Search

Much of the statistical literature regarding categorical data focuses on the odds ratio, yet in many epidemiological and clinical trial settings, the relative risk is the quantity of interest. Recently, Spiegelman and Hertz- mark illustrated modeling and SAS programming for modeling relative risk in contrast to the logistic model's odds ratio. The focus of their work is on a single relative risk, i.e., for one binary response variable. Herein, we outline two methods for estimating relative risks for two correlated binary outcomes. The first method is weighted least squares estimation for categor- ical data modeling. The second method is based on generalized estimating equations. The two methods are readily implemented using common statis- tical packages, such as SAS. The methods are illustrated using clinical trial data examining the relative risks of nausea and vomiting for two different drugs commonly used to provide general anesthesia.

Background and Purpose— Interventional Management of Stroke (IMS) III is a randomized, parallel arm trial comparing the approach of intravenous tissue-type plasminogen activator followed by endovascular treatment with intravenous tissue-type plasminogen activator alone in patients with acute ischemic stroke presenting Methods— Conditional power was defined as the likelihood of finding statistical significance at the end of the study, given the accumulated data to date and with the assumption that a minimum hypothesized difference of 10% truly exists between the 2 groups. The evolution of study data leading to futility determination is described, including the interaction between the unblinded study statisticians and the Data and Safety Monitoring Board in the complex deliberation of analysis results. Results— The futility boundary was crossed at the trial’s fourth interim analysis. At this point, based on the conditional power criteria, the Data and Safety Monitoring Board recommended termination of the trial. Conclusions— Even in spite of prespecified interim analysis boundaries, interim looks at data pose challenges in interpretation and decision making, underscoring the importance of objective stopping criteria. Clinical Trial Registration— URL: http://www.clinicaltrials.gov . Unique identifier: NCT00359424.

Mixed-effects regression models have extraordinary capacity to capture the essence of complex longitudinal data. In this paper we present analyses from a longitudinal study on 157 patients with osteoarthritis (OA) where real-time momentary self-reported pain and physical activity level were collected. We illustrate the application of heterogeneous mixed effect (HME) models to examine the effect of time-varying activity level assessments (sedentary vs. moderate/vigorous) during the 30 min prior to the pain recording on mean pain levels and variability. We also examine the variation in pain associated with activity level and the degree that participant characteristics (weight-bearing vs. non-weight bearing joint w/OA, body mass index, white vs. non-white race, gender) and psychosocial factors (mood, tension, pain catastrophizing) influence pain variation. We fit separate HME models for each characteristic or factor. These HME models are fit with standard statistical software, but there are challenges to fitting and interpreting these models. Once fit the models provide a powerful tool to help guide researchers in understanding highly individualized pain trajectories.

Missing data, e.g. patient attrition, are endemic in sleep disorder clinical trials. Common approaches for dealing with this situation include complete-case analysis (CCA) and last observation carried forward (LOCF). Although these methods are simple to implement, they are deeply flawed in that they may introduce bias and underestimate uncertainty, leading to erroneous conclusions. There are alternative principled approaches, however, that are available in statistical software namely mixed-effects models and multiple imputation. In this paper we introduce terminology used to describe different assumptions about missing data. We emphasize that understanding reasons for missingness is a critical step in the analysis process. We describe and implement both linear mixed-effects models and an inclusive multiple imputation strategy for handling missing data in a randomized trial examining sleep outcomes. These principled strategies are compared with "complete-case analysis" and LOCF. These analyses illustrate that methodologies for accommodating missing data can produce different results in both direction and strength of treatment effects. Our goal is for this paper to serve as a guide to sleep disorder clinical trial researchers on how to utilize principled methods for incomplete data in their trial analyses.

Analysis of longitudinal data is often complicated by the presence of informative right censoring. This type of censoring should be accounted for in the analysis so that valid slope estimates are attained. In this study, we developed a new likelihood-based approach wherein the likelihood function is integrated over random effects to obtain a marginal likelihood function. Maximum likelihood estimates for the population slope were acquired by direct maximisation of the marginal likelihood function and empirical Bayes estimates for the individual slopes were generated using Gaussian quadrature. The performance of the model was assessed using the geometric and Poisson distributions to model the number of observations for every individual subject. Our model generated valid estimates for the slopes under both distributions with minimal bias and mean squared errors. Our sensitivity analysis confirmed the robustness of the model to assumptions pertaining to the underlying distribution and demonstrated its insensitivity to normality assumptions. Moreover, superiority of the model in terms of accuracy of slope estimates was consistently shown across the different levels of censoring in comparison to the naïve and bootstrap approaches. This model was illustrated using the cohort of renal transplant patients and estimates of the slopes that are adjusted for informative right censoring were acquired.

Background and Purpose: Ischemic stroke lesion volumes have proven difficult to analyze due to the extremely skewed shape of their underlying distribution. We introduce an extension of generalized linear models, beta regression, as a possible method of modeling extremely skewed distributions as evidenced in ischemic stroke lesion volumes. Methods: The NINDS rt-PA clinical trials measured ischemic stroke lesion volume as a secondary trial outcome. Three-month lesion volumes from these trials were analyzed using beta regression. A multi-variable regression model associating explanatory variables with ischemic stroke lesion volumes was constructed using accepted model building strategies and compared with the previously published volumetric analysis. Results: Beta regression produced a similar model when compared to the previous analysis published by the study group. All previously identified variables of importance were detected in the model building process. The age by treatment interaction described in previous studies was also found in this analysis, confirming the strong effect age has on stroke outcomes. Further, a treatment effect was elicited in terms of odds ratios, yielding a previously unknown quantification of the effect of rt-PA on lesion volumes. Conclusions: Beta regression proved adept in modeling ischemic stroke lesions and offered the interpretation of covariates in terms of odds ratios. Beta regression is seen as a legitimate alternative to analyze ischemic stroke volumes.

The notion that eating chocolate would prevent or treat hypertension is appealing to many who produce and enjoy chocolate. Several studies have documented beneficial effects of dark chocolate on insulin action and endothelial function. However, the published studies on chocolate and blood pressure include a relatively small number of subjects, and results are conflicting. In addition, because of secrecy surrounding the production of chocolate and the unique sociocultural context of this popular food, research on efficacy and effectiveness is complex. This commentary summarizes 13 peer-reviewed studies on dark chocolate and blood pressure and raises questions relevant to its future as an evidence-based lifestyle intervention.

"Loss to follow-up" can be substantial in substance abuse clinical trials. When extensive losses to follow-up occur, one must cautiously analyze and interpret the findings of a research study. Aims of this project were to introduce the types of missing data mechanisms and describe several methods for analyzing data with loss to follow-up. Furthermore, a simulation study compared Type I error and power of several methods when missing data amount and mechanism varies. Methods compared were the following: Last observation carried forward (LOCF), multiple imputation (MI), modified stratified summary statistics (SSS), and mixed effects models. Results demonstrated nominal Type I error for all methods; power was high for all methods except LOCF. Mixed effect model, modified SSS, and MI are generally recommended for use; however, many methods require that the data are missing at random or missing completely at random (i.e., "ignorable"). If the missing data are presumed to be nonignorable, a sensitivity analysis is recommended.

