A qualitative study evaluating causality attribution for serious adverse events during early phase oncology clinical trials

Background In early phase oncology trials, novel targeted therapies are increasingly being tested in combination with traditional agents creating greater potential for enhanced and new toxicities. When a patient experiences a serious adverse event (SAE), investigators must determine whether the event is attributable to the investigational drug or not. This study seeks to understand the clinical reasoning, tools used and challenges faced by the researchers who assign causality to SAE’s. Methods Thirty-two semi-structured interviews were conducted with medical oncologists and trial coordinators at six Canadian academic cancer centres. Interviews were recorded and transcribed verbatim. Individual interview content analysis was followed by thematic analysis across the interview set. Findings Our study found that causality assessment tends to be a rather complex process, often without complete clinical and investigational data at hand. Researchers described using a common processing strategy whereby they gather pertinent information, eliminate alternative explanations, and consider whether or not the study drug resulted in the SAE. Many of the interviewed participants voiced concern that causality assessments are often conducted quickly and tend to be highly subjective. Many participants were unable to identify any useful tools to help in assigning causality and welcomed more objectivity in the overall process. Interpretation Attributing causality to SAE’s is a complex process. Clinical trial researchers apply a logical system of reasoning, but feel that the current method of assigning causality could be improved. Based on these findings, future research involving the development of a new causality assessment tool specifically for use in early phase oncology clinical trials may be useful.

We thank Kalpana Nair and Sheri Burns for review of the interview guide, Nancy Bordignon for transcribing the interviews and Emmy Arnold and Susan Dimitry for providing invaluable edits to the manuscript.

Authors and Affiliations

1. Juravinski Cancer Centre, McMaster University, 699 Concession Street, Hamilton, ON, Canada, L8V 5C2

   Som D. Mukherjee, Megan E. Coombes, Brenda Kowaleski & Andrew Arnold

2. Centre for Evaluation of Medicines, St. Joseph’s Health Care, McMaster University, Hamilton, ON, Canada

   Mitch Levine

3. Brock University, St. Catherine’s, ON, Canada

   Jarold Cosby

Corresponding author

Correspondence to Andrew Arnold.

Presented, in part, at the 43 ^rd Annual Meeting of the American Society Clinical Oncology, Chicago, USA, 2007 June 1–5 and at the 23 rd International Conference on Pharmacoepidemiology & Therapeutic Risk Management, Quebec City, Canada, August 19–22, 2007

Research Support

This research was supported by a grant-in-aid from AstraZeneca, Canada Inc.

Appendix 1: Interview guide

Semi-structured questions:

1. 1)

   Clinical Reasoning:

   1. a)

      Explanation of how a report of a serious adverse event is handled/processed

   2. b)

      Factors considered when assessing causality?

   3. c)

      General guidelines followed when assigning causality?

2. 2)

   Information Resources (e.g. Investigator’s Brochure):*

   1. a)

      Resources referred to when assigning causality?

3. 3)

   Tools (e.g. decision trees):

   1. a)

      Awareness of tools to help assign causality?

   2. b)

      Tools used to help assign causality?

4. 4)

   Challenges / Concerns:

   1. a)

      Problems or challenges with assigning causality?

   2. b)

      Concerns about how clinicians currently assign causality?

   3. c)

      External influences/pressures from third parties when assigning causality?

   4. d)

      What would make assigning causality easier?

5. 5)

   Education:*

   1. a)

      Formal and informal training received with respect to assigning causality?

   2. b)

      Educational needs around causality assessment?

* Although the domains of Information Resources and Education were explored in the interviews the findings are beyond the scope of what is presented here.

Reprints and permissions