Intensive Versus Standard Treatment of Hyperglycemia in Acute Ischemic Stroke Patient: A Randomized Clinical Trial Subgroups Analysis.

BACKGROUND: Benefit from blood glucose (BG) control during acute ischemic stroke may depend on glycemic parameters. We evaluated for associations between the SHINE (Stroke Hyperglycemia Insulin Network Effort) randomized treatment group and the SHINE predefined 90-day functional outcome, within-patient subgroups defined by various glycemic parameters. METHODS: The SHINE Trial randomized 1151 patients within 12 hours with acute ischemic stroke and hyperglycemia to standard (target BG 80-179 mg/dL) or intensive (target BG 80-130 mg/dL) BG control for 72 hours. We predefined 6 glycemic parameters: acute BG level, absence versus presence of diagnosed and undiagnosed diabetes, hemoglobin A1c, glycemic gap (acute BG-average daily hemoglobin A1c based BG), stress hyperglycemia ratio (acute BG/average daily hemoglobin A1c based BG), and BG variability (SD). Favorable functional outcome was defined by the SHINE Trial and based on the modified Rankin Scale score at 90 days, adjusted for stroke severity. We computed relative risks adjusted for baseline stroke severity and thrombolysis use. RESULTS: Likelihood for favorable outcome was lowest among patients with undiagnosed diabetes compared to patients with true nondiabetes (adjusted relative risk, 0.42 [99% CI, 0.19-0.94]). We did not find any relationship between the favorable outcome rate and baseline BG or any of the glycemic parameters. No differences between SHINE treatment groups were identified among any of these patient subgroups. CONCLUSIONS: In this exploratory subgroup analysis, intensive versus standard insulin treatment of hyperglycemia in acute ischemic stroke patient subgroups, did not influence the 90-day functional outcomes, nor did we identify associations between these glycemic parameters and 90-day functional outcomes.

BiMM forest: A random forest method for modeling clustered and longitudinal binary outcomes.

Clustered binary outcomes and datasets with many predictor variables are frequently encountered in clinical research (e.g. longitudinal studies). Generalized linear mixed models (GLMMs) typically employed for clustered endpoints have challenges for some scenarios, particularly for complex datasets which contain many interactions among predictors and nonlinear predictors of outcome. We propose a new method called Binary Mixed Model (BiMM) forest, which combines random forest and GLMM methodology. BiMM forest offers a flexible and stable method which naturally models interactions among predictors and can be employed in the setting of clustered data. Simulation studies show that BiMM forest achieves similar or superior prediction accuracy compared to standard random forest, GLMMs and its tree counterpart (BiMM tree) for clustered binary outcomes. The method is applied to a real dataset from the Acute Liver Failure Study Group. BiMM forest offers an alternative method for modeling clustered binary outcomes which may be applied in myriad research settings.

Random forest classification of etiologies for an orphan disease.

Classification of objects into pre-defined groups based on known information is a fundamental problem in the field of statistics. Although approaches for solving this problem exist, finding an accurate classification method can be challenging in an orphan disease setting, where data are minimal and often not normally distributed. The purpose of this paper is to illustrate the application of the random forest (RF) classification procedure in a real clinical setting and discuss typical questions that arise in the general classification framework as well as offer interpretations of RF results. This paper includes methods for assessing predictive performance, importance of predictor variables, and observation-specific information.

Clinical trialist perspectives on the ethics of adaptive clinical trials: a mixed-methods analysis.

