The implementation of an integrated observation chart with Newborn Early Warning Signs (NEWS) to facilitate observation of infants at risk of clinical deterioration Chan Man Yi, NC (Neonatal Care) Dept. of Paed. & A.M., PMH 16 May 2017
Background Different observation charts using in Special Care Baby Unit (SCBU) BP, Temperature & Pulse Chart Newborn Observation Chart Neonatal Skin Assessment Record Pain Assessment Record
Background No reference range of significant clinical signs in traditional observation chart Affect early recognition of infants at risk of clinical deterioration Hinder effective team communication and take appropriate interventions promptly Early identify signs of deterioration Prompt intervention and step up further care Reduce morbidity (Duncan, 2007)
Objectives Combine into one chart: Integrated Observation Chart with Newborn Early Warning Signs (NEWS) To develop an integrated observation chart with NEWS to facilitate nurses for identifying infants at risk of clinical deterioration for taking appropriate interventions in SCBU To evaluate the discrimination ability of NEWS in screening the infant for Neonatal Intensive Care Unit (NICU) admission 25 July 2017
Methodology Study Design Retrospective chart review diagnostic study Settings Special Care Baby Unit (SCBU) in PMH Period 1 Feb 2015-31 Jan 2016 Subjects All eligible cases and 1:2 matched controls in the 1 year study period 25 July 2017
Subjects Inclusion Criteria Case Aged 28 days or 44 weeks post-conception Transferred from SCBU to NICU for clinical deterioration and step-up care Matched control Admitted to SCBU within ± 3 days of the case s SCBU transferal date Length of stay in SCBU length of stay of the case Without NICU admission Exclusion Criteria Had previous NICU admission Electively transferred to NICU for procedure, monitoring or management without clinical signs of deterioration
Data Collection Subjects were identified from the admission records of NICU and SCBU, as well as Electronic Patient Record (epr). Demographics and clinical information were collected from medical records. Example for data collection 25 July 2017
Data Analysis & Development of NEWS Step 1: Draft of NEWS Developed by a clinical expert panel including neonatologist and neonatal nurses Included five physiological items, temperature, cardiovascular status, respiratory status, neurological status and severity of abdominal distension Three colour zone (white, yellow and red) to indicate the urgency in intervention Red zone to indicate the need of immediate intervention for clinical deterioration 25 July 2017
Data Analysis & Development of NEWS Step 2: Data analysis for NEWS modification Demographics and clinical information of the subjects were reported as mean ± SD, range or count (%). Chi-square test, Fisher's exact test, Mann-Whitney U test, independent samples t-test were used for comparing variables among groups. The subjects were divided into training set (~70%) for developing modified versions of NEWS and testing set (~30%) for testing the modified charts. Sensitivity, specificity, accuracy, positive and negative predictive values of red zone in NEWS for screening infants with NICU 25 July 2017 admission were calculated.
Data Analysis & Development of NEWS Step 3: Finalizing NEWS The modified versions of NEWS were discussed by the expert panel to develop the finalized version of NEWS. Sensitivity, specificity, accuracy, positive and negative predictive values of red zone in NEWS for screening infants with NICU admission were calculated using all recruited subjects. Statistical analysis was performed using the Statistical Package for the Social Sciences (SPSS version 22.0; SPSS, Chicago, IL, USA) and Open Source Epidemiologic Statistics for Public Health (OpenEpi version 3.01). P value <0.05 was considered as statistical significant. 25 July 2017
Sample Size Estimation First study to evaluate NEWS No prior information about its screening ability Assume: Sensitivity = 90%; Specificity = 90% Precision = 10% Ratio of case to matched control = 1:2 Prevalence = 33% (=1/3) Confidence level = 95% At least 104 samples would be needed 25 July 2017
