This article was downloaded by: [University of California Santa Cruz]On: 08 October 2014, At: 19:02Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: MortimerHouse, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Biopharmaceutical StatisticsPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/lbps20

BIOPHARMACEUTICAL STATISTICS BEYOND 2000A. Lawrence Gould aa Merck Research Laboratories , West Point, Pennsylvania, U.S.A.Published online: 05 Oct 2011.

To cite this article: A. Lawrence Gould (2001) BIOPHARMACEUTICAL STATISTICS BEYOND 2000, Journal ofBiopharmaceutical Statistics, 11:1-2, 1-8, DOI: 10.1081/BIP-100104193

To link to this article: http://dx.doi.org/10.1081/BIP-100104193

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose ofthe Content. Any opinions and views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be reliedupon and should be independently verified with primary sources of information. Taylor and Francis shallnot be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and otherliabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to orarising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

JOURNAL OF BIOPHARMACEUTICAL STATISTICS, 11(1&2), 1–8 (2001)

EDITORIAL

BIOPHARMACEUTICAL STATISTICSBEYOND 2000

One of the duties of a new Editor is that of articulating a vision of wherethe focus of the journal should be under his (or her) editorship; this is the purposeof the comments that follow. The Journal of Biopharmaceutical Statistics hasbecome an important organ for addressing issues pertinent to biopharmaceuticalstatistics in its first decade, and I certainly intend this to continue. However, thescope of activities in which biopharmaceutical statisticians engage has grown con-siderably; I believe it would be useful to review where the field has been andwhere it should be going in order to encourage biopharmaceutical statisticians tothink, and write, about the issues that need to be addressed.

The responsibilities and roles of biopharmaceutical statisticians havechanged dramatically over the past few decades. Many new techniques and toolshave been developed, and there have been important changes in the environmentin which medical statistics is applied. Some of these changes reflect therapeuticareas that have not been explored previously, and some reflect changing social,economic, political, and competitive priorities. The developments recently affect-ing the biopharmaceutical world include a changing healthcare industry in whichmanaged care organizations influence decisions about drug prescriptions andwhere patients are more engaged in choices about their medical treatment; pres-sure from advocacy groups for speedier drug approvals; competition among com-panies with a rapidly decreasing span of therapeutic exclusivity; and regulatoryinsistence on a broader demonstration of the effect of a new therapy on the well-being of the patient rather than just its efficacy in treating a disease.

The regulatory environment has evolved and matured over time, leading tocoordinated regulatory requirements such as the recently finalized ICH E-9 guide-lines (1,2), which provide a common set of reasonably rigorous standards for thedesign, conduct, and analysis of clinical trials. Safety and tolerability, especiallyover the long term, are becoming increasingly important criteria for assessingtherapies and for comparing alternative active treatments (2–7). This is a particu-larly important development because of the emergence of many new therapies fortreating chronic diseases and because of the aging of the populations in the devel-oped countries, who increasingly will need these therapies.

Economic and non-physiological considerations are becoming an important

1

Dow

nloa

ded

by [

Uni

vers

ity o

f C

alif

orni

a Sa

nta

Cru

z] a

t 19:

02 0

8 O

ctob

er 2

014

ORDER REPRINTS

2 EDITORIAL

yardstick for evaluating therapies. Managed care providers and third-party payerssuch as insurance companies and national formularies can determine therapeuticstrategies when resources are limited. (8–19). The economic considerations arenot trivial.

More new models for clinical experimentation will be necessary because theold models of fairly static trials carried to completion will become less acceptablein an environment of changing therapies and changing (larger!) expectations. Butother issues need to be addressed. How should studies of prophylaxis or treatmentfor potentially disastrous epidemics be carried out? In the face of an aging popula-tion in developing countries, it would seem to make good public health policy toencourage preventive medicine, changing lifestyles to lower the risk of the dis-eases that afflict older people. But what should the specific elements of that policybe? And how should they be evaluated?