BACKGROUND: In an adaptive clinical trial (ACT), key trial characteristics may be altered during the course of the trial according to predefined rules in response to information that accumulates within the trial itself. In addition to having distinguishing scientific features, adaptive trials also may involve ethical considerations that differ from more traditional randomized trials. Better understanding of clinical trial experts' views about the ethical aspects of adaptive designs could assist those planning ACTs. Our aim was to elucidate the opinions of clinical trial experts regarding their beliefs about ethical aspects of ACTs. METHODS: We used a convergent, mixed-methods design employing a 22-item ACTs beliefs survey with visual analog scales and open-ended questions and mini-focus groups. We developed a coding scheme to conduct thematic searches of textual data, depicted responses to visual analog scales on box-plot diagrams, and integrated findings thematically. Fifty-three clinical trial experts from four constituent groups participated: academic biostatisticians (n = 5); consultant biostatisticians (n = 6); academic clinicians (n = 22); and other stakeholders including patient advocacy, National Institutes of Health, and U.S. Food and Drug Administration representatives (n = 20). RESULTS: The respondents recognized potential ethical benefits of ACTs, including a higher probability of receiving an effective intervention for participants, optimizing resource utilization, and accelerating treatment discovery. Ethical challenges voiced include developing procedures so trial participants can make informed decisions about taking part in ACTs and plausible, though unlikely risks of research personnel altering enrollment patterns. CONCLUSIONS: Clinical trial experts recognize ethical advantages but also pose potential ethical challenges of ACTs. The four constituencies differ in their weighing of ACT ethical considerations based on their professional vantage points. These data suggest further discussion about the ethics of ACTs is needed to facilitate ACT planning, design and conduct, and ultimately better allow planners to weigh ethical implications of competing trial designs.

BiMM tree: A decision tree method for modeling clustered and longitudinal binary outcomes.

Clustered binary outcomes are frequently encountered in clinical research (e.g. longitudinal studies). Generalized linear mixed models (GLMMs) for clustered endpoints have challenges for some scenarios (e.g. data with multi-way interactions and nonlinear predictors unknown a priori). We develop an alternative, data-driven method called Binary Mixed Model (BiMM) tree, which combines decision tree and GLMM within a unified framework. Simulation studies show that BiMM tree achieves slightly higher or similar accuracy compared to standard methods. The method is applied to a real dataset from the Acute Liver Failure Study Group.

Re-formulating non-inferiority trials as superiority trials: The case of binary outcomes.

Non-inferiority trials are conducted for a variety of reasons including to show that a new treatment has a negligible reduction in efficacy or safety when compared to the current standard treatment, or a more complex setting of showing that a new treatment has a negligible reduction in efficacy when compared to the current standard yet is superior in terms of other treatment characteristics. The latter reason for conducting a non-inferiority trial presents the challenge of deciding on a balance between a suitable reduction in efficacy, known as the non-inferiority margin, in return for a gain in other important treatment characteristics/findings. It would be ideal to alleviate the dilemma on the choice of margin in this setting by reverting to a traditional superiority trial design where a single p -value for superiority of both the most important endpoint (efficacy) and the most important finding (treatment characteristic) is provided. We discuss how this can be done using the information-preserving composite endpoint (IPCE) approach and consider binary outcome cases in which the combination of efficacy and treatment characteristics, but not one itself, paints a clear picture that the novel treatment is superior to the active control.

Predicting daily outcomes in acetaminophen-induced acute liver failure patients with machine learning techniques.

BACKGROUND/OBJECTIVE: Assessing prognosis for acetaminophen-induced acute liver failure (APAP-ALF) patients during the first week of hospitalization often presents significant challenges. Current models such as the King's College Criteria (KCC) and the Acute Liver Failure Study Group (ALFSG) Prognostic Index are developed to predict outcome using only a single time point on hospital admission. Models using longitudinal data are not currently available for APAP-ALF patients. We aim to develop and compare performance of prediction models for outcomes during the first week of hospitalization for APAP-ALF patients. METHODS: Models are developed for the ALFSG registry data to predict longitudinal outcomes for 1042 APAP-ALF patients enrolled 01/1998-02/2016. The primary outcome is defined as daily low versus high coma grade. Accuracy in prediction of outcome (AC), sensitivity (SN), specificity (SP) and area under the receiver operating curve (AUC) are compared between the following models: classification and regression tree, random forest, frequentist generalized linear mixed model (GLMM), Bayesian GLMM, BiMM tree, and BiMM forest using original and imputed datasets. RESULTS: BiMM tree offers predictive (test set) 63% AC, 72% SP and 53% SN for the original dataset, whereas BiMM forest offers predictive (test set) 69% AC, 63% SP and 74% SN for the imputed dataset. BiMM tree has the highest AUC for the original testing dataset (0.697), whereas BiMM forest and standard random forest have the highest AUC for the imputed testing dataset (0.749). The three most important predictors of daily outcome for the BiMM tree are pressor use, bilirubin and creatinine. The BiMM forest model identifies lactate, ammonia and ALT as the three most important predictors of outcome. CONCLUSIONS: BiMM tree offers a prognostic tool for APAP-ALF patients, which has good accuracy and simple interpretation of predictors which are consistent with clinical observations. BiMM tree and forest models are developed using the first week of in-patient data and are appropriate for predicting outcome over time. While the BiMM forest has slightly higher predictive AC, the BiMM tree model is simpler to use at the bedside.

