Bayesian designs with frequentist and Bayesian error rate considerations

So far, most Phase II trials have been designed and analysed under a frequentist framework. Under this framework, a trial is designed so that the overall Type I and Type II errors of the trial are controlled at some desired levels. Recently, a number of articles have advocated the use of Bavesian designs in practice. Under a Bayesian framework, a trial is designed so that the trial stops when the posterior probability of treatment is within certain prespecified thresholds. In this article, we argue that trials under a Bayesian framework can also be designed to control frequentist error rates. We introduce a Bayesian version of Simon's well-known two-stage design to achieve this goal. We also consider two other errors, which are called Bayesian errors in this article because of their similarities to posterior probabilities. We show that our method can also control these Bayesian-type errors. We compare our method with other recent Bayesian designs in a numerical study and discuss implications of different designs on error rates. An example of a clinical trial for patients with nasopharyngeal carcinoma is used to illustrate differences of the different designs.

An optimal design for screening trials

Yao, Begg, and Livingston (1996, Biometrics 52, 992-1001) considered the optimal group size for testing a series of potentially therapeutic agents to identify a promising one as soon as possible for given error rates. The number of patients to be tested with each agent was fixed as the group size. We consider a sequential design that allows early acceptance and rejection, and we provide an optimal strategy to minimize the sample sizes (patients) required using Markov decision processes. The minimization is under the constraints of the two types (false positive and false negative) of error probabilities, with the Lagrangian multipliers corresponding to the cost parameters for the two types of errors. Numerical studies indicate that there can be a substantial reduction in the number of patients required.

Induced smoothing for rank regression with censored survival times

Adaptions of weighted rank regression to the accelerated failure time model for censored survival data have been successful in yielding asymptotically normal estimates and flexible weighting schemes to increase statistical efficiencies. However, for only one simple weighting scheme, Gehan or Wilcoxon weights, are estimating equations guaranteed to be monotone in parameter components, and even in this case are step functions, requiring the equivalent of linear programming for computation. The lack of smoothness makes standard error or covariance matrix estimation even more difficult. An induced smoothing technique overcame these difficulties in various problems involving monotone but pure jump estimating equations, including conventional rank regression. The present paper applies induced smoothing to the Gehan-Wilcoxon weighted rank regression for the accelerated failure time model, for the more difficult case of survival time data subject to censoring, where the inapplicability of permutation arguments necessitates a new method of estimating null variance of estimating functions. Smooth monotone parameter estimation and rapid, reliable standard error or covariance matrix estimation is obtained.

Unbiased estimating equations from working correlation models for irregularly timed repeated measures

The method of generalized estimating equation-, (GEEs) has been criticized recently for a failure to protect against misspecification of working correlation models, which in some cases leads to loss of efficiency or infeasibility of solutions. However, the feasibility and efficiency of GEE methods can be enhanced considerably by using flexible families of working correlation models. We propose two ways of constructing unbiased estimating equations from general correlation models for irregularly timed repeated measures to supplement and enhance GEE. The supplementary estimating equations are obtained by differentiation of the Cholesky decomposition of the working correlation, or as score equations for decoupled Gaussian pseudolikelihood. The estimating equations are solved with computational effort equivalent to that required for a first-order GEE. Full details and analytic expressions are developed for a generalized Markovian model that was evaluated through simulation. Large-sample ".sandwich" standard errors for working correlation parameter estimates are derived and shown to have good performance. The proposed estimating functions are further illustrated in an analysis of repeated measures of pulmonary function in children.

Analysing commercial catch and effort data from a penaeid trawl fishery: A comparison of linear models, mixed models, and generalised estimating equations approaches

