Biostatisticians investigate ‘adaptive’ study designs to maximise success of clinical trials

Written by Web News Editor on . Posted in News

The benefits and practical feasibility of utilising ‘adaptive’ designs in clinical trials are currently being investigated by biostatisticians, and it’s hoped that they could be used in investigating diseases as diverse as tuberculosis, malaria and Alzheimer’s.
 
Adaptive trials allow pre-planned adaptation during the course of the trial in order to maximise its success and the World Health Organization (WHO) have also shown an interest in using the technique in its vaccine trials.
 
Former RSS president Andy Grieve, who will speak at the forthcoming Adaptive Designs in Clinical Trials 2013 conference in London on 8 & 9 April, sets out some of the issues:
 
‘Since 1948, the gold-standard for clinical trials has been the double–blind randomised and controlled clinical trial. A trait of such trials is that they are static – the key elements driving the design are specified in advance and do not change. An obvious problem with this approach is that the planning of such trials is based on a fixed set of assumptions. Incorrect choices for these parameters can lead to trials which are under-powered or over-powered, and both of these outcomes are ethically questionable and undesirable.
 
Adaptive trials employ methodologies that allow study sponsors to monitor the data being gathered during the course of a trial with the objective of implementing a pre-planned adaptation to maximise the success of the trial.
 
Perhaps the simplest adaptive designs utilise interim analyses, long associated with group sequential trials, to test the assumptions of trials and to make decisions about their future conduct. For example, at an interim if the patient-to-patient variability is much larger than anticipated at the planning stage, the sample size can be increased to ensure that the trial has the desired power. Conversely, if the estimated effect size at the interim is non-existent or much smaller than required then the trial can be stopped for so-called futility. A number of large pharmaceutical companies have implemented such strategies with considerable savings.
 
A second area for consideration is in phase II dose-response studies. In adaptive dose-finding studies the post promising doses are followed, with ineffective doses being dropped. These designs allow a wider range of doses to be investigated without unduly increasing the number of patients studied.’
 
The use of many of these designs is being considered outside of the pharmaceutical industry. The WHO has shown an interest in their use in vaccine trials and in diseases such as tuberculosis, HVC and malaria. It is also hoped that adaptive strategies will play a role in developing treatments for diseases such as Alzheimer’s.
 

Statistics Authority issue statement outlining strategic priorities

Written by Web News Editor on . Posted in News

The UK Statistics Authority earlier this month (14 February 2013) published a statement of strategy in which it sets out its key aims, functions and strategic priorities for the next couple of years.
 
The statement (opens as pdf) lists five strategic priorities: quality, impact, efficiency, coverage and trustworthiness. Under these headings the Authority outlines how it intends to achieve progress in each area (presented in a table in the statement document).
 
The Authority details a number of reviews and consultations it intends to conduct in 2013. The governance of inflation statistics will be reviewed as part of a wider look at economic statistics and how they meet user needs. The document also confirms plans to lead a public debate on the long-term vision for official statistics in the UK, including a public consultation by the National Statistician to explore possible alternatives to the Census.
 
Other strategic priorities include increased data sharing and the use of administrative data in order to reduce costs and improve efficiency. Regarding the communication of statistics, more focus will be given to the explanation of statistics being produced, as well as enhancing the ONS website capability.
 
Finally, in addressing public trust in official statistics, the Authority indicates it will continue to argue against pre-release access to official statistics, as well as investigate any concerns about political involvement or misrepresentation of official statistics – something which Andrew Dilnot has continued to do since taking up his post as chair of the Authority. Finally, in order to gauge public trust in official statistics, an exercise measuring confidence in official statistics will be conducted, to report back publicly in 2014.
 

ODI launches scheme to create business opportunities from open data

Written by Web News Editor on . Posted in News

The Open Data Institute (ODI) has unveiled its Open Data Immersion Programme, a £850k new initiative aimed at assisting open data-related business start-ups.
 
The programme run for three years and will comprise a series of events bringing together small businesses and start-ups with data providers, industry experts and business leaders, in order to help turn data sets into business opportunities.
 
Events will be grouped into sector specific series which focus on challenges formulated by industry experts. Winning projects at each event will receive a prize of up to £25k and be eligible to take their concepts into early product development.
 
