Optimal Cutoff Value for Lung Injury Prediction Score and Potential Confounders for Identifying the Risk of Developing Acute Respiratory Distress Syndrome
However, several factors potentially affecting the results in this study should be discussed. First, the cutoff value of 4 for LIPS seems to be inadequate for satisfactory prediction of the development of ARDS in non-ED patients. We agree with the attempt of the authors in validating LIPS greater than or equal to 4, a validated cutoff value in ED (2–4), for predicting ARDS in non-ED. However, the specificity (30.9%) and positive predictive values (17.3%) of LIPS for ARDS were too low compared with the sensitivity (90.3%), suggesting that the cutoff value of 4 for LIPS was not suitable in this cohort. Provision of the receiver operating characteristic curve would be helpful for identifying a better cutoff value of LIPS for predicting ARDS in non-ED patients.
Second, the mortality rate of the patients in this study seems to be higher (48%) than that in previous studies (5). According to the Berlin definition for ARDS, the mortality rates of mild, moderate, and severe ARDS were 27%, 32%, and 45%, respectively (5). As the authors mentioned, the present study may have included sicker patient population compared with previous studies. Additional information of the severity of ARDS in the present cohort and therapeutic regimens applied would be helpful for validating the utility of LIPS. In addition, a subanalysis by stratifying the severity of ARDS for predicting mortality would be informative.
Third, the number of patients who died before developing ARDS is unclear, which could have potentially affected the analysis. A total of 101 patients (11%) died within 7 days of critical care contact in the present study. Some of these patients could have developed ARDS, had they survived longer.
In conclusion, we appreciate the additional data provided by the authors for facilitating the precise evaluation of the utility of LIPS in non-ED patients at high risk of developing ARDS.