Search

Several methods are available for analyzing the results of caries trials when response to treatment is measured on an ordered categorical scale, such as Grainger 's Severity Index and modifications of it. These methods are applied to data from a randomized clinical caries trial. As the methods become more informative, they also become more difficult computationally. The phenomenon of reversals and the selection of patients for study are also discussed.

We examined the relationships among ventricular arrhythmias, left ventricular dysfunction, and mortality after the occurrence of myocardial infarction in 766 patients who enrolled in a nine-hospital study and underwent two special tests. Frequency and repetitiveness of ventricular premature depolarizations (VPDs) were determined by computer analysis of predischarge 24 hr electrocardiographic recordings. The left ventricular ejection fraction (LVEF) was determined by radionuclide ventriculography and dichotomized at its optimal value of 30%. Frequency of VPDs was divided into three categories: (1) less than one per hour, (2) one to 2.9 per hour, and (3) three or more per hour. Repetitiveness of VPDs was also divided into three categories: (1) no repetitive VPDs, (2) paired VPDs, and (3) VPD runs. These variables were related, one at a time and jointly, to total mortality and to deaths caused by arrhythmias. The hazard ratios for dying in the higher or highest risk stratum vs the lower or lowest stratum for each variable (adjusted for the effects of the others) were: LVEF below 30%, 3.5; VPD runs, 1.9; and VPD frequency of three or more per hour, 2.0. There were no significant interactions among the three variables with respect to effects on the risk of mortality. There was a suggestion of an interaction between each risk variable and time after infarction. LVEF below 30% was a better predictor of early mortality (less than 6 months) and the presence of ventricular arrhythmias was a better predictor of late mortality (after 6 months).(ABSTRACT TRUNCATED AT 250 WORDS)

Plasma dopamine beta hydroxylase (DBH) activity was determined in MZ twins and in siblings of both sexes. The enzyme activities were found to be virtually identical in MZ twins. DBH activity was less similar in same-sex sibs, but still significantly correlated. These results indicate that plasma DBH activity is genetically determined.

Reliability coefficients often take the form of intraclass correlation coefficients. In this article, guidelines are given for choosing among six different forms of the intraclass correlation for reliability studies in which n target are rated by k judges. Relevant to the choice of the coefficient are the appropriate statistical model for the reliability and the application to be made of the reliability results. Confidence intervals for each of the forms are reviewed.

One hundred geriatric psychiatric patients were examined with the Geriatric Mental State Schedule in New York and London, and a correlation procedure involving both clinical and statistical operations was carried out on the psychopathological data thus collected. Twenty-one factors were produced, including three dealing with cognitive impairment. Although it was found that elderly depressives show a profile of psychopathology quite different from that shown by patients with organic disorder, it was also found that patients with an apparently functional disorder may sometimes be diagnosed as an organic disorder, that subjective complaints of intellectual impairment are not good indicators of organic disorders, and may be associated with a depressive factor, and that complaints that could be dismissed as attributes of aging may actually be indicative of a depressive disorder in the elderly. The methodological implications, as well as the limitations of the sample size, are discussed.

A simple approximation is provided to the formula for the sample sizes needed to detect a difference between two binomial probabilities with specified significance level and power. The formula for equal sample sizes was derived by Casagrande, Pike and Smith (1978, Biometrics 34 , 483-486) and can be easily generalized to the case of unequal sample sizes. It is shown that over fairly wide ranges of parameter values and ratios of sample sizes, the percentage error which results from using the approximation is no greater than 1%. The approximation is especially useful for the inverse problem of estimating power when the sample sizes are given.

Seventy-one heavy smokers who had failed in previous attempts to stop smoking participated in a randomized clinical trial to test the efficacy of clonidine as an aid in smoking cessation. The success rate in clonidine-treated subjects (verified by serum cotinine concentration) was more than twice that in the placebo-treated subjects. When the data were stratified by gender, a strong effect present in women was not apparent in men. After six months, cessation rates remained significantly higher among smokers treated with clonidine than those receiving placebo. The data also revealed an unexpectedly high prevalence (61%) of a history of major depression in this sample and a significant negative effect of such a history on cessation regardless of treatment. These findings, highly suggestive of an important role of clonidine in smoking cessation, warrant further studies to establish the long-term (greater than or equal to 12 months) efficacy of this drug and to replicate the association between nicotine dependence and depression.

