Quality of Evidence

Apart from the scientific evidence, the information consumer should know how to do research appraisal, which is different from the level of evidence; it is like the diligent and lazy person are different from graduating with certificate or bachelor’s or master’s degree. Research appraisal normally done in the following five aspects:

1. Research question is clear or not or the research aims to answer which question.

2. Research conclusion answer research question or not. It means that the indicator is set to suit question or not. This is because sometimes research result and conclusion are inconsistent. The reader who cannot read the research appraisal is fooled by the conclusion.

3. Research methodology is good or not. The good characteristics of research methodology are

4. The research result is published in the standard medical magazine has the peer-reviewed journal or not. Good research results that are acceptable by the medical science worldwide would be accepted to publish in the good journal such as New England Journal of Medicine, JAMA, Lancet, Circulation and BMJ. The cited research result that never been publishing in medical journal is considered as the fake research. The bad research result or the fake result is always published on internet as it is daily issue.

5. Research result is in contrast with the overall information or not or it is contrast with the common sense or not. If it does, the research should be redone and see the overall results. The result that is consistent with the common sense shall be the reliable result.

Quality of research is the deliberate issue that the journalists worldwide cannot understand it deeply. This is because what they can do is only read the research abstract on the website of PubMed which only presents the articles that are insufficient for analyzing research quality. They have to read the original in the journal to understand and be able to analyze the research quality. I would give one example of research which has been published in Lipid. It questioned that coconut oil consumption which was the saturated vegetable oil would better help to lose weight and belly than unsaturated vegetable oil or not. The sample group was 40 female who had the waist over 88 cm. divided into two groups. The first group ate coconut oil 30 cc. every day while the second group ate soy bean oil 30 cc. every day for 12 weeks. Both groups were forced to decrease calorie in food for 200 calories per day and do exercise by taking a walk for 50 minutes every morning. After 12 weeks, the females from both group lost their weight at the same rate (2 pounds per person). Moreover, it was found that the group that ate coconut oil had slightly higher HDL and LDL than those who ate soy bean oil. Both groups had less waist size which the group that ate coconut oil was slightly smaller with the very small significance in the small sample group like this p=0.05. However, the researcher concluded that coconut oil was better for losing weight and reducing belly while increasing HDL than soy bean oil. This conclusion was enough for the merchant who sells coconut to apply this conclusion to sell coconut products.

My focus is that though this research is the level 1 research, it is the low quality research because it is designed to have the confounding factors to force the results as designed by the researcher. That is to say while the objective wants to measure that coconut oil can help to reduce belly size or not, the research is designed to force daily calorie for 200 calories and doing exercise for 50 minutes. Only the two factors can help losing weight without eating oil. Thus, it cannot be concluded that coconut helps to lose the weight.

Regarding the comparison of research result and common sense as well as the overall epidemiology, it absolutely contrasts. This because in terms of epidemiology, people who eat more fat will be fatter while the common sense realizes that as well. There is none fat that helps to lose weight if it is consumed everyday. Thus, the second result is required. However, there is none result such this after several years so it is the unreliable research.