How Assessment-Schedule Matching Limits Bias When Comparing Progression-Free Survival in Single-Arm Studies: An Application in Second-Line Urothelial Carcinoma Treatments.

OBJECTIVES: Population-adjusted comparisons of progression-free survival (PFS) from single-arm trials of cancer treatments can be derived using matching-adjusted indirect comparisons (MAICs); however, results are still susceptible to bias, particularly if the trials had different tumor assessment schedules. This study aims to assess the effects of assessment-schedule matching (ASM) on the relative effectiveness on the PFS of avelumab versus approved comparator immunotherapies or chemotherapy after population matching in the second-line (2L) setting for metastatic urothelial carcinoma. METHODS: The MAIC used patient-level data for avelumab from the JAVELIN Solid Tumor trial (NCT01772004). PFS was compared with published curves for other treatments to obtain population-adjusted hazard ratios (HRs). The MAIC was repeated after conducting ASM for differences in tumor assessment scheduled first at 6 weeks for avelumab and durvalumab and at 8 or 9 weeks for other treatments. RESULTS: MAIC adjustment alone altered the HR estimates up to 23%, whereas MAIC plus ASM resulted in up to 32.7% reductions from naive comparisons. Even in cases in which MAIC had little effect, ASM brought an additional change of 11.1% to 15.4%. Overall, the HR range of avelumab versus other treatments changed from 0.83 to 1.25 for naive comparisons to 0.76 to 0.99 after ASM plus MAIC, numerically favoring avelumab. CONCLUSIONS: Small variations in assessment schedules can introduce bias in unanchored indirect treatment comparisons of interval-censored time-to-event outcomes. In this study, adjusted PFS was comparable across second-line urothelial carcinoma treatment options, numerically favoring avelumab versus immunotherapies and chemotherapy agents. Correcting this bias is especially important when HRs are applied in cost-effectiveness models to transition patients between states.

Assessing patient PREFERence between the dulaglutide pen and the semaglutide pen: A crossover study (PREFER).

AIM: When selecting treatments for type 2 diabetes (T2D), it is important to consider not only efficacy and safety, but also other treatment attributes that have an impact on patient preference. The objective of this study was to examine preference between injection devices used for two weekly GLP-1 receptor agonists. MATERIALS AND METHODS: The PREFER study was an open-label, multicentre, randomized, crossover study assessing patient preference for dulaglutide and semaglutide injection devices among injection-naïve patients receiving oral medication for type 2 diabetes. After being trained to use each device, participants performed all steps of injection preparation and administered mock injections into an injection pad. Time-to-train (TTT) for each device was assessed in a subset. RESULTS: There were 310 evaluable participants (48.4% female; mean age, 60.0 years; 78 participants in the TTT subgroup). More participants preferred the dulaglutide device than the semaglutide device (84.2% vs. 12.3%; P < 0.0001). More participants perceived the dulaglutide device to have greater ease of use (86.8% vs. 6.8%; P < 0.0001). After preparing and using the devices, more participants were willing to use the dulaglutide device (93.5%) than the semaglutide device (45.8%). Training participants to use the dulaglutide device required less time than the semaglutide device (3.38 vs. 8.14 minutes; P < 0.0001). CONCLUSIONS: Participants with type 2 diabetes preferred the dulaglutide injection device to the semaglutide injection device. If patients prefer a device, they may be more willing to use the medication, which could result in better health outcomes. Furthermore, a shorter training time for injection devices may be helpful in busy clinical practice settings.

Using Bayesian Adaptive Trial Designs for Comparative Effectiveness Research: A Virtual Trial Execution.