Statistical considerations have come to comprise a major component of clini-cal drug development and the design and analysis of clinical trials. There has beenan increasing emphasis on clear specification of the questions the analyses are toaddress, the need to define anticipated analyses in detail, and the need to identifywhat data are to be captured, and how. Many study design issues are essentiallystatistical, for example, follow-up strategies, interim looks at the data either formodifying the trial design or changing the sample size, or for reaching an interimdecision about accepting or rejecting a null hypothesis. These strategies reflect agrowing desire to reach a conclusion about a drug’s acceptability as soon as possi-ble, or to ensure as early as possible that the trial will be sensitive enough todetect a worthwhile clinical effect.

Political, social, and competitive pressures to reduce the time and cost ofbringing a drug to market have led to the need for compressing the time requiredto reach a decision about the marketability of a drug and obtain information tosupport its approval for marketing. The stakes are quite high. Delaying marketingof a product that generates a billion dollars in sales per year costs about $3 millionin sales per day. Evolving technologies for data capture, storage, and retrievalusing, for example, the internet, present opportunities for shortening drug develop-ment cycle time. The evaluation of AIDS drugs also has emphasized issues sur-rounding the use of surrogate markers, which are used as leading indicators of atreatment’s clinical effects. These issues are extremely important because the ef-fect of therapy on ‘‘indicators’’ can be determined by smaller, shorter-durationtrials than are needed for evaluating the effect of treatment on clinical endpoints.Allowing regulatory approval on the basis of findings for indicators means thatproducts can be brought to market sooner.

The need to ensure the safety of patients in the trial by monitoring accumulat-ing data can be contradictory to the need to maintain the scientific integrity ofthe trial by carefully controlling the unblinding of the treatment assignments. DataSafety Monitoring Boards (DSMBs) provide a way to reconcile these needs (20–24). These bodies operate in various ways, and their roles can be relatively limited

Dow

nloa

ded

by [

Uni

vers

ity o

f C

alif

orni

a Sa

nta

Cru

z] a

t 19:

02 0

8 O

ctob

er 2

014

ORDER REPRINTS

EDITORIAL 3

or quite extensive. The scope of action of a DSMB, how it operates, how it uses thedata to make recommendations or decisions, who provides the data, and whether itmakes recommendations or decisions all have important statistical implicationsfor the trial.

A considerable number of therapeutic target areas have become increasinglyimportant over the past few decades, and have provided a stimulus to the develop-ment of new statistical methods and strategies, for example, trials in AIDS, evalua-tions of vaccines, trials in chronic disease, trials of genetic therapies. The emer-gence of these and other new therapeutic areas has been fueled by the enormousincreases in knowledge and techniques in medicinal chemistry. Statistical develop-ments in chemometrics have contributed to this advance and have, in turn, beendriven by it (25–35). The conduct of scientifically valid, comprehensive clinicaltrials in these and other therapeutic areas is a development that has occurredlargely over the past twenty to thirty years. The need to tease out subtle treatmenteffects has led to interesting developments in epidemiologic and statistical tech-niques, including survival analysis, Bayesian modeling, Quality of Life assess-ment, trial design, and endpoint surrogacy, just to name a few areas of endeavor(36–70).

The process by which candidates for development are identified has changedvery dramatically over the past decade, with the increasing industrialization ofthe drug screening process to enable efficient screening of thousands of candidatecompounds. There are clearly profound issues of design that need to be addressedto make this process proceed effectively. Related, but not identical, issues arisein the application of genetic technologies to the determination of potential targetsfor new therapies and in evaluating their effects, possibly in identifying subpopula-tions of patients in whom new therapies are likely to be particularly effective.

Increases in the number, size, and complexity of drug development programsand other clinical studies often entail a corresponding increase in the size andvisibility of the statistical staff. Statisticians have a key role to play in assuringthe scientific integrity of global development programs, so it is necessary to thinkabout how statistical and other staff can be used effectively, and how moderntechnology can be part of this process. The process of medical experimentationand data management and analysis needs to be rethought carefully; technologyshould not be incorporated for its own sake to fix parts of the process without atleast some thought about the effect on the whole process.

The development of medical statistics almost certainly will be tied to thecontinuing development of computing capability, including hardware and soft-ware. This will provide increasing opportunities for medical statisticians to thinkin terms of how best to draw correct insights from observed data, and how bestto communicate them to the consuming audience. This last point is vital. Statistics,medical and otherwise, has the goal of providing insight about observed data tothe people who must translate these insights into decisions and actions, for exam-ple, who to treat, when, with what, and for how long.

