I am writing a critical appraisal regarding a radiological study using an anthropomorphic imaging phantom. The phantom is constructed of materials that mimic human tissue but negate the need to account for factors like age, sex, health status etc and also avoid the need to consider ethical issues surrounding the use of ionizing radiation on humans for research purposes. In the study, 9 images were taken - 3 from each of 3 different body areas, and for each body area exposure factors set according to 3 different 'rules' - namely system pre-sets, 10kVp rule, and the 15% rule. Doses and exposure indices are recorded for each image and then later compared to find out which rule offers the greatest reduction in dose whilst maintaining perceptible image quality.

This is the study in question:

Coffey H, Chanopensiri V, Ly B, Nguyen D. Comparing 10 kVp and 15% Rules in Extremity Radiography. Radiol Technol. 2020;91(6):516-524.

My issue is that I don't know how to appraise the sampling methodology or study design, as there is no sample population of people involved.

Several hours of googling and research article searches has not helped much either. My best understanding so far is that this study is a quantitative quasi-experimental study:

Quantitative because it involves the generation of quantifiable and statistically analyzable data.

Experimental because a causal relationship is being investigated (i.e. the 'rules' vs dose/image quality).

Quasi because it does not involve a control group (seemingly to me) and is not feasible due to the above-mentioned ethical issues.

How do I describe the sampling methodology and study design when there are no humans involved - only a single artificial surrogate phantom?

Thank you in advance!

  • You have cross posted this question on several sites. This is considered rude here, see: meta.stackexchange.com/questions/64068/… – Bryan Krause May 9 at 12:57
  • Apologies - I have now deleted the questions off the other sites. – Rossco May 9 at 14:47
  • Maybe I'm misunderstanding something but what sampling methodology are you talking about? Do you have multiple phantoms to choose from? – Carey Gregory May 9 at 15:03
  • Hi Carey. Regarding sampling methodology, have they used convenience sampling, purposive sampling, snowball sampling etc? Phantom-wise, there is just a single phantom. 3 body parts imaged - hand, knee and shoulder. For each body part, 3 images were taken - 1 with system pre-set exposure values, 1 using exposure factors derived from the '10kVp rule' and another using the exposure factors derived from the '15% rule'. 9 images thus in total. – Rossco May 9 at 16:37
  • In my appraisal, it might be the case that I just have to state that there is no sampling methodology that has been employed, but I don't know if that is correct, hence the question. The fact that there is no sample population (e.g. such as in a randomised controlled trial involving sample populations of people) is confusing me because the phantom, to me, should be the equivalent of the sample population, but there is only a single phantom that is used, which carries implications for external validity and representativeness. – Rossco May 9 at 16:47

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.