A Novel Patent Ductus Arteriosus Severity Score to Predict Clinical Outcomes in Premature Neonates

Abstract

Background: Patent Ductus Arteriosus (PDA) in premature neonates has been associated with comorbidities including chronic lung disease (CLD), and death. However, the treatment of PDA remains controversial. There have been several echocardiographic variables previously used to determine the hemodynamic significance of PDA but their utility in early prediction of clinical outcomes is not well studied. Objective: The objective of our study was to evaluate the use of a severity scoring system incorporating markers of systemic under perfusion, pulmonary over perfusion and left ventricular (LV) function in predicting clinical outcomes in premature neonates. Methods: It is a single center prospective observational study involving newborns < 32 weeks’ gestation. An echocardiogram was done within seven days of life to measure variables previously known to predict severity of shunting in PDA including pulmonary perfusion index (PPI). Predictors of CLD/death were identified using multivariate logistic regression. A severity score was derived and its ability to predict clinical outcomes was tested using a receiver operating characteristic curve. Results: We studied 98 infants with a mean (SD) gestation of 28.9 ± 1.91 weeks and birth weight of 1228.06 ± 318.94 g, respectively. We identified five echocardiographic variables along with gestational age that was independently associated with the outcome variable (PPI, LV output, Superior Mesenteric Artery [SMA] Velocity Time Integral [VTI], Peak diastolic flow velocity in Pulmonary Vein [PV Vd], and reversal of flow in diastole in descending aorta [DFR]). The range of severity score was 0 (low risk) to 12 (high risk). A higher score was associated with the primary outcome variable of CLD/death (7.5 [1.2] vs. 3.6 [1.5], p < 0.001). Our severity score had an area under the curve of 0.97 (95% CI 0.93−0.99, p < 0.001) for predicting CLD/death. Conclusion: Our new PDA severity score of 5.5 has a sensitivity and specificity of 94% and 93%, and positive and negative predictive values of 94% and 93%, respectively.

Keywords: chronic lung disease; patent ductus arteriosus; prematurity; severity score.

Figure 1

Comparison of echocardiography measurements in infants with and without CLD/death. (a): PPI; (b): LV Output; (c): LV a’; (d): Pulmonary Vein diastolic flow velocity; (e): Flow reversal in descending aorta; (f): SMA VTI; (g): Celiac artery VTI; (h); PDA diameter; (i): PDA Vmax; (j): Mitral Valve E/A ratio; (k): LA:AO ratio. The unfilled dots are shown in Figure 1b,c,j to indicate outliers.

Figure 1

Comparison of echocardiography measurements in infants with and without CLD/death. (a): PPI; (b): LV Output; (c): LV a’; (d): Pulmonary Vein diastolic flow velocity; (e): Flow reversal in descending aorta; (f): SMA VTI; (g): Celiac artery VTI; (h); PDA diameter; (i): PDA Vmax; (j): Mitral Valve E/A ratio; (k): LA:AO ratio. The unfilled dots are shown in Figure 1b,c,j to indicate outliers.

Figure 1

Comparison of echocardiography measurements in infants with and without CLD/death. (a): PPI; (b): LV Output; (c): LV a’; (d): Pulmonary Vein diastolic flow velocity; (e): Flow reversal in descending aorta; (f): SMA VTI; (g): Celiac artery VTI; (h); PDA diameter; (i): PDA Vmax; (j): Mitral Valve E/A ratio; (k): LA:AO ratio. The unfilled dots are shown in Figure 1b,c,j to indicate outliers.

Figure 1

Comparison of echocardiography measurements in infants with and without CLD/death. (a): PPI; (b): LV Output; (c): LV a’; (d): Pulmonary Vein diastolic flow velocity; (e): Flow reversal in descending aorta; (f): SMA VTI; (g): Celiac artery VTI; (h); PDA diameter; (i): PDA Vmax; (j): Mitral Valve E/A ratio; (k): LA:AO ratio. The unfilled dots are shown in Figure 1b,c,j to indicate outliers.

Figure 1

Comparison of echocardiography measurements in infants with and without CLD/death. (a): PPI; (b): LV Output; (c): LV a’; (d): Pulmonary Vein diastolic flow velocity; (e): Flow reversal in descending aorta; (f): SMA VTI; (g): Celiac artery VTI; (h); PDA diameter; (i): PDA Vmax; (j): Mitral Valve E/A ratio; (k): LA:AO ratio. The unfilled dots are shown in Figure 1b,c,j to indicate outliers.

Figure 2

Difference in PDAss between infants with and without CLD/death.

Figure 3

Relationship between PDAss and predicted probability of CLD/death of the entire cohort.

Figure 4

Receiver operating characteristics curve of the ability of PDAss to predict CLD/death.