Dow

nloa

ded

by [

Uni

vers

ity o

f C

alif

orni

a Sa

nta

Cru

z] a

t 19:

02 0

8 O

ctob

er 2

014

ORDER REPRINTS

4 EDITORIAL

It is becoming practical to address operational trial issues such as modelingthe effects of therapeutic or protocol noncompliance, including taking prohibitedconcomitant medication, and addressing withdrawals, especially those related tothe measurements not obtained, in sensible ways. An important consequence willbe the ability to design better trials by incorporating the information from earlypharmacology trials with basic knowledge about how the body processes the drug,and using this information to simulate what might be expected in phase 3 trials.This will require a deeper practical knowledge about the relationship betweenpharmacokinetics and pharmacodynamics. The advantage of this approach is thatthe phase 3 trials can be designed to study the appropriate populations and dosages,thereby efficiently using patient and economic resources to obtain informationneeded for a regulatory submission. One interesting side effect of this developmentmay be pressure to do larger and more thorough phase 2 trials to obtain a betteridea of the spectrum of drug activity across dosages and populations.

Data mining is increasingly being used to explore large, complicated data-bases to identify ‘‘interesting’’ relationships, i.e., high-order interactions or verynon-linear relationships that ordinarily would not be detected by standard statisti-cal analyses (71,72). This process is an elaborate version of subgroup analysis,and is subject to the same interpretational difficulties: How does one distinguishbetween real effects and merely coincidental ones? This is an issue that needsattention by statisticians because many large, complex databases are being accu-mulated, especially by managed care institutions and national health plans, amongothers, and the temptation to explore these databases will be irresistible. The impli-cations for public health are profound, and it is important to have in place method-ology for assuring the validity of conclusions from such efforts.

The advancement of statisticians in the administrative organizations wherethey work has enhanced their opportunity to affect the scientific requirements ofpharmaceutical development and thereby marry good medicine and good science.This change in the scheme of things raises many issues, of which communicationis one of the most important. Many non-statisticians such as clinicians, regulators,people involved in marketing and advertising, and people involved in evaluatinghealth care options use what statisticians produce. They can do this most effec-tively if statisticians understand what these people need to know in planning statis-tical reports and analyses.

Having better tools to do analyses that address the really relevant biologicaland medical issues frees statisticians, biological scientists, and clinicians from theshackles of computational limitations and encourages thinking about what thereally relevant issues are and how they should be addressed. However, it is impor-tant in doing so to avoid specialization to the point of losing sight of developmentsin other areas of specialization, such as econometrics, that may be appropriateand useful in medical applications. Thinking about new methods for addressingthe right questions creates a need for technical development to study the propertiesof these methods and perhaps to create new paradigms in statistics.

Dow

nloa

ded

by [

Uni

vers

ity o

f C

alif

orni

a Sa

nta

Cru

z] a

t 19:

02 0

8 O

ctob

er 2

014

ORDER REPRINTS

EDITORIAL 5

REFERENCES

1. Hirokawa, K. Clinical pharmacologists and drug regulation—future perspective inJapan. British Journal of Clinical Pharmacology 1996, 42, 63–71.

2. O’Neill, R.T. Statistical concepts in the planning and evaluation of drug safety fromclinical trials in drug development: issues of international harmonization. Statisticsin Medicine 1995, 14, 117–127.

3. Buyse, M.; Poplavsky, J.L. Clinical evaluation of virus safety and inhibitor inci-dence: statistical considerations. Blood Coagul. Fibrinolysis 1995, 6(Suppl 2), S86–S92.

4. Lan, K. K.; Friedman, L. Monitoring boundaries for adverse effects in long-termclinical trials. Controlled Clinical Trials 1986, 7, 1–7.

5. Freedman, L.; Anderson, G.; Kipnis, V.; Prentice, R.; Wang, C.Y.; Rossouw, J.;Wittes, J.; DeMets, D. Approaches to monitoring the results of long-term diseaseprevention trials: examples from the Women’s Health Initiative. Controlled ClinicalTrials 1996, 17, 509–525.

