1981 Gore’s Statistics in Question: series in British Medical Journal

In 1981, Sheila Gore published articles in the British Medical Journal for 23 consecutive weeks. The purpose was to illustrate the misuse of statistics in medical research and to show how this could be put right. This dissemination of statistical methods reflected a long standing, and still current, goal of the MRC Biostatistics Unit, one that began with the early publication of books on medical statistics by Woods and Russell (1931) and Hill (1937).

And, similarly to the situation with Hill’s book which was based on articles in the Lancet, Sheila Gore’s articles were subsequently published in book form, alongside a similarly motivated series of articles by Douglas Altman. In the preface to Statistics in Practice, Stephen Lock, the Editor of the British Medical Journal, wrote that “only a foolish researcher would start a study without thinking carefully about the statistical principles” and that “this book will help those who wish to understand and apply statistics to their daily work.”

References

1. Gore SM. Statistics in Question. British Medical Journal 1981; 282: weekly for 23 weeks.
2. Gore SM, Altman DG. Statistics in Practice. British Medical Association, London, 1982.
3. Altman DG, Gore SM, Gardner MJ, Pocock SJ. Statistical guidelines for contributors to medical journals. British Medical Journal 1983; 286: 1489-1493.

 

1982 Statistics in Medicine: Colton, Freedman and Johnson founding editors

In 1979, Laurence Freedman provided the initial concept, which was developed with his Cambridge MRC colleague, Tony Johnson (who would remain as Editor-in-Chief for 18 years) into the journal Statistics in Medicine. These two individuals, along with Ted Colton from the USA, were the founding editors of a journal which surely must be one of the fastest growing journals.

The first volume of Statistics in Medicine in 1982 comprised 4 issues and 381 pages. When Tony Johnson retired as an Editor in 1999, the journal published 24 issues and 3 548 pages. The growth continued, although more recently stabilising, with the 2013 volume having published 30 issues and 5 489 pages.

The growth of this journal devoted to medical statistics reflects the ample scope for a journal which emphasized the dualistic nature of medical statistics. This was emphasized by the founding editors who quoted Sir Austin Bradford Hill’s comment that a statistician in a medical project “must endeavour to learn and understand its medical as well as its statistical aspects, and should, I believe, be as intensely absorbed in the one as the other”.

References

1. Colton T, Freedman L, Johnson AL. Editorial. Statistics in Medicine 1982; 1: 1 – 3.

 

1983 Bayesian methods in clinical trials

Uncertainty in parameter estimates arises in the design, monitoring and analysis of clinical trials – Bayesian statistical methods allow this uncertainty to be accounted for easily.

Bayesian statistical methods are especially useful for early phase trials, wherein there is larger uncertainty, and for choosing a sample size for a phase III trial when there are still-limited data on parameters that influence the required sample size.

A current research area of great interest is adaptive designs, in which the BSU is currently playing an important part – Bayesian methods are useful, and sometimes vital, in this area.

References

1. Spiegelhalter DJ, Freedman LS. A predictive approach to selecting the size of a clinical trial, based on subjective clinical opinion. Statistics in Medicine 1986; 5: 1-13.
2. Freedman LS, Spiegelhalter DJ. Comparison of Bayesian with group sequential methods for monitoring clinical trials. Controlled Clinical Trials 1989; 10: 357-367.
3. Spiegelhalter DJ, Freedman LS, Parmar MKB. Applying Bayesian ideas in drug development and clinical trials. Statistics in Medicine 1993; 12: 1501-1517.
4. Spiegelhalter DJ, Freedman LS, Parmar MKB. Bayesian approaches to randomized trials. Journal of the Royal Statistical Society Series A (Statistics in Society) 1994; 157: 357-387.
5. Spiegelhalter DJ, Abrams K, Myles JP. Bayesian Approaches to Clinical Trials and Health Care Evaluation. Chichester: John Wiley & Sons, 2004.

 

1984 The statistics of transplantation in the UK

In collaboration with UK Transplant and the Department of Health, and with Papworth and Addenbrooke’s Hospitals, MRC Biostatistics Unit developed statistical methods for the equitable sharing of UK cadaveric solid organs to maximize their years of function. “Beneficial matching” was the basis of UK’s kidney allocation for a decade before being updated by the same techniques and simulations being applied to later data.

