We would first like to thank the authors of the letter “A commentary on ‘C-Reactive protein and SOFA scale: A simple score as early predictor of critical care requirement in patients with COVID-19 pneumonia in Spain’” for their interest in the results of our study.1
We are very pleased to see that the authors of subsequent studies2–4 have found that the variables included in our multivariate model, which only included patients admitted between March 2020 and June 2020, effectively predict the need for critical care in newly admitted patients, which was the aim of our study.
However, the authors of the letter comment that the C-reactive protein (CRP) values used to discriminate the need for critical care in Smilowitz et al.2 differ from those used in our analysis. There are many reasons for this discrepancy, including differences in design and objective. One potential limitation of the aforementioned study2 is that the authors, for the purpose of calculating sample size before recruitment, did not determine the expected difference in CRP values between the group that required critical care and the one that did not, making it impossible to be certain that the sample size was adequate for the purpose of the study. Smilowitz et al., furthermore, do not use measures to find the best cut-off point (why did they choose 10 mg/dL instead of 9.1 mg/dL, as used in our study?), and the variables used in their multivariate models differ from those used by us in a different population. In addition, the authors do not describe the clinical variables can modify CRP as a marker of the degree of inflammation (time of evolution, previous treatments) nor the distribution among the study groups, making their results unsuitable for comparison with ours. Despite this, Smilowitz et al., like us, found that CRP values are higher in patients that are likely to need critical care.
The same limitations apply to the evaluation of D-dimer data, which, although not the main outcome variable of the study,2 is analysed in subgroups. This methodology limits the potential validity of the results. Furthermore, the authors2 do not mention whether the patients identified as candidates for critical care on admission to hospital presented higher D-dimer levels than those not requiring this level of care, or whether or not this presumed difference is due to chance, as was the case in our sample.1
Regarding the use of the SOFA scale as a predictor of admission to critical care units, the study mentioned by the authors3 does not evaluate this aspect in particular, although it does indicate that a SOFA score of 2 or more (similar to that found in our study1) is as useful as the NEWS scale for predicting respiratory failure. This suggests that our model may also be useful in these patients.
The other review and meta-analysis mentioned4 do not include an analysis of variables or clinical scales, which probably explains, at least in part, why their results differ from ours.
Nevertheless, despite all these limitations, we observed that the studies cited show that the variables included in the scale used to predict the need for admission to critical care units in patients presenting in the hospital emergency room with pneumonia described in our study are fit for purpose.
FundingThe authors declare that they have not received funding for this study.
Please cite this article as: Vaquero Roncero LM, Sánchez Barrado E, Sánchez Hernández MV. Réplica a la carta al director: «Un comentario sobre proteína C reactiva y escala SOFA: una simple escala como factor predictivo temprano de la necesidad de cuidados críticos en los pacientes con neumonía causada por COVID-19 en España». Rev Esp Anestesiol Reanim. 2022;69:442–443.
