The results presented by Fran Paradiso-Hardy and colleagues[1][1] are an excellent example of formal Bayesian causality assessment[2][2] of a series of reported cases of suspected adverse drug reactions to ticlopidine. A sensible reader might ask a number of questions. For instance, wouldn't listing

We present a method for predicting the rate of adverse reactions to a drug. The approach employs a graphical model, as previously used for assessing causality in individual cases, but instead of attempting to interpret what did happen in an individual case we use it to predict what will happen in the next case or series of cases treated. The approach is illustrated on the adverse reactions of pseudomembranous colitis due to antibiotics. Based on a representation of the process by which this adverse event is recognized and reported, and expert opinion on the probabilities of key signs given diagnosis, we demonstrate how the model predicts the rate of true and reponed adverse events. We also demonstrate how an elaboration of this approach can allow the estimates to be modified continuously and updated as new cases are reported. Although further work needs to be done to make the system practically usable, we see it as a demonstration of a new and radically different approach to the main concern of pharmaco-epidemiology-the assessment of drug safety. The crucial shift is from a primary concern with estimating incidence rates in patients treated in the past to predicting incidence rates in future patients to be treated with a drug.

Probabilistic expert systems are intended to provide reasoned guidance in complex environments characterized by extensive uncertainty. An explicit 'causal' model is constructed for the process being observed, in which an acyclic directed graph is used to express conditional independence assumptions about variables, and probability assessments specify a full joint probability distribution. The resulting graphical structure can cope with a range of issues that arise in realistic modelling. Here we consider a particular example of assessing the chance that a suspected adverse reaction is due to a particular drug under suspicion. The background biological knowledge provides an appropriate model and probability assessments are obtained from expert microbiologists. The model allows a variety of interpretations for 'causality'. Details of the graphical and computational algorithms used to perform efficient calculations of conditional probabilities on complex graphical structures are provided and illustrated with the example. Further developments should allow updating of the risk parameters in the light of a series of case reports, and may form the basis for a flexible expert system for causality assessment and post-marketing surveillance.

No AccessJournal of UrologyRadiology, Nuclear Medicine and Sonography1 Jun 1985Renal Function Following Infusion of Radiologic Contrast Material: A Prospective Controlled Study B.C. Cramer, P.S. Parfrey, T.A. Hutchinson, D. Baran, D.M. Melanson, R.E. Ethier, and J.F. Seely B.C. CramerB.C. Cramer More articles by this author , P.S. ParfreyP.S. Parfrey More articles by this author , T.A. HutchinsonT.A. Hutchinson More articles by this author , D. BaranD. Baran More articles by this author , D.M. MelansonD.M. Melanson More articles by this author , R.E. EthierR.E. Ethier More articles by this author , and J.F. SeelyJ.F. Seely More articles by this author View All Author Informationhttps://doi.org/10.1016/S0022-5347(17)49427-4AboutPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareFacebookLinked InTwitterEmail "Renal Function Following Infusion of Radiologic Contrast Material: A Prospective Controlled Study." The Journal of Urology, 133(6), pp. 1135–1136 © 1985 by The American Urological Association Education and Research, Inc.FiguresReferencesRelatedDetails Volume 133Issue 6June 1985Page: 1135-1136 Advertisement Copyright & Permissions© 1985 by The American Urological Association Education and Research, Inc.MetricsAuthor Information B.C. Cramer More articles by this author P.S. Parfrey More articles by this author T.A. Hutchinson More articles by this author D. Baran More articles by this author D.M. Melanson More articles by this author R.E. Ethier More articles by this author J.F. Seely More articles by this author Expand All Advertisement PDF DownloadLoading ...

