AMR Critical Appraisal Example

Q1: What’s your suggestions to improve the scientific background and rationale for the investigation being reported in the introduction section?

A:

-          It is lacking data about AMR as a problem right now (prevalence, mortality rate, etc), they only show the future burden of the conditions. à They should present both actual burden and potential burden.

-          Not presenting clear correlation of knowledge and practice of antibiotic use to actual AMR prevalence. à They should give convincing evidence that knowledge and practice of antibiotic use is affecting the actual case of AMR.

 

Q2: Is there any flaw(s) in the table 3 for the interaction effect?

A:

-          The wide 95% CI of “Interaction between Knowledge of AMR and 35–49 Years Old” and “Interaction between Knowledge of AMR and 21–34 Years Old”.

-          Perhaps it is caused by very imbalance data regarding “Knowledge of AMR”, as we know that only 3% of respondents were considered having good knowledge.

 

Q3. Is there any flaw(s) in the table 4 (also in the table5) for the interaction effect?

A:

-          If only for the interaction effect it seems no flaws, but when we see for each age group, we can see that some age groups showed wide 95% CI. And it was getting wider as the age group being younger.

-          We should aware that younger respondents tend to having inappropriate use of antibiotics. The author did not mention specifically about this issue, they just focusing on the effect size.

 

Q4 If the authors got results from 18-49-year-old female subjects indicated that good accessibility to primary care could reduce the risk of self-medication (one of the inappropriate use of antibiotics), please illustrate your potential explanation(s) with supportive references.

A:

-          In that scenario, 18-49-year-old female having less access (compared to male and older age groups) is contra-intuitive, as female is assumed to be more time spent at home and older people tend to have more limitation on healthcare access. And regarding the paper, the access to primary care did not show as a problem, since most Singaporean get their antibiotic from primary care.

-          The potential explanation is more about women’s mobility, regardless of age group.

-          In the other setting, the explanation is also including the funding, but only when it related to hospital, not primary care since it involving intravenous antibiotic: “Patients who were female, aged 18 years or more (compared to those aged ≤5 years), received antibiotics intravenously and received financial assistance from an insurance scheme were more likely to be prescribed antibiotics inappropriately.”[1] It will be interesting to see the effect of health insurance as explanatory factor of better access for 18-49-year-old female will decrease the self-medication.

-          “This trend is observable in the data for HICs where lower education, less employment, and lower income leads to higher antibiotic use. However, the opposite is true for data from LMICs.”[2] We should also considering broader socioeconomic characteristic of the respondents to explain the finding.

 

Q5 What are the policy implications?

A:

-          Education on antibiotic should focused on younger population.

-          Since they mostly got the antibiotic from GP clinic and polyclinic (83%), so the primary clinics should be the place of promoting awareness about good antibiotic use and AMR.

 

Q6 Please provide at least two of your questions and your suggestions and comments for each question.

A:

-          Why the author set the cut-off point for “Knowledge of AMR” too high (8 out of 8 question should be right answered to be considered as “good”)? If they (at least) try to use different cut-off point, perhaps it will give a different conclusion. It is perhaps the biggest flaw in this paper, since only 3% of respondents were considered having good knowledge of AMR à the proportion between groups were highly imbalance. The author did not mention that they have tried something to solve this problem.

-          Why they were not using the knowledge score as continuous variables? Since there are no validated cut-off points for Singaporean, it is better to also treat the knowledge variables as continuous variables. The different approach with using the raw knowledge score without categorization can show if any association was existing in more sensitive manner.

 

 

References:

1.     Chang, Y., Chusri, S., Sangthong, R., McNeil, E., Hu, J., Du, W., Li, D., Fan, X., Zhou, H., Chongsuvivatwong, V. and Tang, L., 2019. Clinical pattern of antibiotic overuse and misuse in primary healthcare hospitals in the southwest of China. PLoS One, 14(6), p.e0214779.

2.     Schmiege, D., Evers, M., Kistemann, T. and Falkenberg, T., 2020. What drives antibiotic use in the community? A systematic review of determinants in the human outpatient sector. International journal of hygiene and environmental health, 226, p.113497.