BACKGROUND: Bayesian and adaptive clinical trial designs offer the potential for more efficient processes that result in lower sample sizes and shorter trial durations than traditional designs. OBJECTIVE: To explore the use and potential benefits of Bayesian adaptive clinical trial designs in comparative effectiveness research. DESIGN: Virtual execution of ALLHAT (Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial) as if it had been done according to a Bayesian adaptive trial design. SETTING: Comparative effectiveness trial of antihypertensive medications. PATIENTS: Patient data sampled from the more than 42 000 patients enrolled in ALLHAT with publicly available data. MEASUREMENTS: Number of patients randomly assigned between groups, trial duration, observed numbers of events, and overall trial results and conclusions. RESULTS: The Bayesian adaptive approach and original design yielded similar overall trial conclusions. The Bayesian adaptive trial randomly assigned more patients to the better-performing group and would probably have ended slightly earlier. LIMITATIONS: This virtual trial execution required limited resampling of ALLHAT patients for inclusion in RE-ADAPT (REsearch in ADAptive methods for Pragmatic Trials). Involvement of a data monitoring committee and other trial logistics were not considered. CONCLUSION: In a comparative effectiveness research trial, Bayesian adaptive trial designs are a feasible approach and potentially generate earlier results and allocate more patients to better-performing groups. PRIMARY FUNDING SOURCE: National Heart, Lung, and Blood Institute.

Do Bayesian adaptive trials offer advantages for comparative effectiveness research? Protocol for the RE-ADAPT study.

BACKGROUND: Randomized clinical trials, particularly for comparative effectiveness research (CER), are frequently criticized for being overly restrictive or untimely for health-care decision making. PURPOSE: Our prospectively designed REsearch in ADAptive methods for Pragmatic Trials (RE-ADAPT) study is a 'proof of concept' to stimulate investment in Bayesian adaptive designs for future CER trials. METHODS: We will assess whether Bayesian adaptive designs offer potential efficiencies in CER by simulating a re-execution of the Antihypertensive and Lipid Lowering Treatment to Prevent Heart Attack Trial (ALLHAT) study using actual data from ALLHAT. RESULTS: We prospectively define seven alternate designs consisting of various combinations of arm dropping, adaptive randomization, and early stopping and describe how these designs will be compared to the original ALLHAT design. We identify the one particular design that would have been executed, which incorporates early stopping and information-based adaptive randomization. LIMITATIONS: While the simulation realistically emulates patient enrollment, interim analyses, and adaptive changes to design, it cannot incorporate key features like the involvement of data monitoring committee in making decisions about adaptive changes. CONCLUSION: This article describes our analytic approach for RE-ADAPT. The next stage of the project is to conduct the re-execution analyses using the seven prespecified designs and the original ALLHAT data.

Accounting for the relationship between per diem cost and LOS when estimating hospitalization costs.

BACKGROUND: Hospitalization costs in clinical trials are typically derived by multiplying the length of stay (LOS) by an average per-diem (PD) cost from external sources. This assumes that PD costs are independent of LOS. Resource utilization in early days of the stay is usually more intense, however, and thus, the PD cost for a short hospitalization may be higher than for longer stays. The shape of this relationship is unlikely to be linear, as PD costs would be expected to gradually plateau. This paper describes how to model the relationship between PD cost and LOS using flexible statistical modelling techniques. METHODS: An example based on a clinical study of clevidipine for the treatment of peri-operative hypertension during hospitalizations for cardiac surgery is used to illustrate how inferences about cost-savings associated with good blood pressure (BP) control during the stay can be affected by the approach used to derive hospitalization costs.Data on the cost and LOS of hospitalizations for coronary artery bypass grafting (CABG) from the Massachusetts Acute Hospital Case Mix Database (the MA Case Mix Database) were analyzed to link LOS to PD cost, factoring in complications that may have occurred during the hospitalization or post-discharge. The shape of the relationship between LOS and PD costs in the MA Case Mix was explored graphically in a regression framework. A series of statistical models including those based on simple logarithmic transformation of LOS to more flexible models using LOcally wEighted Scatterplot Smoothing (LOESS) techniques were considered. A final model was selected, using simplicity and parsimony as guiding principles in addition traditional fit statistics (like Akaike's Information Criterion, or AIC). This mapping was applied in ECLIPSE to predict an LOS-specific PD cost, and then a total cost of hospitalization. These were then compared for patients who had good vs. poor peri-operative blood-pressure control. RESULTS: The MA Case Mix dataset included data from over 10,000 patients. Visual inspection of PD vs. LOS revealed a non-linear relationship. A logarithmic model and a series of LOESS and piecewise-linear models with varying connection points were tested. The logarithmic model was ultimately favoured for its fit and simplicity. Using this mapping in the ECLIPSE trials, we found that good peri-operative BP control was associated with a cost savings of $5,366 when costs were derived using the mapping, compared with savings of $7,666 obtained using the traditional approach of calculating the cost. CONCLUSIONS: PD costs vary systematically with LOS, with short stays being associated with high PD costs that drop gradually and level off. The shape of the relationship may differ in other settings. It is important to assess this and model the observed pattern, as this may have an impact on conclusions based on derived hospitalization costs.

