Intermittent intravenous paracetamol versus continuous morphine in infants undergoing cardiothoracic surgery: a multi-center randomized controlled trial

We performed a prospective, multi-center, randomized double-blinded, controlled trial in children aged 0–3 years undergoing cardiac surgery with the use of CPB. The aim of the study was to test the hypothesis that intermittent IV paracetamol administration as primary analgesic after cardiac surgery will result in a reduction of at least 30% of the median weight-adjusted cumulative morphine dose (in mcg/kg) during the first 48 h after cardiac surgery.

A prospective, multi-center, randomized double-blinded controlled trial conducted in four level-3 PICUs in the Netherlands and Belgium (Erasmus MC-Sophia Rotterdam, Wilhelmina Children’s Hospital UMC Utrecht, Beatrix Children’s Hospital UMC Groningen, the Netherlands and University Hospital Leuven, Belgium).

The study was approved by the Erasmus MC Medical Ethics Committee and was registered in the Dutch trial registry under the code NTR5448 as well at the Central Committee on Research Involving Human Subjects (NL 53085.078.15), EudraCT (2015-001835-20-NL/BE), and clinicaltrials.gov (identifier: NCT05853263).

Parents or legal guardians of the participating children provided written informed consent prior to any study procedures. The study protocol is published in Trials and available online [17].

Blocked randomization with randomly chosen block sizes and stratification by center was applied. A biostatistician (JvR) carried out the randomization in advance, with a randomization schedule for each participating center. Participants were assigned a consecutive trial number on the randomization schedule. The randomization schedule was safely stored in the local pharmacy at every center. The hospital pharmacies of all participating centers, but not the physicians, had access to the randomization schedule to ensure concealed allocation. Study medication was prepared at the participating centers by the pharmacy. In case of a medical emergency the pharmacists could be consulted on treatment allocation. To ensure blinding in both groups, a double dummy (intermittent placebo bolus of Sodium chloride (NaCl) 0.9% in a similar volume as the IV paracetamol dose or a continuous placebo infusion of NaCl 0.9% at the same rate as an equivalent morphine continuous infusion) was used.

Trained pediatric ICU nurses applied the Numeric Rating Scale-11 (NRS-11) pain and COMFORT-Behavior Scale (COMFORT-B) every 2 h. Both instruments are validated in critically ill children [18].

Pediatric delirium was assessed thrice daily in children requiring sedatives or analgesics for more than 48 h using the validated SOS-Pediatric Delirium (SOS-PD) scale [19].

All patients were classified according to the Risk-Adjusted Classification for Congenital Heart Surgery (RACHS-1) score [20]. The Pediatric Risk of Mortality (PRISM) lll score [21], combined with the Pediatric Logistic Organ Dysfunction 2 (PELOD-2) [22] were assessed on the day of surgery postoperatively, and postoperative days 1 and 2. To evaluate inotropic support in the first 48 h after surgery we calculated the vasoactive inotropic score (VIS) using the highest recorded inotropic and vasopressor doses. VIS was calculated as follows: dobutamine dose (mg/kg/min) + dopamine dose (mg/kg/min) + norepinephrine dose (mg/kg/min) *100 + epinephrine dose (mg/kg/min) *100 + milrinone dose (mg/kg/min) *10 + vasopressin dose (U/kg/min) *10,000 [23, 24].

To assess renal injury the lowest glomerular filtration rate (GFR) in the first 48 h after surgery was recorded and categorized using the KDIGO criteria [25].

A loading dose of morphine 100 mcg/kg IV was administered to all patients after separation from CPB, directly after surgery. Hereafter, patients were randomized to receive either continuous morphine or intermittent IV paracetamol. Paracetamol was dosed according to the Dutch Pediatric Formulary (loading dosage 20 mg/kg in all patients, maintenance dosage 40 mg/kg/day in neonates and 60 mg/kg/day in all other patients, 4 times daily) [26]. Morphine dosing was based on a population PK model-derived dosing regimen, range 3.9 to 16.0 mcg/kg per hour, resulting in similar morphine concentration across children’s age and bodyweight ranges [9, 17, 27, 28].

