Unity from diversity: the evidential use of anecdotal reports of adverse drug reactions and interactions

WITHDRAWN: Diclofenac for acute pain in children

At December 2014, this review has been withdrawn from the Cochrane Library. This review is out of date, although it is correct at the date of publication. The review may be misleading as new studies could alter the original conclusions. All previous versions of the review can be found in the ‘Other versions’ tab. We are seeking additional authors to support the updating of this review. For further information, please contact PaPaS Managing Editor, Anna Hobson [Contact Person].

The editorial group responsible for this previously published document have withdrawn it from publication.

Mere anecdote: evidence and stories in medicine

In evidence-based medicine, randomized controlled trials are said to be the highest evidence of what works, while anecdotes have low value or are not even considered to be medical evidence. Similar hierarchical views of evidence have infected other disciplines, including evidence-based education and evidence-based government. Here, I explore the artificial divisions of acceptable from unacceptable evidence, numbers from narrative and sciences from humanities. I challenge the deprecation of stories in medicine. Some stories are based on experiments while others are based on more or less plausible theories. Some stories offer vast and impressive statistics gathered from many observations while others present one noteworthy event. Published reports are themselves stories of what experimenters did. Systematic reviewers generate their own observations of collected stories of experiments. Reviewers of systematic reviews in turn report their observations of systematic reviews. All of these stories become evidence of what works in medicine.

The existence of a relationship between increased serum alanine aminotransferase levels detected in premarketing clinical trials and postmarketing published hepatotoxicity case reports

BACKGROUND

Drug-induced liver injury (DILI) profile in most drugs' available information is based on both the incidence of alanine aminotansferase (ALT) elevations in clinical trials and published case reports.

AIM

To assess the relationship between ALT elevations in clinical trials and the number of published case reports in the postmarketing setting.

METHODS

Hepatotoxic drugs were identified from product labelling and classified in high-medium risk (Black Box Warning or Precautions section) or low risk (a statement in the Adverse Reactions section). Incidence of ALT elevations (> or = 3 x ULN) for drug (I(D)) and placebo (I(C)) treated patients in premarketing clinical trials and DILI published case reports were retrieved from product labelling and MEDLINE.

RESULTS

The median I(C) was 10/1000. The high-medium-risk drugs' median I(D) was significantly higher compared with low-risk drugs (17/1000 vs. 10/1000; P = 0.046). Chi-squared test, absolute difference and odds ratio comparing I(D) and I(C) identified 35%, 51% and 77% of high-medium-risk drugs respectively. Less number of case reports were associated with low- than high-medium-risk drugs (1 vs. 7; P = 0.001). A high odds ratio in clinical trials (I(D) vs. I(C)) was the strongest predictor of published DILI case reports.

CONCLUSION

A relationship between increased ALT incidence in premarketing clinical trials and postmarketing published case reports exists.

Diclofenac for acute pain in children

BACKGROUND

Diclofenac is commonly used for acute pain in children, but is not licensed for this indication in all age groups.

OBJECTIVES

1) Assess the efficacy of diclofenac for acute pain in children. 2) Assess the safety of diclofenac for short-term use in children. 3) Identify gaps in the evidence to direct future research.

SEARCH STRATEGY

Seventeen databases indexing clinical trial reports were searched in February 2005 (with an update search as part of this first review in May 2008). A hand search of Paediatric Anaesthesia was undertaken and summaries obtained of adverse reaction reports from the UK Yellow Card Scheme and World Health Organization (WHO) Monitoring Centre. The reference lists of included studies were also searched.

SELECTION CRITERIA

Any published report, in any language, involving the administration of diclofenac to a patient aged 18 years or younger for acute pain and detailing either monitoring of efficacy or safety.

DATA COLLECTION AND ANALYSIS

Two review authors independently assessed study quality and extracted the data. Authors were contacted where necessary. Review Manager version 5 was used for analysis.

MAIN RESULTS

1) EFFICACY: randomised controlled trials (RCTs) comparing diclofenac with placebo/any other treatment by using pain scores (assessed or reported), or need for rescue analgesia.2) SAFETY: any type of study seeking adverse events (regardless of cause). An adverse event was defined as any reported adverse or untoward happening to a patient being treated with diclofenac for acute pain.Seven publications on diclofenac efficacy and 79 on safety (74 studies plus five case reports) were included in the final analysis. Compared with placebo/no treatment, diclofenac significantly reduced need for post-operative rescue analgesia (relative risk [RR] 0.6; number needed to treat to benefit [NNT] 3.6; 95% confidence interval [CI] 2.5 to 6.3).Compared with any other non-NSAID, patients receiving diclofenac suffered less nausea or vomiting, or both (RR 0.6; NNT 7.7 [5.3 to 14.3]). There appeared to be no increase in bleeding requiring surgical intervention in patients receiving diclofenac in the peri-operative period. Serious diclofenac adverse reactions occurred in fewer than 0.24% of children treated for acute pain. The types of serious adverse reactions were similar to those reported in adults.

AUTHORS' CONCLUSIONS

Diclofenac is an effective analgesic for perioperative acute pain in children. It causes similar types of serious adverse reactions in children as in adults, but these are rare. More research on optimum dosing and safety in asthmatic children is required.

Gold standards in pharmacovigilance: the use of definitive anecdotal reports of adverse drug reactions as pure gold and high-grade ore

Anecdotal reports of adverse drug reactions are generally regarded as being of poor evidential quality. This is especially relevant for postmarketing drug safety surveillance, which relies heavily on spontaneous anecdotal reports. The numerous limitations of spontaneous reports cannot be overemphasised, but there is another side to the story: these datasets also contain anecdotal reports that can be considered to describe definitive adverse reactions, without the need for further formal verification. We have previously defined four categories of such adverse reactions: (i) extracellular or intracellular tissue deposition of the drug or a metabolite; (ii) a specific anatomical location or pattern of injury; (iii) physiological dysfunction or direct tissue damage demonstrable by physicochemical testing; and (iv) infection, as a result of the administration of an infective agent as the therapeutic substance or because of demonstrable contamination. In this article, we discuss the implications of these definitive ('between-the-eyes') adverse effects for pharmacovigilance.

Dietary supplements, herbs and oral anticoagulants: the nature of the evidence

In the US, the use of dietary supplements, including vitamins, minerals, amino acids, and herbal products, is extensive. Nonetheless, the majority of patients report that they have little information about the risks, benefits, and adverse effects of medicines, or about their potential interactions with prescription drugs. Patients taking warfarin are at particular risk of interactions with dietary supplements, yet approximately 30% use herbal or natural product supplements on a regular basis. No current governmental regulations or voluntary programs address dietary supplement interactions with prescription drugs. Case reports represent the majority of the evidence surrounding drug interactions between warfarin and dietary supplements. Those of the highest quality include, as an assessment of causality, a modification of the recently published Drug Interaction Probability Scale. Despite positive case reports, formal drug interaction studies are often negative, suggesting that numerous patient-specific influences other than the suspected interaction alone may be responsible for a particular observation. The cranberry-juice/warfarin interaction is a recent example of such a discrepancy. Healthcare providers can play an active role in improving quantity and the quality of case reports of interactions involving warfarin and dietary supplements. A registry of anticoagulant interactions with dietary supplements has been proposed, and is currently being developed through Clotcare Online Resource (http://www.clotcare.com). The goal of this registry is to obtain high quality case-based evidence of drug interactions between anticoagulants and dietary supplements, to define these interactions based on clinical and monitoring outcomes, and to analyze likelihood of causation using a modification of the Drug Interaction Probability Scale.