Health state utilities associated with weight loss in type 2 diabetes and obesity.

INTRODUCTION: Health state utilities associated with weight change are needed as inputs for cost-utility analyses (CUAs) examining the value of treatments for obesity and type 2 diabetes (T2D). Although some pharmaceutical treatments currently in development are associated with substantial weight loss, little is known about the utility impact of weight decreases greater than 10%. The purpose of this study was to estimate utilities associated with body weight decreases up to 20% based on preferences of individuals with obesity, with and without T2D. METHODS: Health state vignettes were developed to represent respondents' own current weight and weight decreases of 2.5, 5, 10, 15, and 20%. Health state utilities were elicited in time trade-off interviews in two UK locations (Edinburgh and London) with a sample of participants with obesity, with and without T2D. Mean utility increases associated with each amount of weight decrease were calculated. Regression analyses were performed to derive a method for estimating utility change associated with weight decreases. RESULTS: Analyses were conducted with data from 405 individuals with obesity (202 with T2D, 203 without T2D). Utility increases associated with various levels of weight decrease ranged from 0.011 to 0.060 in the subgroup with T2D and 0.015 to 0.077 in the subgroup without T2D. All regression models found that the percentage of weight decrease was a highly significant predictor of change in utility (p < .0001). The relationship between weight change and utility change did not appear to be linear. Equations are recommended for estimating utility change based on the natural logarithm of percentage of weight decrease. DISCUSSION: Results of this study may be used to provide inputs for CUAs examining and comparing the value of treatments that are associated with substantial amounts of weight loss in patients with obesity, with or without T2D.

Costs and cost-effectiveness of carotid stenting versus endarterectomy for patients at increased surgical risk: results from the SAPPHIRE trial.

OBJECTIVES: To evaluate the cost-effectiveness of carotid stenting vs. carotid endarterectomy using data from the SAPPHIRE trial. BACKGROUND: Carotid stenting with embolic protection has been introduced as an alternative to carotid endarterectomy for prevention of cerebrovascular and cardiovascular events among patients at increased surgical risk. METHODS: Between August 2000 and July 2002, 310 patients with an accepted indication for carotid endarterectomy but at high risk of complications were randomized to and subsequently underwent either carotid stenting (n = 159) or endarterectomy (n = 151). Clinical outcomes, resource use, costs, and quality of life were assessed prospectively for all patients over a 1-year period. Life expectancy, quality-adjusted life expectancy, and health care costs beyond the follow-up period were estimated for patients alive at 1 year, based on observed clinical events during the first year of follow-up. RESULTS: Although initial procedural costs were significantly higher for stenting than for endarterectomy (mean difference: $4,081/patient; 95% CI, $3,849-$4,355), mean post-procedure length of stay was shorter for stenting (1.9 vs. 2.9 days; P < 0.001) with significant associated cost offsets. As a result, initial hospital costs were just $559/patient higher with stenting (95% CI, $3,470 less to $2,289 more). Neither follow-up costs after discharge nor total 1-year costs differed significantly. The incremental cost-effectiveness ratio for stenting compared with endarterectomy was $6,555 per quality-adjusted life year (QALY) gained, with over 98 percent of bootstrap estimates < $50,000/QALY gained. CONCLUSIONS: Although carotid stenting with embolic protection is more costly than carotid endarterectomy, by commonly accepted standards, stenting is an economically attractive alternative to endarterectomy for patients at high surgical risk.

