An advanced pediatric early warning system: a reliable sentinel, not annoying extra work

Article information

Acute Crit Care. 2022;37(4):667-668
Publication date (electronic) : 2022 November 29
doi :
Department of Pediatrics, Haeundae Paik Hospital, Inje University College of Medicine, Busan, Korea
Corresponding author: Young Joo Han Department of Pediatrics, Haeundae Paik Hospital, Inje University College of Medicine, 75 Bokji-ro, Busanjin-gu, Busan 47392, Korea Tel: +82-51-797-0527 Fax: +82-51-797-0527 E-mail:
Received 2022 November 24; Accepted 2022 November 25.

Traditionally, changes in a symptom or sign in a hospitalized patient are reported from the patient or caregiver to the nurse and then to the physician. In other words, when a patient or a caregiver informs a nurse of a problem, the nurse checks the patient’s status, then informs the doctor, and the doctor then checks the patient and takes necessary action. However, since patients or caregivers without medical expertise cannot give sufficient reports, periodic visits by nurses to check symptoms and objective signs and, perhaps less frequently, periodic visits by physicians are routinely performed. For patients at a high risk of deterioration, the frequency of these visits and checks should be increased, which is best achieved through monitoring in the intensive care unit. For patients outside the intensive care unit, the frequency of checking symptoms or signs and the interpretation of those data play a very important role in enabling appropriate and timely interventions [1].

Situations in which rapid interventions for possible in-hospital cardiac arrests through a rapid warning system can be helpful mostly involve conditions related to disease progression. In other words, these situations include the functional deterioration or a major organ or circulatory failure as a disease progresses, rather than being an accident that is difficult to predict, such as a sudden obstruction of the upper respiratory tract, or a critical medical mistake. Because children do not describe their symptoms well and have a low compensatory ability, they may require more frequent assessments and application of a lower threshold than adults during the course of disease progression [2].

Shin et al. [3] reported the excellent performance of a deep-learning-based pediatric early-warning system (pDEWS) in a retrospective multicenter cohort study. The findings of this study suggest that the pDEWS can be introduced to other institutions without substantial difficulty because it only uses basic vital signs (respiratory rate, heart rate, systolic blood pressure, diastolic blood pressure, and body temperature) and showed similarly good performance regardless of the institution.

For an early warning system to show real effects, it must be connected to proper actions by the rapid response team [4]. One of the biggest hurdles to introducing early warning systems to many healthcare institutions is the lack of personnel. Nevertheless, medical staff should remember that proper and timely responses by the rapid response team through an early warning system can not only improve patient outcomes and prevent medical-legal disputes, but also significantly reduce the medical burden itself. Automated systems are useful for objective detection and display sufficient sensitivity, but the risk of false alarms is relatively high. The repetition of false alarms causes considerable exhaustion among medical personnel and can easily lead to intentional ignoring of alarms. After all, if an alarm is ignored under an automated system, the patient’s risk is actually increased compared to not applying the system. Therefore, the frequency of false alarms that trigger unnecessary reactions from the rapid response team must be kept low [5].

In order to improve the quality of critical care, early detection and response must now be viewed as essential, and these systems also increase access to critical care through an effective distribution of critical care opportunities. Ideally, an early warning system and rapid response team should be introduced at all medical institutions that may treat critically ill pediatric patients. The participation of more medical institutions should be preceded by the accumulation of experience centered on major tertiary medical institutions and advances in minimizing false alarms, and it is expected that Shin's research [3] will be a useful reference in that regard.











1. Vincent JL, Moreno R. Clinical review: scoring systems in the critically ill. Crit Care 2010;14:207.
2. Bonafide CP, Roberts KE, Priestley MA, Tibbetts KM, Huang E, Nadkarni VM, et al. Development of a pragmatic measure for evaluating and optimizing rapid response systems. Pediatrics 2012;129:e874–81.
3. Shin Y, Cho K, Lee Y, Choi YH, Jung JH, Kim SY, et al. Multicenter validation of a deep-learning-based pediatric early-warning system for prediction of deterioration events. Acute Crit Care 2022;37:654–66.
4. Sharek PJ, Parast LM, Leong K, Coombs J, Earnest K, Sullivan J, et al. Effect of a rapid response team on hospital-wide mortality and code rates outside the ICU in a Children's Hospital. JAMA 2007;298:2267–74.
5. Olsen SL, Soreide E, Hillman K, Hansen BS. Succeeding with rapid response systems--a never-ending process: a systematic review of how health-care professionals perceive facilitators and barriers within the limbs of the RRS. Resuscitation 2019;144:75–90.

Article information Continued