On February 3, 2020, CodeBlue reported on antibiotic shortages in Ministry of Health (MOH) hospitals over the past year in an article titled: “Survey Claims Public Hospitals Often Lack Antibiotics, But MOH Says Controlling Usage”.
As an academic, I feel compelled to pen this letter because of my concern over the effect that the allegations has on the Ministry of Health and public.
Apart from the fact that only 56 respondents took part in the study, at the time of writing, the study methodology has not been fully shared.
Health care studies are usually oriented by research problems or questions, which should be clearly defined in the study project from the beginning.
Sample size calculation is an essential item to be included in a study to reduce the probability of error, respect ethical standards, define the logistics of the study and, last but not least, improve its reliability when analyses are performed.
Sample size calculation is a crucial step in designing studies. Insufficient or a small sample size may not be able to demonstrate the desired difference or estimate the frequency of the event of interest with acceptable precision.
A common cause of sampling bias lies in the design of the study or in the data collection procedure, either of which may favour or disfavour data collection from certain classes or individuals or in certain conditions.
In this study, only doctors were included, and no other stakeholders. Sampling bias is also particularly prevalent when researchers use sampling strategies based on their judgment or convenience, in which the criterion used to select samples is somehow related to the variables of interest.
Furthermore, the significance level of the study was unclear. The significance level is the probability that the expected prevalence will be within the error margin being established. The larger the sample size, the higher the confidence level, the greater expected precision. This parameter is usually set at 95 per cent. The smaller the sample size, the tougher it is to draw any definitive conclusion.
In this study, a sample size of 56 respondents may be too small to draw any final conclusions, let alone to generalise the findings to a much bigger population.
It is worthwhile to note that procurement, supply, prescribing, and dispensing of all medicines in government facilities consist of many other stakeholders across the health care landscape.
This report showed a severe sampling bias, which means that the distribution of samples was inappropriate and did not represent the true distribution.
If we only poll the opinion of doctors, the views of other stakeholders across the health care landscape are likely to be underrepresented, hence compromising the ability in predicting the outcome.
From an academic standpoint, ill-designed studies that are poorly conducted and analyses that lack robustness will not provide the highest level of evidence, and most often lead to unclear and misleading conclusions.
June Choon is a lecturer and researcher in Health Economics at the School of Pharmacy, Monash University Malaysia.
- This is the personal opinion of the writer or publication and does not necessarily represent the views of CodeBlue.