Dear Sir,

I have read the article by Gasque et al.¹ which, in my opinion, has methodological flaws that affect its reproducibility and internal consistency which merit discussion:

1. The statistical power of the sample was not calculated. The incidence of acute appendicitis does not provide a reason for a small sample size and deprives readers of a valuable element for the interpretation of the results.

2. The authors compared the proportions of positive diagnoses between men and women and at different age intervals but without statistical inference, so the reader does not know if the results can be generalized.

3. The lack of statistical significance (p = 0.11) in the mean duration of symptoms shows that a hypothesis test was used for this variable, but the test used, the level of significance established and the software used to perform the calculations are not mentioned.

4. The authors reported the mean overall duration of symptoms with its standard deviation, while the standard deviation of duration of symptoms in the groups with and without appendicitis, which was just one point lower, is not mentioned.

5. The screening values were reported without the corresponding confidence intervals. Due to the small sample size, especially in the group without acute appendicitis, the confidence interval of specificity is so wide (95% CI = 0.2 -0.78) that the authors calculated an uncertainty of 0.4 (40%).

6. The sample is poorly representative of the target population of the RIPASA score, since it includes more than 85% of positive diagnoses, a proportion that magnifies the incidence observed in patients visiting the emergency departments with “right iliac fossa pain” (Chong reported 52.60% in his prospective study)², producing a distortion of the predictive values of the test, which are essential to estimate its performance in the healthcare scenario.

7. Newman et al.³, among many other authors, ban the estimation of predictive values in non-consecutive samples.

8. The evaluation of patient records with a known final diagnosis by a single observer involves potential interpretation bias. To reduce the risk, a minimum of two independent blinded assessors are recommended with estimation of interobserver agreement⁴.

9. The conclusion is typical of a review of the literature rather than an analysis of primary data, as it is based on studies by other authors.

The assessment of the performance of a score for the diagnosis of acute appendicitis offers opportunities for research that, unfortunately, were not considered:

1. There are data from previous studies to calculate, with reasonable approximation, the number of patients to obtain appropriate estimates per interval.

2. The disease is common, which would allow to reach the required sample size in a short recruitment period.

3. There is no need for prolonged follow-up, which ensures few dropouts and facilitates prospective studies.