Paxton et al describe an approach for involving different stakeholders in processes to improve the quality of mental health services. Such initiatives can surely be powerful tools to change clinical practice and have been successfully applied in different healthcare systems. The approach has been developed within the context of quality improvement, but the authors have relabelled it as research in their paper.

‘Real-world’ research

Paxton et al place their processes in ‘the real world’, implicitly and explicitly suggesting that other service research does not happen in the same real world. I am not sure what type of research they refer to. There are numerous publications every week reporting service research studies that were conducted in the real world, that is, interviewing real patients in real services with real outcomes after real treatments. I am not aware of service research that happens in laboratories, ‘in vitro’or with fictitious patients. Also, I wonder which ‘ university clinics’ Paxton et al refer to in their paper as carrying out mental health service research. The UK does not have specific ‘ university clinics’ and (unless it is an experimental study on an innovative treatment) research is exclusively conducted in ‘ordinary services’, something that Paxton et al claim as specific to their approach. The fact that many rigorous studies have shortcomings that limit the generalisability of the findings is certainly true, but will not be totally overcome by the approach of Paxton et al. For example, the often restrictive inclusion criteria and drop-out rates in research studies are a problem, but quality improvement initiatives will also need inclusion criteria and encounter patients who are unable to give informed consent, refuse to be interviewed or drop out of follow-up assessments.

The world of academic research has been changing rapidly over the last 10 years. Funding depends on success in an increasingly globalised competition arena and requires researchers who are more or less dedicated to full-time research. Researchers who spend much of their time on local quality improvement initiatives may struggle to generate the necessary income to continue with their work - an implication that one may regard as problematic but is nevertheless very ‘real’.

Globalisation of research

The globalisation of research also means that most information on studies is available worldwide. Researchers need to keep up to date on what other researchers in the world are doing (for example to avoid unethical duplication of similar studies). It is hard to imagine how clinicians and other stakeholders can remain fully aware of the research literature so that they can competently decide on the most relevant and timely research question. Their involvement in such decisions is certainly welcome, and who would not want that ‘the research questions are fully understood and owned by service personnel as well as the academics’? Yet, service personnel are likely to expect research that has an immediate and direct impact on service improvement. This may apply to the quality improvement initiatives of Paxton et al but is unfortunately often unrealistic for rigorous scientific research. Studies that can actually contribute to the knowledge about how services can be improved often take several years and eventually only contribute a tiny amount to a worldwide growing body of evidence on a specific research issue. This evidence will then be reviewed and as a whole, one hopes, influence guidelines, clinical practice and decisions on service development. Thus, the connection between a given piece of research and practical improvements in services is more complex than those who are less familiar with research understandably tend to hope. If the processes described by Paxton et al are honest attempts to include stakeholders in decisions on research, they will have to compromise - sometimes on the smallest common denominator - on the research questions to address and the methods to use. High-quality research frequently requires independent thinking and unusual ideas. Compromises between many stakeholders are rarely the way forward in this.

Final comments

I wholeheartedly welcome the initiatives that Paxton et al propose. They can help to improve services directly - through the immediate results - as well as indirectly - through increasing commitment and enthusiasm of staff, patients and carers involved. Like any other provider of healthcare, the National Health Service should be encouraged to fund these initiatives. Yet, when Paxton et al call their approach ‘research’, they should at least acknowledge that this is not without difficulty and that there may be a tension between the interest of local stakeholders and those of the global research community. Also, if the terms ‘quality improvement’, ‘reflective practice’ and ‘ research’ are not synonymous, their different connotations should be clarified for a useful debate on what ‘research’ in the real world should encompass.

Declaration of interest

None.