Vasoactive Inotropic Score and Outcome Assessment in Cyanotic Infants after Cardiovascular Surgery

Introduction

Infants undergoing open-heart surgery with cardiopulmonary bypass (CPB) support are at an increased risk of morbidity and mortality.¹ However, in the last few decades, improvement in surgical and perioperative management has helped to improve outcomes.¹, ² Objective scoring systems may help to predict outcomes and to compare surgical modalities and treatment protocols in the early postoperative period.³, ⁴ An ideal outcome measure should be easily measurable, reproducible, and most importantly independently co-relatable with short and long-term outcomes.⁵, ⁶

In 1995, Wernovsky et al proposed inotropic score (IS) in the Boston circulatory arrest study⁷ to quantify the amount of pharmacologic cardiovascular support required during the postoperative period. IS has been used in other clinical studies as a measurement of illness severity.³, ⁸, ⁹ In 2010, Gaies et al proposed expanding the IS to include other commonly used vasoactive medications, such as milrinone, vasopressin, and norepinephrine, and called it a vasoactive inotropic score (VIS).¹⁰ In their study, a higher VIS was associated with postoperative morbidity and mortality in patients > 6 months of age undergoing open-heart surgery with CPB support. However, limited data are available to support the use of either scoring system as a predictor of outcomes in the pediatric intensive care unit (ICU).⁷, ⁸, ⁹, ¹⁰

We performed this study to determine the role of VIS in predicting postoperative outcome of infants undergoing open-heart surgery for cyanotic congenital heart disease.

Patients and Methods

Between September 2014 and April 2016, 100 patients, 1 year of age or less undergoing surgery for cyanotic congenital heart disease were prospectively studied. We excluded premature infants with gestational age less than 34 weeks and weight less than 1.5 kg at surgery. Informed consent was obtained from the parents of all patients, and the study protocol was duly approved by the institutional ethics committee (IESC/T-319/08.08.2014).

The diagnosis and delineation of cardiac anatomy was established by transthoracic echocardiography, and if needed, by cardiac catheterization. Preoperative demographic and clinical data were recorded. Intraoperative data included details of the operation, use of inotropic and vasoactive medications, CPB, aortic cross clamp, and circulatory arrest times. Inotropic and vasoactive medications were initiated in the operating room as per standard protocol. These were modified in the ICU depending on the hemodynamic status by the attending surgeon and the ICU team. The hourly doses of the following inotropic and vasoactive medications were recorded: dopamine, dobutamine, epinephrine, norepinephrine, milrinone, and vasopressin.

The IS was calculated as described by Wernovsky et al⁷:

Wernovsky IS = dopamine dose (μg/kg/min) + dobutamine dose (μg/kg/min) + 100 × epinephrine dose (μg/kg/min).

The VIS was calculated as described by Gaies et al¹⁰:

Gaies VIS = IS +10 × milrinone dose (μg/kg/min) + 10,000 × vasopressin dose (U/kg/min) + 100 × norepinephrine dose (μg/kg/min),

These scores were obtained at five separate time periods—before surgery, within 1 hour of the completion of surgery, 24 hours after the completion of surgery, 48 hours after the completion of surgery, and just before discharge from the ICU. The maximum IS and VIS for both the first 24 hours and the subsequent 24 hours were calculated.

The following parameters were recorded: time for peripheral rewarming (hours) > 32°C, serum lactate levels, time to first negative fluid balance (hours), for every 12-hour interval, output including both urine output and chest tube drainage, time to first extubation (hours), ICU stay (days), and length of hospital stay (days).

Results

Mean CPB time was 78.4 ± 39.3 minutes (23–339, median 69 minutes) and mean aortic cross clamp time was 39.8 ± 24.1 minutes (5–95, median 42 minutes).Thirteen procedures were performed on CPB without arresting the heart. VIS in the operating room was 7.57 ± 5.60 (0–30, median 5). No inotropes were required in 17 patients. Mean VIS at 1 hour, 24 hours, and 48 hours was 15.07, 13.5, and 8.02, respectively. We divided all our patients into five groups based on VIS at various time intervals. The classification system we used was as proposed by Gaies et al¹⁰ (Table 2).

At time intervals of 1 hour, 24 hours, and 48 hours, the number of patients in group 1 increased from 36 to 56, while that in group 5 decreased from 28 to 16 (Table 3), but even at the end of 48 hours, a significant number of patients had high inotropic requirement (VIS > 20).

Based on our institute protocol and ICU team clinical decision, inotropes were tapered off according to each patient’s clinical condition. The mean duration of mechanical ventilatory support was 28.9 ± 39.5 hours (range, 4–228 hours, median 12 hours), with mean time to negative balance 12.5 ± 16.2 hours (range, 1–90 hours, median 6 hours) and mean peak lactate 3 ± 2.13 mEq/L (range, 1–10 mEq/L, median 2.1 mEq/L), and mean time to peripheral rewarming was 11.5 ± 15.1 hours (range, 1–102 hours, median 6 hours). Based on Spearman analysis, we found that maximum VIS at 24 and 48 hours were positively correlated with adverse outcome variables (intubation time, time to negative balance, and peak lactate levels) with p value > 0.001 (Table 4 Table 5 Table 6 Table 7 Table 8).

