Workshop on Future Radstats Publications

Come to York for the Radstats conference on Saturday, 27 February 2016

Statistics of crisis in UK and EU

and one day earlier for a workshop to discuss

Future Radical Statistics publications

Friday, 26 February 12:00 – 6:00 pm (in the same venue as the conference)
Please email if you would like to come for the meeting.There will also be evening social events open to all in the same venue.


The Radical Statistics 2015 conference triggered several suggestions regarding future RadStats publications. A followup discussion took place in October. View the report of this meeting and the papers discussed.

The proposals have now been distilled down to just two. Click on links for detail.

  1. Radical Statistics in the 2020s (RS2020)
  2. Statistics for Radical Change (SRC): A Handbook for Community and Political Activists

Future developments will be posted on the Activity page of this website.

Posted on behalf of John Bibby, Jeff Evans, Humphrey Southall and Rachel Cohen.


Radical Statistics, Issue 110

This issue begins with two articles about the effect of demographic
change. Maria Sol Torres Minoldo turns a spotlight on Argentina,
while Alan Marshall, John Read, and James Nazroo focus on the
United Kingdom. Both pieces take issue with the argument, common
in the media and amongst politicians of all stripes, that an ageing
population is increasingly likely to constitute an unsustainable drain
on national resources. Specifically Minoldo produces considerable
data to show that the ‘dependency’ ratio (of ‘working adults’ to
‘pensioners’) is seriously flawed at assessing levels of real material
dependency in society. Marshall, Read and Nazroo decompose estimates of
population ageing. They show that, contrary to public discourse, older
age longevity accounts for only a small part of expected demographic
change, with the far larger part due to the ageing of the baby boom cohort.
As such population ageing may largely be a temporary, not permanent,

We follow this article with two reflections on the research process. In
the first Alan Sloan provides us with some qualitative reflections from
his work as a survey interviewer. He highlights the social context of
non-response and the emotional and practical ways that interviewers
respond. For many of us who work regularly with survey data this
reflection from the messy and human side of data collection serves as
a salutary reminder of the social uncertainty that data retain.

Following this, Stephen Gorard addresses the contentious issue of
how to assess the trustworthiness of evidence. His article produces a
framework to be used both by users and producers of research
evidence that enables a judgement-based star-rating of research
evidence. The framework emphasises design, sample size and quality,
data quality, fidelity of intervention, and threats to validity.

We finish the issue with four comment pieces addressing a diverse
range of contemporary issues – all of which in different ways highlight
the ways in which statistics and social policy are interwoven. The first
piece by Alison Macfarlane provides an overview of what has happened
with, the proposed data linkage between GP and hospital
records. She shows that poor handling of the process and the huge
public resistance engendered has produced serious obstacles for
academic health research. Ludi Simpson then offers a cogent critique
of the ways in which segregation measures are used. He points out
that since there will always be some segregation these measures
provide ready grist for politicians seeking to ignite moral panics over
racial segregation.

Boycott Workfare provides us with an important discussion of the
impact of and use of statistics to support the workfare policies forming
part of the government’s social policies. The comment is a critique of
workfare, forming part of the government’s austerity politics, which
involves a toughening of the treatment of and sanctions put on welfare
claimants. The edition ends with a discussion of abortion and abortion
statistics in Ireland by Frank Houghton. His comment exposes the
ways in which the Irish government and public institutions shy away
from openness around the actual number of abortions taking place in
a country which has been criticised for its restrictive abortion

With this issue of Radical Statistics we welcome Trude Sundberg, a
Social Policy expert from the University of Kent to the editorial team
and say goodbye and thank you to Alistair Greig who has been part of
the team for the past two years.

If you have an article or short comment piece that you think would be
suitable for a future issue of Radical Statistics or ideas about a
themed set of articles please get in touch with us at editors @

Trude Sundberg
Rachel Lara Cohen
Larry Brownstein

Results of the Radstats 2014 Critical Essay Competition

The judges have chosen the following winners of the 2014 Critical Essay Competition which closed in July with decisions in Oct.