PURPOSE Recent evidence supports a signifi cant association between the intake of dietary fi ber and levels of infl ammatory markers. The objective of this study was to determine whether daily fi ber supplementation would reduce levels of infl ammatory markers. METHODS This study was a prospective randomized controlled trial at a single university medical center. Participants were overweight or obese adults with no history of heart disease. The intervention was psyllium supplementation at either 7 or 14 g/d for 3 months compared with no supplements in a control group. The main outcome measure was change in level of high-sensitivity C-reactive protein (hsCRP) concentration; secondary outcomes included changes in interleukin-6 (IL- 6) levels, fi brinogen levels, and white blood cell (WBC) count. Protocol completers attended at least 2 visits and took more than 75% of the prescribed fi ber dose. RESULTS In this intent-to-treat analysis (n = 158), there were no signifi cant dif- ferences between either of the 2 treatment groups and the control group in the amount of change in CRP, fi brinogen, or IL-6 levels or in WBC count (P >.05). In the analysis of protocol completers (n = 132), there also were no signifi cant dif- ferences between the groups except for a small decrease in fi brinogen level in the high-fi ber group (-6 mg/dL (-0.18 μmol/L) compared with 13 mg/dL (0.38 μmol/L) in the control group, P

We analyzed cardiovascular disease mortality risks associated with obesity using participant-level meta-analysis of data from the Black Pooling Project for Black and White individuals. The adjusted relative risks (ARRs) were stronger among White participants than among Black participants for coronary heart disease AAR=1.21 (95% confidence interval [CI]=1.07, 1.36) versus 0.87 (95% CI=0.69, 1.09), respectively, and cardiovascular disease ARR=1.18 (95% CI=1.07, 1.29) versus 0.91 (95% CI=0.77, 1.05), repectively. The results suggest that obesity is an independent risk factor in White people, and additional study of body size and disease progression is necessary in the assessment of racial disparities.

We sought to analyze the self-reported hospitalization, emergency department visits, and outpatient visits of Persian Gulf War (deployed; n=1896) and Persian Gulf War–era (nondeployed; n=1799) military personnel 5 years postconflict to determine whether these groups had different rates of health care use. Compared with personnel who had not been deployed, personnel who had been deployed were more likely to have visited an emergency department (25% vs 21%; odds ratio [OR]=1.24; 95% confidence interval [CI] = 1.06, 1.51]). Among these groups, the National Guard and Reserve personnel were more likely to have been hospitalized than were the regular military personnel (OR= 1.65; 95% CI=1.21, 2.26).

Background. Teaching cancer prevention and detection is important in health professional education. It is desirable to select a comprehensive framework for teaching oral cancer (OC) prevention and detection skills. Methods. The PRECEDE-PROCEED model was used to design a randomized pretest and posttest study of the OC prevention and detection skills of dental students (n = 104). OC knowledge, opinions, and competencies were evaluated. Results. Second year students in the intervention group were more competent than those in the control group. Conclusions. The novel use of PRECEDE-PROCEED sets a precedent for designing a standardized OC curriculum for a wide range of health professional disciplines. J Cancer Educ. 2007;22:250–253.

Our study sought to 1) determine if deployment status is associated with chronic widespread pain (CWP), and 2) evaluate whether veterans with CWP have greater psychiatric comorbidity, higher health care utilization, and poorer health status than veterans without CWP. Five years after the conclu- sion of the first Gulf War (August 1990 to June 1991), we conducted a cross-sectional study of veterans who listed Iowa as the home of record using a stratified sampling design to determine their health status. We compared the prevalence of CWP between deployed and nondeployed veterans. Logistic and multiple linear regression models were constructed to test whether CWP was associated with comorbidities and health-related outcomes of interest. Five hundred ninety of 3695 veterans interviewed (16%) had CWP. Gulf deployment was associated with higher prevalence of CWP than deployment elsewhere (OR 2.03, 95%CI 1.60-2.58), after adjustment. Both deployed and nondeployed veterans with CWP reported more health care utilization and comorbidities and lower health-related quality of life scores than veterans without CWP. Deployed veterans were more likely to have CWP than nondeployed veterans, and CWP was associated with poor health outcomes. Military and medical personnel should be aware that efforts to prevent, identify, and treat CWP in veterans returning from the current war may be needed. Perspective: This article indicates that deployed veterans may have an increased risk for develop- ment of CWP, which is associated with greater healthcare utilization and comorbidity and lower quality of life. The risk of poor health outcomes suggests that veterans returning from the present conflict should be screened for CWP on their return. © 2007 by the American Pain Society

Background. Nephrogenic systemic fibrosis is a debilitating disease occurring exclusively in patients with renal failure. The aetiology of nephrogenic systemic fibrosis is unclear, but recent reports suggest that exposure to gadolinium for enhancement of magnetic resonance imaging may play a role. In the present study, we assessed the association of exposure to gadolinium with the development of nephrogenic systemic fibrosis in patients with various stages of chronic kidney disease. Methods. We analysed the exposure to gadolinium and development of nephrogenic systemic fibrosis in 849 patients on renal replacement therapy over 5 years. We also performed inquiry of development of the nephrogenic systemic fibrosis in 592 patients exposed to gadolinium and estimated to be in stages 3 and 4 of chronic kidney disease. Results. In 849 patients undergoing chronic dialysis from 2001 through 2006 time period, four of the 261 who had received gadolinium (1.5%) and none of the 588 not exposed to gadolinium developed clinically apparent disease. The odds ratio for developing nephrogenic systemic fibrosis was 6.671 [95% confidence interval (CI) 1.537–53.97] in patients with a single gadolinium exposure compared to patients without gadolinium exposure. This ratio increased to 44.5 (95% CI 2.362–2913) in patients with multiple gadolinium exposures compared to patients not receiving gadolinium. None of the 592 patients estimated to be in stage 3 or 4 of chronic kidney disease developed nephrogenic systemic fibrosis after exposure to gadolinium. Conclusion. Gadolinium exposure is associated with nephrogenic systemic fibrosis in patients on chronic renal replacement therapy at a low rate. This association appears to increase with repeated exposure to gadolinium. Since nephrogenic systemic fibrosis may be clinically occult, its prevalence may be higher than reported. Despite this association, it is unclear if gadolinium is the sole or most important factor in the pathogenesis of the disease.

We address the problem of tests of homogeneity in two-way contingency tables in case-control studies when the case category is subdivided into k subcategories. In this situation, we have two cells with large frequencies and 2 X k cells with frequencies that become small as k increases. We propose two ad hoc statistics in which a statistic for the sparse cells is combined with a statistic for the cells with large frequencies. We will study these tests along with the Pearson test (using a chi-square approximation) in a Monte Carlo simulation study. Two sets of null hypothesis models and two sets of alternative hypothesis models are considered. The best test for the models considered is the usual Pearson test (using an approximate chi-square distribution) although the ad hoc models are more powerful under one alternative model considered.

