Authors (including presenting author) :
Ngai SL(1), Chui KI(2), Cheng YK(1), Choi SL(1), Ng E(2), Wong PP(1), Yan MK(2), Yim KL(1)
Affiliation :
(1) Accident and Emergency Department, Queen Elizabeth Hospital
(2) Accident and Emergency Department, Kwong Wah Hospital
Introduction :
The increasing service demand of Accident and Emergency Departments (AED) may impose adverse patient outcomes. Streaming aims to enhance patient safety for patients at higher risk of deterioration while queuing for triage. By applying streaming, patients at triage are allocated to cubicles where physicians and nurses are readily available to cater for their needs. However, there has been no studies on selection criteria on streaming.
Objectives :
(1) To explore the streaming criteria for Category 3 patients in AED
(2) To evaluate the screening tool for streaming used in AED
Methodology :
A retrospective chart review was conducted to explore the common symptoms and characteristics of Category 3 patients. 480 medical records were retrieved in Queen Elizabeth Hospital and Kwong Wah Hospital. Univariate analysis and backward binary logistic regression were used for data analysis. Different combinations of the analyzed factors were tested with ROC curve. Several tests were performed to evaluate the streaming criteria: (1) Content validity test; (2) Inter-rater reliability test; (3) Accuracy test; and (4) Usability test.
Result & Outcome :
Of the 60 factors identified in the retrospective chart review, only 18 variables were included after univariate analysis (each of p<0.25). Subsequent analysis using backward binary logistic regression revealed that only 8 factors were included in the streaming criteria, including mode of transport, age 70, old-aged home residents, fever, shortness of breath, using oxygen, lower limb edema, and numbness. Further testing these 8 factors with ROC curve showed that attendants who transported to AED by ambulance presenting with any one of the remaining 7 factors had the best prediction of patients at higher risk of deterioration.
The CVI of the 7 factors were all above 0.8. The percentage agreement in inter-rater reliability test was 90%. The percentage of accuracy was 84%. Staff commented that the streaming tool was easy to use. 70% of them agreed that the streaming tool targets Category 3 patients precisely.
The streaming tool serves as an assistive role in enhancing the efficiency of patient flow at triage and consultation. It can bring about patient safety, quality care, and fully utilize available resources and manpower.