Two prizes each were awarded in the Student and the Open category.

The winning essays will appear in an upcoming issue of Radical Statistics in 2015.

Student category
1st: Clara Musto – On the gateway hypothesis
2nd: Geraldine Clark – Yearning to Earn or Yearning To Learn?

Open category
1st: Elisabeth Garrat – The rise in UK foodbanks: What can statistics tell us about the current landscape of food insecurity and food aid in the UK?
2nd: Pauline McGovern – Path Analysis for People who Hate Statistics

The award for first prize is £50 and the second of £30, both in book vouchers.

Congratulations to the prize winners!

Also, many thanks to the judges for making the 2014 Radstats Critical Essay Competition a success.

Lee Williamson

Radstats Critical Essay Competition

CALLING EVERYONE to enter the Radical Statistics Critical Essay Competition!

GET your essay published & WIN a book token

Submit an original essay that addresses a current social research/policy question with critical use & interpretation of relevant data sources (3,000 words max) by 1st July 2014

F i r s t p r i z e £ 5 0
S e c o n d p r i z e £ 3 0

Plus 1 year subscription membership to Radical Statistics (RadStats).

All winning essays will be published in the RadStats newsletter/journal.

Please tell students and others and post this  flyer in your department or community.

There are two categories:

  • Student*: undergraduate or postgraduate
  • Open: any non-student

Your submission must be unpublished & unaided original work:

  • either specifically produced for the competition
  • or originating from your course of study or dissertation

Your essay could address a current social research policy/question with critical use & interpretation of relevant data sources, or be a critique of statistical methodology in an applied context.

Prizes will be awarded on the basis of readability, clear presentation of statistical material, critical perspective & convincing argument.

Essays do not necessarily have to involve statistical modelling, only critical use & interpretation of relevant data sources.

The deadline for submission is 1st July 2014
Judge’s decisions by 1st October 2014

Winning essays will be featured on the website & published in a special issue of Radical Statistics.

Enter by sending your essay, including full name, email & postal address, to

* Note: if entering in the student category, please provide university & details of essay origin when submitting (ie module details for coursework essays or whether the essay stems from an undergraduate, masters or PhD thesis).

Reduced Statistics Redux

In 2012 ‘Reduced Statistics’, a working group from Radical Statistics, took a detailed look at cuts to official statistics under the coalition  since 2010, and what they might mean in local government, housing, health and education. This led to a draft report as well as a debate with the UK Statistics Authority and ONS at the Royal Statistical’s Society’s conference. In a recent blog post and seminar at the London School of Economics, Alex Fenton looks at the latest information from the UK Statistics Authority on official statistics, and considers what it means in the light of the government’s proclaimed enthusiasm for “open data” and “transparency”.

View the LSE British Politics and Policy blog pos by Alex Fenton (3/11/14): Austerity stats: Making sense of cuts and changes to official statistics under the coalition

Is Britain Pulling Apart? Today!

Radical Statistics is gathering in Manchester today to take a fresh look at inequality in Britain for the 2014 conference, Is Britain Pulling Apart?RScrowd2014

Look on twitter for #Radstats to follow along. Presentations will be collected, proceedings are being filmed, and of course there will be a special issue of the journal.

If any delegate would like to contribute a blog post please get in touch with me at

Robin Rice

Is the UK too complacent on drinking water safety?

On 7 November,  Water Minister Dan Rogerson argued in response to a written question from Chris Ruane, Labour MP for the Vale of Clwyd, that water filtering within households is unnecessary, as monitoring by drinking water regulators shows tap water quality in the UK is “equivalent to the best in Europe”. (I’m not aware of a data series that would underpin this claim). 

On 9 December, a statistical study by engineer Marc Edwards was published showing that cheap water filters certified to remove lead, together with advice to run taps and consider bottled water in older buildings, prevented hundreds of miscarriages in Washington DC in 2004-2006, when a change in water treatment caused a city wide spike in lead in water levels (links below).   The paper also examines a miscarriage cluster in an office undergoing renovation, where lead in water from solder on copper pipes may have been wrongly excluded as a factor, because water was tested too late. 