Adjusting for patient crossover in clinical trials using external data: a case study of lenalidomide for advanced multiple myeloma.

OBJECTIVES: In some trials, particularly in oncology, patients whose disease progresses under the comparator treatment are crossed over into the experimental arm. This unplanned crossover can introduce bias in analyses because patients who crossover likely have a different prognosis than those who do not cross over; for instance, sicker patients not responding to standard therapy or those expected to benefit the most may be selectively chosen to receive the experimental treatment. Standard statistical methods cannot adequately correct for this bias. We describe an approach designed to minimize the impact of crossover, and illustrate this by using data from two randomized trials in multiple myeloma (MM). METHODS: The MM-009/010 trials compared lenalidomide and high-dose dexamethasone (Len+Dex) with dexamethasone alone (Dex). Nearly half (47%) of the patients randomized to Dex crossed over to Len with or without Dex (Len+/-Dex) at disease progression or study unblinding. Data from these trials was used to predict survival in an economic model evaluating the cost-effectiveness of lenalidomide. To adjust for crossover, the prediction equations were calibrated to match survival with Dex or Dex-equivalent therapies in trials conducted by the Medical Research Council (MRC) in the United Kingdom. To adjust for differences between the MM and MRC trial populations, a prediction equation was developed from the MRC data and used to predict survival by setting predictors to mean values for patients in the MM-009/010 trials. The expected survival with Dex without crossover was then predicted from the calibrated MM-009/010 equation (i.e., adjusted to match survival predicted from the MRC equation). RESULTS: The adjusted median overall survival predicted by the MRC equation was 19.5 months (95%CI, 16.6-22.9) for patients with one prior therapy, and 11.6 months (95% CI, 9.5-14.2) for patients with >1 prior therapy. These estimates are considerably shorter than was observed in the clinical trials: 33.6 months (27.1-NE) and 27.3 months (95% CI, 23.3-33.3) as of December 2005. CONCLUSION: The calibration method described here is simple to implement, provided that suitable data are available; it can be implemented with other types of endpoints in any therapeutic area.

Bayesian meta-analyses for comparative effectiveness and informing coverage decisions.

BACKGROUND: Evidence-based medicine is increasingly expected in health care decision-making. The Centers for Medicare and Medicaid have initiated efforts to understand the applicability of Bayesian techniques for synthesizing evidence. As a case study, a Bayesian analysis of clinical trials of implantable cardioverter defibrillators was undertaken using patient-level data not typically available for analysis. PURPOSE: Conduct Bayesian meta-analyses of the defibrillator trials using published results to demonstrate a Bayesian approach useful to policy makers. DATA SOURCES, STUDY SELECTION, DATA EXTRACTION: We reconsidered trials in a 2007 systematic review by Ezekowitz et al (Ann Intern Med. 2007;147:251-262) and extracted information from the original published articles. Employing a Bayesian hierarchical approach, we developed a base model and 2 variants, and modeled hazard ratios separately within each year of follow-up. We considered sequential meta-analyses over time and found the predictive distribution of the results of the next trial, given its sample size. DATA SYNTHESIS: For the most robust of 3 models, the probability that the mean defibrillator effect (in the population of trials) is beneficial is greater than 0.999. In that model, about 5% of trials in the population of trials would have a detrimental effect. Despite the moderate amount of heterogeneity across the trials, there was stability of conclusions after the first 3 of the 12 total trials had been conducted. This stability enabled reasonable predictions for the results of future trials. LIMITATIONS: Inability to assess treatment effects within subsets of patients. CONCLUSIONS: Bayesian meta-analyses based on literature surveys can effectively inform coverage decisions. Bayesian modeling for endpoints such as mortality can elucidate treatment effects over time. The Bayesian approach used in a sequential manner over time can predict results and help assess the utility of future clinical trials.