Because the clinics in a multiclinic randomized clinical trial represent neither fixed stratification effects nor random classificatory effects, the appropriate analysis of data from such a trial has been the subject of controversy and debate. The following are some of the elements of controversy that are discussed and for which some bases for resolution are proposed. Is it ever valid to ignore the effects of clinics in the analysis? Is it ever valid to drop clinics from the analysis? Is a multiclinic clinical trial similar in structure or not to a single-clinic clinical trial in which patients have been stratified on a classificatory factor? Assuming that clinics will be taken account of in the analysis, should it be the weighted or the unweighted average of within-clinic treatment differences that is to be taken as the best estimate of the overall difference between the treatments? How should the data be analyzed if there is evidence of treatment-by-clinic interaction?

Distributions of seasonal average levels of sulfur dioxide and smoke-shade over the five boroughs of New York City are described in terms of the distance between stations and the heights of the stations above ground level and sea level. An assessment of the suitability of the data for epidemiologic studies of acute and chronic health effects of air pollution is made.

Some statistical techniques for analyzing the kinds of studies typically reported in Circulation Research are described. Particular emphasis is given to the comparison of means from more than two populations, the joint effect of several experimentally controlled variables, and the analysis of studies with repeated measurements on the same experimental units.

Psychiatric illness in spouses of patients with primary affective disorder was determined and compared to psychiatric illness in spouses of a nonpsychiatrically ill control group. An increase in affective illness in wives of bipolar male patients with affective disorder was found. There was no increase in affective illness among husbands of female patients. Marital status of these patients was evaluated and the percentages of patients who had never married or who had married but had ever been divorced or separated were similar to control data. Several of the marriages were quite stable over long time periods in spite of the severe recurrent affective illness experienced by these patients.

At least a dozen indexes have been proposed for measuring agreement between two judges on a categorical scale. Using the binary (positive-negative) case as a model, this paper presents and critically evaluates some of these proposed measures. The importance of correcting for chance-expected agreement is emphasized, and identities with intraclass correlation coefficients are pointed out.

The effects of random errors of measurement on a variety of basic multivariate procedures are discussed. For some procedures, the effect is one of attenuation (i.e., simply reducing the estimated strength of association). For others, the effect may be to alter the apparent direction of association. Some proposed solutions to the problems caused by errors of measurement are presented, and areas requiring further research are pointed out.

The Global Assessment Scale (GAS) is a rating scale for evaluating the overall functioning of a subject during a specified time period on a continuum from psychological or psychiatric sickness to health. In five studies encompassing the range of population to which measures of overall severity of illness are likely to be applied, the GAS was found to have good reliability. GAS ratings were found to have a greater sensitivity to change over time than did other ratings of overall severity or specific symptom dimensions. Former inpatients in the community with a GAS rating below 40 had a higher probability of readmission to the hospital than did patients with higher GAS scores. The relative simplicity, reliability, and validity of the GAS suggests that it would be useful in a wide variety of clinical and research settings.

A 24-hour continuous electrocardiographic recording was made 11 +/- 3 days after acute myocardial infarction (AMI) in 820 of the 867 participants in the Multicenter Post-Infarction Program. Ninety patients (11%) had unsustained ventricular tachycardia (VT) and 2 (0.2%) had sustained VT (more than 15 seconds). In 53 of the 92 patients (58%) with VT, only 1 episode of VT was in the recording. In 26 patients the longest episode of VT was 3 consecutive complexes (28%); in 56 patients (61%) it was 4 to 10 complexes; and in only 10 patients (11%) were there more than 10 consecutive complexes. The average rate of VT was 121 +/- 34 complexes per minute (range 75 to 240). Most episodes of VT started well after the T wave. Occurrence of VT was strongly related to the frequency of ventricular premature complexes in the 24-hour recording; 46% of the patients with at least 100 ventricular premature complexes per hour had VT. The 92 patients who had VT were compared to the 728 who did not with respect to relevant clinical characteristics. Several variables were significantly (p less than 0.05) more common in the VT group: age older than 60 years, previous AMI, history of angina pectoris, occurrence of VT or ventricular fibrillation in the coronary care unit, left ventricular ejection fraction less than 30%, rales greater than bibasilar in the coronary care unit, and use of antiarrhythmic drugs, digitalis or diuretics at the time of discharge from hospital. Based on Kaplan-Meier survival curves, the cumulative probability of surviving 3 years was 0.67 for patients with VT and 0.85 for patients without VT (p less than 0.001). There were no statistically significant associations between individual VT characteristics and mortality. However, patients with longer runs of VT tended to have a higher mortality rate, and both patients with sustained VT died in the first month after the index infarct. VT had a strong and statistically significant association (p less than 0.05) with all-cause and arrhythmic mortality independent of other risk variables that were associated with VT. Adjusted for other risk indicators, VT nearly doubled the risk of dying during an average follow-up of 31 months.