The study compared fetal death rates and birth rates in Washington DC against a similar sized city with the same water supply (Baltimore), and the US as a whole, before, during and after two incidents when lead “rust” from lead pipes and from solder on copper pipes was disturbed. The first lead spike was caused by a change in disinfection chemicals, the second by attempts to solve the first problem by replacing sections of lead pipe work. 

Appropriate domestic water filters can reduce chlorination by-products present in tap water from chemical disinfection, together with the herbicides, other farm chemicals and pharmaceutical residues present in some river water (but not in water from boreholes favoured for bottled supplies). Filtering tap water was recommended in 2010 by the US President’s Cancer Panel, along with washing fruit and vegetables to remove pesticide residues and avoiding charred meat. 

Concerned over the known toxicity of disinfection by-products, the Netherlands has abandoned the use of chlorine in a water safety system which prioritizes choosing the safest raw water sources for drinking water (link below). Edwards’ statistical analysis shows that the US change in disinfection practice – initiated and persisted with in an attempt to reduce miscarriages from chlorination byproducts – did not achieve its health goal. Possibly the by-products of the replacement disinfectant regime, though less studied, are equally noxious. A troubling pattern for regulatory changes. 

Edwards’ key recommendations  – supported by his laboratory test simulating lead release from solder in response to physical disturbance – are that pregnant women should always receive advice to use filters, flushing or bottled water in older premises, and that everyone should receive warnings and take precautions when plumbing is disturbed. Currently no such advice is issued in the UK, despite a predominance of older properties (lead free solder was only available from the late ‘80s),  on-going lead pipe replacement to meet a stricter lead in water standard from the end of this year, and programmes to retrofit water meters.

It would be interesting to review fetal death rate trends within the Netherlands, to see whether the package of changes they made to how river water was treated (areas with better water sources never used chlorination) measurably improved maternal health. There is no easy switch to the Dutch method – the need for residual disinfection here goes hand in hand with the c. 25% leakage rate from poorly maintained pipework (vs. 4% for the Netherlands). Bottled water and home/office filtration are the policy options currently available in the UK to safeguard fetal health. 

Lucy Borland


Lucy Borland’s paper on Drinking Water Regulation in Rad Stats:

Hansard, Written Answers 7 Nov 2013:

Marc Edwards: in miscarriages coincided with high … – Washington Post

2008-2009 – Environmental Factors in Cancer: Reducing Environmental Cancer Risk, What We Can Do Now available at

The Dutch Secret: how to provide safe water without chlorine in the Netherlands, P. W.M. H. Smeets, G. J. Medema, J. C. van Dijk:  available to download at

The AllTrials campaign calls for all past and present clinical trials to be registered and their results reported.

Radical Statistics is a supporter of the AllTrials campaign. This paper is reprinted from under licence: CC BY-NC-ND 3.0.

Download this as a PDF here (310Kb)

Clinical trials are investigations designed to assess the effects – wanted and unwanted – of healthcare interventions in people. The Declaration of Helsinki, which is the World Medical Association’s statement of principles for medical research involving people, states that every investigator running a clinical trial should register it and report its results. Millions of volunteers have participated in clinical trials to help find out more about the effects of treatments on disease, yet that important ethical principle about reporting has been widely ignored. Information on what was done and what was found in these trials could be lost forever to doctors and researchers, leading to bad treatment decisions, missed opportunities for good medicine, and trials being repeated. This is what led to the AllTrials campaign in January 2013, a campaign which is now supported by thousands of individual patients, clinicians and researchers across the world, and by hundreds of organisations representing millions of people.

This document sets out more information about achieving a situation globally where all trials are registered and results reported. It is an achievement that will involve regulators and registries, clinical trial funders, universities and institutes, professional and learned societies and medical journals, patients and researchers.

This document is part of a continuing discussion which many different organisations are working on elaborating further over coming weeks and months. Please email views and contributions to:

What trial information needs to be registered and reported?