Assessment-Schedule Matching in Unanchored Indirect Treatment Comparisons of Progression-Free Survival in Cancer Studies.

BACKGROUND: The timing of efficacy-related clinical events recorded at scheduled study visits in clinical trials are interval censored, with the interval duration pre-determined by the study protocol. Events may happen any time during that interval but can only be detected during a planned or unplanned visit. Disease progression in oncology is a notable example where the time to an event is affected by the schedule of visits within a study. This can become a source of bias when studies with varying assessment schedules are used in unanchored comparisons using methods such as matching-adjusted indirect comparisons. OBJECTIVE: We illustrate assessment-time bias (ATB) in a simulation study based on data from a recent study in second-line treatment for locally advanced or metastatic urothelial carcinoma, and present a method to adjust for differences in assessment schedule when comparing progression-free survival (PFS) against a competing treatment. METHODS: A multi-state model for death and progression was used to generate simulated death and progression times, from which PFS times were derived. PFS data were also generated for a hypothetical comparator treatment by applying a constant hazard ratio (HR) to the baseline treatment. Simulated PFS times for the two treatments were then aligned to different assessment schedules so that progression events were only observed at set visit times, and the data were analysed to assess the bias and standard error of estimates of HRs between two treatments with and without assessment-schedule matching (ASM). RESULTS: ATB is highly affected by the rate of the event at the first assessment time; in our examples, the bias ranged from 3 to 11% as the event rate increased. The proposed method relies on individual-level data from a study and attempts to adjust the timing of progression events to the comparator's schedule by shifting them forward or backward without altering the patients' actual follow-up time. The method removed the bias almost completely in all scenarios without affecting the precision of estimates of comparative effectiveness. CONCLUSIONS: Considering the increasing use of unanchored comparative analyses for novel cancer treatments based on single-arm studies, the proposed method offers a relatively simple means of improving the accuracy of relative benefits of treatments on progression times.

Risk of discontinuation of risperidone after exposure to potentially interacting drugs: a nested case-control study in patients with schizophrenia.

BACKGROUND: The cytochrome P450 (CYP) 2D6 and 3A4 isozymes play an important role in the metabolism of risperidone. Concurrent use of drugs that inhibit or induce the action of these enzymes may increase or decrease levels of risperidone, thereby increasing the risk for discontinuation of risperidone or the need for supplemental treatment to control disease symptoms. OBJECTIVE: This study examined the association between exposure to potentially interacting drugs and nonpersistence in a cohort of patients with schizophrenia newly starting treatment with risperidone. METHODS: The data for this nested case-control study were obtained from the administrative health databases of the Régie de l'Assurance Maladie de Québec. The base cohort included patients aged > or = 15 years who began treatment with risperidone between July 2001 and December 2004. Cases consisted of those who were nonpersistent with risperidone either through drug discontinuation or the addition of/switch to a different atypical antipsychotic; noncases were those who persisted with risperidone through the end of their follow-up period. Exposure to CYP-inhibiting or CYP-inducing medications in the 1-, 3-, and 6-month windows before nonpersistence was compared between cases and noncases. The association between exposure to interacting medications and nonpersistence was analyzed using conditional logistic regression models to account for matching by time on treatment. RESULTS: The base cohort included 20,840 patients, of whom 59.2% were female and 57.7% were aged > or = 65 years. Nonpersistence occurred in 10,913 patients (52.4%) during the study period. Between 40% and 50% of patients were exposed to potential CYP inhibitors, and 6% to 10% were exposed to potential CYP inducers. Exposure to CYP inhibitors over 3 and 6 months was associated with an increase in risk for nonpersistence of approximately 10% (odds ratio [OR] for 3-month exposure = 1.10 [95% CI, 1.06-1.14]; OR for 6-month exposure = 1.11 [95% CI, 1.07-1.15]), whereas exposure to CYP inducers was not associated with a significant change in risk. Exposure to potentially interacting drugs was more likely to lead to nonpersistence while patients were still new to risperidone (ie, in the first month of treatment). For instance, the OR for 6-month exposure to inhibiting drugs was 1.20 (95% CI, 1.04-1.39) in the 1st month of treatment and 1.11 (95% CI, 1.01-1.20) between the 6th and 12th months. CONCLUSION: In this study, use of medications that are potential inhibitors of CYP2D6 and CYP3A4 appeared to be associated with an increased risk of nonpersistence with risperidone, particularly while patients were still new to treatment.

