Internal and external validity of cluster randomised trials: systematic review of recent trials

With the development of society and the progress of medicine, clinical research methods have increased day by day, and design of clinical trials is deliberate increasingly, to a certain extent, all of these ensure that the results of clinical research are objective and scientific, but these results and by the guidelines and consensus have much practical significance on clinical practice are still controversial [1,2,3]. The author is engaged in both clinical practice and clinical research, deeply feel that it is necessary to discuss the relationship between the above two, and now address it as follows.

1. Refine scientific question from clinical practice. As a doctor, save lives in clinical practice, while has faced a series of clinical problems; on these problems to be timely summarized and extracted have become a source of scientific question[4]. For example, aspirin is a cornerstone of prevention of cardiovascular events, but the 5 to 40% of clinical patients will still have cardiovascular events after they applied it, and this is aspirin resistance. So bring some clinical problems, such as whether it has difference between the patients of aspirin resistance and sensitive patients on incidence rate of cardiovascular events or not, and how to solve clinical problems aspirin resistance[5].

The scientific question's explanation is the purpose of clinical trials. Clinicians learn to extract, refine scientific question in practice, and then to carry out clinical trials, these will not only promote their clinical practice level enhancement, but also will bring benefits on patients.

2. Clinical trial design serves clinical practice. After World War II, randomized controlled trial was used to assess the effectiveness of streptomycin treatment on TB, which was accepted and applied widely by the medical profession. But we should consider different types of methods of clinical trial design, when we carry it out. Every research method has strengths and weaknesses which cannot be resolved within that method itself, so according to problems of clinical practice we should confirm the purpose of the research, while select the method [6].

There are different methods to answer different scientific questions. A poorly designed and badly implemented RCT is, as a rule, less valuable than well conducted studies using other designs. Sometimes even non- randomized studies can produce more reliable and useful information than a well conducted randomized study, and better services for clinical practice. It is not important whether study randomizes or not, but whether the adopted method is fit for us to answer the raised question in research, and whether perform this method in a scientific and strict manner.

3. To implement clinical trial protocol should be strict. According to the forthcoming purpose, clinical trial has its relevant design protocol. Investigators, especially as the study clinicians in the front lines, often incompletely implement the protocol due to the objective and subjective reasons, resulting in clinical trial design trend to formality. For instance, with regard to design of blindness, some investigator often has to guess groups intentionally or unintentionally to lead blindness failed; with regard to restrictive combination of using drugs, owing to clinical concern, some investigator has restriction on subjects who not do in terms of the protocol, eventually only to remove it.

So specially emphasized on investigators should be strict in the implementation of clinical trials protocol. As the study clinicians should avoid tending to the common fault, lest the research occur bias, resulting in the poor quality or even the failure of the trial.

4. Trial results in clinical practice should be applied moderately. One of the clinician` s job is to extract knowledge from colony research, and arrange, transform it into relevant data about individual patients, which is one of the core skills of evidence-based clinical practice. According to how to sum up the important result and use it on individual patients, David Sackett once suggested doctors contemplate it as follow [7]:

A.Whether it is much difference between our patients and patients in the study, and whether the results are applicable to our patients?

B.In the current health care environment, the treatment method in the study is feasible?

C.What are the potential harms and benefits about performing the treatment on our patients?

D.What are our patients` values and expectations for us want to avoid the outcome and adopt treatment measures?

Doctors must arrange the research evidence into the form of the requirement of clinical decision fitting the patients, just as a tailor must fit the fabric into the clothes wearing on person [8]. For example, a clinician meet a patient who has got a disease of gastric cancer or migraine, but in many respects this patient has some differences with the allowing patient in the literature, so clinician must arrange the research evidence to ensure that the collection of information fits for the current patient. This is only the reason of moderate application for the outcome of clinical trials in clinical practice [9]. At the same time we consider that less emphasized consensus on evidence-based and more concerned about the personality in clinical practice might not be a bad thing.

CONCLUSION As the team of designers of clinical research and implementers, we need to carefully consider what kind of research can really help us treat patients. It is a reason that the patients are randomly assigned to use the same drug for a different strategy group to accept treatment; maybe find a way to make dose settings of blindness study more variation, to a greater extent, which would be beneficial to our clinical practice. Therefore we should cautiously design clinical trials to answer the problem that is really in need of solution in clinical work.

REFERENCES

1. Birgitta S, Kjell A, Kerstin H, et al. Stroke Units in Their Natural Habitat Can Results of Randomized Trials Be Reproduced in Routine Clinical Practice? Stroke.1999; 30:709-14

2. Jacqueline A. French. Can Evidence-Based Guidelines and Clinical Trials Tell us How to Treat Patients? Epilepsia.2007; 48(7):1264-67

3. Steg PG, Lopez SJ, Lopez SaE, et al. External validity of clinical trials in acute myocardial infarction. Archives of Internal Medicine 2007; 167(1):68-73

4. Richard S. Doctors are not scientists. BMJ.2004; 328:0

5. George K, Stephanie JB, W Scott B, et al. Aspirin "resistance" and risk of cardiovascular morbidity: systematic review and meta-analysis. BMJ.2008; 336:195-8

6. Harald W, Torkel F, Vinjar F, et al. Circular instead of hierarchical: methodological principles for the evaluation of complex interventions. BMC Medical Research Methodology 2006; 6:29

7. Sackett D.L., Straus S, Richardson W.S., et al. Evidence-based Medicine. Harcourt Brace. 1999

8. J.A. Muir Gray, Ying Qin, Jinling Tang translate. The Resourceful Patient. Peking University Medical Press.2006

9. Peter M Rothwell. External validity of randomised controlled trials:To whom do the results of this trial apply? Lancet. 2005; 365:82-93

Competing interests: None declared