6. Monzani, F.; Caraccio, N.; Goletti, O.; Lippolis, P.V.; Casolaro, A.; Del, G.P.; Cav-ina, E.; Miccoli, P. Five-year follow-up of percutaneous ethanol injection for thetreatment of hyperfunctioning thyroid nodules: a study of 117 patients. Clin. Endo-crinol. Oxf. 1997, 46, 9–15.

7. DCCT Research Group. A screening algorithm to identify clinically significantchanges in neuropsychological functions in the Diabetes Control and ComplicationsTrial. Journal of Clinical and Experimental Neuropsychology 1994, 16, 303–316.

8. Claxton, K.; Posnett, J. An economic approach to clinical trial design and researchpriority-setting. Health Econ. 1996, 5, 513–524.

9. Drummond, M.F. The future of pharmacoeconomics: bridging science and practice.Clin.Ther. 1996, 18, 969–978.

10. Elliott, S.L.; Harris, A.H. The methodology of cost-effectiveness analysis: avoidingcommon pitfalls. Med. J. Aust. 1997, 166, 636–639.

11. Ellwein, L.B.; Drummond, M.F. Economic analysis alongside clinical trials. Biasin the assessment of economic outcomes. Int. J. Technol. Assess. Health Care 1996,12, 691–697.

12. Garber, A.M.; Romer, P.M. Evaluating the federal role in financing health-relatedresearch. Proceedings of the National Academy of Sciences, USA 1996, 93, 12717–12724.

13. Gorsky, R.D.; Farnham, P.G.; Straus, W.L.; Caldwell, B.; Holtgrave, D.R.; Simonds,R.J.; Rogers, M.F.; Guinan, M.E. Preventing perinatal transmission of HIV—costsand effectiveness of a recommended intervention. Public Health Rep. 1996, 111,335–341.

14. Gray, A.M.; Marshall, M.; Lockwood, A.; Morris, J. Problems in conducting eco-nomic evaluations alongside clinical trials. Lessons from a study of case manage-ment for people with mental disorders. Br. J. Psychiatry 1997, 170, 47–52.

15. Hunink, M.G. Outcomes research and cost-effectiveness analysis in radiology. Eur.Radiol. 1996, 6, 615–620.

16. Johannesson, M. The cost-effectiveness of hypertension treatment in Sweden: ananalysis of the criteria for intervention and the choice of drug treatment. J. Hum.Hypertens. 1996, 10(Suppl 2), S23–S26.

Dow

nloa

ded

by [

Uni

vers

ity o

f C

alif

orni

a Sa

nta

Cru

z] a

t 19:

02 0

8 O

ctob

er 2

014

ORDER REPRINTS

6 EDITORIAL

17. Johannesson, M.; Jonsson, B.; Kjekshus, J.; Olsson, A.G.; Pedersen, T.R.; Wedel,H. Cost effectiveness of simvastatin treatment to lower cholesterol levels in patientswith coronary heart disease. Scandinavian Simvastatin Survival Study Group. NewEngland Journal of Medicine 1997, 336, 332–336.

18. Keck, P. E.; McElroy, S. L.; Bennett, J. A. Health-economic implications of theonset of action of antimanic agents. J. Clin. Psychiatry 1996, 57(Suppl 13), 13–18.

19. Drummond, M.F.; O’Brien, B.; Stoddart, G.L.; Torrance, G.W. Methods for theEconomic Evaluation of Health Care Programmes, 2nd Edition; Oxford UniversityPress: United Kingdom, 1997.

20. Armstrong, P.W.; Furberg, C.D. Clinical trial data and safety monitoring boards.The search for a constitution [editorial]. Circulation 1995, 91, 901–904.

21. DeMets, D.L.; Fleming, T.R.; Whitley, R.J.; Childress, J.F.; Ellenberg, S.S.; Foulkes,M.; Mayer, K.H.; O’Fallon, J.; Pollard, R.B.; Rahal, J.J.; et al. The data and safetymonitoring board and acquired immune deficiency syndrome (AIDS) clinical trials.Controlled Clinical Trials 1995, 16(6), 408–421.

