Hi,

if you can provide a [ideally pre-specified] reason to expect why and on what basis there might be a difference, then the analysis of heterogeneity is meaningful. If you look for explanations that arise after the analysis, it might appear to be a 'fishing expedition'. A couple of examples for VAP incidence;

Hurley JC. World-wide variation in incidence of Acinetobacter associated ventilator associated pneumonia: a meta-regression. BMC infectious diseases. 2016 Dec;16(1):1-2.

Hurley JC. Worldwide variation in Pseudomonas associated ventilator associated pneumonia. A meta-regression. Journal of Critical Care. 2019 Jun 1;51:88-93.

James Hurley

Dear Professor Hurley,

Thank you very much for your invaluable feedback.

I read your recommended articles. Despite the (in-) explainability of variation, the (in-) considerability for analysis of several confounders, and the (non-) significance of variations have been mentioned appropriately, I did not find anything regarding the pooled mean incidence in your articles. Therefore, my question still stands and I will be glad to hear from you about it.

Best,

The other side of your question is heterogeneity. Remember that heterogeneity can be driven by outlying study results. The measurement of heterogeneity is as much recognition of studies that are dissimilar on general as much as on statistical criteria. Like all statistical tests, heterogeneity measurement has are problems with power and interpretation.

James Hurley

Many thanks for your helpful guidance.

It’s difficult and not necessarily possible to pool results of significant heterogeneity. Its one of the limitations in pooling studies with high heterogeneity.

Pooled prevalence or incidence refers to the practice of combining data from multiple studies to estimate the overall prevalence or incidence of a particular outcome or condition. There are several benefits to using pooled prevalence or incidence in descriptive studies:

Osinachi Ekeagwu

Dear Dr. Ekeagwu,

I am sorry for my delayed reply.

Thank you for spending your worth time on feedback, which not only assisted me in reaching acceptable answers but also made me look at the issue from a different perspective. However, I do not agree with your points 2 and 5 so much.

In general, I think it is true that due to significant heterogeneity, the pooled prevalence or incidence may not be considered too correct and logical; however, sometimes the policymaker wants to make a series of comparisons between countries on a broad level or make decisions supposedly for the provision of medicine based on the results from pooled means. Therefore, pooled prevalence or incidence can maybe be useful in this regard.

Best,