Editorial: Evaluation of CAM interventions


Increasingly, complementary and alternative medicine (CAM) research is being conducted by clinical nurses; this is partly evidenced by the evolution of CAM special issues and sections in JCN and previous editorials on the subject (Smith 2008a,b,c). There is a growing awareness amongst CAM researchers and practitioners of the importance of employing appropriate and rigorous research methods when studying an intervention (Broom & Adams 2007). Presently, CAM research faces unique challenges in relation to the design and execution of studies, with specific respect to generalisability and internal validity. Furthermore, research on CAM interventions is usually only considered once the therapy is widely used, by which time personal experiences and expectation biases may well be established. As such, many published CAM studies are methodologically flawed; they suffer from poor design and inappropriate analysis (O’Mathuna 1998). Watson (2008) identified the need for more scientific rigour in CAM research.

CAM research design is further complicated when a clinical nurse is delivering the intervention, where characteristic and incidental components of the study may be intertwined and often difficult to define. This very issue was highlighted by Paterson and Dieppe (2005) in their examination of placebo effects in acupuncture. What is clear in CAM research is that interventions are often very complex in nature and that evaluation can be problematic.

Saks (2005) has identified large gaps in CAM research, in relation to the mechanisms of action, safety issues and the synergies between conventional medicine and CAM in clinical practice. The most fundamental – and perhaps most pertinent – question for clinical nurses in CAM research is related to the efficacy of an intervention and it is this question that requires the most attention. How can nurse researchers show that CAM therapies work in their clinical practice? It is important that efficacy of CAM is evaluated as rigorously as conventional medicine to protect our patients from unsafe and unethical practices.

With increasing focus on evidence-based practice, a rigorously designed randomised controlled trial would be considered the most robust way to obtain evidence about the effectiveness of any health care intervention. This approach is well respected as the ‘gold standard’ in medical research. For interventions with one component, like clinical drug trials, standardisation is straightforward by measuring optimal dose of a drug against placebo or standard treatment. However, standardisation is more complicated when we deal with the evaluation of nurse-led CAM interventions. Such interventions are complex as they consist of many interdependent and inter-related components. The delivery of a nurse-led hypnotherapy service for irritable bowel syndrome patients highlights complex intervention highlights this point (Smith 2006).

Complex interventions (CIs) in health care comprise several separate elements which seem essential for the proper function of the intervention, although the ‘active ingredient’ or ‘X factor’ of the intervention that is effective is difficult to specify. They have been defined as ‘built up from several components, which may act both independently and interdependently’ (MRC 2000, p. 2). Much nursing and CAM research fulfils these criteria. Campbell et al. (2000) viewed complex interventions as often difficult to measure using conventional RCT’s as by there very nature they have multiple variants that can affect the outcomes. There have been a variety of approaches that have been developed to evaluate complex treatment, including Medical Research Council (MRC) guidance from the UK.

The MRC originally produced a framework for the development and evaluation of RCT’s for complex interventions to improve health. They proposed an approach to evaluation which was similar to that adopted for evaluating new drugs. Their five-phase sequential and step-wise model framework for the evaluation of complex interventions, from preclinical theoretical work through to long-term implementation, concluded that the RCT was the optimal study design for CIs.

Hawe et al. (2004) raised concerns about the standardisation in CI research. Despite recommending the use of RCT’s, they proposed that researchers should be lees concerned about the actual content of a CI and more about its function. Blackwood (2006) also supported the use of the MRC framework in clinical nursing research; however, she argued that each component of a CI should have a separate evaluation and highlighted the potential role of qualitative research methods in CI evaluation. This use of qualitative methods alongside randomised controlled trials has been further elaborated by Lewin et al. (2009).

The MRC original report has now been updated (MRC 2008, http://www.mrc.ac.uk/complexinterventions, retrieved 27/09/10) to reflect developments since its original publication. To deal with criticism of the original report, the new guidance has a broader scope, and it covers observational methods as well as RCT’s and includes a broader definition of complex interventions beyond the core of having several components. In their summary of the new framework, Craig et al. (2008) identified a less linear model than the original guidance.

The need for both quantitative and qualitative approaches in CAM research has been identified (Saks 2005). It may be that some modification to research methodology is appropriate in CAM research to ensure that what is being examined under experimental conditions is consistent with everyday practice CAM. Nursing research in clinical practice can often have several components.

In this special section, there are several papers which highlight the complex nature of evaluating CAM interventions. Music therapy is one form of intervention which receives attention and the use of three different research designs highlights the methodological challenges that nurse researcher faces evaluating music therapy. Lin et al. (2010a,b) used a quasi-experimental design to evaluate the effects of music therapy on anxiety and postoperative pain in patients undergoing spine surgery. They concluded that music therapy had a positive impact on these patients. In a Turkish study, Korhan et al. (2010) examined the effect of music on patients receiving mechanical ventilatory support using a study case control experimental repeated measures design. They demonstrated that music therapy was effective in reducing anxiety levels in this patient group. Finally, a randomised control trial design was employed to determine the effect of music therapy on state anxiety and anxiety-induced physiological manifestations in cancer patients following chemotherapy (Lin et al. 2010). In this Taiwanese study, music therapy was found to be effective in reducing chemotherapy-induced anxiety. These studies demonstrate the positive impact that music therapy can play in health care and suggest that, when used appropriately, it can have a positive impact on quality of nursing care.

In summary, the real challenge for CAM researchers is to derive a research methodology to address philosophical concerns and maintain high standards of methodological rigour. The updated MRC framework highlights several key issues that help nurse researchers develop and evaluate their CAM study. First, ensuring relevant research evidence is used systematically to develop all components of the intervention. Second, improve definition and measurement of potentially complex interventions. Finally, adapt the most relevant research design/evaluation for their study. This may involve the use of both quantitative and qualitative or a mixed method approach. There exists many ways of collecting CAM research data, and nurse researchers should make best use of the MRC guidance to identify the most appropriate and feasible approach to research in this area.