When evaluating potential therapies for acute or progressing ischemic stroke, we should discard our preconceived notions about either efficacy or lack of efficacy of individual interventions. With carefully designed and performed clinical trials, we have the opportunity to change the present feeling of helplessness when it comes to the acute management of patients with ischemic stroke. The alternative to expensive, time-consuming, controlled clinical trials of promising therapies is to continue to treat on the basis of opinions or bias.

Purpose: To examine the patterns of coexisting (comorbid) mental disorders and whether comorbidity influences quality of life ratings in a sample of U.S. veterans. Patients and Methods: The Iowa Gulf War Study Case Validation study evaluated 602 military personnel, two-thirds of whom had symptoms of depression, cognitive dysfunction, or chronic widespread pain, who were activated or on active duty sometime during the first Gulf War (GW). Mental health disorders were defined using the SCID-IV, and the Health Utilities Index Mark 3 (HUI3) was used to measure health-related quality of life (HRQoL). Comorbidity was defined as having two or more mental disorders that spanned across at least two separate categories (e.g., depressive disorders and anxiety disorders). Results: Over 35% of veterans with a current mental disorder had at least one other comorbid mental disorder. Those with mental disorder comorbidity had lower HUI scores than veterans with only one or less mental disorders (mean 0.41 ± 0.30 vs. 0.72 ± 0.25, p < 0.0001). Conclusion: The co-occurrence of mental disorders that span at least two mental disorder categories is associated with impaired HRQoL in this veteran population. Early identification of mental health comorbidity may lead to interventions to enhance HRQoL among military personnel.

Background and Purpose— Most large, randomized phase III efficacy trials of therapeutic agents in ischemic stroke have failed to find treatment benefit. We determined whether some phase III studies could have been avoided if preceded by smaller single-arm phase II studies to evaluate the futility of proceeding to phase III. Methods— To provide examples of the application of phase II methodology, we obtained primary outcome data for the active treatment group of 6 phase III ischemic stroke therapy trials. For each study, we estimated the sample size number required for a multistage single-arm study using parameters specified in the original study. We evaluated outcome data for the first number of subjects in the phase III study treatment arm ordered by enrollment dates. We compared the proportion of favorable outcomes to prespecified stopping criteria derived from a single-arm phase II futility design. If the observed proportion of favorable outcomes was less than the stopping criterion, we declared the treatment not sufficiently effective to warrant further evaluation in phase III. Results— We identified 3 trials as futile in phase II; none of 3 showed treatment efficacy in phase III. In the 3 remaining phase II trials in which we did not show futility, one showed efficacy in phase III. Conclusion— Single-arm phase II futility studies have been underused in stroke research, but provide a strategy for discarding treatments likely to be ineffective in phase III trials.

Although naltrexone has been shown to be effective in the treatment of alcohol dependence, less is known about its efficacy when combined with different behavioral therapies. Previous work has suggested that naltrexone works best when combined with weekly cognitive behavioral therapy (CBT). This study examined the efficacy of naltrexone when combined with CBT or a motivational enhancement therapy involving less patient contact. Outpatient alcoholics (N = 160) were randomly assigned to either naltrexone (50 mg/d) or placebo and either CBT (12 sessions) or motivational enhancement therapy (4 sessions), in a 4-cell design, and treated over a 12-week period. Subjects were evaluated periodically for alcohol consumption, craving, and biologic markers of drinking (carbohydrate-deficient transferrin and gamma-glutamyltransferase). There was high retention and adherence to therapy and medication in the trial with no significant difference across the treatment groups. Naltrexone, independent of therapy assignment, increased the time to first relapse. However, the CBT-naltrexone group did better than the other groups on a variety of outcome measures. Fewer CBT-naltrexone-treated subjects relapsed, and those that did had both fewer, and more time between, subsequent relapses. This randomized controlled trial is consistent with previous reports about the utility of combining naltrexone with CBT. Despite being more efficient to administer, the combination of motivational enhancement therapy and naltrexone is less effective than CBT and naltrexone. Because CBT and naltrexone share common mechanisms of action, such as craving reduction and relapse prevention, these therapies are likely to be well suited to use in combination.

Often, two or more binary endpoints are used to assess the impact of a new treatment in a randomized phase III clinical trial. In this article, we focus on two binary endpoints. In assessing treatment effects, we are interested in positive treatment effects on the endpoints, and a positive correlation between the two endpoints, i.e., we expect to see positive treatment effects on the same patient. Four estimation methods are reviewed and contrasted for this problem: weighted least squares (WLS), generalized estimating equations (GEE), maximum likelihood (ML), and Mantel-Haenszel (MH) type estimator. In order to test the efficacy of the new treatment, four testing strategies are proposed: a global test and three multi-stage tests. Simulation studies are conducted to study the performances of the four estimation methods and the four testing strategies. Different combinations of significance levels for multi-stage tests are also evaluated. Empirical test size, test power, and estimation precision (i...

This study investigated the prevalence of incarceration and the association with deployment among veterans of the first Persian Gulf War (GW).A structured telephone interview of military personnel from Iowa deployed to the Persian Gulf and a comparison sample of nondeployed military personnel was conducted. The interview consisted of validated questions, validated instruments, and investigator-derived questions to assess relevant medical and psychiatric conditions. A total of 4,886 subjects were randomly drawn from one of four study domains, i.e., GW regular military, GW National Guard/Reserve, non-GW regular military, or non-GW National Guard/Reserve. Symptoms of medical conditions, psychiatric disorders, and health care utilization were the main outcome measures.Nearly one-quarter (845 of 3,695 subjects, 22.9%) had been incarcerated at some point before the interview ("ever incarcerated"). Ever incarcerated veterans had a higher frequency of psychiatric and medical comorbidity and higher rates of health care utilization. Ever incarcerated status was associated with male gender, enlisted rank, lower educational levels, low levels of military preparedness, discharge from service, cigarette smoking, antisocial traits, court martial and/or other military discipline, having seen a mental health professional, and having used illegal drugs. GW veterans who participated in combat had a modestly higher risk for incarceration after the GW than did noncombatants (odds ratio, 1.6; 95% confidence interval, 1.0-2.5).Military recruits with a history of incarceration more often displayed problematic behaviors, more often developed psychiatric/medical conditions, and had high rates of health care utilization. A history of incarceration may be a behavioral marker for substance abuse, antisocial behavior, and mental illness. Importantly, GW deployment carried no increased risk of subsequent incarceration overall.

SummaryO'Brien (1984, Biometrics40, 1079–1087) introduced a simple nonparametric test procedure for testing whether multiple outcomes in one treatment group have consistently larger values than outcomes in the other treatment group. We first explore the theoretical properties of O'Brien's test. We then extend it to the general nonparametric Behrens–Fisher hypothesis problem when no assumption is made regarding the shape of the distributions. We provide conditions when O'Brien's test controls its error probability asymptotically and when it fails. We also provide adjusted tests when the conditions do not hold. Throughout this article, we do not assume that all outcomes are continuous. Simulations are performed to compare the adjusted tests to O'Brien's test. The difference is also illustrated using data from a Parkinson's disease clinical trial.