Results Recruitment 20 cases not eligible - 1 with age > 28 days or 44 weeks post-conception - 19 electively transferred for 56 cases transferred to NICU procedure, monitoring or + management 36 cases recruited 72 matched controls 108 patients recruited Recruitment flow chart Training set (n=75) 25 cases 50 matched controls Testing set (n=33) 11 cases 22 matched controls
Results Baseline Characteristics
Results Baseline Characteristics
Results Baseline Characteristics
Results Baseline Characteristics
Results NEWS Modification Training Set Version 1 (Draft) Version 2 Version 3 Version 4 Accuracy (%;95%CI) 85.3 (75.6-91.6) 93.3 (85.3-97.1) 94.7 (87.1-97.9) 100 (75.1-100) Sensitivity (%;95%CI) 72.0 (52.4-85.7) 96.0 (80.5-99.3) 100 (86.7-100) 100 (86.7-100) Specificity (%;95%CI) 92.0 (81.2-96.8) 92.0 (81.2-96.8) 92.0 (81.2-96.8) 100 (92.9-100) PPV (%;95%CI) 81.8 (61.5-92.7) 85.7 (68.5-94.3) 86.2 (69.4-94.5) 100 (86.7-100) NPV (%;95%CI) 86.8 (75.2-93.5) 97.9 (88.9-99.6) 100 (92.3-100) 100 (92.9-100) Items / Color Zone Version 1 Version 2 Version 3 Version 4 Mild & Moderate Retraction Mild & Moderate Abdominal Distension Temperature ( C) 35-35.9 <35
Results NEWS Modification Training Set Version 1 (Draft) Version 2 Version 3 Version 4 Accuracy (%;95%CI) 85.3 (75.6-91.6) 93.3 (85.3-97.1) 94.7 (87.1-97.9) 100 (75.1-100) Sensitivity (%;95%CI) 72.0 (52.4-85.7) 96.0 (80.5-99.3) 100 (86.7-100) 100 (86.7-100) Specificity (%;95%CI) 92.0 (81.2-96.8) 92.0 (81.2-96.8) 92.0 (81.2-96.8) 100 (92.9-100) PPV (%;95%CI) 81.8 (61.5-92.7) 85.7 (68.5-94.3) 86.2 (69.4-94.5) 100 (86.7-100) NPV (%;95%CI) 86.8 (75.2-93.5) 97.9 (88.9-99.6) 100 (92.3-100) 100 (92.9-100) Testing Set Version 1 (Draft) Version 2 Version 3 Version 4 Accuracy (%) 75.8 84.8 90.9 100 Sensitivity (%) 54.5 81.8 100 100 Specificity (%) 86.4 86.4 86.4 100 PPV (%) 66.7 75.0 78.6 100 NPV (%) 79.2 90.5 100 100
Results Finalized NEWS All Subjects Version 1 (Draft) Finalized Version of NEWS Accuracy (%;95%CI) 82.4 (74.2-88.4) 87.0(79.4-92.1) Sensitivity (%;95%CI) 66.7 (50.3-79.8) 80.6(65.0-90.3) Specificity (%;95%CI) 90.3 (81.3-95.2) 90.3(81.3-95.2) PPV (%;95%CI) 77.4 (60.2-88.6) 80.6(65.0-90.3) NPV (%;95%CI) 84.4 (74.7-90.9) 90.3(81.3-95.2) Items / Color Zone Draft NEWS Finalized version of NEWS Retraction Abdominal Distension Mild Moderate Mild Moderate Temperature ( C) 35-35.9 <35
Discussion The NEWS chart with action pathway and reference range of significant clinical signs Facilitate early identification of infants at risk of clinical deterioration and take appropriate interventions promptly Enhance effective team communication especially inexperienced nurses on the recognition of the unwell infant in SCBU
Discussion The NEWS chart is only a set of observations, but the subsequent full clinical evaluation of the infant is needed. The other elements of assessment is also significant. For respiratory assessment, chest retraction is one of clinical signs for respiratory distress. The observation of respiratory pattern, SpO2 and O2 requirement are also very important for infants. The chest retraction scale is simply defined as mild or severe in new version instead of mild and moderate or marked. 25 July 2017
Discussion For abdominal distension assessment, it is associated with feeding tolerance, nature and frequency of stool in daily intake and output chart, and soft or tense in abdominal palpation. The abdominal distension scale is simply defined as mild or severe in new version instead of mild and moderate or marked. 25 July 2017
Discussion Hypothermia is common in newborn infants despite measures to prevent hypothermia <36 C immediate after birth, resuscitation, and warm transportation. Neonatal hypothermia is associated with increased morbidity and mortality especially among sick newborn infants who need medical attention in SCBU. Nurses have to step up appropriate supportive intervention, and notify concerned parties. The temperature scale is kept status quo in new version. 25 July 2017
Clinical Practice A new version of the NEWS chart is developed and will be implemented in 2017 when available. The NEWS chart is only one component of our assessment for early identifying newborn infants at risk of clinical deterioration in SCBU. 25 July 2017
Conclusion NEWS can facilitate early identification of infant s clinical deterioration in SCBU and provide guidance for nurses to initiate appropriate intervention for improving patient outcome.