There are four levels of information in clinical trial reporting: (1) knowledge that a trial has been conducted, from a clinical trials register; (2) a brief summary of the trial’s results; (3) full details about the trial’s methods and results; (4) individual patient data from the trial.

The AllTrials campaign is concerned with the first three. There are now initiatives in many countries to work out how individual patient data can be shared with other researchers.

 1. Registration

In brief: Planned clinical trials should be registered, with a summary of the trial protocol, before the first participant is recruited. Past trials that were not registered should now be registered retrospectively. This is essential if the trial was on medicines or interventions that we currently use (this includes some trials conducted before registries were established).

Checks on the registration status of published trials, show that around 40% of clinical trials concerning treatments in current use were not registered[1]. This figure does not include unregistered trials that have never been published.

The situation is improving: increasingly, funders and research organisations are insisting that trials are registered and it is a legal requirement for trials on some medicinal products in the EU, USA and five other countries[2].

The World Health Organization (WHO) has set out a 20 item Trial Registration Data Set[3]of the minimum information that should be included when registering a trial. Registration covers rationale and background to the trial; information on study participants and informed consent; the intervention under investigation, primary and key secondary outcomes; the method of data collection and statistical analysis plans.

For further information see the SPIRIT[4] guidelines published in 2013.

Prospective registration is the gold standard for the reasons set out in the 2005 Ottawa Statement[5]. All trials that were not prospectively registered should still be registered now, i.e. retrospectively. This is particularly important for trials conducted to evaluate the efficacy and safety of a treatment in current use, some of which were done before trial registration was possible. Many registrations are incomplete against the WHO data set and registries should advise on which aspects of these could reasonably be completed.

There is no excuse for not registering planned or completed clinical trials. is the world’s largest register. It accepts registration from anywhere in the world and allows retrospective registration of trials. There are numerous national and regional registries, and others held by funders, institutions and corporations. About 20 of these are collected in the WHO’s Registry Network[6].

The proliferation of registries with different requirements is, though, limiting the usefulness and transparency of reported information. A trial can be registered in multiple registries but the entries are not always connected together. It is not currently possible for researchers or patients to find all trials that have been done on a particular intervention, even if all the trials have been registered somewhere.

The registration system could be streamlined and standardised internationally. There are now discussions about how to achieve this. Drawing down central information to multiple destinations may be more achievable than drawing it in from multiple sources to a central place, which has so far been the model. A small number of global centres would make it possible to standardise the way the data are structured so that entries can be linked and searched. Another option is to ensure that registries require trials to give all other registries’ ID numbers for trials that appear on multiple registers. Both strategies would help to ensure that trials can be linked and tracked from registration to publication of results.

Enforcement and Monitoring

It should be impossible to obtain funding for a trial, including funding from Government, or to sell a product, or to obtain permission to do a clinical trial, without proving registration.

Regulatory routes: In some regions the registration requirement has become or will become law for trials related to new marketing authorisation of drugs. The proposed EU Clinical Trials Regulation will require registration as part of approval for any new trial of a medicinal product. The US TEST Act, tabled in 2013, would require trials used to support licensing applications to have been registered before they have started. The FDA Amendments Act 2007 already requires trials with at least one site in the US to be registered within 21 days of the first patient being enrolled. The regulatory and ethical approval processes for clinical trials in every country can be developed to incorporate and monitor compliance with registration.

Funders: Applications for reimbursement and funding could include explicit statements that the trials will be registered and results reported. Some funders have already started to do this. Trial registration IDs should be requested and compliance monitored to the best of an organisation’s ability. A declaration that all past trials conducted by the investigator were registered could also be requested.

Journals and professional societies: The International Committee of Medical Journal Editors (ICMJE) committed in 2005 to publish only reports from trials that had been registered at inception. Requesting the trial number will help to monitor compliance with this more effectively. However, in order to overcome the historic gap in trial registration and reporting, journals should look at how they deal with any previously unreported trials that weren’t registered or pre-dated the registration requirement. For the future, journals could also ask for disclosure of registration details of all linked trials and commit to making it clear on a trial report if previously undisclosed trials come to light after publication. Professional bodies and learned societies should make it explicit in their codes of conduct that members must register clinical trials, and they can lobby for this to become an international standard.