22. Ellenberg, S.S.; Myers, M.W.; Blackwelder, W.C.; Hoth, D.F. The use of externalmonitoring committees in clinical trials of the National Institute of Allergy and Infec-tious Diseases. Statistics in Medicine 1993, 12, 461–467.

23. Fleming, T.R.; Neaton, J.D.; Goldman, A.; DeMets, D.L.; Launer, C.; Korvick, J.;Abrams, D. Insights from monitoring the CPCRA didanosine/zalcitabine trial, TerryBeirn Community Programs for Clinical Research on AIDS. J. Acquired ImmuneDeficiency Syndrome and Human Retrovirology 1995, 10(Suppl 2), S9–18.

24. Wittes, J. Behind closed doors: the data monitoring board in randomized clinicaltrials. Statistics in Medicine 1993, 12, 419–424.

25. Zhao, P.L.; Nachbar, R.B.; Bolognese, J.A.; Chapman, K. Two new criteria forchoosing sample size in combinatorial chemistry. J. Med. Chem. 1996, 39, 350–352.

26. Metting, H.J.; Coenegracht, P.M. Neural networks in high-performance liquid chro-matography optimization: response surface modeling. J. Chromatogr. A 1996, 728,47–53.

27. Brown, S.D.; Blank, T.B.; Sum, S.T.; Weyer, L.G. Chemometrics. Anal. Chem.1994, 66, 315R–359R.

28. Belvisi, L.; Bravi, G.; Scolastico, C.; Vulpetti, A.; Salimbeni, A.; Todeschini, R.A 3D QSAR approach to the search for geometrical similarity in a series of non-peptide angiotensin II receptor antagonists. J. Comput. Aided Mol. Des. 1994, 8,211–220.

29. Berridge, J.C.; Jones, P.; Roberts, M.A. Chemometrics in pharmaceutical analysis.Journal of Pharmaceutical and Biomedical Analysis 1991, 9, 597–604.

30. Wold, S. Chemometrics, why, what and where to next? Journal of Pharmaceuticaland Biomedical Analysis 1991, 9, 589–596.

31. Karjalainen, E.J. The role of chemometrics in medical decision making. Scand. J.Clin. Lab. Invest. Suppl. 1990, 202, 109–111.

32. Castaneda, M.K. Chemometrics: measurement reliability. Clin.Chem. 1988, 34,2494–2498.

33. Boyd, J.C. Perspectives on the use of chemometrics in laboratory medicine. Clin.Chem. 1986, 32, 1726–1733.

34. Deming, S.N. Chemometrics: an overview. Clin. Chem. 1986, 32, 1702–1706.

Dow

nloa

ded

by [

Uni

vers

ity o

f C

alif

orni

a Sa

nta

Cru

z] a

t 19:

02 0

8 O

ctob

er 2

014

ORDER REPRINTS

EDITORIAL 7

35. Beck, J.R. Laboratory decision science applied to chemometrics: strategic testing ofthyroid function. Clin. Chem. 1986, 32, 1707–1713.

36. Zeger, S.L.; See, L.C.; Diggle, P.J. Statistical methods for monitoring the AIDSepidemic. Statistics in Medicine 1989, 8, 3–21.

37. Zangerle, R.; Reibnegger, G.; Klein, J.P. Survival differences in Austrian patientswith the acquired immunodeficiency syndrome. Eur. J. Epidemiol. 1995, 11, 519–526.

38. Zackin, R.; Wei, L.J. Analysis of repeated virological measurements based on celldilution assays. Statistics in Medicine 1997, 16, 571–582.

39. Yao, Q.; Wei, L.J. Play the winner for phase II/III clinical trials. Statistics in Medi-cine 1996, 15, 2413–2423.

40. Williams, P.L. Sequential monitoring of clinical trials with multiple survival end-points. Statistics in Medicine 1996, 15, 2341–2357.

41. Whyte, B.M.; Carr, J.C. Comparison of AIDS in women in rural and urban Georgia.South. Med. J. 1992, 85, 571–578.

42. Wang, Y.; Taylor, J.M. Inference for smooth curves in longitudinal data with appli-cation to an AIDS clinical trial. Statistics in Medicine 1995, 14, 1205–1218.