Data from non-rapid-cycling bipolar (manic-depressive) patients who were receiving long-term treatment with lithium carbonate were analysed by the life table method to determine when lithium carbonate prophylaxis failures occurred. Forty-four of 96 patients failed to keep well in spite of maintenance lithium therapy. The analysis revealed an early, rapid failure rate during the first six months of treatment, which was followed by a slower rate of failure. Several clinical factors were assessed to determine if any of them predicted which patient would experience their initial failure in the early or late interval, but none of these factors, which included age, sex, age of onset, rate of affective attacks, family history, and the nature of the preceding episode, were found to have any predictive value regarding lithium prophylaxis failure. We found, however, that patients who had early failures tended to have a subsequent early failure in spite of continued maintenance with lithium carbonate.

The life table is presented as the method of choice for analyzing data from longitudinal studies in which the outcome under study occurs randomly and in which patients are followed up varying lengths of time. We discuss the superiority of the life table to methods typically used, the calculation of its entries, and some of the clinical uses that can be made of its results. The method is applied to follow-up data on manic-depressive patients maintained with prophylactic lithium carbonate or with control regimens, and it is shown to disclose mathematical regularities in the parameters of longitudinal course.

Plasma dopamine beta hydroxylase (DBH) activity was measured in patients with affective disorders and in their relatives. The groups studied had wide distributions of values for plasma DBH activity. No significant difference of plasma DBH activity was found between unipolar and bipolar patients, nor between patients given lithium or placebo. Exercise on a treadmill at 40 degrees or 10 degrees C elicited a different pattern of response for plasma DBH activity in three patients as compared to control subjects. In familial studies we found the values of plasma DBH activity to be almost identical in monozygotic twin pairs and quite similar in dizygotic twin pairs. All pairs, however, were discordant for affective illness. There was also a marked similarity of plasma DBH activity in 15 pairs of the same sex sibs discordant for affective illness. These studies suggest that the resting level of plasma DBH activity is not related to affective illness but is genetically determined.

Data from the Veterans Administration Dental Longitudinal Study of 736 men each examined four times were analyzed to determine how representative the six "Ramfjord Teeth" (RT) were of the rest of the dentition for epidemiologic studies of gingivitis and periodontitis. Correlation coefficients were calculated between scores for gingival inflammation generated by the RT and scores for the teeth in the rest of the mouth; the correlation coefficients between the two scores tended to be high. Sensitivity, specificity and predictive values were calculated for periodontal pathology classified on the basis of the pocket depths of the six RT. Reliance on the RT resulted in systematic underestimation of the prevalence and incidence of deep periodontal pockets. The RT are more than adequately representative of the rest of the dentition for epidemiologic studies of gingivitis, but are inadequate surrogates of the rest of the mouth for epidemiologic studies of periodontitis.

It is most important that the proper design of a clinical trial be developed prior to the start of the study. Such a design is one that will not tend to find a treatment to be effective when it really is ineffective, and will not tend to miss a good treatment when it really is effective. The questions to be answered by a clinical trial must be decided in advance and will help determine the design. Trials should take into consideration how the agent is to be used: carefully controlled, ordinary use, subject or professionally applied, etc. The double-blind parallel randomized trial is preferred to the double blind cross-over design, except possibly for a professionally applied treatment which does not leach into the saliva or affect the oral microflora. Blocking or stratification tends to increase the efficiency of a clinical trial, and also may produce more information on the agent's effectiveness. Stratification is preferred to blocking, particularly when subjects enter the trial serially over time. Subjects should be stratified by age, sex and baseline level of plaque or gingivitis. Stratification permits a check on whether the effect of treatment is atypically strong or weak for certain types of patients. Where possible, the same examiner should examine the same patients throughout the study. Examiners should be self-calibrated to reduce intra-examiner error. Plaque and gingivitis trials should be at least of 6 months duration and may be preventive or therapeutic in nature. The preferred statistical analysis when 2 treatments are being compared is a comparison of mean changes in plaque or gingivitis scores by the independent sample t-test. An analysis of covariance using the baseline scores as the covariate may improve power somewhat. When pairwise t-tests are performed in comparing several treatments, multiple comparison criteria are indicated. Parametric rather than nonparametric tests are preferred since they usually have more power. The square root of whole mouth means in gingivitis clinical trials is the preferred transformation to obtain greater power and to produce approximately normal distributions.

