Richard Lilford

The stepped wedge trial design: a systematic review

BackgroundStepped wedge randomised trial designs involve sequential roll-out of an intervention to participants (individuals or clusters) over a number of time periods. By the end of the study, all participants will have received the intervention, although the order in which participants receive the intervention is determined at random. The design is particularly relevant where it is predicted that the intervention will do more good than harm (making a parallel design, in which certain participants do not receive the intervention unethical) and/or where, for logistical, practical or financial reasons, it is impossible to deliver the intervention simultaneously to all participants. Stepped wedge designs offer a number of opportunities for data analysis, particularly for modelling the effect of time on the effectiveness of an intervention. This paper presents a review of 12 studies (or protocols) that use (or plan to use) a stepped wedge design. One aim of the review is to highlight the potential for the stepped wedge design, given its infrequent use to date.MethodsComprehensive literature review of studies or protocols using a stepped wedge design. Data were extracted from the studies in three categories for subsequent consideration: study information (epidemiology, intervention, number of participants), reasons for using a stepped wedge design and methods of data analysis.ResultsThe 12 studies included in this review describe evaluations of a wide range of interventions, across different diseases in different settings. However the stepped wedge design appears to have found a niche for evaluating interventions in developing countries, specifically those concerned with HIV. There were few consistent motivations for employing a stepped wedge design or methods of data analysis across studies. The methodological descriptions of stepped wedge studies, including methods of randomisation, sample size calculations and methods of analysis, are not always complete.ConclusionWhile the stepped wedge design offers a number of opportunities for use in future evaluations, a more consistent approach to reporting and data analysis is required.

The eVALuate study: two parallel randomised trials, one comparing laparoscopic with abdominal hysterectomy, the other comparing laparoscopic with vaginal hysterectomy

Objective To compare the effects of laparoscopic hysterectomy and abdominal hysterectomy in the abdominal trial, and laparoscopic hysterectomy and vaginal hysterectomy in the vaginal trial. Design Two parallel, multicentre, randomised trials. Setting 28 UK centres and two South African centres. Participants 1380 women were recruited; 1346 had surgery; 937 were followed up at one year. Primary outcome Rate of major complications. Results In the abdominal trial laparoscopic hysterectomy was associated with a higher rate of major complications than abdominal hysterectomy (11.1% v 6.2%, P = 0.02; difference 4.9%, 95% confidence interval 0.9% to 9.1%) and the number needed to treat to harm was 20. Laparoscopic hysterectomy also took longer to perform (84 minutes v 50 minutes) but was less painful (visual analogue scale 3.51 v 3.88, P = 0.01) and resulted in a shorter stay in hospital after the operation (3 days v 4 days). Six weeks after the operation, laparoscopic hysterectomy was associated with less pain and better quality of life than abdominal hysterectomy (SF-12, body image scale, and sexual activity questionnaires). In the vaginal trial we found no evidence of a difference in major complication rates between laparoscopic hysterectomy and vaginal hysterectomy (9.8% v 9.5%, P = 0.92; difference 0.3%, -5.2% to 5.8%), and the number needed to treat to harm was 333. We found no evidence of other differences between laparoscopic hysterectomy and vaginal hysterectomy except that laparoscopic hysterectomy took longer to perform (72 minutes v 39 minutes) and was associated with a higher rate of detecting unexpected pathology (16.4% v 4.8%, P = < 0.01). However, this trial was underpowered. Conclusions Laparoscopic hysterectomy was associated with a significantly higher rate of major complications than abdominal hysterectomy. It also took longer to perform but was associated with less pain, quicker recovery, and better short term quality of life. The trial comparing vaginal hysterectomy with laparoscopic hysterectomy was underpowered and is inconclusive on the rate of major complications; however, vaginal hysterectomy took less time.