43. Wallenstein, S.; Berger, A. Weighted logrank tests to detect a transient improvementin survivorship. Biometrics 1997, 53, 736–744.

44. Tu, X.M.; Meng, X.L.; Pagano, M. Survival differences and trends in patients withAIDS in the United States. J. Acquir. Immune. Defic. Syndr. 1993, 6, 1150–1156.

45. The European Collaborative Study. Natural history of vertically acquired humanimmunodeficiency virus-1 infection. Pediatrics 1994, 94, 815–819.

46. Taylor, J.M. Models for the HIV infection and AIDS epidemic in the United States.Statistics in Medicine 1989, 8, 45–58.

47. Taylor, J.M.; Kim, D.K. Marker values at the time of an AIDS diagnosis. Statisticsin Medicine 1994, 13, 2059–2066.

48. Sun, J. Empirical estimation of a distribution function with truncated and doublyinterval-censored data and its application to AIDS studies. Biometrics 1995, 51,1096–1104.

49. Schoenfeld, D.A. Long-term follow-up in AIDS clinical trials. Statistics in Medicine1996, 15, 2359–2366.

50. Raab, G.M.; Fielding, K.L.; Allardice, G. Incorporating HIV test data into forecastsof the AIDS epidemic in Scotland. Statistics in Medicine 1994, 13, 2009–2020.

51. Neaton, J.D.; Wentworth, D.N.; Rhame, F.; Hogan, C.; Abrams, D.I.; Deyton, L.Considerations in choice of a clinical endpoint for AIDS clinical trials, Terry BeirnCommunity Programs for Clinical Research on AIDS (CPCRA). Statistics in Medi-cine 1994, 13, 2107–2125.

52. Law, C.G.; Brookmeyer, R. Effects of mid-point imputation on the analysis of dou-bly censored data. Statistics in Medicine 1992, 11, 1569–1578.

53. Kiuchi, A.S.; Hartigan, J.A.; Holford, T.R.; Rubinstein, P.; Stevens, C.E. Changepoints in the series of T4 counts prior to AIDS. Biometrics 1995, 51, 236–248.

54. Kaplan, R.M.; Anderson, J.P.; Patterson, T.L.; McCutchan, J.A.; Weinrich, J.D.;Heaton, R.K.; Atkinson, J.H.; Thal, L.; Chandler, J.; Grant, I. Validity of the Qualityof Well-Being Scale for persons with human immunodeficiency virus infection,HNRC Group, HIV Neurobehavioral Research Center. Psychosom. Med. 1995, 57,138–147.

Dow

nloa

ded

by [

Uni

vers

ity o

f C

alif

orni

a Sa

nta

Cru

z] a

t 19:

02 0

8 O

ctob

er 2

014

ORDER REPRINTS

8 EDITORIAL

55. Gelber, R.D.; Lenderking, W.R.; Cotton, D.J.; Cole, B.F.; Fischl, M.A.; Goldhirsch,A.; Testa, M.A. Quality-of-life evaluation in a clinical trial of zidovudine therapyin patients with mildly symptomatic HIV infection, The AIDS Clinical Trials Group.Ann. Intern. Med. 1992, 116, 961–966.

56. Fleming, T.R.; Prentice, R.L.; Pepe, M.S.; Glidden, D. Surrogate and auxiliary end-points in clinical trials, with potential applications in cancer and AIDS research.Statistics in Medicine 1994, 13, 955–968.

57. Fleming, T.R. Surrogate markers in AIDS and cancer trials. Statistics in Medicine1994, 13, 1423–1435.

58. Finkelstein, D.M.; Schoenfeld, D.A. Analysing survival in the presence of an auxil-iary variable. Statistics in Medicine 1994, 13, 1747–1754.

59. Finkelstein, D.M.; Schoenfeld, D.A.; Stamenovic, E. Analysis of multiple failuretime data from an AIDS clinical trial. Statistics in Medicine 1997, 16, 951–961.