Nosocomial infections due to methicillin-resistant Staphylococcus aureus and vancomycin-resistant enterococcus: relationships with antibiotic use and cost drivers.

BACKGROUND: Increased incidence of nosocomial infections due to methicillin-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant Enterococcus (VRE) has been associated with the use of certain antibiotics and has resulted in increased morbidity, mortality, and costs of care. OBJECTIVE: To describe relationships between vancomycin and linezolid use and incidence of these nosocomial infections over time and to determine factors associated with the increased costs of care (cost drivers) associated with affected patients. METHODS: The association between institution-wide antibiotic use and the rate of nosocomial MRSA and VRE infections was assessed using segmented regression analysis for interrupted time series. The effect that patient characteristics and procedures, as well as certain antibiotic use, had on costs and length of stay of patients with MRSA or VRE nosocomial infection was also assessed and cost drivers for the 2 types of infections were compared. RESULTS: Our analysis included 206 patients who developed MRSA (n = 187) or VRE (n = 19) nosocomial infection. Although small numbers of VRE nosocomial infection may limit generalizations from our results, we found no significant relationship between vancomycin or linezolid use and the rate of either infection. While mean hospital costs were similar, cost drivers varied somewhat between infection types. CONCLUSIONS: The incidence of MRSA or VRE infections does not appear to be related to the use of vancomycin or linezolid. Costs of care are quite high in some affected patients and, while mean total hospital costs are similar, cost drivers appear to differ between the 2 infection types.

Noninferiority and equivalence designs: issues and implications for mental health research.

The terms noninferiority and equivalence are often used interchangeably to refer to trials in which the primary objective is to show that a novel intervention is as effective as the standard intervention. The use of these designs is becoming increasingly relevant to mental health research. Despite the fundamental importance of these designs, they are often poorly understood, improperly applied, and subsequently misinterpreted. In this article, the authors explain noninferiority and equivalence designs and key methodological and statistical considerations. Decision points in using these designs are discussed, such as choice of control condition, determination of the noninferiority margin, and calculation of sample size and power. With increasing utilization of these designs, it is critical that researchers understand the methodological issues, advantages, disadvantages, and related challenges.

Evaluation of treatment with direct thrombin inhibitors in patients with heparin-induced thrombocytopenia.

STUDY OBJECTIVE: To evaluate the efficacy, safety, and associated costs of anticoagulation with argatroban, bivalirudin, and lepirudin for managing patients with heparin-induced thrombocytopenia (HIT) or presumed HIT. DESIGN: Retrospective medical record review. SETTING: University-affiliated teaching hospital. PATIENTS: Forty-two patients who were hospitalized between January 1 and December 31, 2004, and who were treated with bivalirudin, argatroban, or lepirudin for at least 24 hours. MEASUREMENTS AND MAIN RESULTS: The primary outcome was the time to reach the desired goal for activated partial thromboplastin time (aPTT). Secondary outcomes were the number of aPTT measurements within the therapeutic range, costs, treatment duration, clinical outcomes, and adverse events. Of the 42 patients who met the inclusion criteria, 24 received bivalirudin, 13 received argatroban, and 5 received lepirudin. Patients receiving bivalirudin who reached therapeutic aPTTs attained them sooner than those receiving either argatroban or lepirudin (8.5 vs 14 and 24 hrs, respectively, p=0.124). Average percentage of therapeutic aPTTs/patient was greatest in the argatroban group (62%), followed by the bivalirudin (57%) and lepirudin (29%) groups (p=0.062). Average drug cost/day/patient was greater in the lepirudin group than the other groups, whereas average laboratory costs were similar among groups. Treatment duration was longer with argatroban than with bivalirudin or lepirudin. Bleeding rates were similar in the argatroban and bivalirudin groups, but higher than in the lepirudin group. A composite of clinical outcomes (deep vein thrombosis, nonfatal myocardial infarction, nonfatal stroke, limb amputation, and all-cause mortality) were similar among the three groups. CONCLUSION: All three drugs were effective as anticoagulants for patients with HIT or presumed HIT. Based on average use and average wholesale price, bivalirudin cost less per day than the other two agents. Although not yet approved by the United States Food and Drug Administration for management of HIT, bivalirudin appears to be a viable treatment alternative for anticoagulation therapy.