A prospective study to determine prognostic factors for risk stratification in 867 patients surviving the coronary care unit phase of acute myocardial infarction (AMI) is reported. During a 48-month follow-up, 144 patients (17%) died. The deaths were examined for the chronology, cause, mechanism, location and presence of myocardial ischemia in the terminal event. A classification previously proposed by Hinkle and Thaler was used to define the mechanism of cardiac death and the presence of ischemia. There were 113 deaths due to coronary atherosclerotic coronary artery disease, including 5 due to complications of coronary artery bypass graft surgery. Of the remaining 108 of these deaths, 74% were classified as due to an arrhythmic mechanism and 26% as myocardial failure. Of the deaths due to an arrhythmia or to myocardial failure, 56 (52%) occurred out of hospital. The ratio of arrhythmic: myocardial failure deaths was not different for the patients who died within 3 months after the index AMI compared with later deaths. Sudden death (less than or equal to 1 hour of new symptoms) was strongly associated with arrhythmic death but 32 (54%) of patients who died greater than 1 hour after the onset of symptoms were also classified as having an arrhythmic cause of death. Previously described risk factors, including an ejection fraction less than 0.40 and greater than or equal to 10 ventricular premature complexes/hour, were independent predictors of mortality but did not differentially predict the mechanism of cardiac death. Evidence of myocardial ischemia before the terminal event was found in about 50 (60%) patients whose deaths were witnessed and who died from an arrhythmia or myocardial failure.

Logistic analysis methodology is being increasingly used in the analysis of categorical data from the biomedical sciences. We describe how the relationships between a dichotomous outcome variable and several qualitative classificatory or experimental variables can be modeled in a format similar to that used for quantitative data. The technique of weighted least squares is used to estimate the underlying parameters in the model and to perform various tests of statistical significance. The methodology is illustrated using data from a randomized dental anaesthetic study comparing three local anaesthetic solutions under two operative procedures. The similarities between the logistic analysis of dichotomous data and the analysis of variance applied to quantitative data are stressed.

Ridit analysis is presented as an appropriate method of analyzing dental clinical data which fall somewhere between the purely categorical (e.g., improved vs. not improved) and the bona fide quantitative (e.g., mg % salivary calcium) scales of measurement. The key feature of the method is the estimation of the probability that a randomly-selected patient from one treatment group is "better-off" than a randomly-selected patient from another. Methods are presented for testing statistical significance and constructing confidence intervals. The method is illustrated on data from a comparative clinical trial of ibuprofen, aspirin and placebo in the relief of post-extraction pain. There were no significant differences in efficacy among the active treatments, but each was significantly superior to placebo.

A simple performance test was developed and used to evaluate a sample of psychiatric geriatric patients on their capacity for self-care. Evaluations were made by asking the patient to demonstrate certain essential activities of daily living. The objective test, compared with patients' subjective reports, was found to be a more valid measuring instrument of functional impairment and more useful as a diagnostic and prognostic tool Test scores correlated with independent clinical assessments of mental status and physical conditions, as well as with short-term outcome of hospitalization. The scale is easy and quick to administer and has many applications for the treatment and management planning of patients.

Most previous studies of long-term maintenance therapy for bipolar illness simply compared proportions of patients experiencing recurrences. This method is inadequate because the length of time until a recurrence is not taken into account and because data from patients who withdraw prematurely from the study without a recurrence are either ignored or analyzed improperly. More appropriate survival analytic techniques were applied to the data from the National Institute of Mental Health Collaborative Study of bipolar patients treated with lithium carbonate, imipramine hydrochloride, or both. An interaction between treatment and the nature of the index episode (the episode that brought the patients into the study) was found. In patients with manic index episodes, both lithium and the combination were superior to imipramine. In patients with depressive index episodes, the combination was significantly superior to imipramine, whereas lithium was indistinguishable from imipramine. The latter finding differs from the original analysis, which found the combination to be no different from the other two treatments for patients with a depressive index episode.

Replicate measurements of probing depth and attachment level were made by three examiners on the six Ramfjord teeth of 34 patients. Within-mouth correlations were high for single measurements, and yet higher for means across replicate measurements. Inter- and intra-examiner variability were both found to be low relative to between-patient variability. Conclusions are that the low within-mouth correlations reported by others may be the result of poor reliability, and that it is inappropriate to take the individual site as the unit of statistical analysis in clinical studies of periodontitis.