60. Faucett, C.L.; Thomas, D.C. Simultaneously modelling censored survival data andrepeatedly measured covariates: a Gibbs sampling approach. Statistics in Medicine1996, 15, 1663–1685.

61. Ewell, M. Comparing methods for calculating confidence intervals for vaccine effi-cacy. Statistics in Medicine 1996, 15, 2379–2392.

62. Etzioni, R.D.; Kadane, J.B. Bayesian statistical methods in public health and medi-cine. Annu. Rev. Public Health 1995, 16, 23–41.

63. Epstein, L.D.; Munoz, A. A bivariate parametric model for survival and intermediateevent times. Statistics in Medicine 1996, 15, 1171–1185.

64. Elashoff, M.; Lagakos, S. HIV treatment strategies utilizing virologic and immuno-logic markers as criteria for changing treatments. Statistics in Medicine 1996, 15,2425–2443.

65. De Angelis, D.; Day, N.E.; Gore, S.M.; Gilks, W.R.; McGee, M.A. AIDS: the statis-tical basis for public health. Statistical Methods in Medical Research 1993, 2, 75–91.

66. Commenges, D.; Etcheverry, B. An empirical Bayes approach to the estimation ofthe incidence curve of HIV infection. Statistics in Medicine 1993, 12, 1317–1324.

67. Chiarotti, F.; Palombi, M.; Schinaia, N.; Ghirardini, A.; Bellocco, R. Median timefrom seroconversion to AIDS in Italian HIV-positive haemophiliacs: different para-metric estimates. Statistics in Medicine 1994, 13, 163–175.

68. Carlin, B. P.; Sargent, D. J. Robust Bayesian approaches for clinical trial monitoring.Statistics in Medicine 1996, 15, 1093–1106.

69. Bailey, N.T. Prediction and validation in the public health modelling of HIV/AIDS.Statistics in Medicine 1994, 13, 1933–1943.

70. Bacchetti, P. Incidence of HIV-related deaths in the United States: seasonality andtrend. Statistics in Medicine 1997, 16, 645–652.

71. Szolovits, P. Uncertainty and decisions in medical informatics. Methods Inf. Med.1995, 34, 111–121.

72. Borok, L.S. Data mining: sophisticated forms of managed care modeling throughartificial intelligence. J. Health Care Finance 1997, 23, 20–36.

A. Lawrence GouldMerck Research LaboratoriesWest Point, Pennsylvania

Dow

nloa

ded

by [

Uni

vers

ity o

f C

alif

orni

a Sa

nta

Cru

z] a

t 19:

02 0

8 O

ctob

er 2

014

Order now!

Reprints of this article can also be ordered at

http://www.dekker.com/servlet/product/DOI/101081BIP100104193

Request Permission or Order Reprints Instantly!

Interested in copying and sharing this article? In most cases, U.S. Copyright Law requires that you get permission from the article’s rightsholder before using copyrighted content.

All information and materials found in this article, including but not limited to text, trademarks, patents, logos, graphics and images (the "Materials"), are the copyrighted works and other forms of intellectual property of Marcel Dekker, Inc., or its licensors. All rights not expressly granted are reserved.

Get permission to lawfully reproduce and distribute the Materials or order reprints quickly and painlessly. Simply click on the "Request Permission/Reprints Here" link below and follow the instructions. Visit the U.S. Copyright Office for information on Fair Use limitations of U.S. copyright law. Please refer to The Association of American Publishers’ (AAP) website for guidelines on Fair Use in the Classroom.

The Materials are for your personal use only and cannot be reformatted, reposted, resold or distributed by electronic means or otherwise without permission from Marcel Dekker, Inc. Marcel Dekker, Inc. grants you the limited right to display the Materials only on your personal computer or personal wireless device, and to copy and download single copies of such Materials provided that any copyright, trademark or other notice appearing on such Materials is also retained by, displayed, copied or downloaded as part of the Materials and is not removed or obscured, and provided you do not edit, modify, alter or enhance the Materials. Please refer to our Website User Agreement for more details.

Dow

nloa

ded

by [

Uni

vers

ity o

f C

alif

orni

a Sa

nta

Cru

z] a

t 19:

02 0

8 O

ctob

er 2

014