Alzheimer's disease Archimedes condition-event simulator: Development and validation.

INTRODUCTION: Several advances have been made in Alzheimer's Disease (AD) modeling, however, there remains a need for a simulator that represents the full scope of disease progression and can be used to study new disease-modifying treatments for early-stage and even prodromal AD. METHODS: We developed AD Archimedes condition-event simulator, a patient-level simulator with a focus on simulating the effects of early interventions through changes in biomarkers of AD. The simulator incorporates interconnected predictive equations derived from longitudinal data sets. RESULTS: The results of external validations on AD Archimedes condition-event simulator showed that it provides reasonable estimates once compared to literature results on transition to dementia AD, institutionalization, and mortality. A case study comparing a disease-modifying treatment and a symptomatic treatment also showcases the benefits of early treatment. DISCUSSION: The AD Archimedes condition-event simulator is designed to perform economic evaluation on various interventions through close tracking of disease progression and the related clinical outcomes.

Relative effectiveness of sunitinib versus everolimus in advanced pancreatic neuroendocrine tumors: an updated matching-adjusted indirect comparison.

AIM: A matching-adjusted indirect comparison (MAIC) of sunitinib and everolimus has been previously reported based on the RADIANT-3 everolimus trial. We performed an analysis using updated overall survival (OS) data based on sunitinib's trial (A6181111). METHODS: The MAIC matched on all baseline characteristics available from both studies. An anchored MAIC was performed for progression-free survival (PFS); an unanchored analysis was deemed more appropriate for OS due to crossover in both trials. A hazard ratio for sunitinib versus everolimus was derived from adjusted (weighted) sunitinib effects compared with the observed results for everolimus. RESULTS: The adjusted hazard ratio for sunitinib versus everolimus was 0.85 (0.39-1.89) for PFS and 0.82 (0.53-1.27) for OS. CONCLUSION: Findings indicate comparable PFS and OS with sunitinib and everolimus.

Simulation and matching-based approaches for indirect comparison of treatments.

Estimates of the relative effects of competing treatments are rarely available from head-to-head trials. These effects must therefore be derived from indirect comparisons of results from different studies. The feasibility of comparisons relies on the network linking treatments through common comparators; the reliability of these may also be impacted when the studies are heterogeneous or when multiple intermediate comparisons are needed to link two specific treatments of interest. Simulated treatment comparison and matching-adjusted indirect comparison have been developed to address these challenges. These focus on comparisons of outcomes for two specific treatments of interest by using patient-level data for one treatment (the index) and published results for the other treatment (the comparator) from compatible studies, taking into account possible confounding due to population differences. This paper provides an overview of how and when these approaches can be used as an alternative or to complement standard MTC approaches.

Canadian economic comparison of extended-release oxybutynin and immediate-release tolterodine in the treatment of overactive bladder.

