An Acute Heart Failure Symptom Score For Children Hospitalized with Acute Decompensated Heart Failure

ABSTRACT BACKGROUND Currently there are no simple tools to evaluate acute heart failure (HF) symptom severity in children hospitalized with acute decompensated HF (ADHF). We sought to develop an inpatient HF score that could be used as clinical tool and for clinical trials. METHODS Pediatric HF clinicians at Stanford reviewed the limitations of existing HF scores, which include lack of calibration to the inpatient setting, omission of gastrointestinal symptoms, need for multiple age-based tools, and scores that reflect treatment intensity over patient symptoms. To address them, we developed an acute HF score corresponding to the three cardinal symptoms of HF: difficulty with breathing, feeding and activity. The score was iteratively improved over a 3-year pilot phase until no further changes were made. The inter-rater reliability (IRR) across a range of providers was assessed using the final version. Peak HF scores were analyzed against mortality and length of stay (LOS) for all pediatric HF discharges between July-October 2019. RESULTS The final HF score was a 4-point ordinal severity score for each of the 3 symptom domains (total score 0-12). Among clinicians who scored 12 ADHF inpatients simultaneously, the intra-class correlation was 0.94 (respiratory ICC 0.89, feeding ICC 0.85, activity ICC 0.80). Score trajectory reflected our clinical impression of patient response to HF therapies across a range of HF syndromes including 1- and 2-ventricle heart disease and reduced or preserved ejection fraction. Among 28 patients hospitalized during a 3 months period (N=28), quartiles of peak score were associated with LOS (P CONCLUSION This simple acute HF score may be a useful tool to quantify and monitor day-to-day HF symptoms in children hospitalized with ADHF regardless of etiology or age group. The score has excellent inter-rater reliability across provider levels and is associated with major hospital outcomes supporting its clinical validity. Validation in a multicenter cohort is warranted.