Preclinical models of cocaine dependence have shown favorable reductions in cocaine use using dihydropyridine calcium channel antagonists. This is one of the first reports testing the efficacy of the long-acting calcium channel antagonist, amlodipine, for the treatment of cocaine dependence. This was a 12-week, double-blind, randomized, placebo-controlled, parallel patient group trial of amlodipine vs. placebo for the treatment of cocaine dependence. One hundred and sixteen subjects participated in a 12-week medication trial in which 60 subjects received medication and 56 received placebo. Subjects in both groups received up to 12 standard manual-driven cognitive behavioral therapy sessions. Overall, drop-out rate for both groups was high, with only about 20% of subjects completing all 12 weeks of treatment. Both groups showed comparable levels of medication compliance and therapy attendance. In the end, amlodipine was no more effective than placebo in reducing craving or measured levels of cocaine use.

To delineate the frequency, course, risk factors, and neuroanatomy of hemispatial neglect in a large stroke cohort.One thousand two hundred eighty-one patients with acute stroke were enrolled in a multicenter trial of an anticoagulant. Presence and severity of neglect were assessed with the NIH Stroke Scale (NIHSS) neglect item, assessing tactile extinction and visuospatial neglect at entry, daily for 1 week, and at 3 months. Head CT scans were obtained on day 7, and infarct location and size were characterized.Neglect was common at presentation, occurring in 43% of right brain-lesioned (RBL) patients and 20% of left brain-lesioned (LBL) patients (p0.001). At 3 months, neglect was present in 17% of RBL patients and in 5% of LBL patients (p0.001). In RBL patients, neglect was most frequently associated with lesions involving the (in descending order) temporal, parietal, frontal, occipital lobes, basal ganglia, and thalamus. Neglect was more common and persistent with cortical than with subcortical lesions. Increasing age was associated with increased risk of neglect in RBL patients, whereas gender and handedness did not significantly affect neglect frequency.This series confirms that hemispatial neglect may occur with damage to several supratentorial structures but is most common and persistent with lesions of the right temporoparietal cortex. Increasing age is associated with neglect, particularly after right brain lesions. Gender and handedness do not exert a marked effect on the likelihood of the occurrence of neglect.

Depression is a common mental disorder associated with poor health outcomes. The purpose of this study is to examine the prevalence of depression, mental health comorbidity, illness variables, and quality of life in a sample of military veterans serving during the first Gulf War. The Iowa Gulf War Case Validation Study involved face-to-face evaluations in 1999--2002 of 602 military personnel--either deployed ("deployed veterans") or eligible but not deployed ("non-deployed veterans") to the Gulf. Subjects were sampled by conducting a series of case-control studies nested within a population-based survey of 4,886 military personnel. All subjects were interviewed using the Structured Clinical Interview for DSM-IV (SCID-IV), and a series of semi-structured interviews and validated questionnaires. Best estimate psychiatric diagnoses were assigned based on all available data. One-hundred-ninety-two (32%) of the 602 surveyed veterans met criteria for a current or lifetime depressive disorder (major depression, dysthymia, depressive disorder--not otherwise specified). Depressed non-deployed veterans were more likely to be female and to have served in the Air Force than depressed deployed veterans. There were few significant differences between the depressed deployed veterans and the depressed non-deployed veterans. Depressed deployed veterans had significantly higher lifetime rates of comorbid cognitive dysfunction (55% vs. 35%), and anxiety disorders (59% vs. 33%)--mainly accounted for by specific phobias (12% vs. 2%) and posttraumatic stress disorder (33% vs. 10%)--than did depressed non-deployed veterans. Lifetime substance use disorders were significantly more frequent in deployed veterans than non-deployed veterans (70% vs. 52%), particularly alcohol disorders (68% vs. 52%). There were no differences in rates of personality characteristics, family psychiatric history, stressors, hypochondriasis, and level of functioning between the two study groups showed no significant differences. Depressive illness is frequent in military samples, as it is in the general population. The prevalence, pattern of comorbidity, and illness features were similar in deployed veterans and non-deployed veterans, suggesting that the depression suffered by both groups of veterans is qualitatively comparable. The main difference between study, groups was that depressed deployed veterans had higher rates than depressed non-deployed veterans of comorbid anxiety disorders, hypothesized to be part of the stress-related syndromes seen in those who experience combat.

Background Unsedated esophagoscopy with ultrathin endoscopes is a valuable screening modality for Barrett's esophagus, but the stomach and the duodenum cannot be examined completely with the smallest and best tolerated of these endoscopes. There are no data as to how often disease in the stomach and the duodenum would be missed when using this screening strategy. Our hypothesis is that patients with reflux symptoms, in the absence of daily abdominal pain, nausea, or history of ulcer, were unlikely to have clinically significant gastroduodenal disease. Methods Patients scheduled for upper endoscopy at a single outpatient endoscopy unit in a tertiary referral center were screened. The inclusion criterion was reflux symptoms as the primary indication for upper endoscopy. Patients with another valid indication were excluded. A detailed history was recorded and symptom questionnaire completed for each patient before endoscopy; these data were compared with the endoscopy findings. Results A total of 175 patients were included. Indications for upper endoscopy were the following: worsening symptoms (n=74), ongoing reflux despite therapy (n=27), and long-standing reflux (n=74). Major esophageal findings were discovered in 95 patients. In 10 patients, major gastric or duodenal findings were detected as follows: erosive gastritis (n=8), gastric ulcer (n=2), duodenal ulcer (n=2), erosive duodenitis (n=2), and duodenal polyp (n=1). Daily abdominal pain ( p =0.014) or possibly daily nausea ( p =0.028 unadjusted, 0.197 adjusted) was associated with major gastric/duodenal disease. Patients without daily abdominal pain, nausea, or a history of gastric/duodenal ulcer were much less likely to have major disease (0.9%) than patients with one of these predictors (13.2%, p =0.00097). Conclusions Daily abdominal pain and nausea, in combination with a history of ulcer disease, are strong predictors of major gastric or duodenal disease. Patients with reflux without these predictors are highly unlikely to have a major disease involving the stomach or duodenum, and are suitable candidates for esophagoscopy alone.

We assessed resistance rates and trends for important antimicrobial-resistant pathogens (oxacillin-resistant Staphylococcus aureus [ORSA], vancomycin-resistant Enterococcus species [VRE], ceftazidime-resistant Klebsiella species [K-ESBL], and ciprofloxacin-resistant Escherichia coli [QREC]), the frequency of outbreaks of infection with these resistant pathogens, and the measures taken to control resistance in a stratified national sample of 670 hospitals. Four hundred ninety-four (74%) of 670 surveys were returned. Resistance rates were highest for ORSA (36%), followed by VRE (10%), QREC (6%), and K-ESBL (5%). Two-thirds of hospitals reported increasing ORSA rates, whereas only 4% reported decreasing rates, and 24% reported ORSA outbreaks within the previous year. Most hospitals (87%) reported having implemented measures to rapidly detect resistance, but only approximately 50% reported having provided appropriate resources for antimicrobial resistance prevention (53%) or having implemented antimicrobial use guidelines (60%). The most common resistant pathogen in US hospitals is ORSA, which accounts for many recognized outbreaks and is increasing in frequency in most facilities. Current practices to prevent and control antimicrobial resistance are inadequate.