Computed tomographic colonography (virtual colonoscopy): a multicenter comparison with standard colonoscopy for detection of colorectal neoplasia.

CONTEXT: Conventional colonoscopy is the best available method for detection of colorectal cancer; however, it is invasive and not without risk. Computed tomographic colonography (CTC), also known as virtual colonoscopy, has been reported to be reasonably accurate in the diagnosis of colorectal neoplasia in studies performed at expert centers. OBJECTIVE: To assess the accuracy of CTC in a large number of participants across multiple centers. DESIGN, SETTING, AND PARTICIPANTS: A nonrandomized, evaluator-blinded, noninferiority study design of 615 participants aged 50 years or older who were referred for routine, clinically indicated colonoscopy in 9 major hospital centers between April 17, 2000, and October 3, 2001. The CTC was performed by using multislice scanners immediately before standard colonoscopy; findings at colonoscopy were reported before and after segmental unblinding to the CTC results. MAIN OUTCOME MEASURES: The sensitivity and specificity of CTC and conventional colonoscopy in detecting participants with lesions sized at least 6 mm. Secondary outcomes included detection of all lesions, detection of advanced lesions, possible technical confounders, participant preferences, and evidence for increasing accuracy with experience. RESULTS: A total of 827 lesions were detected in 308 of 600 participants who underwent both procedures; 104 participants had lesions sized at least 6 mm. The sensitivity of CTC for detecting participants with 1 or more lesions sized at least 6 mm was 39.0% (95% confidence interval [CI], 29.6%-48.4%) and for lesions sized at least 10 mm, it was 55.0% (95% CI, 39.9%-70.0%). These results were significantly lower than those for conventional colonoscopy, with sensitivities of 99.0% (95% CI, 97.1%->99.9%) and 100%, respectively. A total of 496 participants were without any lesion sized at least 6 mm. The specificity of CTC and conventional colonoscopy for detecting participants without any lesion sized at least 6 mm was 90.5% (95% CI, 87.9%-93.1%) and 100%, respectively, and without lesions sized at least 10 mm, 96.0% (95% CI, 94.3%-97.6%) and 100%, respectively. Computed tomographic colonography missed 2 of 8 cancers. The accuracy of CTC varied considerably between centers and did not improve as the study progressed. Participants expressed no clear preference for either technique. CONCLUSIONS: Computed tomographic colonography by these methods is not yet ready for widespread clinical application. Techniques and training need to be improved.

The Stroke Hyperglycemia Insulin Network Effort (SHINE) trial protocol: a randomized, blinded, efficacy trial of standard vs. intensive hyperglycemia management in acute stroke.