The Unit‘s collaboration with Papworth Hospital on heart and lung transplantation identified time-specific prognostic influences on progression, supported surgical trials and assessed the cost-effectiveness of devices.

The 1989-90 confidential audit of all deaths in intensive care units averted ‘required request’ legislation because the problem was not doctors’ failure to ask families about organ donation but relatives’ refusal when asked – then by 30% of families, now by 40%.

The Corneal Transplant Study established the UK’s registry on corneal graft outcomes and the determinants of corneal graft survival and visual quality.

References

1. Gilks WR, Bradley BA, Gore SM, Koulda PT for the Users of the UK Transplant Service. Substantial benefits of tissue matching in renal transplantation. Transplantation 1987; 43: 669-674.
2. Gilks WR, Gore SM, Bradley BA. Matchability in kidney transplantation. Tissue Antigens 1988; 32: 121-129.
3. Gilks WR, Gore SM, Bradley, BA. Predicting match grade and waiting time to kidney transplantation. Transplantation 1991; 51: 618-624.
4. Gore SM, Taylor RMR, Wallwork J. Availability of transplantable organs from brain stem dead donors in intensive care units. British Medical Journal 1991; 302: 149-153.
5. Gore SM, Cable DJ, Holland AJ. Organ donation from intensive care units in England and Wales: Two year confidential audit of deaths in intensive care. British Medical Journal 1992; 304: 349-355.
6. Sharples LD, Caine N, Mullins P, Scott JP, Solis E, English TAH, Large SR, Schofield PM, Wallwork J. Risk factor analysis for the major hazards following heart transplantation – Rejection, infection, and coronary occlusive disease. Transplantation 1991; 52: 244-252.
7. Sharples LD, Scott JP, Dennis C, Higenbottam TW, Stewart S, Wreghitt T, Large SR, Wells FC, Wallwork J. Risk factors for survival following combined heart-lung transplantation: The first 100 patients. Transplantation, 1994; 57: 218-223.
8. Sharples LD, Briggs A, Caine N, McKenna M, Buxton M. A model for analyzing the cost of main clinical events after cardiac transplantation. Transplantation 1996; 62: 615-621.
9. Jackson CH, Sharples LD. Hidden Markov models for the onset and progression of bronchiolitis obliterans syndrome in lung transplant recipients. Statistics in Medicine 2002; 21: 113-128.
10. Sharples L, Buxton M, Caine N, Cafferty F, Demiris N, Dyer M, Freeman C. Evaluation of the ventricular assist device programme in the UK. Health Technology Assessment 2006; 10: 1-119, iii-iv.
11. Vail A, Gore SM, Bradley BA, Easty DL, Rogers CA. Corneal graft survival and visual outcome: a multicenter study. Ophthalmology 1994; 101: 120-127.
12. Vail A, Gore SM, Bradley BA, Easty DL, Rogers CA, Armitage WJ. Influence of donor and histocompatibility factors on corneal graft outcome. Transplantation 1994; 58: 1210-1216.
13. Gore SM, Barroso E, White DJG. Risk factors in orthotopic first liver transplantation. In R Calne (ed) Liver Transplantation, 2nd edition, Grune and Stratton, London, 1987, pages 513-530.
14. Friend PJ, Hale G, Waldmann H, Gore S, Thiru S, Joysey V, Evans DB, Calne RY. Campath-1M – prophylactic use after kidney transplantation: a randomized controlled clinical trial. Transplantation 1989: 48; 248-253.

 

1986 The statistics of screening

The Unit was the statistical and epidemiological centre for the Swedish Two-County Trial of breast cancer screening and for the UK trial of breast screening frequency, both of which informed the NHS Breast Screening Programme.

MRC Biostatistics Unit carried out pioneering work on modelling the progression of breast cancer and its arrest by screening, which has been applied to screening for other cancers and has been influential worldwide.

The Unit was also the statistics centre for the Multicentre Aneurysm Screening Study (see later), as a result of which UK now has a national screening programme for abdominal aortic aneurysm.