Statistical methods are often used to analyse commercial catch and effort data to provide standardised fishing effort and/or a relative index of fish abundance for input into stock assessment models. Achieving reliable results has proved difficult in Australia's Northern Prawn Fishery (NPF), due to a combination of such factors as the biological characteristics of the animals, some aspects of the fleet dynamics, and the changes in fishing technology. For this set of data, we compared four modelling approaches (linear models, mixed models, generalised estimating equations, and generalised linear models) with respect to the outcomes of the standardised fishing effort or the relative index of abundance. We also varied the number and form of vessel covariates in the models. Within a subset of data from this fishery, modelling correlation structures did not alter the conclusions from simpler statistical models. The random-effects models also yielded similar results. This is because the estimators are all consistent even if the correlation structure is mis-specified, and the data set is very large. However, the standard errors from different models differed, suggesting that different methods have different statistical efficiency. We suggest that there is value in modelling the variance function and the correlation structure, to make valid and efficient statistical inferences and gain insight into the data. We found that fishing power was separable from the indices of prawn abundance only when we offset the impact of vessel characteristics at assumed values from external sources. This may be due to the large degree of confounding within the data, and the extreme temporal changes in certain aspects of individual vessels, the fleet and the fleet dynamics.

Early stopping by using stochastic curtailment in a three-arm sequential trial

Summary. Interim analysis is important in a large clinical trial for ethical and cost considerations. Sometimes, an interim analysis needs to be performed at an earlier than planned time point. In that case, methods using stochastic curtailment are useful in examining the data for early stopping while controlling the inflation of type I and type II errors. We consider a three-arm randomized study of treatments to reduce perioperative blood loss following major surgery. Owing to slow accrual, an unplanned interim analysis was required by the study team to determine whether the study should be continued. We distinguish two different cases: when all treatments are under direct comparison and when one of the treatments is a control. We used simulations to study the operating characteristics of five different stochastic curtailment methods. We also considered the influence of timing of the interim analyses on the type I error and power of the test. We found that the type I error and power between the different methods can be quite different. The analysis for the perioperative blood loss trial was carried out at approximately a quarter of the planned sample size. We found that there is little evidence that the active treatments are better than a placebo and recommended closure of the trial.

Does the size of subsamples taken from multispecies trawl catches affect estimates of catch composition and abundance?

Although subsampling is a common method for describing the composition of large and diverse trawl catches, the accuracy of these techniques is often unknown. We determined the sampling errors generated from estimating the percentage of the total number of species recorded in catches, as well as the abundance of each species, at each increase in the proportion of the sorted catch. We completely partitioned twenty prawn trawl catches from tropical northern Australia into subsamples of about 10 kg each. All subsamples were then sorted, and species numbers recorded. Catch weights ranged from 71 to 445 kg, and the number of fish species in trawls ranged from 60 to 138, and invertebrate species from 18 to 63. Almost 70% of the species recorded in catches were "rare" in subsamples (less than one individual per 10 kg subsample or less than one in every 389 individuals). A matrix was used to show the increase in the total number of species that were recorded in each catch as the percentage of the sorted catch increased. Simulation modelling showed that sorting small subsamples (about 10% of catch weights) identified about 50% of the total number of species caught in a trawl. Larger subsamples (50% of catch weight on average) identified about 80% of the total species caught in a trawl. The accuracy of estimating the abundance of each species also increased with increasing subsample size. For the "rare" species, sampling error was around 80% after sorting 10% of catch weight and was just less than 50% after 40% of catch weight had been sorted. For the "abundant" species (five or more individuals per 10 kg subsample or five or more in every 389 individuals), sampling error was around 25% after sorting 10% of catch weight, but was reduced to around 10% after 40% of catch weight had been sorted.