RATIONALE: Patients with acute ischemic stroke and hyperglycemia have worse outcomes than those without hyperglycemia. Intensive glucose control during acute stroke is feasible and can be accomplished safely but has not been fully assessed for efficacy. AIMS: The Stroke Hyperglycemia Insulin Network Effort trial aims to determine the safety and efficacy of standard vs. intensive glucose control with insulin in hyperglycemic acute ischemic stroke patients. DESIGN: This is a randomized, blinded, multicenter, phase III trial of approximately 1400 hyperglycemic patients who receive either standard sliding scale subcutaneous insulin (blood glucose range 80-179 mg/dL, 4·44-9·93 mmol/L) or continuous intravenous insulin (target blood glucose 80-130 mg/dL, 4·44-7·21 mmol/L) for up to 72 h, starting within 12 h of stroke symptom onset. The acute treatment phase is single blind (for the patients), but the final outcome assessment is double blind. The study is powered to detect a 7% absolute difference in favorable outcome at 90 days. STUDY OUTCOMES: The primary outcome is a baseline severity adjusted 90-day modified Rankin Scale score, defined as 0, 0-1, or 0-2, if the baseline National Institutes of Health Stroke Scale score is 3-7, 8-14, or 15-22, respectively. The primary safety outcome is the rate of severe hypoglycemia (<40 mg/dL, <2·22 mmol/L). DISCUSSION: This trial will provide important novel information about preferred management of acute ischemic stroke patients with hyperglycemia. It will determine the potential benefits and risks of intensive glucose control during acute stroke.

The effect of covariate adjustment for baseline severity in acute stroke clinical trials with responder analysis outcomes.

BACKGROUND: Traditionally in acute stroke clinical trials, the primary clinical outcome employed is a dichotomized modified Rankin Scale (mRS). New statistical methods, such as responder analysis, are being used in stroke studies to address the concern that baseline prognostic variables, such as stroke severity, impact the likelihood of a successful outcome. Responder analysis allows the definition of success to vary according to baseline prognostic variables, producing a more clinically relevant insight into the actual effect of investigational treatments. It is unclear whether or not statistical analyses should adjust for prognostic variables when responder analysis is used, as the outcome already takes these prognostic variables into account. This research aims to investigate the effect of covariate adjustment in the responder analysis framework in order to determine the appropriate analytic method. METHODS: Using a current stroke clinical trial and its pilot studies to guide simulation parameters, 1,000 clinical trials were simulated at varying sample sizes under several treatment effects to assess power and type I error. Covariate-adjusted and unadjusted logistic regressions were used to estimate the treatment effect under each scenario. In the case of covariate-adjusted logistic regression, the trichotomized National Institute of Health Stroke Scale (NIHSS) was used in adjustment. RESULTS: Under various treatment effect settings, the operating characteristics of the unadjusted and adjusted analyses do not substantially differ. Power and type I error are preserved for both the unadjusted and adjusted analyses. CONCLUSIONS: Our results suggest that, under the given treatment effect scenarios, the decision whether or not to adjust for baseline severity when using a responder analysis outcome should be guided by the needs of the study, as type I error rates and power do not appear to vary largely between the methods. These findings are applicable to stroke trials which use the mRS for the primary outcome, but also provide a broader insight into the analysis of binary outcomes that are defined based on baseline prognostic variables. TRIAL REGISTRATION: This research is part of the Stroke Hyperglycemia Insulin Network Effort (SHINE) trial, Identification Number NCT01369069.

Rationale and design: telephone-delivered behavioral skills interventions for Blacks with Type 2 diabetes.