To establish the impact of transplantation on the course of chronic hepatitis B liver disease we performed a prospective study of the clinical and pathological sequelae of hepatitis B disease in all 22 patients who had renal allografts that functioned for more than 1 year and who were hepatitis B surface antigen (HBsAg)-positive following transplantation. No patient converted to HBsAg-negative. During a mean follow-up of 83 months serial liver biopsies were performed in 20 patients and 1 liver biopsy was available in the remaining 2 patients. Eleven patients died of liver disease, 5 of whom died of hepatic failure, 3 with hepatoma, 2 of gastrointestinal hemorrhage, and 1 of ascites with pleuroperitoneal fistula. Aggressive liver disease was observed in the vast majority of patients: 12 ultimately developed cirrhosis, (mean follow-up 81 months), 6 chronic active hepatitis (mean follow-up 93 months), 3 chronic persistent hepatitis (mean follow-up 89 months), and in 1 patient the presence of HB virus in hepatocytes was the sole morphologic alteration (follow-up 42 months). There was a marked tendency to progression in that 82% of patients with virus only, reactive hepatitis, or chronic persistent hepatitis on initial biopsy subsequently developed chronic active hepatitis or cirrhosis. For comparison, 10 HBsAg-positive patients whose renal failure had been treated by hemodialysis were also studied over a comparable period. Four patients converted to the negative state. Biochemical evidence of persistent liver dysfunction occurred in only 1 patient and no patient has died from complications of liver disease. We conclude that in the immunosuppressed renal transplant patient HB infection often results in the development of cirrhosis, leading to death from hepatoma and hepatic failure. This course is worse than that in dialysis patients. Renal transplantation of HBsAg-positive patients with end-stage renal failure may be inadvisable.

A prospective study of the clinical and pathological sequelae of hepatitis B disease in 22 immunosuppressed renal transplant patients is reported. All patients had allografts that functioned for more than 1 year, and all were hepatitis B surface antigen (HB8Ag)-positive following transplantation. None of the 18 patients who had serial HB8Ag tests converted to HB8Ag negative. Serial liver biopsies were performed in 19 patients and one liver biopsy was available in the remaining three patients. Follow-up ranged from 12 to 93 months. Seven patients ultimately developed cirrhosis, 6 developed chronic active hepatitis, 5 developed chronic persistent hepatitis, and in 4 the presence of HB virus in hepatocytes was the sole morphologic alternation. The initial liver biopsy was not an accurate predictor of ultimate severity of liver disease because 5 of the 12 patients with virus only or chronic persistent hepatitis subsequently developed chronic active hepatitis or cirrhosis. Clinical liver dysfunction occurred in 8 patients, all of whom had chronic active hepatitis or cirrhosis. Three patients died with hepatic failure and 2 with hepatoma. The risk of death from liver disease in HB8Ag-positive renal transplant patients was 5% per patient-year. For comparison, 10 HB8Ag-positive patients whose renal failure had been treated by hemodialysis were also studied over a comparable period. Biochemical evidence of persistent liver dysfunction recurred in 1 patient only; 4 patients converted to the HB8Ag-negative state; and no patient has died from complications of liver disease. We conclude that in the immunosuppressed renal transplant patient HB infection often results in the development of chronic active hepatitis, leading to cirrhosis and death from hepatoma and hepatic failure.

The assessment of causality in drug-event associations depends on the setting and purpose of such an assessment. Epidemiologists are primarily interested in population-based inferences about whether a given drug can cause a certain adverse drug reaction (ADR), and if so, how often it does so. Pharmaceutical industries and regulatory agencies are also concerned with population-based risks, but in addition must worry about individual cases. Clinicians are primarily interested in the individual, ie, whether a given drug did cause a certain adverse event in a particular patient. The authors describe an algorithm that provides specific, detailed criteria for ranking the probability that an observed untoward clinical manifestation was caused by a given drug. The criteria are subdivided into six axes of decision strategy with a built-in scoring system that ordinally ranks the probability of an adverse drug reaction as definite, probable, possible, or unlikely. To illustrate the use of the algorithm, the authors assess a reference case of pancreatitis occurring after administration of methyldopa.

At the Royal Victoria Hospital in Montreal, 22 patients received two successive cadaver renal transplants. The results were analysed to determine which factors have the best predictive value for success or failure in renal retransplantation. The fate of a second cadaver renal allograft was found to be about the same as the first if the initial transplant has been lost because of rejection and not technical failure. The duration of survival of the initial transplant serves as the best guide to potential outcome of retransplantation after rejection of the initial graft.

Despite widespread clinical and epidemiologic attention to adverse drug reactions (ADRs), their clinical identification has been a nonreproducible act of unspecified subjective judgment; adequate operational criteria have not been available for diagnostic decisions about the cause of an observed untoward clinical manifestation. To improve scientific precision in the diagnosis of ADRs, we have developed an algorithm that provides detailed operational criteria for ranking the probability of causation when ADR is suspected between a drug and a clinical manifestation. The algorithm provides a scoring system for six axes of decision strategy: previous general experience with the drug, alternative etiologic candidates, timing of events, drug levels and evidence of overdose, dechallenge, and rechallenge. The sum of the scores is ordinally partitioned to rate the candidate ADR as definite, probable, possible, or unlikely.