BACKGROUND: Overactive bladder (OAB) is a condition characterized by urgency, increased frequency of micturition, or urge incontinence. It affects a considerable segment of the population, particularly with increasing age. Pharmacotherapy is one of the most common approaches to the treatment of OAB. OBJECTIVE: This article describes the development and results of a model comparing health-economic outcomes for the new extended-release (XL) formulation of oxybutynin and immediate-release (IR) tolterodine in a population of community-dwelling Canadian adults with OAB. METHODS: A Markov model was developed to compare health-economic outcomes over the course of 1 year. Effectiveness and treatment-persistence data were derived from the OBJECT (Overactive Bladder: Judging Effective Control and Treatment) trial, a 3-month comparison of oxybutynin XL 10 mg and tolterodine IR 4 mg, and were used, together with data from the literature (identified through a MEDLINE search of articles published between 1990 and 2003), to project outcomes beyond the trial period. Severity-specific cost profiles for incontinence were developed. In the principal analyses, cost items were limited to drug therapy, physician visits, use of pads or other protection, and laundry costs. Costs are reported in 2002 Canadian dollars. RESULTS: Costs after 1 year were estimated to be an average of $32 less per patient for oxybutynin XL compared with tolterodine IR, and 3.1 additional patients in every 100 who received oxybutynin XL were expected to attain complete continence compared with those who received tolterodine. During the course of 1 year, patients receiving oxybutynin XL were expected to have a mean 16.5 additional incontinence-free days compared with those receiving tolterodine IR. The results were sensitive to relative drug prices. In the other sensitivity analyses, however, oxybutyrin XL maintained its advantage over a wide range of inputs. CONCLUSION: The results of these analyses suggest that when priced equivalently, oxybutynin XL would reduce costs and provide better results than tolterodine IR over 1 year of treatment.

Methods for adjusting for bias due to crossover in oncology trials.

Trials of new oncology treatments often involve a crossover element in their design that allows patients receiving the control treatment to crossover to receive the experimental treatment at disease progression or when sufficient evidence about the efficacy of the new treatment is achieved. Crossover leads to contamination of the initial randomized groups due to a mixing of the effects of the control and experimental treatments in the reference group. This is further complicated by the fact that crossover is often a very selective process whereby patients who switch treatment have a different prognosis than those who do not. Standard statistical techniques, including those that attempt to account for the treatment switch, cannot fully adjust for the bias introduced by crossover. Specialized methods such as rank-preserving structural failure time (RPSFT) models and inverse probability of censoring weighted (IPCW) analyses are designed to deal with selective treatment switching and have been increasingly applied to adjust for crossover. We provide an overview of the crossover problem and highlight circumstances under which it is likely to cause bias. We then describe the RPSFT and IPCW methods and explain how these methods adjust for the bias, highlighting the assumptions invoked in the process. Our aim is to facilitate understanding of these complex methods using a case study to support explanations. We also discuss the implications of crossover adjustment on cost-effectiveness results.

Overview of parametric survival analysis for health-economic applications.

Health economic models rely on data from trials to project the risk of events (e.g., death) over time beyond the span of the available data. Parametric survival analysis methods can be applied to identify an appropriate statistical model for the observed data, which can then be extrapolated to derive a complete time-to-event curve. This paper describes the properties of the most commonly used statistical distributions as a basis for these models and describes an objective process of identifying the most suitable parametric distribution in a given dataset. The approach can be applied with both individual-patient data as well as with survival probabilities derived from published Kaplan-Meier curves. Both are illustrated with analyses of overall survival from the Sorafenib Hepatocellular Carcinoma Assessment Randomised Protocol trial.

Smoking cessation treatment and outcomes patterns simulation: a new framework for evaluating the potential health and economic impact of smoking cessation interventions.