Use and misuse of process and outcome data in managing performance of acute medical care: avoiding institutional stigma

The history of monitoring the outcomes of health care by external agencies can be traced to ancient times. However, the danger, now as then, is that in the search for improvement, comparative measures of mortality and morbidity are often overinterpreted, resulting in judgments about the underlying quality of care. Such judgments can translate into performance management strategies in the form of capricious sanctions (such as star ratings) and unjustified rewards (such as special freedoms or financial allocations). The resulting risk of stigmatising an entire institution injects huge tensions into health-care organisations and can divert attention from genuine improvement towards superficial improvement or even gaming behaviour (ie, manipulating the system). These dangers apply particularly to measures of outcome and throughput. We argue that comparative outcome data (league tables) should not be used by external agents to make judgments about quality of hospital care. Although they might provide a reasonable measure of quality in some high-risk surgical situations, they have little validity in acute medical settings. Their use to support a system of reward and punishment is unfair and, unsurprisingly, often resisted by clinicians and managers. We argue further that although outcome data are useful for research and monitoring trends within an organisation, those who wish to improve care for patients and not penalise doctors and managers, should concentrate on direct measurement of adherence to clinical and managerial standards.

Evaluation and stages of surgical innovations

Surgical innovation is an important part of surgical practice. Its assessment is complex because of idiosyncrasies related to surgical practice, but necessary so that introduction and adoption of surgical innovations can derive from evidence-based principles rather than trial and error. A regulatory framework is also desirable to protect patients against the potential harms of any novel procedure. In this first of three Series papers on surgical innovation and evaluation, we propose a five-stage paradigm to describe the development of innovative surgical procedures.

Using hospital mortality rates to judge hospital performance: a bad idea that just won’t go away

Standardised mortality rates are a poor measure of the quality of hospital care and should not be a trigger for public inquiries such as the investigation at the Mid Staffordshire hospital, say Richard Lilford and Peter Pronovost

The stepped wedge cluster randomised trial : rationale, design, analysis, and reporting

The stepped wedge cluster randomised trial is a novel research study design that is increasingly being used in the evaluation of service delivery type interventions. The design involves random and sequential crossover of clusters from control to intervention until all clusters are exposed. It is a pragmatic study design which can reconcile the need for robust evaluations with political or logistical constraints. While not exclusively for the evaluation of service delivery interventions, it is particularly suited to evaluations that do not rely on individual patient recruitment. As in all cluster trials, stepped wedge trials with individual recruitment and without concealment of allocation (or blinding of the intervention) are at risk of selection biases. In a stepped wedge design more clusters are exposed to the intervention towards the end of the study than in its early stages. This implies that the effect of the intervention might be confounded with any underlying temporal trend. A result that initially might seem suggestive of an effect of the intervention may therefore transpire to be the result of a positive underlying temporal trend. Sample size calculations and analysis must make allowance for both the clustered nature of the design and the confounding effect of time. The stepped wedge cluster randomised trial is an alternative to traditional parallel cluster studies, in which the intervention is delivered in only half the clusters with the remainder functioning as controls. When the clusters are relatively homogeneous (that is, the intra-cluster correlation is small), parallel studies tend to deliver better statistical performance than a stepped wedge trial. However, if substantial cluster-level effects are present (that is, larger intra-cluster correlations) or the clusters are large, the stepped wedge design will be more powerful than a parallel design, even one in which the intervention is preceded by a period of baseline control observations.

The Top Patient Safety Strategies That Can Be Encouraged for Adoption Now

Over the past 12 years, since the publication of the Institute of Medicines report, “To Err is Human: Building a Safer Health System,” improving patient safety has been the focus of considerable public and professional interest. Although such efforts required changes in policies; education; workforce; and health care financing, organization, and delivery, the most important gap has arguably been in research. Specifically, to improve patient safety we needed to identify hazards, determine how to measure them accurately, and identify solutions that work to reduce patient harm. A 2001 report commissioned by the Agency for Healthcare Research and Quality, “Making Health Care Safer: A Critical Analysis of Patient Safety Practices” (1), helped identify some early evidence-based safety practices, but it also highlighted an enormous gap between what was known and what needed to be known.