The programme’s first series, starting in March, will be themed crime and justice. Participants will be able to work with a range of specialists and industry experts such as police forces, government departments, private security companies and charities, using datasets that may include crime data, sentencing statistics, poverty measures and rates of re-offending.
 
The total budget for each series is £70k and the ODI is inviting suitable individuals and organisations to lead on each series. Further planned Immersion Programme series in 2013 will look at energy and the environment and personal data. Stephan Shakespeare, chair of the Data Strategy Board, said he was ‘looking forward to real, innovative ideas coming through’.
 

Statistics Authority clarifies Miliband’s use of youth unemployment figures

Written by Web News Editor on . Posted in News

Andrew Dilnot, chair of the UK Statistics Authority has this week written to Labour leader Ed Miliband over his claim that ‘only crisis-hit Spain has higher numbers of young unemployed than the UK’ in Europe.
 
The claim was made in The Sun newspaper on January 31 and Harriett Baldwin, Conservative MP for West Worcestershire, wrote to the Authority to complain about the leader of the opposition’s use of the statistic. She pointed out using numbers rather than percentages painted ‘a more damning comparison than is actually the case’.
 
In his response, Dilnot acknowledged that while Miliband’s claim was factually correct (the UK has 951,000 unemployed youths and Spain 959,000), it didn’t reflect relative unemployment as the size of each country’s labour force was not taken into account. ‘Accepted statistical practice is that comparing unemployment rates between countries is preferable to comparing absolute numbers,’ he wrote. ‘The latter, of course, may be heavily affected by the relative sizes of the general population.’
 
He included a two tables of Eurostat data to accompany his response, one showing numbers of unemployed and a second showing percentages of the population. With the UK’s youth unemployment rate at 20.7 percent (compared to Spain’s 55.4 percent) it also showed that there are 17 countries with a higher rate of youth unemployment than the UK.
 
However, Dilnot also cautioned that even using percentages, comparing unemployment between countries is not without difficulties. ‘The normal basis to derive unemployment rates is to use the percentage of the “economically active” population and in the case of youth unemployment such a concept can sometimes be inconsistent between countries,’ he noted.
 

Royal Statistical Society’s 2013 honours announced

Written by Web News Editor on . Posted in News

The RSS has announced its honours for 2013, having been decided by Council in January. The awards will be presented to the recipients at a ceremony held during the Society’s annual conference in Newcastle on 3 September 2013.
 
The Guy Medal in Gold is awarded to Sir John Kingman for his fundamental contributions to stochastic processes in four main areas: heavy traffic queueing theory; regenerative phenomena; subadditive ergodic theory; and mathematical population genetics. His work has been simultaneously fundamental to the mathematics of stochastic processes and highly influential for a wide range of application areas.
 
The Guy Medal in Silver is awarded to Brian Ripley for his pioneering contributions to spatial statistics and deep insight in stochastic simulation, his two highly influential papers ‘Modelling spatial patterns’ and ‘Neural networks and related methods for classification’, read to the Society in 1977 and 1994 respectively, and his pivotal role in the open-source R environment for statistical computing and data analysis.
 
The Guy Medal in Bronze is awarded to Piotr Fryzlewicz for his significant contribution to time series research as the originator of the Haar-Fisz transform, in particular for his two papers, ’Haar-Fisz estimation of evolutionary wavelet spectra’ and ‘GOES-8 X-ray sensor variance stabilization using the multiscale data-driven Haar-Fisz transform’ published in the Society’s journals in 2006 and 2007.
 
The Bradford Hill Medal is awarded to Paddy Farrington for his extensive contributions to the theory and practice of medical research, in particular the development of the self-controlled case-series method to estimate adverse reactions to vaccines and the development of a widely used algorithm for detecting infectious disease outbreaks.
 
The Greenfield Industrial Medal is awarded to Ron Kenett for his extensive work in the development and application of statistics in business and industry, and in recognition of his significant and sustained contributions as author, teacher and active practitioner.  Through his consulting work he has promoted the use of statistical methods across a wide range of strategic business activities, including Six Sigma quality systems, risk management and surveys of customers and employees.
 
The Royal Statistical Society Research Prize is awarded to Haeran Cho for her contribution to the paper ‘High-dimensional variable selection by tilting’ (JRSS B, (2012) 74, 593-622, with P. Fryzlewicz) and to other published work on the methodology and applications of statistics.
 