BACKGROUND: African Americans with Type 2 diabetes (T2DM) have higher prevalence of diabetes, poorer metabolic control, and greater risk for complications and death compared to American Whites. Poor outcomes in African Americans with T2DM can be attributed to patient, provider, and health systems level factors. Provider and health system factors account for <10% of variance in major diabetes outcomes including hemoglobin A1c (HbA1c), lipid control, and resource use. Key differences appear to be at the patient level. Of the patient level factors, consistent differences between African Americans and American Whites with T2DM have been found in diabetes knowledge, self-management skills, empowerment, and perceived control. A variety of interventions to improve diabetes self-management have been tested including: 1) knowledge interventions; 2) lifestyle interventions; 3) skills training interventions; and 4) patient activation and empowerment interventions. Most of these interventions have been tested individually, but rarely have they been tested in combination, especially among African Americans who have the greatest burden of diabetes related complications. This study provides a unique opportunity to address this gap in the literature. METHODS/DESIGN: We describe an ongoing four-year randomized clinical trial, using a 2 x 2 factorial design, which will test the efficacy of separate and combined telephone-delivered, diabetes knowledge/information and motivation/behavioral skills training interventions in high risk African Americans with poorly controlled T2DM (HbA1c >or= 9%). Two-hundred thirty-two (232) male and female African-American participants, 18 years of age or older and with an HbA1c >or= 9%, will be randomized into one of four groups for 12-weeks of phone interventions: (1) an education group, (2) a motivation/skills group, (3) a combined group or (4) a usual care/general health education group. Participants will be followed for 12-months to ascertain the effect of the interventions on glycemic control. Our primary hypothesis is that among African Americans with poorly controlled T2DM, patients randomized to the combined diabetes knowledge/information and motivation/behavioral skills training intervention will have significantly greater reduction in HbA1c at 12 months of follow-up compared to the usual care/general health education group. DISCUSSION: Results from this study will provide important insight into how best to deliver diabetes education and skills training in ethnic minorities and whether combined knowledge/information and motivation/behavioral skills training is superior to the usual method of delivering diabetes education for African Americans with poorly controlled T2DM. TRIAL REGISTRATION: National Institutes of Health Clinical Trials Registry (ClinicalTrials.gov identifier# NCT00929838).

An open-label assessment of aripiprazole in the treatment of PTSD.

BACKGROUND: Recent studies suggest that atypical antipsychotics may be effective augmentation strategies for the treatment of posttraumatic stress disorder (PTSD). Limited data were available on the newest agent, aripiprazole, so we aimed to evaluate its efficacy and tolerability in the treatment of PTSD. METHODS: A 12-week, prospective, open-label, flexible-dose, adjunctive trial of aripiprazole was conducted in military veterans meeting DSM-IV criteria for PTSD. Concomitant psychiatric medications continued unchanged, except for other neuroleptics which were not allowed. The primary outcome variable was change from baseline in the Clinician Administered PTSD scale (CAPS). RESULTS: All 17 subjects were male, with an average age of 57 years. Total CAPS scores decreased from 78.2 (SD = 17.8) at baseline to 60.0 (23.5) at study end (p = 0.002). Re-experiencing (CAPS-B) and avoidance/numbing symptoms (CAPS-C) were significantly improved, and trend level reductions were observed in hyperarousal symptoms (CAPS-D). Fifty-three percent (9/17) were considered responders, as defined by a decrease in total CAPS scores of at least 20%. Reductions in the Positive and Negative Symptom Scale (PANSS) total score and positive and general psychopathology subscale scores were statistically significant. The final average dose of aripiprazole was 13.06 (SD = 6.45) mg daily. Nine patients discontinued because of side effects. The most common adverse events consisted of gastro-intestinal disturbances, sedation, and psychomotor activation. Tolerability was improved with lower starting doses (e.g., 5 mg daily) and slow titration. CONCLUSIONS: Addition of aripiprazole to ongoing treatment further reduced PTSD symptoms in military veterans with severe PTSD. These preliminary findings await confirmation in randomized, controlled trials.

Open-label trial of escitalopram in the treatment of posttraumatic stress disorder.