Ethical issues in the design and conduct of cluster randomised controlled trials

In most randomised controlled trials, individual patients are randomised to a treatment or control group, but sometimes this is undesirable or even impossible and groups (clusters) of people may be randomised instead. These are called cluster randomised controlled trials, and although they have been around for a long time, the need for them is likely to increase in line with growing concern to evaluate the delivery of health services, public education, and policy on social care. ### Summary points Need for cluster trials will increase with concern over health service evaluation, but issues of ethics and guardianship must be addressed In some cluster trials the intervention can be targeted at individuals (individual-cluster); where this would be too difficult or expensive the intervention is targeted at the whole group (cluster-cluster) Autonomy is important in individual-cluster trials, while the utilitarian welfare of the cluster as a whole is of paramount importance in cluster-cluster trials In individual-cluster trials the participants should give consent; cluster-cluster trials need procedural safeguards appropriate to the risks carried by the cluster intervention Guardians should sign a consent form that sets out their duties before they volunteer a cluster for a trial The ethical aspects of medical practice and medical research are most often discussed in the context of two main moral traditions—utilitarianism and Kantian ethics. Broadly speaking, utilitarianism is concerned with increasing social utility (value), which usually means that the individuals maximise their expected utility and so act in their own best interests. In the long run social utility will not be served by demanding that individuals be self sacrificing for the common good. This leads to matters of distributive justice whereby utility and disutility, benefits and costs, are distributed as fairly and evenly as possible across society. The Kantian tradition shows why we are duty bound to respect a persons …

Allelic drop-out and preferential amplification in single cells and human blastomeres: implications for preimplantation diagnosis of sex and cystic fibrosis.

Abstract Previously the diagnosis of sex and cystic fibrosis status has been studied on single cells using the polymerase chain reaction (PCR). It has been suggested that allelic drop-out (PCR failure of one allele) and/or preferential amplification (hypo-amplification of one allele) may contribute to poor reliability and misdiagnosis, although this remains controversial as some reports suggest that allelic drop-out does not occur. We investigated an improved method of diagnosing sex and cystic fibrosis in single cells using a new technology (fluorescent PCR) to determine the base level of PCR artefacts (allelic drop-out and preferential amplification) which, in combination with improved sensitivity, should improve PCR reliability and accuracy. Fluorescent PCR gives high reliability (approximately 97%) and accuracy rates (approximately 97%) in somatic cells for both sex and cystic fibrosis diagnosis and its lower detection threshold allows allelic drop-out and preferential amplification to be easily distinguished. We also achieved high reliability and accuracy in diagnosing cystic fibrosis in human blastomeres. This study confirms earlier reports of both allelic drop-out and preferential amplification in single cell analysis. We demonstrate that both allelic drop-out and preferential amplification occur in somatic cells and suggest these are separate phenomena. Preferential amplification appeared common in single cell PCR while allelic drop-out apparently occurred at random in each allele. Preferential amplification was mainly amplification of the larger allele. We suggest that some inaccuracy/misdiagnosis may be due to both preferential amplification as well as allelic drop-out. Other findings were variability in drop-out between PCR and that amplification of signals from human blastomeres may be linked to embryo quality. We suggest that allelic drop-out is dependent on the number of cells within the sample.

An epistemology of patient safety research: a framework for study design and interpretation. Part 2. Study design

This is the second in a four-part series of articles detailing the epistemology of patient safety research. This article concentrates on issues of study design. It first considers the range of designs that may be used in the evaluation of patient safety interventions, highlighting the circumstances in which each is appropriate. The paper then provides details about an innovative study design, the stepped wedge, which may be particularly appropriate in the context of patient safety interventions, since these are expected to do more good than harm. The unit of allocation in patient safety research is also considered, since many interventions need to be delivered at cluster or service level. The paper also discusses the need to ensure the masking of patients, caregivers, observers and analysts wherever possible to minimise information biases and the Hawthorne effect. The difficulties associated with masking in patient safety research are described and suggestions given on how these can be ameliorated. The paper finally considers the role of study design in increasing confidence in the generalisability of study results over time and place. The extent to which findings can be generalised over time and place should be considered as part of an evaluation, for example by undertaking qualitative or quantitative measures of fidelity, attitudes or subgroup effects.