E-DRUG: Critical appraisal of articles
-------------------------------
My name is Praveen Kumar, i am pursuing my Pharm D, in Annamalai University, and i required information about Critical appraisal of recently published articles in the biomedical literature which deals with a drug.
regards
Praveen Kumar,
Pharm D student
Annamalai University
India
Praveen Kumar <praveenkumar_nvrt@yahoo.com>
E-DRUG: Critical appraisal of articles (2)
-----------------------------------
Praveen
I gather that you want to appraise mainly quantitative research type of articles. What I am not sure of is what type of research papers you want to appraise (e.g. RCTs only, or RCTs plus quasi-experimental designs), and whether you plan to do a meta-analysis or a comprehensive narrative-type synthesis....
This has implications; scoring articles (ranking them in terms of quality) is more relevant for meta-analysis (weighting can be done), although there is disagreement on relevance of giving index scores to articles, unless of course they are based on very similar populations, report similar interventions and comparisons, and have common outcomes.....a very rare occurrence.
Assuming from your question that your interest is drug trial type of studies (Randomized Clinical Trials), your appraisal will look to determine validity (study quality e.g. randomization processes, concealment of allocation, blinding etc), importance (precision, effect sizes) and generalizability (applicability of findings outside the drug trial setting).
Again depending on whether you want to attach scores to papers or not, there are checklists available for appraising literature. The Jadad checklist can be useful if you are interested in scoring (and ranking) studies, but you have to be very careful, as quality scores may not always reflect treatment effect (since each item on the checklist carries similar weight, and yet some affect validity of study findings more).
Alternatively, you could use Cochrane tools. I have used the EPOC critical appraisal checklist, and found it quite useful, especially for narrative-synthesis type of reviews. What am unsure about is how useful it would be for drug trial studies, and meta-analyses.
Remember, it is perfectly acceptable to discuss the quality of articles without attaching definite scores; it's certainly more scientific than scoring studies that differ in populations, interventions, comparisons or outcomes (i.e. comparing apples and oranges)
Good luck
Frank Wafula
Consortium for Research on Equitable Health Systems (CREHS)
KEMRI-University of Oxford-Wellcome Trust Collaborative Programme
P.O. Box 43640 – 00100
Nairobi, Kenya
Tel: +254 20 2715160 or 2720163 or 2719936
Fax: +254 20 2711673
Cell: +254 722 679467
"Frank Wafula" <fwafula@nairobi.kemri-wellcome.org>
E-DRUG: Critical appraisal of articles (6)
------------------------------------------
[discussion closed - thanks for all contributions! WB]
Dear Praveen,
I have reviewed literature and received instruction of highly experienced on Critical Thinking. Many of them have been involved in developing New Molecules for Drug Development.
I have also read books on how to read research reports, etc. But i don't find many helpful to students.
Out of these experiences i decided to come up with a way of how to do things.
In addition to this way, I also recommended 3 chapters in the following Book:
"Calculations for Pharmaceutical Practice by Arthur Winfield and Ivan Edafiogho.
ISBN: 0-443-10019-5. Publisher: Elsevier Churchill Livingstone."
The chapters of interest are on pages 223 to 295.
Also, I have written about Power, and Statistical significance which, are lacking in that Book.
Power and Statistical Significance are actually the cornerstone of statistics and research and review of these reports. Also remember to always examine the assumptions of the statistical tests used in the reports.
Cross-examine these with the parameters measured in the report.
Then, consider the technical, established facts about the variables and how they influence the parameter the researchers measured.
When you do these things in a group in your class, you will enjoy.
And you will discover over 50% of research reports make wrong conclusions.
Instructions for Critical Study of a Research Report: Follow Instructions Like a Sheep
Read the results section.
Read the methods section.
Re-read the methods section.
Read the results section, again.
Read the aims of the study.
Read the conclusions.
Read the title.
Re-read the title, the aim, the methods section, the results section and the conclusions.
Define unfamiliar terms.
Say the theme, what the problem is about.
Discuss prior knowledge about issues raised in the problem.
Do a concept map: A concept map is where you begin from, and the process you go through to reach a goal.
Make your development goals.
The basics of statistics needed in our Universities in order that unnecessary expenditure is avoided
In statistics, there are two critical words: Power and significance. These form the basis of making conclusions. These conclusions are the ones which result in social and economic transformation of societies.
Power and significance empower people, and make them get significance in their lives: feeding themselves, their children; accessing safe drinking water; and getting clothed.
The power of 100% is the best. Significance of 100% is the best. 100% means there ‘s no room for doubt that what you discovered in a study will never change for all time.
Power: The ability of a study design, and statistical test to examine the numbers for precision, accuracy, reproducibility, and every hew in between. This is so as to avoid mix-up of distributions of numbers.
These numbers pertain to things in the groups being considered. The groupings are made to define things the population you are considering. The level of scatter of these things is measured. The measurements are given values.
They are cost, with adjustment for any possible changes, like time. If you measured or valued these things and converted them to monetary terms, then the possible changes would be alterations in demand, inflation, purchasing power of your people, etc.
Different groups have different numbers. Different means. Different averages. Then you go ahead to compute another figure. This is called standard variation of the various values in the distribution from the average.
This number, also called standard deviation is important because it defines the degree of scatter. This scatter is helpful to tell you how clear the numbers are with association to the average.
This is to avoid haziness, blurring so as to see things clearly; the way they truly are, everywhere. For example if we talk of income of Ugandans as a thing to measure, and we made a distribution.
We measure income per household, per year, and we plot this on the x-axis.
And we plotted number of people earning corresponding incomes, incomes corresponding to the points on the x-axis. And we put the number of people on the y-axis; we get a distribution of numbers showing what is on the ground clearly.
This would avoid false assurances as to how the population in general is fairing in terms of prosperity for all.
That is if you have extensively, highly skewed programmes causing unacceptable inequality, then you will see clearly concerning this matter. The picture, pictogram of the normal distribution, showing what ‘s normal on the ground will display reality for us.
If there ‘s excessive income variability, the spread of the normal distribution will be unacceptably wide. This means the rich are rich, the poor are poor, when you don ‘t change you ‘re approach, it will only get worse.
This is because; you will be fueling the distribution, flattening it further, and further. When you change, from the poor‘s perspective, and address corruption on the other hand, you will stage one side of the distribution.
This side is the extreme, the excesses. And it will stop bulging. In the meantime, the left side will get up, and we will achieve the precision, we need. Everybody will be happy. There will be cohesion. People will stick together. This is social transformation.
Significance at: 100%. This is the ability of a statistical test to say what you talking of power is indeed true, sure, and exact as you set. And that the differences you observed of the distribution of numbers in the different study groups isn’t by chance.
It‘s real. It‘s repeatable everywhere for all peoples, races, tribes, languages, nations, countries under Heaven in which the God of everything is indeed in charge.
Kibumba George
kibumba@yahoo.com