2. Summary results reporting

In brief: A summary of results should be publicly available where the trial was registered, within one year of completion of the trial. Summary results from all past trials of medicines currently in use should be made publicly available on a register now. Summary results include information on the primary and any secondary outcomes measured and statistical analysis. This is part of the structured information that global registries should support.

An audit published in 2012 found that only a fifth of trials registered on had reported results within one year of completion[7] and different research found that trials which produced negative results are twice as likely to remain unreported than positive trials[8].  Publication of all results will reduce reporting biases and help researchers and policymakers produce more reliable systematic reviews of the safety and effectiveness of medical interventions.

Millions of patients have volunteered for clinical trials in the expectation that the findings generated by their effort will contribute to the body of knowledge about their conditions and future treatments. Publishing results fulfils clinical trialists’ ethical responsibility to patients in clinical trials, as set out in the Helsinki Declaration.

Summary results should be posted publicly within a year of the completion of the trial[9]where the trial was registered. Current discussions about registry development are looking at how to provide a clearer timeline of updates made to each entry and to indicate more clearly where information about the results is missing. As well as helping to improve compliance this will raise awareness among investigators about what is expected.

All past trials which have not reported results for medicines in current use should do this now. Registers should provide space for reporting of requests for results by third parties and include a log of requests for overdue information sent to trial sponsors, as well as responses to such requests.

Reports of clinical trial summary results on a register should at least contain the items on results page (which includes summary participant information, protocol and amendments, summary results for pre-specified primary and secondary end points, details of adverse events and statistical analyses)[10]. If they don’t, they should be supplemented with extra information added to the register the trial was registered on. Results are produced in a variety of formats – in peer reviewed journal papers, clinical study reports in the case of drugs for which marketing authorisation applications are being made, reports to grant giving bodies, and so on. These may contain all or some of the summary results information required. Links and documents can be uploaded directly onto registers.

Registers currently have different formats for reporting results. Results cannot be uploaded to as PDFs for example. Ideally every major register could require results to be uploaded in a format that allows the main reported items to be searchable and enables sharing of information between registries. Some registries are curated to ensure there is internal logic in entries. Global registries would certainly have to do this to be useful and manageable.

The ICMJE has stated that prior publication of results on a register is not a barrier to publication in a journal. Journal reports on trials should be linked to the clinical trial unique identifier.

Enforcement and Monitoring

Regulation: The US FDA Amendment Act 2007 requires that results must be posted on within a year of the completion of the trial for all trials with at least one site in the US. The FDA has the power to fine trial sponsors who do not comply but rarely does this. Whether or not a trial is required to post results – or has been granted an extension – is often the subject of legal discussion, and as a consequence there is no clarity about whether a trial is truly overdue by the terms of the Act. The proposed EU Clinical Trials Regulation will require that summary results for every registered trial must be posted within one year of the completion of the trial, and the European Commission is discussing how to enforce this properly. Trial approval bodies in each country should consider expanding their monitoring of reporting, and ensure there is routine and open public audit of compliance for each individual trial.

Funders: Trials approval, processes such as marketing authorisation and reimbursement for medicinal products, and applications for funding could require an explicit statement that the results of the trial will be made available on a register within a year of trial’s end. Some funders have started this, and begun withholding funds until results are shared. A declaration that results from all past trials conducted by the investigator have been reported could also be required.

Journals and professional societies: Journals should state clearly that there are no bars to subsequent publication of a trial report when summary results are posted to a register. A number of journals have supported the Restoring Invisible and Abandoned Trials statement which gives trialists an amnesty of one year to publish results of previously unreported trials[11]. Professional societies should ensure that their professional codes of conduct reflect the requirement to report summary results.

3.  A full report

In brief: Trial sponsors or others who produce a full report for marketing authorisation or any other purpose should make this publicly available. The narrative reports of adverse events and individual patient data in a full report can be redacted and available on request to researchers, in the same way that reports of adverse incidents currently are, with a commitment that no reasonable request will be refused.

