ Was follow-up sufficiently long and complete to ensure development and ascertainment of outcomes? ■ Was the study design such that there was a correct temporal relationship between the exposure and subsequent development of the outcome? ■ Is there evidence a dose-response relationship exists between the magnitude of the exposure or factor and the subsequent development and magnitude of the outcome? Are the results reported and interpreted appropriately? ■ Are the measures of association reported appropriate for the study design and the level of measurement of the exposure or factor and the outcome? ■ Are confidence intervals provided for all measures of association? Are the results of sufficient magnitude and relevant? ■ Is the observed association strong? ■ Are observed measures of association sufficiently reliable? Are the findings and implications of the report applicable to my own clinical practice? ■ Do my patients meet the entry criterions described in the report, and do they have similar characteristics to the study subjects? ■ Are the results such that they provide valid and important information regarding counseling about prognosis for my patients should they have the exposure or factor? ■ Is the association of sufficient magnitude that it would alter management? Diagnosis The clinical practice of those involved with pediatric and congenital cardiac disease relies heavily on diagnostic testing A good diagnostic test is valid, reliable, quick, safe, simple, associated with minimal discomfort, inexpensive, and ethical Studies about optimal and novel methods for making measurements are common Often the measures in these studies are compared with an accepted methodology Other studies aim to look at the contribution of new tests to the definition of clinical conditions and their prognosis or response to treatments, such as defining the relationship of brain natriuretic peptide levels with clinical markers of heart failure The classic study of a diagnostic test is one designed to determine the performance characteristics of a given test in making or excluding a specific diagnosis The methodology also applies to studies about screening We are interested in the accuracy of the methods themselves, particularly if they require subjective assessment We wish to know features such as test-retest reliability and intra- and interobserver variability Test-retest reliability is the amount of variation in pairs of measurements performed in the same individuals under the same conditions by the same observers but at different times It represents variation of the test itself Intraobserver variability is the amount of variation when the same observer performs or interprets the same test in the same individual at the same point in time Interobserver variability is the amount of variation when two or more different observers perform or interpret the same test in the same individual at the same point in time Each of these is defined by determining the degree and pattern of agreement, not correlation or association In relation to defining or excluding a diagnosis, we wish to know the performance characteristics of the test itself in both subjects with and without the condition of interest Sensitivity is the ability of the test to accurately identify those subjects with the disease, or the true positive rate It is the proportion of subjects who test positive and have the disease (the numerator, true positives) among all subjects who have the disease (the denominator) If the subject had the disease but the test was negative, he or she is referred to as a false negative and would have the implications of having a missed disease Specificity is the ability ... ■ Is the association of sufficient magnitude that it would alter management? Diagnosis The clinical practice of those involved with pediatric and congenital cardiac disease relies heavily on diagnostic testing A good diagnostic test is valid,