Download Full Text (602 KB)
Our Cardiac High Acuity Monitoring program (CHAMP) utilizes daily videos as an adjunct to monitor interstage single ventricle infants (SVI) at home. The current study was designed to develop and validate an objective clinical scoring system for video analysis, in order to identify SVI at risk for clinical deterioration and predict unplanned hospital admissions (UHA).
Six candidate items (respiratory rate and effort, color, behavior, skin and general appearance) were selected using local expert consensus to develop a pragmatic score for standardized video analysis. Observations were scored from 0 to 2 or 4, with higher numbers for increasing levels of concern. Thus, respiratory rate was graded 0-4 with increasing scores for increasing tachypnea or bradypnea. Respiratory effort was graded based on retractions and perceived use of accessory muscles. Color was graded as pink, pale and gray. Behavior was graded as appropriate, somnolent or irritable. Skin was dichotomous for the presence or absence of rash and/or infection. Appearance was graded as expected, mild concern or significant concern.
All SVI monitored at home by CHAMP between March 2014-March 2018 at our center were included. The primary outcome variable was UHA. Videos obtained within 48 hours prior to UHA (case videos) were compared to videos obtained at baseline (control videos). A subset of 30 videos was reviewed by a second blinded rater for inter-rater reproducibility. Independent t-, paired t- and Mcnemar’s tests were used for comparisons. Inter-class coefficients (ICC) were used to quantify reproducibility. Receiver operating characteristic (ROC) curve was used to establish a cut off score for predicting UHA.
Thirty-nine subjects with 64 UHA were included. We compared 64 case videos to 64 paired controls. Video scoring was feasible for 91.6% of all observations with 64 non scorable- items, due to inadequate lighting, distance or clothing. Of those, 16 were for skin. ICC for inter-reader reproducibility of video score items ranged from 0.73-0.98. Individual score items were significantly higher in case videos than in controls, with the exception of skin scores. Hence, skin was removed from the total video score. The resulting 5-item video composite score provided values ranging from 0 to 15. Mean composite scores were significantly higher for cases (6.9 ± 2.1) than for controls (1.7 ± 1.3), P< 0.001. Area under the ROC curve was 0.97. Score of 4.5 provided sensitivity of 89% and specificity of 97% to predict UHA.
We developed a reproducible video telemetry score that can serve as a tool to predict UHA in SVI. Future directions involve prospective, multicenter validation of this tool.
Aly, Doaa; Erickson, Lori; Hancock, Hayley S.; Apperson, Johnathan; Reid, Kimberly J.; Marshall, Jennifer A.; Shirali, Girish S.; and Goudar, Suma, "Ability of Video Telemedicine to Predict Unplanned Hospital Readmission for Single Ventricle Infants" (2019). Research at Children's Mercy Month 2019. 3.