References

1. Tabar L, Fagerberg G, Gad A, Baldetorp L, Holmberg LH, Grontoft O, Ljungquist U, Lundstrom B, Mansson JC, Eklund G, Day NE, Pettersson F. Reduction in breast cancer mortality by mass screening with mammography: first results of a randomized trial in two Swedish counties. Lancet 1985; 1: 829-XXX.
2. Tabar L, Fagerberg G, Duffy SW, Day NE, Gad A, Grontoft O. Update of the Swedish two- county program of mammographic screening for breast cancer. Radiological Clinics of North America 1992; 30: 187-210.
3. Chen HH, Duffy SW, Tabar L, Day NE. Markov chain models for progression of breast cancer. Part I: Tumour attributes and the preclinical screen-detectable phase. Journal of Epidemiology and Biostatistics 1997; 2: 9-23.
4. Ashton HA, Buxton MJ, Day NE, Kim LG, Marteau T M, Scott RA, Thompson SG, Walter NM. The Multicentre Aneurysm Screening Study (MASS) into the effect of abdominal aortic aneurysm screening on mortality in men: a randomised controlled trial. Lancet 2002; 360, 1531-1539.
5. Breast Screening Frequency Trial Group. The frequency of breast cancer screening: results from the UKCCCR randomised trial. European Journal of Cancer 2002; 38: 1458-1464.

 

1987 Breslow & Day’s Statistical Methods in Cancer Research: The Design and Analysis of Cohort Studies

It is difficult to overstate the influence of the books by Breslow and Day on case-control and cohort studies. Epidemiologists around the world regard them as the definitive sources for analytic methods. Within the Unit, they have not only influenced study design and analysis, but have also spurred further methodology.

The Director of the International Agency for Research on Cancer wrote in his foreward to Statistical Methods in Cancer Research: The Design and Analysis of Cohort Studies by Norman Breslow and Nick Day (then Director of the MRC Biostatistics Unit) that: “Epidemiological studies provide the only definitive information on the degree of cancer risk to man. Since malignant diseases are clearly of multi-factorial origin, their investigation in man has become increasingly complex and epidemiological and statistical studies on cancer require a correspondingly complex and rigorous methodology.” This volume on cohort studies was a comprehensive presentation of the needed methodology and remains a primary reference in the field.

The need for this book, expressed in the foreward, grew out of long-standing work in international cancer epidemiology, important aspects of which were initiated in the MRC Biostatistics Unit in the late 1960s by then director, Richard Doll, and which continued for many years. The methodology developed in conjunction with The Design and Analysis of Cohort Studies had immediate application in studies of other diseases besides cancer.

References

1. Breslow NE, Day NE. Statistical Methods in Cancer Research I: The Analysis of Case-Control Studies, IARC, Lyon, 1980. [published before Day became Director of MRC Biostatistics Unit]
2. Breslow NE, Day NE. Statistical Methods in Cancer Research II: The Design and Analysis of Cohort Studies, IARC, Lyon, 1987.
3. Day NE, Duffy SW (1996). Trial design based on surrogate end-points – application to comparison of different breast screening frequencies. Journal of the Royal Statistical Society Series A (Statistics in Society) 1996; 159: 49-60.
4. Day N, Oakes S, Luben R, Khaw K-T, Bingham S, Welch A, Wareham N. EPIC in Norfolk: Study design and characteristics of the cohort. British Journal of Cancer 1999; 80: 95-103.
5. Clayton DG and Hills M. Statistical Models in Epidemiology. Oxford: Oxford University Press, 1993.
6. Duffy SW, South MC, Day NE. Cluster randomization in large public health trials: The importance of antecedent data. Statistics in Medicine 1992; 11: 307-316.

 

1988 AIDS/HIV: UK’s projections, immunological surveillance, prisons and progression

The year 1988 marked the start of a long and successful involvement of the MRC Biostatistics Unit in HIV epidemiology (Day and Gore, 1988).

Work carried out in the Unit has included development and application of statistical and epidemiological approaches to: estimate and project the evolution of the HIV epidemic (De Angelis et al, 1998); understand HIV disease progression (Gore and Farewell, 1994; Cascade Collaboration, 2000); establish immunological surveillance (CD4 collaborative Group, 1997); and quantify the magnitude of HIV burden, including in prisons, using data from multiple sources (Bird et al, 1995; Goubar et al, 2008).

The Unit’s contribution in this area has significantly influenced the statistical and epidemiological literature, and has provided and continues to provide quantitative underpinning to the Department of Health’s policies on HIV.