The expanding role of women in the military raises questions related to the military experiences of women serving in major conflicts. We assess the military experiences and postwar health care use of women who served during the Gulf War. Data from a population-based survey of military personnel serving between August 1990 and July 1991 assessing military preparedness, combat experience, occupational and other service-related exposures, and health care use were analyzed. Deployed women were more often in the Army, single, without children, college educated, and reported fewer vaccinations. Deployed men and women had similar military experiences; however, men more often participated in combat. Deployed women had more outpatient and inpatient health care use 5 years after deployment and more often received Department of Veterans Affairs compensation than men. If these important differences in exposures and health care use are confirmed in other studies, optimal training and deployment preparedness strategies should be reconsidered.

In this paper, we present a modified Kelly and Rice method for testing synergism. This approach is consistent with Berenbaum's1-3 framework for additivity. The delta method[4] is applied to obtain the estimated variance for the predicted additivity proportion. A Monte Carlo simulation study for the evaluation of the method's performance, i.e., global overall tests for synergism, is also discussed. Kelly and Rice[5] do not provide a correct test statistic because the variance is underestimated. Hence, the performance of the Kelly–Rice[5] method is generally anti-conservative, based on the simulation findings. In addition, the overall test of synergism with χ2(r) from the modified Kelly and Rice method for larger sample sizes is better than that with χ2(1) from the modified Kelly and Rice method.

To identify consistent provision of clinical preventive services, we sought to benchmark all acute care Veterans Affairs Medical Centers (VAMCs) against each other nationally on the basis of multiple evidence-based, performance measures to identify facilities performing consistently higher and lower than expected.The 1998 Veterans Health Survey assessed the self-reported delivery of evidence-based clinical preventive services in a stratified national sample of 450 ambulatory care patients seen at each VAMC. Proportions appropriately receiving each service within the recommended time interval were calculated for 138 VAMCs. Percentile ranks for each outcome were assigned. Two approaches were used for benchmarking performance. First, a scaled score for each facility was calculated across the set of 12 measures. Second, facilities were ranked based on the sum of the percentile ranks over a range of specific high cutoffs (eg, 70-80%) and above a range of lower cutoffs (eg, 40-50%). Ranking was validated by comparing with deciles of ranks on chart audit (External Peer Review Program, EPRP) data using Kendall's tau-b and chi2 quality-of-fit test. Differences between consistently high adherence (CHA) and low adherence (CLA) facilities were compared using the Wilcoxon rank sum test on 14 VHS and 11 EPRP outcomes.Data from 39,939 patients (67% response rate) were examined. In combination, cutoffs of greater than 50th percentile and greater than 75th percentile rank yielded 12 of 14 VHS and 6 of 11 EPRP measures different between CHA and CLA facilities. The scaled-score approach resulted in 20 CHA and 14 CLA facilities. The sum of outcomes ranked above 50th percentile and over 75th percentile for CHA facilities (n = 17) was 15 or more. The sum of outcomes ranked above the same cutoffs for CLA facilities (n = 16) was 3 or less. EPRP and 1998 VHS data demonstrated that the survey measures and benchmarking approaches were both reliable and valid. Both approaches resulted in multiple differences between CHA and CLA facilities; differences were greater using the percentile rank approach.The VA has successfully encouraged adoption of evidence-based clinical preventive services throughout its health care system. However, facilities show wide variation in their levels of delivery and can be distinguished on the basis of their consistently high or low levels of adherence. Examining service delivery across multiple performance indicators allows identification of opportunities to improve clinical practice guideline implementation and the delivery of preventive services. This approach identifies model institutions where focused investigation of factors associated with consistent performance may be particularly fruitful.

The objective of this study was to investigate the relation between posttraumatic stress disorder (PTSD) and perceived physical health. Participants included 3,682 Gulf War veterans and control subjects of the same era who completed a telephone survey about their health status. PTSD was assessed using the PTSD Checklist-Military Version. Veterans screening positive for PTSD reported significantly more physical health symptoms and medical conditions than did veterans without PTSD. They were also more likely to rate their health status as fair or poor and to report lower levels of health-related quality of life. The results of this study are consistent with studies of other combat veterans and provide further support for an association between PTSD and adverse physical health outcomes. Stressful or traumatic life events, such as those encountered during a rapid military deployment and conflict, are associated with a variety of adverse health effects. These health effects may manifest themselves in both psychological and physical outcomes. Health care providers must be attentive to recognize and evaluate both of these dimensions.

Rationale, aims and objectives Although clinical practice guidelines have been promoted widely, there is considerable concern that physicians have not incorporated them into their practice. Models suggest that a ‘knowledge‐attitude‐behaviour’ sequence is important in modifying physician practice patterns. To address this, we examined physicians’ knowledge of, attitudes towards and compliance with a widely implemented guideline ‐ the Agency for Health Care Policy and Research (AHCPR) smoking cessation guideline. Methods A survey was mailed to a random sample of physicians at 128 veterans health administration facilities, which had implemented the guideline 2 years previously. Results Completed surveys were received from 879 physicians (50.3% response rate). Only 26% of physicians reported receiving a copy or summary of the guideline, 44% reported little or no familiarity with it, 42% did not know if they agreed with it, 40% did not know if they complied with it and 46% did not know if it was effective. However, a high percentage of physicians reported that they always or usually explain the health risks of smoking (86%) and that they always or usually suggest that their smoking patients stop (94%). Conclusions In spite of little familiarity with the guideline, the responding physicians reported practice patterns consistent with adherence to it. Knowledge is only one of a spectrum of barriers that affects physician adherence to guidelines. There are numerous opportunities for health care organizations to overcome the barriers to physician adoption of clinical practice guidelines in their day-to-day practice.

We examined the relationship between physician adherence with a smoking cessation guideline and organizational structures, policies, leadership support, and physician knowledge and attitudes. A random sample of 844 physicians practicing in 127 VHA hospitals was surveyed. Survey results were aggregated to the hospital level and linked with data on organizational characteristics from the 1998 annual survey of hospitals by the American Hospital Association. Significant predictors of adherence included organizational policies related to nicotine replacement prescriptions, the timing and effectiveness of guideline implementation, physicians' knowledge of VHA clinical guidelines generally, and physicians' beliefs about the receptiveness of patients to smoking cessation.

Background and Purpose — Although the efficacy of aspirin in reducing stroke incidence is clear, its role in reducing stroke severity is disputed. This study compares stroke severity between patients who did or did not take aspirin in the week before stroke and enrollment in the Trial of Org 10172 in Acute Stroke Treatment (TOAST). Methods — Of 1275 patients randomized, 509 reported aspirin use in the week before stroke; 766 did not. Clinical stroke severity was assessed with the National Institutes of Health Stroke Scale (NIHSS) and the Supplementary Motor Examination (SME) at trial entry and at 3 months. Using these scales, we compared the categorization of stroke severity (mild, moderate, and severe) and mean scores between aspirin users and nonusers. Results — The difference in distribution of baseline NIHSS scores was statistically significant between aspirin users and nonusers ( P =0.006), with a greater percentage of milder strokes among aspirin users. The difference in mean baseline NIHSS scores was also significantly lower in aspirin users (8.2) and nonusers (9.3) ( P =0.003). The distribution of baseline SME scores and mean SME scores also showed lower stroke severity in aspirin users than in nonusers ( P =0.048 and P =0.004, respectively). At 3 months, differences in stroke severity measured by the SME but not the NIHSS remained statistically significant. Seven-day and 3-month mortality did not differ significantly. Conclusions — In this study aspirin use is associated with milder clinical deficits at stroke onset. These deficits may affect prognosis and influence response to treatment. Future clinical trials should ensure that prestroke aspirin use is comparable in study groups.

