Unified severity and organ dysfunction scoring system in pediatric intensive care unit: A pressing priority

Abstract

Several scoring systems have been used to objectively assess the severity of illness and to predict the short-term mortality among critically ill children. The Pediatric Risk of Mortality-III (PRISM-III) and Pediatric Index of Mortality-3 (PIM-3) are the commonly used severity scores; and Pediatric Logistic Organ Dysfunction-2 (PELOD-2), Sequential Organ Failure Assessment (SOFA), and pediatric SOFA (pSOFA) are organ dysfunction scoring systems.[1–6] In addition, these scoring systems are also useful in assessing the performance of different units, monitoring the quality of pediatric intensive care and benchmarking, and further improvement in performance. SOFA was introduced in 1996 by Vincent et al.[6] to assess the severity of organ dysfunction in critically ill adult patients with sepsis. SOFA score objectively evaluates the organ dysfunction using six organ system variables (clinical and laboratory) that measure the disease severity during the stay in the intensive care unit. Recently, pSOFA score was devised and validated by adapting the original SOFA score with two additional changes: age-adjusted cutoffs for the cardiovascular and renal systems and inclusion of noninvasive surrogates of lung injury (SpO2/FiO2 ratio in addition to PaO2/FiO2 ratio) in the respiratory criteria.[5,7,8] Since SOFA score requires multiple clinical and laboratory data, its use may be potentially challenging, especially in resource-limited settings. Keeping in mind the limitations of SOFA, quick SOFA (qSOFA) score was developed to help clinicians to identify patients at risk of sepsis, by assessing predictive validity using mortality as an outcome more likely occur in patients with sepsis. qSOFA requires only three clinical examination components (i.e., systolic blood pressure, respiratory rate, and Glasgow Coma Scale).[9] qSOFA has been used in low-resource settings.[10] More recently, it has been demonstrated that addition of point-of-care venous lactate to qSOFA was superior to qSOFA alone to predict sepsis-related mortality among adults.[11] Similarly, Kumbar and Chandrashekhara[12] evaluated pSOFA with lactate (pSOFA-L) to predict the mortality among critically ill children (n = 75) and demonstrated that pSOFA-L score had good ability to predict mortality (area under the curve [AUC] = 0.92, cutoff value 10.5, P < 0.001). The mortality rate in children with pSOFA-L <9, 9–11, and >11 was 26.1%, 38.9%, and 50%, respectively. The PRISM-III, PIM-3, and PELOD-2 scores are commonly used to predict mortality and these were validated and calibrated in large populations. However, pSOFA and pSOFA-L are newer ones and these need to be validated in large studies. There is some evidence that SOFA or pSOFA has better accuracy to predict mortality in critically ill children than the PRISM-III or PELOD-2 score or other organ dysfunction scores.[5,13–15] The performance of available severity and organ dysfunction scoring systems to predict mortality among critically ill children is variable and it depends on hospital settings, patients’ population selected, and resources.[16,17] However, the majority of these scoring systems need a lot of clinical and laboratory data. There is felt need for simple, unified, and well-validated scoring systems to predict outcome among critically ill children. We read with interest the recently published article titled, “A study to assess the role of pSOFA-L score in predicting the clinical outcome of critically ill children admitted to the pediatric intensive care unit (PICU) at a tertiary care centre.[18]” We would like to make a few important comments about this study. In this cohort of critically ill children admitted to PICU, the mortality was very high (44%). In contrast, the usual mortality rate in most of the PICUs including that in developing countries ranges from 2% to 20%.[5,14,19] The information about admission diagnosis; PRISM-III/PIM-3 score; percentage of children who received mechanical ventilation, vasoactive drugs, and other organ-supportive therapies; and rate of health-care-associated infections could have provided some clues about the possible reasons for high mortality in the index study. The authors have not mentioned the predominant diagnosis in the index study or patient population their PICU caters to. In the index study, the AUC for pSOFA and pSOFA-L to predict mortality is 0.881 and 0.882, respectively, which make pSOFA and pSOFA-L having similar ability to predict mortality. The possible reasons for no difference in pSOFA and pSOFA-L in predicting mortality among critically ill children in the index study are discussed here. Abnormal lactate (>2 mmol/L) adds only 1 point, which does not make much difference in the final pSOFA-L score as compared to other parameters where score for individual parameters ranges from 0 to 4. This is also highlighted by the fact that the best cutoff to predict mortality for both pSOFA and pSOFA-L is 10. Furthermore, it is not clear from the index study whether AUC for mortality was calculated using 24 h or 72 h values of pSOFA or pSOFA-L. In Tables 3, 4, and 6 in the index study, comparison of pSOFA in addition to pSOFA-L with mortality could have given some clue about which one of these parameters has better ability to predict mortality. Thus, as per the results of the index study, pSOFA-L is no better than pSOFA in predicting mortality among critically ill children. Concept of combining pSOFA and lactate is feasible and has strong physiological rationale. However, there are limited studies to support this concept including the index study, which too has a small sample size. This warrants further exploration of this concept to determine outcome in critically ill children. Multicentric studies with large sample size are needed to verify whether pSOFA-L is better than pSOFA in predicting mortality among critically ill children. Also, it has to be investigated whether categorization of increasing lactate values into different scores (0, 1, 2, 3, and 4 scores) may give more weightage to pSOFA-L score and better ability to predict the mortality than pSOFA score. The search for an ideal severity and organ dysfunction scoring system for critically ill children is still elusive. Therefore, there is a need to have a unified severity and organ dysfunction scoring system that is derived and validated in large, heterogeneous international databases of critically ill children.[17]