References

1. Day NE, Gore SM. Prediction of the number of new AIDS cases and the number of new persons infected with HIV up to 1992: the results of ‘back projection’ methods (appendix 12) In Short-term Prediction of HIV infection and AIDS in England and Wales, Report of a Working Group, Department of Health Welsh Office, Her Majesty’s Stationery Office, London, 1988, pages 74-78.
2. Multicohort Analysis Project Workshop (Co-convenors: Gore SM and Farewell VT. Participants: McNeil AJ, Fielding K, De Angelis D) (1994). Part I: Immunologic markers of AIDS progression: Consistency across five HIV-infected cohorts. AIDS 1994; 8: 911-921.
3. Bird AG, Gore SM, Cameron S, Ross AJ, Goldberg DJ. Anonymous HIV surveillance with risk factor elicitation at Scotland’s largest prison, Barlinnie. AIDS 1995; 9: 801-808.
4. Report of a Working Group (Chair N.E. Day; statistician members Gore SM and De Angelis D). The incidence and prevalence of AIDS and prevalence of other severe HIV disease in England and Wales for 1995-1999: Projections using data to the end of 1994. Communicable Disease Report 1996; 6: R1-R24.
5. McNeil AJ, Yap PL, Gore SM, Brettle RP, McColl M, Wyld R, Davidson S, Weightman R, Richardson AM and Robertson JR. Association of HLA types A1-B8-DR3 and B27 with rapid and slow progression of HIV disease. Quarterly Journal of Medicine 1996; 89: 177-185.
6. CD4 Collaborative Group (Gore SM: convenor; Ross AJ and Raab GM: statistician members). CD4 surveillance in Scotland: Perspectives on severe HIV-related immunodeficiency. AIDS 1997; 11: 1509-1517.
7. Collaborative Group on AIDS Incubation and HIV Survival including the CASCADE EC Concerted Action (De Angelis D: Writing Committee member). Time from HIV-1 seroconversion to AIDS and death in 13,000 individuals before the widespread use of highly active anti-retroviral therapy: effect of age at seroconversion and exposure category. Lancet 2000; 355: 1131-1137.
8. De Angelis D, Gilks WR, Day NE. Bayesian projection of the AIDS epidemic (With Discussion). Journal of the Royal Statistical Society Series C (Applied Statistics) 1998; 47: 449-498.
9. Goubar A, Ades AE, De Angelis D, McGarrigle CA, Mercer C, Tookey P, Fenton K, Gill ON. Bayesian multi-parameter synthesis of HIV surveillance data in England and Wales, 2001 (With Discussion). Journal of the Royal Statistical Society Series A (Statistics in Society) 2006; 171: 541-580.

 

1989 Epilepsy: natural history and randomised trials

Epilepsy is a serious neurological disease that affects millions of people worldwide: although some causes have been identified, over 40% of cases are idiopathic. The Unit participated in a collaborative programme of research into epilepsy that extended over 40 years from 1973. The programme included clinical trials investigating when to start anti-epileptic drugs, when to withdraw these drugs, as well as trials comparing older and newer drugs and, in addition, cohort studies examining the natural prognosis of epilepsy.

Tony Johnson who led the programme became a member of the Minister of Transport’s Advisory Panel on Driving and Diseases of the Nervous System as well as of the Expert Neurology panels of both Medicines and Healthcare Regulatory Authority and European Medicines Agency. In 2005, Johnson was awarded the Epilepsy Lifetime Service Award for Research in Epilepsy by the UK Chapter of the International League Against Epilepsy.

References

1. Hart YM, Sander JWAS, Johnson AL, Shorvon SD. National General Practice Study of Epilepsy: recurrence after a first seizure. Lancet 1990; 336: 1271-1274.
2. Medical Research Council Antiepileptic Drug Withdrawal Study Group. Prognostic index for recurrence of seizures after remission of epilepsy. British Medical Journal 1993; 306: 1374-1378.
3. Taylor J, Chadwick D, Johnson T. Risk of accidents in drivers with epilepsy. Journal of Neurology, Neurosurgery and Psychiatry 1996; 60: 621-627.
4. Gaitatzis A, Johnson AL, Chadwick, SD Shorvon, Sander JW. Life expectancy in people with newly diagnosed epilepsy. Brain 2004; 127: 2427-2432.
5. Kim LG, Johnson TL, Marson AG, Chadwick DW on behalf of the MRC MESS Study Group. Prediction of risk of seizure recurrence after a single seizure and early epilepsy: further results from the MESS trial. Lancet Neurology 2006; 5: 317-322.