The role of comfort in child restraint use practices

Suboptimal restraint use, particularly the incorrect use of restraints, is a significant and widespread problem among child vehicle occupants, and increases the risk of injury. Previous research has identified comfort as a potential factor influencing suboptimal restraint use. Both the real comfort experienced by the child and the parent’s perception of the child’s comfort are reported to influence the optimal use of restraints. Problems with real comfort may lead the child to misuse the restraint in their attempt to achieve better comfort whilst parent-perceived discomfort has been reported as a driver for premature graduation and inappropriate restraint choice. However, this work has largely been qualitative. There has been no research that objectively studies either the association between real and parent-perceived comfort, or any association between comfort and suboptimal restraint use. One barrier to such studies is the absence of validated tools for quantifying real comfort in children. We aimed to develop methods to examine both real and parent-perceived comfort and examine their effects on suboptimal restraint use. We conducted online parent surveys (n=470) to explore what drives parental perceptions of their child’s comfort in restraint systems (study 1) and used data from field observation studies (n=497) to examine parent-perceived comfort and its relationship with observed restraint use (study 2). We developed methods to measure comfort in children in a laboratory setting (n=14) using video analysis to estimate a Discomfort Avoidance Behaviour (DAB) score, pressure mapping and adapted survey tools to differentiate between comfortable and induced discomfort conditions (study 3). The DAB rate was then used to compare an integrated booster with an add-on booster (study 4) Preliminary analysis of our recent online survey of Australian parents (study 1) indicates that 23% of parents report comfort as a consideration when making a decision to change restraints. Logistic regression modelling of data collected during the field observation study (study 2) revealed that parent-perceived discomfort was not significantly associated with premature graduation. Contrary to expectation, children of parents who reported that their child was comfortable were almost twice as likely to have been incorrectly restrained (p<0.01, 95% CI 1.24 - 2.77).In the laboratory study (study 3) we found our adapted survey tools did not provide a reliable measurement of real comfort among children. However our DAB score was able to differentiate between comfortable and induced discomfort conditions and correlated well with pressure mapping. Preliminary results from the laboratory comparison study (study 4) indicate a positive correlation between DAB rate and use errors. In experiments conducted to date, we have seen a significantly higher DAB rate in the integrated booster compared to the add-on booster (p < 0.01). However, this needs to be confirmed in a naturalistic setting and in further experiments that take length of time under observation into account. Our results suggest that while some parents report concern about their child’s comfort, parent-reported comfort levels were not associated with restraint choice. If comfort is important for optimal restraint use, it is likely to be the real comfort of the child rather than that reported by the parent. The method we have developed for studying real comfort can be used in naturalistic studies involving child occupants to further understand this relationship. This work will be of interest to vehicle and child restraint manufacturers interested in improving restraint design for young occupants as well as researchers and other stakeholders interested in reducing the incidence of restraint misuse among children.

Multidisciplinary review of teaching and assessing competency attainment by non-medical prescribers

Background Prescribing is a complex task, requiring specific knowledge and skills, and the execution of effective, context-specific clinical reasoning. Systematic reviews indicate medical prescribing errors have a median rate of 7% [IQR 2%-14%] of medication orders [1-3]. For podiatrists pursuing prescribing rights, a clear need exists to ensure practitioners develop a well-defined set of prescribing skills, which will contribute to competent, safe and appropriate practice. Aim To investigate the methods employed to teach and assess the principles of effective prescribing in the undergraduate podiatry program and compare and contrast these findings with four other non-medical professions who undertake prescribing after training at Queensland University of Technology. Method The NPS National Prescribing Competency Standards were employed as the prescribing standard. A curriculum mapping exercise was undertaken to determine whether the prescribing principles articulated in the competency standards were addressed by each profession. Results A range of methods are currently utilised to teach prescribing across disciplines. Application of prescribing competencies to the context of each profession appears to influence the teaching methods used. Most competencies were taught using a multimodal format, including interactive lectures, self-directed learning, tutorial sessions and clinical placement. In particular clinical training was identified as the most consistent form of educating safe prescribers across all five disciplines. Assessment of prescribing competency utilised multiple techniques including written and oral examinations and research tasks, case studies, objective structured clinical examination exercises and the assessment of clinical practice. Effective and reliable assessment of prescribing undertaken by students in diverse settings remains challenging e.g. that occurring in the clinical practice environment. Conclusion Recommendations were made to refine curricula and to promote efficient cross-discipline teaching by staff from the disciplines of podiatry, pharmacy, nurse practitioner, optometry and paramedic science. Students now experience a sophisticated level of multidisciplinary learning in the clinical setting which integrates the expertise and skills of experience prescribers combined with innovative information technology platforms (CCTV and live patient assessments). Further work is required to establish a practical, effective approach to the assessment of prescribing competence especially between the university and clinical settings.