In 1985, the International Study of Extracranial-to-Intracranial Arterial Anastomosis demonstrated no benefit from extracranial-to-intracranial arterial bypass operations in treatment of patients with extensive cerebrovascular disease including those with occlusions of the internal carotid artery. Interest in the potential use of extracranial-to-intracranial arterial bypass operations, however, has been rekindled by evidence that some patients with occlusion of the internal carotid artery have a poor collateral circulation and a high risk for recurrent ischemic events. Other patients with adequate perfusion after occlusion have a low likelihood for recurrent stroke. Restricting surgical treatment to only those patients judged to have a high risk for recurrent stroke might improve the usefulness of the bypass operation. A new clinical trial is proposed, testing the potential usefulness of extracranial-to-intracranial arterial bypass operations for treatment of carefully selected patients with occlusion of the internal carotid artery. Several issues that are being addressed in this new trial are described in this article.

Purpose To provide ophthalmologists and pathologists with expanded criteria for separating patients at high risk of metastatic melanoma from those at low risk on the basis of microcirculation patterns in choroidal and ciliary body melanomas. Design Tissue culture studies and observational case series. Participants The pattern-forming ability of four uveal melanoma cell lines of varying degrees of aggressive behavior was studied in vitro. Histologic sections of 234 eyes removed for choroidal or ciliary body melanoma were studied for the presence of microcirculation patterns. Methods The study was divided into two phases: the study of histologic sections of eyes removed for choroidal and ciliary body melanomas and observations on the in vitro behavior of cultured melanoma cells of varying degrees of invasive behavior. The presence or absence of each of nine microcirculation patterns was recorded from tissue sections, and interrelationships between different patterns were explored statistically. In vitro reconstitution of patterns and a study of the interrelationships of patterns in histologic sections was carried out. In the in vitro studies, uveal melanoma cell lines of varying degrees of aggressive potential were cultured to observe the development of architectural patterns other than loops and networks. Main outcome measures In histologic studies, the outcome measure was the conditional probability of detecting loops or networks given the presence or absence of other patterns positive for periodic acid–Schiff. For tissue culture studies, the outcome measure was either the development or lack of development of patterns of different shapes in vitro. Results Histologic studies disclosed that given the presence of arcs without or with branching in a tissue section, it is likely that loops or networks will be detected in the same section plane, suggesting that the production of these patterns by aggressive tumor cells reflects a spectrum of architectural potential. In vitro studies confirmed this hypothesis by revealing that highly aggressive and metastatic uveal melanoma cell lines, but not poorly aggressive tumor cell lines, generated parallel channels with and without crosslinking and arcs with and without branching as well as loops and networks. Conclusions The criteria for separating patients into low- and high-risk categories for metastasis from uveal melanoma should be expanded to include patterns other than loops or networks. In both the pathology laboratory as well as in a clinical setting, the detection of arcs or arcs with branching and parallel channels should prompt a careful search for loops and networks and for crosslinking parallel channels, respectively.

Objective:To describe hospital practices and policies relating to bloodborne pathogens and current rates of occupational exposure among healthcare workers.Participants and Methods:Hospitals in Iowa and Virginia were surveyed in 1996 and 1997 about Standard Precautions training programs and compliance. The primary outcome measures were rates of percutaneous injuries and mucocutaneous exposures.Results:153 (64%) of 240 hospitals responded. New employee training was offered no more than twice per year by nearly one third. Most (79%-80%) facilities monitored compliance of nurses, housekeepers, and laboratory technicians; physicians rarely were trained or monitored. Implementation of needlestick prevention devices was the most common action taken to decrease sharps injuries. Over one half of hospitals used needleless intravenous systems; larger hospitals used these significantly more often. Protected devices for phlebotomy or intravenous placement were purchased by only one third. Most (89% of large and 80% of small) hospitals met the recommended infection control personnel-to-bed ratio of 1:250. Eleven percent did not have access to postexposure care during all working hours. Percutaneous injury surveillance relied on incident reports (99% of facilities) and employee health records (61%). The annual reported percutaneous injury incidence rate from 106 hospitals was 5.3 injuries per 100 personnel. Compared to single tertiary-referral institution rates determined more than 5 years previously, current injury rates remain elevated in community hospitals.Conclusions:Healthcare institutions need to commit sufficient resources to Standard Precautions training and monitoring and to infection control programs to meet the needs of all workers, including physicians. Healthcare workers clearly remain at risk for injury. Further effective interventions are needed for employee training, improving adherence, and providing needlestick prevention devices.

Background and Purpose: Clinicians have tended to view anterior circulation (AC) and posterior circulation (PC) strokes as separate entities, with different underlying pathogenesis, natural histories, and potential responsiveness to interventions such as anticoagulation. We sought to explore differences between AC and PC stroke in the Trial of ORG 10172 in Acute Stroke Treatment (TOAST). Methods: For patients enrolled in TOAST, prospective clinical information was collected including outcome at 3 months. Data on vascular distribution were obtained from the clinical impression of the investigators. Group comparisons for categorical data were performed using Fisher’s exact test. Independent sample t tests and analysis of covariance were used for all continuous data. Results: The analysis included 1,039 patients with AC stroke and 180 patients with PC stroke. There were fewer women in the PC than in the AC groups, but otherwise there were no differences in demographics, risk factors or stroke subtypes between the two groups. Headache (AC 8.7%, PC 15%, p = 0.013) and vomiting (AC 3.5%, PC 17.8%, p < 0.001) were more common among PC patients. Mean baseline National Institutes of Health Stroke Scale (NIHSS) score was lower (less severe) among PC (6.1) than AC patients (9.5; p < 0.001). On univariate analysis, favorable outcome at 3 months was more common for PC patients in both the placebo group (PC 82%, AC 71%, p = 0.04) and heparinoid group (PC 87%, AC 73%, p = 0.005). However, multivariate analysis, controlling for gender, history of previous stroke and baseline NIHSS score, showed no difference in outcome between PC and AC stroke. For favorable outcome, there was no interaction between vascular distribution and treatment category, suggesting that the effect of heparinoid did not differ between PC and AC strokes. Conclusion: Patients with PC stroke seem to have a better long-term outcome than do AC patients, but this difference is no longer apparent when controlling for important prognostic variables. PC patients did not show any particular benefit from anticoagulation, and the efficacy of heparinoid did not vary between AC and PC stroke. While AC and PC patients do differ in some respects, it may be inappropriate to single out PC patients for anticoagulant treatment.