Full reports (Clinical Study Reports or their equivalent in non-commercial settings) contain a large amount of detailed information about the methods, analysis, results and conclusions of a clinical trial[12]. This information is needed to make and to scrutinise decisions about medicines and to assess published summary findings. Clinical Study Reports are produced for regulatory and licensing purposes and follow a standard structure set out by ICH GCP guidelines[13]. An equivalent for researchers who do not plan to produce a Clinical Study Report is any document that complies with the 25-item CONSORT statement on trial reporting.[14] Full reports should be made publicly available when they have been created.

Full reports sometimes contain narrative descriptions of adverse events experienced by trial participants. This information is important to understanding the trade off between risks and benefits of a treatment. These paragraphs may contain identifiable patient information which may need to be redacted. These paragraphs should be available on request to researchers who provide a protocol of their study plan, with no reasonable request refused by the academic or company who authored the report. This is similar to the system for releasing the full narrative descriptions in spontaneous reports of possible adverse events to prescribed medications, reported by doctors and patients to regulators through the Yellow Card[15] scheme in the UK.

Clinical Study Reports also contain line by line individual patient data on all participants in one carefully specified section. We do not call for individual patient data to be made publicly available though there are extensive discussions at present on how this information could be shared where it is of value to research. The EU Ombudsman has ruled that it is not a significant burden to remove individual patient data from full reports before public sharing. Some organisations (GSK) have committed to making all of their reports publicly available, with this information redacted. Others (Roche) have committed to providing this information on demand.

Enforcement and Monitoring

Regulation: The proposed EU Clinical Trials Regulation will contain guidance that no information in a clinical study report should be considered commercially confidential once a decision about marketing authorisation has been made. The European Medicines Agency’s transparency policy is to release any full report it holds on request. Other regions should adopt a similar approach.


4. Individual patient data

The AllTrials campaign is not calling for individual patient data to be made publicly available.

There are currently initiatives in many countries looking at how to improve sharing of this level of information for the benefit of future research. This offers significant opportunities, such as: improving the accuracy of estimates of benefits from a treatment, through individual patient data meta-analyses; and identifying subgroups of patients who respond better, or worse, to a specific treatment. Patient groups, medical research funders and trialists have raised concerns about the inability to reuse past research. They are keen to develop consent protocols that will optimise the ability to reuse findings, and want legislators to look at whether new data protection regulations impose unnecessary burdens and restrictions on reuse of past research.

The AllTrials campaign is an initiative of Bad ScienceBMJCentre for Evidence-based MedicineCochrane CollaborationJames Lind InitiativePLOS and Sense About Science and is being led in the US by Dartmouth’s Geisel School of Medicine and the Dartmouth Institute for Health Policy & Clinical Practice. AllTrials was launched in January 2013 to call for all clinical trials to be registered and results reported.

[1] Huser 2013, Freshwater 2013, Killeen 2013, van de Wetering 2012, Jones 2012, Scherer 2012, McGee 2011, Mathieu 2009, Rasmussen 2009

[3] WHO Trial Registration Data Set

[4]SPIRIT 2013 (Standard Protocol Items: Recommendations for Interventional Trials) explanation and elaboration: guidance for protocols of clinical trials

[5] Krleža-Jeric K et al for the Ottawa Group 2005 Principles for international registration of protocol information and results from human trials of health related interventions: Ottawa Statement
[6] The WHO Registry Network
[7] Prayle AP 2012 Compliance with mandatory reporting of clinical trial results on cross sectional study
[8] Song et al 2010 Dissemination and publication of research findings: an updated review of related biases
[9] The completion date of a trial is the final date on which data was (or is expected to be) collected.
[10] See Appendix 1 for a suggested list of contents of summary results

[11] Restoring invisible and abandoned trials: a call for people to publish the findings BMJ2013;346:f2865

[12] Doshi & Jefferson, 2012 Clinical study reports of randomised controlled trials: an exploratory review of previously confidential industry reports. BMJ Open

[15] Yellow Card Scheme – MHRA