Changes in depression during and following pregnancy

Rates of depression were studied in a sample of over 9000 women who were participants in the Avon Longitudinal Study of Pregnancy and Childhood. Assessments of depression were made at 18 and 32 weeks gestation, and at 8 and 32 weeks postpartum. Changes in depressive status across time were modelled using latent Markov modelling methods. This analysis showed that when classification errors were taken into account there was relatively high stability in diagnostic status during pregnancy and after pregnancy. However, the transition from late pregnancy to the early postnatal period showed evidence of increased instability and remission of depression. The net effects of this were that rates of depression tended to decline following childbirth. The implications of these results for a series of issues including measurement errors in depression reports and the prevalence of depression before and after childbirth are discussed.

CBCT-guided radiotherapy for locally advanced head and cancer: dosimetric impact of weight loss on VMAT and IMRT plans for selected organs at risk structures (OARs)

Purpose: It is common for head and neck patients to be affected by time trend errors as a result of weight loss during a course of radiation treatment. The objective of this planning study was to investigate the impact of weight loss on Volumetric Modulated Arc Therapy (VMAT) as well as Intensity modulated radiation therapy (IMRT) for locally advanced head and neck cancer using automatic co-registration of the CBCT. Methods and Materials: A retrospective analysis of previously treated IMRT plans for 10 patients with locally advanced head and neck cancer patients was done. A VMAT plan was also produced for all patients. We calculated the dose–volume histograms (DVH) indices for spinal cord planning at risk volumes (PRVs), the brainstem PRVs (SC+0.5cm and BS+0.5cm, respectively) as well as mean dose to the parotid glands. Results: The results show that the mean difference in dose to the SC+0.5cm was 1.03% and 1.27% for the IMRT and VMAT plans, respectively. As for dose to the BS+0.5, the percentage difference was 0.63% for the IMRT plans and 0.61% for the VMAT plans. The analysis of the parotid gland doses shows that the percentage change in mean dose to left parotid was -8.0% whereas that of the right parotid was -6.4% for the IMRT treatment plans. In the VMAT plans, the percentages change for the left and the right parotid glands were -6.6% and -6.7% respectively. Conclusions: This study shows a clinically significant impact of weight loss on DVH indices analysed in head and neck organs at risk. It highlights the importance of adaptive radiotherapy in head and neck patients if organ at risk sparing is to be maintained.

Radical nursing and the emergence of technique as healthcare technology

The integration of technology in care is core business in nursing and this role requires that we must understand and use technology informed by evidence that goes much deeper and broader than actions and behaviours. We need to delve more deeply into its complexity because there is nothing minor or insignificant about technology as a major influence in healthcare outcomes and experiences. Evidence is needed that addresses technology and nursing from perspectives that examine the effects of technology, especially related to increasing demands for efficiency, the relationship of technology to nursing and caring, and a range of philosophical questions associated with empowering people in their healthcare choices. Specifically, there is a need to confront in practice the ways technique influences care. Technique is the creation of a kind of thinking that is necessary for contemporary healthcare technology to develop and be applied in an efficient and rational manner. Technique is not an entity or specific thing, but rather a way of thinking that seeks to shape and organize nursing activity, and manage efficiently individual difference(s) in care. It emphasizes predetermined causal relationships, conformity, and sameness of product, process, and thought. In response is needed a radical vision of nursing that attempts in a real sense to ensure we meet the needs of individuals and their community. Activism and advocacy are needed, and a willingness to create a certain detachment from the imperatives that technique demands. It is argued that our responsibility as nurses is to respond in practice to the errors, advantages, difficulties, and temptations of technology for the benefit of those who most need our assistance and care.

Differences in the accommodation stimulus response curves of adult myopes and emmetropes: A summary and update