PURPOSE: Concerns have been raised about whether veterans of the Gulf War have a medical illness of uncertain etiology. We surveyed veterans to look for evidence of an illness that was unique to those deployed to the Persian Gulf and was not seen in comparable military controls. SUBJECTS AND METHODS: A population-based sample of veterans (n = 1,896 from 889 units) deployed to the Persian Gulf and other Gulf War–era controls (n = 1799 from 893 units) who did not serve in the Gulf were surveyed in 1995–1996. Seventy-six percent of eligible subjects, including 91% of located subjects, answered questions about commonly reported and potentially important symptoms. We used factor analysis, a statistical technique that can identify patterns of related responses, on a random subset of the deployed veterans to identify latent patterns of symptoms. The results from this derivation sample were compared with those obtained from a separate validation sample of deployed veterans, as well as the nondeployed controls, to determine whether the results were replicable and unique. RESULTS: One half (50%) of the deployed veterans and 14% of the nondeployed controls reported health problems that they attributed to military service during 1990–1991. Compared with the nondeployed controls, the deployed veterans had significantly greater prevalences of 123 of 137 (90%) symptoms; none was significantly lower. Factor analysis identified three replicable symptom factors (or patterns) in the deployed veterans (convergent correlations ≥0.85). However, these patterns were also highly replicable in the nondeployed controls (convergent correlations of 0.95 to 0.98). The three factors also accounted for similar proportions of the common variance among the deployed veterans (35%) and nondeployed controls (30%). CONCLUSIONS: The increased prevalence of nearly every symptom assessed from all bodily organ systems among the Gulf War veterans is difficult to explain pathophysiologically as a single condition. Identification of the same patterns of symptoms among the deployed veterans and nondeployed controls suggests that the health complaints of Gulf War veterans are similar to those of the general military population and are not consistent with the existence of a unique Gulf War syndrome.

On 12 June 1991, Mt Pinatubo in the Philippines erupted in what was the largest volcanic disaster of this century. Three hundred and fifty-one tribal and nontribal disaster victims were interviewed 6 years after they were displaced following the eruption. The acute and chronic effects of the eruption and the populations affected by Mt Pinatubo are described. The culturally and ecologically diverse population imposed logistical and methodological challenges that were expressed in ways not ordinarily encountered when conducting a community survey. Translating and administering survey instruments involved modifications of traditional methods to minimize item, method, construct and selection bias. Investigators conducting cross-cultural field studies in developing countries must strive to maintain high standards of scientific rigor, yet remain flexible, efficient and culturally sensitive.

This report presents the initial results of a mental health survey of 351 tribal and non-tribal Mount Pinatubo disaster victims 6 years after they were displaced following the volcanic eruption in the Philippines on June 12, 1991. Mental illness prevalence rates in both Filipino ethnic groups were comparable to those found in a U.S. study using the same assessment instrument. Post-traumatic stress disorder (PTSD; 27.6%) and major depression (14.0%) were the two most frequent diagnoses. Diagnostic test-retest interviewer agreement was good for probable alcohol abuse (kappa = .65, agreement = 97%) and any mood disorder (kappa = .53, agreement = 91%) but was reduced for any anxiety disorder (kappa = .15, agreement = 81%) and separately evaluated PTSD (kappa = .18, agreement = 69%). Diagnostic test-retest agreement was good among typical Filipinos (mean kappa = .66, mean agreement = 93%) but was reduced among tribal aborigines (mean = .30, mean agreement = 86%). Internal consistency of the PTSD rating scale was high within and across both ethnic groups, including total scale (alpha = .91) and DSM-IV Criteria B, C, and D sub-scales (alpha = .80, 81, and .78, respectively). With the exception of probable alcohol abuse, construct and criterion validity was demonstrated among both tribal and non-tribal Filipinos for all classes of psychiatric disorders by comparing diagnostic results with respondents' views of their physical and mental health and level of functional impairment. Overall, DSM-IV mood, anxiety, alcohol use, and PTSDs with adequate reliability and construct and criterion validity were made in this culturally diverse, non-Western, disaster victim population. However, test-retest diagnostic agreement was reduced for anxiety disorders and among aboriginal respondents, and validity was not demonstrated for probable alcohol abuse.

Background and Purpose —While previous studies suggest that the peak time period for the occurrence of ischemic stroke is in the mid- to late-morning hours, detailed information pertaining to circadian variations among the various stroke subtypes has been limited. The purpose of our study was to define the circadian patterns of symptom onset in an acute stroke trial with an established system for stroke subtype classification. Methods —An analysis was conducted on 1272 patients enrolled in the Trial of Org 10172 in Acute Stroke Treatment (TOAST) study. All patients had a documented time of stroke symptom onset, and all stroke subtype determinations were made by a single rater. Results —The greatest portion of atherothrombotic strokes (25.7%), cardioembolic strokes (30.5%), and strokes of other/unknown mechanism (27.1%) occurred between 6:01 am and 12:00 noon. The greatest portion of lacunar strokes (31.6%) were present on awakening. More than one half of the infarcts in this series were either present on awakening or occurred in the mid- to late-morning hours. The correlation between stroke subtype and time of symptom onset did not reach statistical significance ( P =0.07, Pearson’s χ 2 method). Conclusions —Although there is a trend for clustering of ischemic stroke in the morning hours, there is insufficient specificity to predict with any reasonable likelihood the stroke subtype according to the circadian pattern of symptom onset.

We previously demonstrated that whole blood contains significantly more hepatitis C virus (HCV) RNA than plasma. To validate the whole-blood-based HCV RNA detection method, a prospective comparison of HCV RNA detection in whole blood and plasma from 50 patients with chronic liver disease was undertaken. Whole-blood and plasma aliquots were independently tested for HCV RNA by reverse transcriptase (RT) PCR assay, and plasma was tested by the Roche Amplicor assay. HCV RNA was detected in 35 of 50 (70%) whole-blood samples by RT-PCR but in only 26 of 50 (52%) plasma samples tested by the Amplicor assay ( P < 0.01). HCV RNA was detected in 85% of HCV antibody-positive patients by the whole-blood method compared with 74% of plasma samples by the Amplicor method. The five HCV antibody-positive subjects who were negative by whole-blood-based RT-PCR assay were all receiving interferon therapy and had normal transaminases at the time of testing. HCV RNA was detected in 38% of HCV antibody-negative subjects by the whole-blood-based RT-PCR assay compared with 6.25% of these patients by the Amplicor assay ( P < 0.05). There were nine samples in which HCV RNA was detected in whole blood but the Amplicor test was negative. Eight of the nine RNAs prepared from these whole-blood samples tested positive in the Amplicor assay, thus confirming the specificity of our results. This study demonstrates that whole-blood-based HCV RNA detection is more sensitive than currently available commercial tests and that whole-blood RNA is suitable for use in commercial assays.