BACKGROUND: Most existing models of smoking cessation treatments have considered a single quit attempt when modelling long-term outcomes. OBJECTIVE: To develop a model to simulate smokers over their lifetimes accounting for multiple quit attempts and relapses which will allow for prediction of the long-term health and economic impact of smoking cessation strategies. METHODS: A discrete event simulation (DES) that models individuals' life course of smoking behaviours, attempts to quit, and the cumulative impact on health and economic outcomes was developed. Each individual is assigned one of the available strategies used to support each quit attempt; the outcome of each attempt, time to relapses if abstinence is achieved, and time between quit attempts is tracked. Based on each individual's smoking or abstinence patterns, the risk of developing diseases associated with smoking (chronic obstructive pulmonary disease, lung cancer, myocardial infarction and stroke) is determined and the corresponding costs, changes to mortality, and quality of life assigned. Direct costs are assessed from the perspective of a comprehensive US healthcare payer ($US, 2012 values). Quit attempt strategies that can be evaluated in the current simulation include unassisted quit attempts, brief counselling, behavioural modification therapy, nicotine replacement therapy, bupropion, and varenicline, with the selection of strategies and time between quit attempts based on equations derived from survey data. Equations predicting the success of quit attempts as well as the short-term probability of relapse were derived from five varenicline clinical trials. RESULTS: Concordance between the five trials and predictions from the simulation on abstinence at 12 months was high, indicating that the equations predicting success and relapse in the first year following a quit attempt were reliable. Predictions allowing for only a single quit attempt versus unrestricted attempts demonstrate important differences, with the single quit attempt simulation predicting 19 % more smoking-related diseases and 10 % higher costs associated with smoking-related diseases. Differences are most prominent in predictions of the time that individuals abstain from smoking: 13.2 years on average over a lifetime allowing for multiple quit attempts, versus only 1.2 years with single quit attempts. Differences in abstinence time estimates become substantial only 5 years into the simulation. In the multiple quit attempt simulations, younger individuals survived longer, yet had lower lifetime smoking-related disease and total costs, while the opposite was true for those with high levels of nicotine dependence. CONCLUSION: By allowing for multiple quit attempts over the course of individuals' lives, the simulation can provide more reliable estimates on the health and economic impact of interventions designed to increase abstinence from smoking. Furthermore, the individual nature of the simulation allows for evaluation of outcomes in populations with different baseline profiles. DES provides a framework for comprehensive and appropriate predictions when applied to smoking cessation over smoker lifetimes.

Improved perioperative blood pressure control leads to reduced hospital costs.

BACKGROUND: Perioperative hypertension affects 80% of cardiac surgery patients and is associated with an increased risk of complications. OBJECTIVE: To determine the relationship between perioperative blood pressure (BP) control and hospital costs for cardiac surgery in the United States (US) and estimate the potential cost reductions associated with effective therapies. METHODS: The analysis estimated hospitalization costs (2011 US dollars (USD)) for cardiac surgery when BP was controlled with intravenous (IV) antihypertensives. Patient characteristics, hospital length of stay, and clinical event rates during the initial hospitalization and post-discharge 30 days after study drug infusion were based on the ECLIPSE (Evaluation of CLevidipine In the Perioperative Treatment of Hypertension Assessing Safety Events) trials. These clinical trial data were combined with data from the Massachusetts Acute Hospital Case Mix Database 2007 - 2009 (MA Case Mix Database) to estimate total hospitalization costs. RESULTS: Effective perioperative BP control in patients requiring IV antihypertensives was associated with a 7% decrease in hospital costs compared with less effective BP control. Reductions in total hospital costs associated with clevidipine versus other IV antihypertensives averaged $394 per patient overall. Cost savings with clevidipine exceeded $500 per patient versus sodium nitroprusside and nitroglycerin, but only $22 compared to nicardipine. CONCLUSION: Improved perioperative BP control may reduce hospital costs. Given the low cost of IV antihypertensives, the total hospital cost reductions may offset any incremental cost increases associated with newer, more effective therapies.

Optimizing the design of pragmatic trials: key issues remain.

Clinical trials have largely focused on whether an intervention can work. To ensure valid and powerful testing of this hypothesis, trials attempt to maximize the effect of the intervention of interest, controlling other factors that can confound comparisons. The benefits observed in these studies are often not sustained once the treatment is used in routine care, leaving regulators, practitioners and patients with a paucity of reliable evidence to assist decision-making. Attempts to address this need have led to 'pragmatic trials' that prioritize applicability of findings to real-world practice by minimizing design features that produce less pertinent information. Minimizing biases in this pragmatic context remains a very difficult task, however. This paper reviews some of these challenges and highlights specific aspects of design that must be approached with a pragmatic attitude.