Purpose To provide a summary of the classic paper "Differences in the accommodation stimulus response curves of adult myopes and emmetropes" published in Ophthalmic and Physiological Optics in 1998 and to provide an update on the topic of accommodation errors in myopia. Summary The accommodation responses of 33 participants (10 emmetropes, 11 early onset myopes and 12 late onset myopes) aged 18-31 years were measured using the Canon Autoref R-1 free space autorefractor using three methods to vary the accommodation demand: decreasing distance (4 m to 0.25 cm), negative lenses (0 to -4 D at 4 m) and positive lenses (+4 to 0 D at 0.25 m). We observed that the greatest accommodation errors occurred for the negative lens method whereas minimal errors were observed using positive lenses. Adult progressing myopes had greater lags of accommodation than stable myopes at higher demands induced by negative lenses. Progressing myopes had shallower response gradients than the emmetropes and stable myopes; however the reduced gradient was much less than that observed in children using similar methods. Recent Findings This paper has been often cited as evidence that accommodation responses at near may be primarily reduced in adults with progressing myopia and not in stable myopes and/or that challenging accommodation stimuli (negative lenses with monocular viewing) are required to generate larger accommodation errors. As an analogy, animals reared with hyperopic errors develop axial elongation and myopia. Retinal defocus signals are presumably passed to the retinal pigment epithelium and choroid and then ultimately the sclera to modify eye length. A number of lens treatments that act to slow myopia progression may partially work through reducing accommodation errors.

CBCT image guidance in head and neck irradiation: The impact of daily and weekly imaging protocols

Purpose This study evaluated the impact of a daily and weekly image-guided radiotherapy protocols in reducing setup errors and setting of appropriate margins in head and neck cancer patients. Materials and methods Interfraction and systematic shifts for the hypothetical day 1–3 plus weekly imaging were extrapolated from daily imaging data from 31 patients (964 cone beam computed tomography (CBCT) scans). In addition, residual setup errors were calculated by taking the average shifts in each direction for each patient based on the first three shifts and were presumed to represent systematic setup error. The clinical target volume (CTV) to planning target volume (PTV) margins were calculated using van Herk formula and analysed for each protocol. Results The mean interfraction shifts for daily imaging were 0·8, 0·3 and 0·5 mm in the S-I (superior-inferior), L-R (left-right) and A-P (anterior-posterior) direction, respectively. On the other hand the mean shifts for day 1–3 plus weekly imaging were 0·9, 1·8 and 0·5 mm in the S-I, L-R and A-P direction, respectively. The mean day 1–3 residual shifts were 1·5, 2·1 and 0·7 mm in the S-I, L-R and A-P direction, respectively. No significant difference was found in the mean setup error for the daily and hypothetical day 1–3 plus weekly protocol. However, the calculated CTV to PTV margins for the daily interfraction imaging data were 1·6, 3·8 and 1·4 mm in the S-I, L-R and A-P directions, respectively. Hypothetical day 1–3 plus weekly resulted in CTV–PTV margins of 5, 4·2 and 5 mm in the S-I, L-R and A-P direction. Conclusions The results of this study show that a daily CBCT protocol reduces setup errors and allows setup margin reduction in head and neck radiotherapy compared to a weekly imaging protocol.

Children’s personal exposure to air pollution in rural villages in Bhutan

Exposure assessment studies conducted in developing countries have been based on fixed-site monitoring to date. This is a major deficiency, leading to errors in estimating the actual exposures, which are a function of time spent and pollutant concentrations in different microenvironments. This study quantified school children’s daily personal exposure to ultrafine particles (UFP) using real-time monitoring, as well as volatile organic compounds (VOCs) and NO2 using passive sampling in rural Bhutan in order to determine the factors driving the exposures. An activity diary was used to track children’s time activity patterns, and difference in mean exposure levels across sex and indoor/outdoor were investigated with ANOVA. 82 children, attending three primary schools participated in this study; S1 and S2 during the wet season and S3 during the dry season. Mean daily UFP exposure (cm-3) was 1.08 × 104 for children attending S1, 9.81 × 103 for S2, and 4.19 × 104 for S3. The mean daily NO2 exposure (µg m-3) was 4.27 for S1, 3.33 for S2 and 5.38 for S3 children. Likewise, children attending S3 also experienced higher daily exposure to a majority of the VOCs than those attending S1 and S2. Time-series of UFP personal exposures provided detailed information on identifying sources of these particles and quantifying their contributions to the total daily exposures for each microenvironment. The highest UFP exposure resulted from cooking/eating, contributing to 64% of the daily exposure, due to firewood combustion in houses using traditional mud cookstoves. The lowest UFP exposures were during the hours that children spent outdoors at school. The outcomes of this study highlight the significant contributions of lifestyle and socio-economic factors in personal exposures and have applications in environmental risk assessment and household air pollution mitigation in Bhutan.