P C B Khong1, M S W Yeo2, C C Goh3. 1. Senior Nurse Manager, Tan Tock Seng Hospital Pte Ltd, 11 Jalan Tan Tock Seng, Singapore 308433. 2. Consultant, Tan Tock Seng Hospital Pte Ltd, 11 Jalan Tan Tock Seng, Singapore 308433. 3. Nurse Clinician, Tan Tock Seng Hospital Pte Ltd, 11 Jalan Tan Tock Seng, Singapore 308433.
Abstract
OBJECTIVE: This pilot study tested the reliability, validity and sensitivity of an innovative prototype iPad app, WoundAide, developed to measure wound dimension (length, width and surface area) using image analysis. METHOD: A prospective evaluation was conducted comparing WoundAide (WA) with two gold standard instruments, Wound Zoom (WZ) and Visitrak (VT). The study was conducted in a tertiary hospital in Singapore. Patients were recruited after obtaining local ethics approval. RESULTS: We recruited six patients with 10 venous wounds. Images were captured by two researchers. Generally, 7/10 (70%) of the wounds had surface areas of less than 16cm². The intraclass correlation coefficient was greater than 0.95 for length, width and surface area for all measuring devices. The mean differences for length, width and surface area were not statistically significant. The average percentage of coefficient of variation for WA ranged from 3% to 33.3%, WZ 1.3% to 19.3%, and VT 2.1% to 43.4%. Length, width and surface area fell within the 95% limit of agreement. The capturing sensitivity was, on average, 75% for WA, 99.4% for VT, and 100% for WZ. CONCLUSION: WA has similar reliability and validity as WZ and VT. However, WA has the least sensitivity in image captures. Nonetheless, with enhancements made to its user-interface and system algorithm, this limitation can be addressed.
OBJECTIVE: This pilot study tested the reliability, validity and sensitivity of an innovative prototype iPad app, WoundAide, developed to measure wound dimension (length, width and surface area) using image analysis. METHOD: A prospective evaluation was conducted comparing WoundAide (WA) with two gold standard instruments, Wound Zoom (WZ) and Visitrak (VT). The study was conducted in a tertiary hospital in Singapore. Patients were recruited after obtaining local ethics approval. RESULTS: We recruited six patients with 10 venous wounds. Images were captured by two researchers. Generally, 7/10 (70%) of the wounds had surface areas of less than 16cm². The intraclass correlation coefficient was greater than 0.95 for length, width and surface area for all measuring devices. The mean differences for length, width and surface area were not statistically significant. The average percentage of coefficient of variation for WA ranged from 3% to 33.3%, WZ 1.3% to 19.3%, and VT 2.1% to 43.4%. Length, width and surface area fell within the 95% limit of agreement. The capturing sensitivity was, on average, 75% for WA, 99.4% for VT, and 100% for WZ. CONCLUSION: WA has similar reliability and validity as WZ and VT. However, WA has the least sensitivity in image captures. Nonetheless, with enhancements made to its user-interface and system algorithm, this limitation can be addressed.
Entities:
Keywords:
measuring device; pilot study; reliability; system app; wound dimension; wound measurement