Statistical approaches for assessment of synergism have been widely used in pharmacological settings to identify beneficial drug combinations. Several commonly used approaches for synergy assessment are studied in this paper. Each of the methods studied can be related to the isobole-based definition of additivity (Berenbaum, 1989). A Monte Carlo study is conducted in which both significance levels and powers for the five methods are estimated and compared. Based on the results from the simulation study, two approaches due to Finney (1971b) perform similarly, with high power and acceptable significance levels. In addition, a global overall test of synergism from the modified Kelly and Rice (1990) method for larger sample sizes and an approach due to Carter et al. (1988) perform well in terms of estimated Type I error rate.

TOAST is a multicenter, randomized, placebo-controlled clinical trial testing the usefulness of a new antithrombotic drug in improving the outcome of persons with acute ischemic stroke. Until recently, no clinical trial testing a treatment for ischemic stroke had demonstrated efficacy in outcome. Design problems of previously conducted trials with inconclusive results may partly explain their failures. During the design of TOAST, the investigators addressed several issues so the trial could test the treatment accurately. We report the strategies used in designing, implementing, and coordinating the trial.

INTRODUCTION: IMS 3 Trial is a multicenter randomized study to compare the effectiveness of the bridging (IV tPA + intra-arterial) therapy to the IV tPA only treatment in patients with acute ischemic stroke presenting within 3 hours of symptom onset. The trial intended to enroll 900 subjects to ensure adequate statistical power to detect an absolute 10% difference in the % of subjects with good outcome, defined as mRS score of 0-2 at 3 months. In April 2012, after 656 subjects were randomized, further enrollment was terminated by the NINDS upon recommendation by the DSMB. The decision was based on the pre-specified criterion for futility using conditional power of METHOD: Conditional power was defined as the likelihood of finding statistical significance at the end of the study given the accumulated data to date and with the assumption that a minimum hypothesized difference of 10% truly exists between the two groups. Hence, it is akin to ordinary statistical power. RESULTS: The futility boundary was crossed at the Trial’s 4 th interim analysis. The details of the evolution of the study that led to the futility determination are presented, including the interaction between the unblinded study statisticians and the DSMB in the complex deliberation of the data analysis results. CONCLUSION: Even in spite of pre-specifying the parameters of interim analysis, interim looks at the data pose challenges in the interpretation and decision-making. It therefore underscores the importance of taking utmost care in the design of the study, including specification of objective stopping guidelines.

The microcirculation of choroidal and ciliary body melanomas is remodeled into architecturally distinctive patterns. The presence of two histologic microvascular patterns, networks and parallel vessels with cross-linking, is strongly associated with metastasis. This study was designed to test the hypothesis that networks and parallel vessels with cross-linking patterns are not distributed evenly throughout the tumor. From a set of 234 eyes removed for ciliary body or choroidal melanoma, 152 tumors contained at least one focus of either vascular networks or parallel vessels with crosslinking. Histological cross-sections were digitized and foci of tumor containing these patterns were pseudocolorized so that their location within the periphery or central tumor zone could be mapped. Ciliary body and choroidal melanomas vary widely in size and shape and it is not appropriate to describe the periphery of a tumor as a fixed value because in a small tumor, the periphery thus defined would occupy a larger percent area than in a larger tumor. In this study, the peripheral and central zones of each tumor were described by a function that was constant from tumor to tumor, allowing the width of the peripheral and central zones to vary proportionally with tumor size. Observed counts of vascular patterns per zone were compared statistically with expected counts based upon the percent area occupied by the peripheral and central zones. Discrete foci of networks and parallel with cross-linking vessels are over-represented in the tumor periphery (p < 0.0001).

A 1-year, double-blind, placebo-controlled, randomized, parallel-group study compared the efficacy and safety of acetyl-L-carnitine hydrochloride (ALCAR) with placebo in patients with probable Alzheimer's disease (AD). Subjects with mild to moderate probable AD, aged 50 or older, were treated with 3 g/day of ALCAR or placebo (1 g tid) for 12 months. Four hundred thirty-one patients entered the study, and 83% completed 1 year of treatment.The Alzheimer's Disease Assessment Scale cognitive component and the Clinical Dementia Rating Scale were the primary outcome measures.Overall, both ALCAR- and placebo-treated patients declined at the same rate on all primary and most secondary measures during the trial. In a subanalysis by age that compared early-onset patients (aged 65 years or younger at study entry) with late-onset patients (older than 66 at study entry), we found a trend for early-onset patients on ALCAR to decline more slowly than early-onset AD patients on placebo on both primary endpoints. In addition, early-onset patients tended to decline more rapidly than older patients in the placebo groups. Conversely, late-onset AD patients on ALCAR tended to progress more rapidly than similarly treated early-onset patients. The drug was very well tolerated during the trial.The study suggests that a subgroup of AD patients aged 65 or younger may benefit from treatment with ALCAR whereas older individuals might do more poorly.However, these preliminary findings are based on post hoc analyses. A prospective trial of ALCAR in younger patients is underway to test the hypothesis that young, rapidly progressing subjects will benefit from ALCAR treatment.NEUROLOGY 1996;47: 705-711

A popular linear regression estimator for censored data is the one proposed by Buckley and James (1979). However, this estimator is not robust to outliers, which is not surprising since it is a modified version of the uncensored data least squares estimator. Lai and Ying (1994) have proposed an M-estimator for censored data that is a generalization of the Buckley- James estimator. In this paper we discuss a weighted least squares algorithm for computing these M-estimates and compare the performance of two Huber M-estimators with the Buckley-James estimator in a simulation study. We find that the Huber M-estimators perform more robustly for a broad range of censoring and error distributions.

Purpose: To study the relation between vascular patterns and the biologically aggressive behavior of Ciliary body melanomas. Methods: The authors compared the frequency distribution of vascular patterns by location for 234 uveal melanomas (54 tumors involving the Ciliary body, and 180 without Ciliary body involvement). Stepwise Cox regressions (for the endpoint of timeto-death due to melanoma), performed separately for melanomas with and without Ciliary body involvement, included the following variables: size, vascular patterns, cell type, mean of the largest nucleoli, mitoses, tumor infiltrating lymphocytes, age, and sex. A separate Cox regression procedure included the variable of tumor location. Kaplan-Meier survival curves were generated for time to melanoma death with Ciliary body involvement and melanomas without Ciliary body involvement for tumors containing or lacking vascular networks. Results: These vascular patterns appear more often in the Ciliary body than in the choroid: parallel vessels ( P = 0.022), arcs ( P = 0.003), and parallel with cross-linking, arcs with branching, and loops and networks (all P = 0.0001). Stepwise regression for tumors confined to the choroid indicated that the presence of networks was the most significant variable ( P = 0.0001); stepwise regression for tumors with Ciliary body involvement suggested that only one variable, networks, was significant ( P = 0.0066). Kaplan-Meier survival estimates indicated that the survival of patients with tumors containing networks in the Ciliary body was comparable to those containing networks in the choroid. Conclusion: Regardless of location, Ciliary body or choroid, the presence of vascular networks shortens survival. The tumor location does not enter a stepwise Cox regression model when vascular patterns are included as variables. Therefore, the aggressive behavior of Ciliary body melanomas appears to be related to the tendency for vascular networks to develop in this location. Ophthalmology 1995;102:844-851