Original ArticleDefining adverse events in manual therapies: A modified Delphi consensus study
Introduction
The incidence of adverse events from manual therapy is of considerable interest to manual therapists and to the general public. Good quality data are sparse, with scientific debate about incidence of adverse events foundering on differences in opinion as to what constitutes a therapy-related adverse event rather than the incidence itself. Defining therapy-related adverse events in manual therapy is difficult as they occur in many guises, contexts and settings. They can range in severity and impact; also, patient and practitioner views and expectations about what constitutes an important adverse event may differ. The literature about manual therapy-related adverse events is dominated by studies concerning manipulation (Stevinson and Ernst, 2002, Kerry et al., 2008); specifically, high velocity thrust techniques used on the cervical spine and consequential cervical artery dissections – vertebral and internal carotid arteries, vertebrobasilar accidents and strokes (Gross al., 2003, Haneline et al., 2005; Kawchuk et al., 2008; Dittrich et al., 2007). There is, however, a large spectrum of adverse events that can occur with varying degrees of severity and duration, from transient muscle aches to bruising to fracture.
The World Health Organisation Adverse Reaction Team (WHO-ART) and the pharmaceutical industry have each been considering the definition of adverse events for decades and have clearer definitions than many other organisations (Leape and Abbokire, 2003). In addition, adverse events, reactions, harm, safety and side effects are defined and used in the revised and extended 2003 CONSORT statement (Ioannidis et al., 2004) for reporting clinical trial data. Whilst these definitions and guidelines are useful to the manual therapy professions, they are not entirely applicable as it is often difficult to assign causality, or to measure the ‘dose’ of a manual therapy.
Malone et al. (2002) defined an adverse ‘effect’ as any detrimental result of a treatment; a ‘reaction’ as a slight or clinically insignificant short lived symptom and an ‘incident’ as an unexpected event resulting in serious impairment, injury or fatality or an irreversible complication. Thiel et al. (2007) used a pharmaceutical definition (Edwards and Aronson, 2000) and applied it pragmatically to a prospective cohort study about adverse events in chiropractic. Serious adverse events were defined as: ‘referred to hospital accident and emergency and/or severe onset or worsening of symptoms immediately after treatment and/or resulted in persistent or significant disability/incapacity’. Other graded definitions have been used such as: ‘certain neurological deficits’; ‘severe neurological deficits’; and ‘serious complications’ (Dvorak and Orelli, 1985). The problem with these definitions is that they do not cover the range of adverse events that may exist in manual therapies.
Manual therapy professions such as chiropractic, osteopathy and physiotherapy are obliged under their codes of conduct to seek consent before administering treatment. Gaining informed consent, however, is difficult as we know little about risks involved with different treatments. As a first step towards quantifying risk, and providing patients with realistic estimates of the incidence of important therapy-related adverse effects, there is a need for a pragmatic definition of adverse events applicable to manual therapy. The aim of this study was, therefore, to seek an expert consensus definition of adverse events in relation to manual therapy by exploring understanding and meaning using a modified Delphi technique (Dalkey and Helmer, 1963).
Section snippets
Modified Delphi consensus study
A Delphi consensus study is a questionnaire survey of expert opinion conducted in ‘rounds’; responses to each round of questionnaires are fed anonymously back to participants until an agreement or consensus is evolved or established. We selected this approach both to avoid key individuals' views dominating any open discussion and to ensure we could achieve international representation on our panel. In all three rounds of this Delphi study, consensus was defined as >74% agreement. We reviewed a
Focus group
The focus group discussed the issues surrounding adverse events in manual therapy and highlighted the need for a hierarchy that could: a) classify adverse events in order of importance and b) take into account ‘non-adverse’ adverse events. The group decided on a hierarchical taxonomy using the terms ‘minor’, ‘moderate’, ‘major’ and ‘not adverse’. The definitions of these terms were to be decided by the Delphi process. The focus group generated constructs that they believed to be important
Discussion
We believe that this Delphi study is the first of its type to address the issue of defining an adverse event in the context of manual therapy in a systematic, non individual and interdisciplinary way. We developed a layered approach for defining adverse events. The first layer identifies duration and severity and the second layer provides context and description about the nature of the adverse event; this enables us to classify any adverse event into a hierarchy of minor, moderate, or major.
Conclusions
The definitions obtained following this Delphi study can be used to categorise or classify adverse events in the context of manual therapy. Not only is a logical hierarchy presented, but also this definition allows for classifying those events that occur that may be regarded as ‘not adverse’. The application of this definition may be useful in both research and clinical settings for recording and documenting the nature and type, prevalence and incidence of adverse events to increase
Acknowledgements
The members of the focus group: Pamela Cross, Sandra Mellors, Haymo Thiel, Steve Vogel and study participants.
The General Osteopathic Council and the National Council for Osteopathic Research for funding this study.
References (18)
- et al.
Adverse drug reactions: definitions, diagnosis and management
Lancet
(2000) - et al.
Manual therapy for mechanical neck disorders: a systematic review
Manual Therapy
(2002) - et al.
An analysis of the etiology of cervical artery dissections: 1994 to 2003
Journal of Manipulative & Physiological Therapeutics
(2005) - et al.
An evidence and consensus based guideline for acute diarrhoea management
Archives of Disease in Childhood
(2001) - et al.
Dysfunctional tear syndrome: a Delphi approach to treatment recommendations
Cornea
(2006) BS EN ISO 14155-1:2003–clinical investigation of medical devices for human subjects – part 1: general requirements
(2003)- et al.
An experimental application of the Delphi method for the use of experts
Management Science
(1963) - et al.
Mild mechanical traumas are possible risk factors for cervical artery dissection
Cerebrovascular Disease
(2007) - et al.
How dangerous is manipulation to the cervical spine?
Manual Medicine
(1985)