Statistical Problems in Monitoring Adverse Drug Reactions

This paper proposes a probabilistic basis for monitoring adverse reactions to drugs, with allowance for the heterogeneities and irregularities inevitable among spontaneously submitted reports. The model partitions the population of reports into subpopulations. Each subpopulation includes clusters, a cluster being defined as a homogeneous group of reports on one drug set from one subpopulation in unit time. Signals from clusters can be more informative than those obtainable from data aggregated over the whole population.

Reports in one cluster on a particular event (drug-reaction association) are assumed to conform to a Bernoulli distribution, with a parameter varying between clusters of the same subpopulation according to a two-parameter beta distribution and a cluster size that follows a negative binomial distribution. The four parameters of the compound model are estimated by maximum likelihood. The one-sided numerical cumulative sum test (NCST) is proposed as a signalling procedure; its performance characteristics are studied by Monte Carlo sampling. Data from the UK Committee on Safety of Medicines are used in an example.

In dieser Arbeit wird ein probabilistisches Modell für die Überwachung von Arzneimittel-Nebenwirkungen unter Berücksichtigung der bei spontaner Meldung unvermeidbaren Heterogenitäten und Unregelmäßigkeiten vorgeschlagen. Das Modell untergliedert die Gesamtheit der Berichte in Teilgesamtheiten. Jede Gesamtheit enthält Cluster, wobei ein Cluster als eine homogene Gruppe von Berichten über eine Medikamentengruppe aus einer Teilgesamtheit in einer festgelegten Berichtszeit definiert wird. Alarmsignale aus Clustern können informativer sein als solche aus der Grundgesamtheit der gesammelten Daten. Es wird davon ausgegangen, daß Berichte in einem Cluster über ein bestimmtes Ereignis (Assoziation von Medikamentenwirkung) einer Bernoulli-Verteilung gehorchen, mit einem zwischen den Clustern der gleichen Teilgesamtheit variierenden, einer Zwei-Parameter-Betaverteilung folgenden Parameter und einer Cluster-Größe, die einer negativen Binomialverteilung entspricht. Die vier Parameter des Compound-Modells werden mittels der Maximum Likelihood geschätzt. Der einseitige numerische kumulative Summentest (NCST) wird als Signalgeber vorgeschlagen; seine Leistungscharakteristika werden mittels Monte Carlo-Stichprobenerhebung untersucht. In einem praktischen Beispiel werden Daten des UK Committee on Safety of Medicines benutzt.

Key-words:

Drug Monitoring - Adverse Reactions - Compound Distributions - Clusters - Cumulative Sum Test - Stochastic Simulation - Monte Carlo Methods

Schlüssel-Wörter:

Medikamentenüberwachung - Arzneimittel-Nebenwirkungen - Compound-Verteilungen - Cluster - kumulativer Summentest - stochastische Simulation - Monte Carlo-Methoden

• References

• 1 Abramowitz M, Stegun I. A. (Eds.) Handbook of Mathematical Functions. National Bureau of Standards, Applied Mathematics Series No. 55. Washington, D.C: U.S. Government Printing Office; 1964
• 2 Barnabd G. A. Control Charts and Stochastic Processes. J. roy. statist. Soc. B 21 1959; 239-270.
• 3 Finney D. J. The Design and Logic of a Monitor of Drug Use. J. chron. Dis 18 1965; 77-98.
• 4 Finney D. J. Statistical Aspects of Monitoring for Dangers in Drug Therapy. Meth. Inform. Med 10 1971; 1-8.
• 5 Finney D. J. The Detection of Causation of Adverse Events. Paper Delivered to the 39th Session of the International Statistical Institute; Vienna: 1973
• 6 Finney D. J. Signalling of Adverse Reactions to Drugs. Memorandum Submitted to WHO. Geneva: 1973
• 7 Finney D. J. Systematic Signalling of Adverse Reactions to Drugs. Meth. Inform. Med 13 1974; 1-10.
• 8 Inman W. H. W. The Reporting of Adverse Reactions to Drugs. The British System. Proc. Europ. Soc. Study Drug Toxicity 07 1966; 191-197.
• 9 Inman W. H. W. Role of Drug-reaction Monitoring in the Investigation of Thrombosis and the »Pili«. Brit. med. Bull 26 1970; 248-256.
• 10 Inman W. H. W, Adelstein A. M. Rise and Fall of Asthma Mortality in England & Wales in Relation to Use of Pressurized Aerosoles. Lancet. 1969. II: 279-285.
• 11 Inman W. H. W, Vessey M. P, Westerholm B, Enoeltjnd A. Thromboembolic Disease and the Steriodal Content of Oral Contraceptives. Brit. med. J 1970; II: 203-209.
• 12 Johnson N. L, Kotz S. Distributions in Statistics — Continuous Univariate Distributions.. Boston: Houghton, Mifflin; 1970
• 13 Mandel S. P. H. Report on Statistical Monitoring of Adverse Reactions to Drugs. Report Prepared for WHO, DEM/74. 7.. Geneva: 1974
• 14 Mandel S. P. H, Levine A, Beleño G. E. Signalling Increases in Reporting in International Monitoring of Adverse Reactions to Therapeutic Drugs. Meth. Inform. Med 15 1976; 1-10.
• 15 Page E. S. Continuous Inspection Schemes. Biometrika 41 1954; 100-114.
• 16 Patwaby K. M. Report on Statistical Aspects of the Pilot Research Project for International Drug Monitoring. Confidential Report Prepared for WHO. Geneva: 1969
• 17 Royall B. W. International Aspects of the Study of Adverse Reactions to Drugs. Biometrics 27 1971; 689-698.
• 18 Royall B. W, Venulet J. Methodology for International Drug Monitoring. Meth. Inform. Med 11 1972; 75-86.
• 19 Venulet J. Adverse Reactions to Drugs. Int. J. clin. Pharmacol 07 1973; 253-264.
• 20 Williams D. A. The Analysis of Binary Responses from Toxicological Experiments Involving Reproduction and Teratogenecity. Biometrics 31 1975; 949-952.