BACKGROUND: Posttraumatic stress disorder (PTSD) is a highly prevalent, disabling illness. Selective serotonin reuptake inhibitors (SSRIs) are considered first-line medication treatment, with sertraline, paroxetine, and fluoxetine being the most studied. More limited but favorable data suggest that citalopram, an SSRI, may also have a role in the treatment of PTSD. Its S-enantiomer escitalopram, which may have faster onset and greater magnitude of effect than citalopram in other conditions, has not yet been investigated in PTSD. OBJECTIVE: To assess the efficacy, safety, and tolerability of escitalopram in the treatment of PTSD. METHOD: A 12-week, prospective, open-label trial of escitalopram was conducted from January 2003 through August 2004 in military veterans with PTSD. Escitalopram was initiated at 10 mg daily for 4 weeks, then increased to 20 mg daily for the remainder of the study. Concomitant psychiatric medications were discontinued at least 2 weeks prior to enrollment. The primary outcome variable was the change from baseline to endpoint in global Clinician-Administered PTSD Scale-Symptom version (CAPS-SX) score. Secondary efficacy measures included the Clinical Global Impressions-Severity of Illness (CGI-S) and -Improvement (CGI-I) scales, the Hamilton Rating Scale for Depression (HAM-D), and the Davidson Trauma Scale (DTS). Posttraumatic stress disorder and comorbid diagnoses were established using the Structured Clinical Interview for Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition. RESULTS: Twenty-four of 25 patients were evaluated for efficacy. The mean global CAPS-SX score decreased from 79.4 (SD = 15.7) at baseline to 61.2 (SD = 24.7) at the end of the study (p = .0002). The CAPS-C avoidance/numbing and CAPS-D hyper-arousal subscale scores decreased significantly from baseline to endpoint (CAPS-C, p = .0171; CAPS-D, p = .0001), with trend-level reductions observed in CAPS-B reexperiencing subscale scores (p = .0593). Forty-five percent of patients (9/20) were much or very much improved at the end of the study (CGI-I of 1 or 2). The HAM-D and DTS also significantly improved (p = .0063 and p = .0004, respectively). Mild to moderate gastrointestinal disturbances were the most common side effects. Only 4 patients discontinued early because of adverse effects. CONCLUSIONS: This preliminary open-label study suggests that escitalopram is both efficacious and well tolerated in PTSD patients. However, randomized controlled studies are needed to confirm these results and to further define its potential role in the treatment of PTSD.

Therapeutic inertia is an impediment to achieving the Healthy People 2010 blood pressure control goals.

Therapeutic inertia (TI), defined as the providers' failure to increase therapy when treatment goals are unmet, contributes to the high prevalence of uncontrolled hypertension (> or =140/90 mm Hg), but the quantitative impact is unknown. To address this gap, a retrospective cohort study was conducted on 7253 hypertensives that had > or =4 visits and > or =1 elevated blood pressure (BP) in 2003. A 1-year TI score was calculated for each patient as the difference between expected and observed medication change rates with higher scores reflecting greater TI. Antihypertensive therapy was increased on 13.1% of visits with uncontrolled BP. Systolic BP decreased in patients in the lowest quintile of the TI score but increased in those in the highest quintile (-6.8+/-0.5 versus +1.8+/-0.6 mm Hg; P<0.001). Individuals in the lowest TI quintile were &33 times more likely to have their BP controlled at the last visit than those in highest quintile (odds ratio, 32.7; 95% CI, 25.1 to 42.6; P<0.0001). By multivariable analysis, TI accounted for &19% of the variance in BP control. If TI scores were decreased &50%, that is, increasing medication dosages on &30% of visits, BP control would increase from the observed 45.1% to a projected 65.9% in 1 year. This study confirms the high rate of TI in uncontrolled hypertensive subjects. TI has a major impact on BP control in hypertensive subjects receiving regular care. Reducing TI is critical in attaining the Healthy People 2010 goal of controlling hypertension in 50% of all patients.

Analysis of trichotomous pharmaceutical endpoints.

Many important clinical trial endpoints are measured on an ordered categorical scale. These include objective tumor response in oncology trials, the Thombolysis in Myocardial Infarction (TIMI) flow in cardiovascular trials, and the American College of Radiology (A CR) criterion in rheumatology trials. A common tendency among researchers is to simplify the ordered outcomes and collapse the data into a 2 x 2 contingency table in order to perform simpler statistical tests, such as the chi-square test or Fisher's exact test. Recently, more appropriate approaches, such as adaptive tests, have been developed for the analysis of ordered categorical endpoints. Each test in the adaptive class of tests is exact and balances good global power with nearly optimal power to detect a specific alternative of most interest. Prior knowledge of the direction of the treatment effect and the level of confidence in this prior information can be used to select a specific test from this class. However, little guidance has been offered regarding the selection of adaptive parameters when prior information is available. The purpose of this paper is to fill this gap by offering an objective approach for parameter selection, and to provide real data examples to illustrate the use of this objective approach.