After surgery, blood transfusion was required in 44 patients with mean transfused blood volume of 132.1 ± 124.90 mL (range, 50–1,300 mL, median 20 mL). Mean ICU stay was 98 ± 81.1 hours (range, 24–432 hours, median 60 hours) with mean hospital stay of 10.3 ± 7.94 days (range, 4–45 days, median 7 days). On logistic regression analysis, VIS at 48 hours correlated best in predicting ICU and hospital stay (Table 7 Table 8 Table 9).

Patients in group 2 were likely to have ICU stay longer by 1.94 days than the group 1 patients were, similarly groups 3, 4, and 5 patients were likely to have longer ICU stay by 3.99, 5.88, and 6.92 days, respectively, than group 1 patients were (Table 7).

Similarly by regression analysis for hospital stay (Table 8), we found patients in group 2 were likely to stay longer by 5.15 days than the group 1 patients, while groups 3, 4, and 5 patients were likely to have longer hospital stay by 8.48, 11.7, and 14.1 days, respectively, than group 1 patients were (Figs. 1 and 2).

Close

Fig. 1

Scatter plot showing correlation of various parameters with VIS. VIS, vasoactive inotropic score.

Export to PPT

Close

Fig. 2

Roc curve showing comparisons and mortality. ROC, receiver-operating characteristic.

Export to PPT

There were 10 early deaths (Table 9). Low birth weight (> 2.5 kg) and preoperative ventilator requirement was seen to be commonly associated with poor outcome. Common causes of death were pulmonary hypertensive crisis, pulmonary hemorrhage, and acute renal failure.

Ninety patients were discharged from the ICU after mean stay of 4 days. Non-fatal complications were seen in 24 patients and included recurrent lung collapse requiring prolonged ventilator support (n = 5), chylothorax (n = 2), wound infection (n = 7), and sternal complications requiring refixation or strapping (n = 4), acute renal failure necessitating peritoneal dialysis (n = 2). Rhythm disturbances such as tachy- and bradyarrhythmias were noted in six patients.

Discussion

To the best of our knowledge, this is the first prospective study to assess the application and validation of VIS in infants with cyanotic heart disease. Infants with cyanotic heart disease have a higher postoperative mortality compared with those who are older and those with acyanotic heart disease. Surgical principles and post-operative management strategies have evolved over the past few decades leading to improvement in patient outcomes. A reliable scoring system that can predict hospital course can be of tremendous help to the surgical and ICU team as they can counsel the family members, compare various protocols, and determine effects of various interventions.

Gaies et al¹⁰ in 2010 proposed VIS in a study group of 174 infants (0–6 months), and they found that maximum VIS score at 48 hours predicts poor outcome, prolonged intubation time, longer ICU stay, and increased time to achieve negative fluid balance. In the last few years, some more authors⁸, ⁹ have validated these findings. However, no study is available on patients with cyanotic congenital heart disease undergoing open-heart surgery. In addition, there is no consensus on the single most important score, as VIS has been analyzed at varying time intervals. In this study, we attempted to obtain VIS at various time intervals and to determine which of these scores correlate best with postoperative outcomes. Gaies et al¹⁰ analyzed the maximum and mean values of VIS at 24 and 48 hours; we, however, assessed absolute values of VIS at 1, 24, and 48 hours. We observed that analyzing absolute values is not only convenient, but also practical, as it does not take into account the initial period of low cardiac output immediately after surgery. Similar findings were proposed by Davidson et al.⁹

Gaies et al¹⁰ proposed a classification system of inotropic score (Table 2) in their study; we extended this work and classified our patients in five groups (Table 3). We performed regression analysis on this and assessed VIS scores at 1, 24, and 48 hours with ICU stay and hospital stay. We found that VIS score at 48 hours correlated best in predicting prolonged ICU and hospital stay (Table 6).

So if a child at the end of 48 hours postoperatively has VIS score of 11 (group 3), he or she is likely to stay 3.99 days more in ICU than a group 1 child, similarly hospital stay would be 8.48 days more than the child with score less than 5 (group 1) at the end of 48 hours.

Davidson et al⁹ first started using the 48 hours VIS value and highlighted its importance as marker of prolonged high intensity therapy. They showed relation of VIS at 48 hours with ICU and hospital stay. They published results of 70 infants (> 3 months of age), but the group was heterogonous as they included only acyanotic patients undergoing surgery without CPB. In contrast to our study, they found relation of only intubation time with maximum VIS at 48 hours. We, however, found that all three poor prognostic variables (intubation time, time to negative balance, and peak lactate levels) were related to the maximum VIS at 48 hours (p > 0.0001).

Study Limitations

Although we have prospectively analyzed a homogenous group, ours is a single-center study, with multiple surgeons contributing to the study. Hence, surgeon-related variations cannot be ruled out. Moreover, for universal application of VIS across various age groups and diagnoses, studies involving larger numbers of patients in multiple set-ups are needed.

Conclusion

Maximum VIS at 24 and 48 hours correlates with poor outcome variables such as intubation time, peak lactate, and time to negative balance (Tables 10 and 11). However, the absolute value of VIS at 48 hours correlates best with prolonged ICU and hospital stay.