 

Lords criticise ‘lack of clarity’ regarding open access policy

Written by Web News Editor on . Posted in News

The House of Lords Science and Technology Committee has criticised Research Councils UK (RCUK) regarding the planned implementation of its open access policy following concern from academics and publishers that they were not adequately consulted.
 
There has been significant confusion around some of the policy’s wording, which does not make clear what embargo periods will allowed after the policy is implemented. The lack of clarity over whether embargo periods of 12 and 24 months were allowed after April 2013 was deemed ‘unacceptable’ by the report, which welcomed RCUK’s recent clarification that it will phase in its open access policy gradually over a five year implementation phase.
 
The Committee conducted a short inquiry in the light of these concerns, and produced a report last week (22 February 2013), titled ‘The implementation of open access’. The RSS contributed to the consultation and among other concerns, commented that the proposed timescale for implementation was ‘too tight, and that insufficient money has been allocated to support transition to open access’.
 
Lord Krebs, Chairman of the House of Lords Science and Technology Committee, commented: ‘RCUK did not consult or communicate effectively with key stakeholders in the publishing and academic communities when implementing its open access policy. While we are delighted that our inquiry has shown that RCUK are proposing to phase in their open access policy during the initial five-year implementation phase, this should have been made clear much earlier.’
 
The report recommends that RCUK should hold regular reviews during the implementation period, and that these reviews examine a number of different aspects regarding the process. Considerations should include: whether different disciplines require different embargo periods; how the UK compares to other countries implementing their own open access policies; what effect the policy has on the number of journals published; any effect on the quality of peer review; and also the impact it has on research and learned societies.
 
‘There are still many unknowns concerning the impact of the open access policy, which is why RCUK must commit to a wide ranging review of its policy in 2014, 2016 and before it expects full compliance in 2018,’ explained Lord Krebs. ‘We heard, and are pleased that RCUK are both aware of these concerns and prepared to act on them.’
 

National Statistician responds to criticism of public sector debt measures

Written by Web News Editor on . Posted in News

The National Statistician has responded to strong criticism from the Financial Times’ economic editor Chris Giles, who has attacked the way in which the Office for National Statistics measures the UK’s public sector borrowing and has questioned its independence from government.
 
Giles’ criticism refers to an ONS announcement made on 21 February 2013 which sets out how the figures for Public Sector Net Borrowing (PSNBex) will be affected by cash transfers from the Bank of England Asset Purchase Facility Fund (BEAPFF) to the treasury.
 
Giles points out that the introduction of PSNBex in 2009 was brought in to show public sector net borrowing but exclude measures taken in the wake of the financial crisis, so as not to distort the underlying state of public sector finances.
 
Giles was also critical of the structure of the body behind this decision, the Public Sector Finances Technical Advisory Group, pointing out that the ONS officials on this body are outnumbered by government officials. ‘It is not independent from ministers’ wishes,’ he concluded. ‘Official figures on the British public finances are no longer to be trusted.’
 
In a letter to the newspaper Jil Matheson, the National Statistician, rebuked the criticism by pointing out that since 1997 UK government finances have been aligned with internationally agreed standards. ‘These provide a clear, trusted basis for the UK’s public finances. To ignore that link risks creating a confusing situation in which the measures of gross debt and deficit prepared under European law included these payments, but the leading domestic UK measure ignored them. This would be very hard to justify’, she said.
 
In response to the criticism of the structure of the Public Sector Finances Technical Advisory Group, she pointed out the group is advisory in nature, and is bound by the Code of Practice for Official Statistics.  She made clear the decisions are her responsibility as National Statistician.  ’It is misleading to imply that the process is in any way tainted by political influence,’ she concluded.
 

Join the RSS

Join the RSS

Become part of an organisation which works to advance statistics and support statisticians

Copyright 2019 Royal Statistical Society. All Rights Reserved.
12 Errol Street, London, EC1Y 8LX. UK registered charity in England and Wales. No.306096

Twitter Facebook YouTube RSS feed RSS feed RSS newsletter

We use cookies to understand how you use our site and to improve your experience. By continuing to use our site, you accept our use of cookies and Terms of Use.