Pain or discomfort was scored with both the NRS pain scale and COMFORT-B. An open label morphine IV rescue dose (10 mcg/kg in neonates < 10 days, 15 mcg/kg in older patients) was administered if NRS-score ≥ 4. Pain was re-evaluated 10 min after the intervention. If pain persisted after three rescue doses, a morphine loading dose of 100 mcg/kg was administered and open label continuous morphine infusion was started at 10 mcg/kg per hour. Open label morphine infusion could be increased to maximum 30 mcg/kg per hour. In case of inadequate analgesia with maximum open label morphine, patients were switched to continuous IV fentanyl. Open label morphine or fentanyl infusion was titrated to effect using the NRS pain scale. In each group, continuous study morphine infusion was decreased on postoperative day 1 if the NRS score < 3 and COMFORT-B score < 10. Study medication was continued until 48 h after surgery. In patients who met the withdrawal criteria, treatment allocation was de-blinded and trial morphine infusions and/or paracetamol were switched to equivalent open label dose infusions of both paracetamol and morphine and analyzed as intention to treat within their treatment arm.

The primary endpoint is median weight-adjusted cumulative morphine dose in mcg/kg during the first 48 h postoperatively.

Secondary outcomes are:

Secondary outcomes were registered until 48 h after stop trial medication (96 h after surgery).

For the comparison of the primary outcome between groups, the nonparametric Van Elteren test with stratification by center was used. Linear regression analysis with group and treatment center as categorical predictor variables was applied for the secondary outcomes.

For comparison of the proportion of patients with one or more NRS scores of 4 and higher between the two groups, a non-inferiority analysis was done, using a non-inferiority margin of 20%. The confidence interval was calculated with the method of Klingenberg[29], with adjustment for center. Adverse effects were compared between groups using Fisher exact tests. The level of significance was set to 5%, and all tests were two-sided.

An international external Data and Safety Monitoring Board (DSMB) composed of an experienced cardiac surgeon, pediatric intensivist and cardio-anesthesiologist together with an independent biostatistician was installed. The study protocol did not contain an interim analysis. The DSMB evaluated inclusion rate and safety of participants (need for rescue morphine and NRS pain scores in both groups) 4 times during the inclusion period and advised us to continue the study without design changes to the protocol. To better assess safety aspects and before patient enrolment, morphine rescue dose in micrograms per kilogram in the first 48 h postoperatively, number of patients receiving rescue morphine doses, number patients needing rescue morphine continuous infusions and average COMFORT-B scores were added as secondary outcome.

Fourteen of the enrolled 208 children were withdrawn from the study before start of study medication leaving 194 patients for final analysis. The main reasons are noted in the flowchart in Fig. 1.

The morphine group contained 100 patients versus 94 in the intravenous paracetamol group. The two groups did not significantly differ in patient characteristics and risk of mortality scores (Table 1. Patient characteristics). Underlying cardiac diagnoses are shown in Table 2.

Twenty-eight patients in the IV paracetamol group and 26 patients in the continuous morphine group were switched to open label IV PCM and continuous morphine during the study period, most often because of fever. Reasons for withdrawal and time to withdrawal are shown in Table 3. All patients were analyzed in their primary allocated group using the Fisher exact test.

The median weight-adjusted cumulative morphine dose in the first 48 h postoperative in the paracetamol group was 5 times lower (79%) than in the morphine group (P < 0.001 (Table 4)).

There were no significant differences between groups in median total rescue morphine consumption in the first 48 h postoperative, proportion of patients receiving rescue morphine boluses and additional continuous morphine infusions.

The median weight-adjusted cumulative morphine consumption in the 194 patients who received the study medication for 48 h was 145.0 (115.0–432.5) mcg/kg in the IV paracetamol group vs 692.6 (532.7–856.1) mcg/kg (p < 0.001) in the continuous morphine group. The rescue morphine consumption was similar in both groups, 29.4 (0–45.7) mcg/kg in the IV paracetamol group vs 30.0 (0–70.9) mcg/kg in the continuous morphine group. The percentage of patients needing continuous open label morphine was 42.3% in the IV paracetamol group vs 42% in the continuous morphine group.

Percentage of adverse effects did not significantly differ between treatment groups (Table 4). Hemodynamic instability as predefined was the most frequently occurring adverse effect in both groups; 22 (23%) in the IV paracetamol group versus 28 (28%) in the continuous morphine group (p = 0.46). Inotropic support expressed as the vasoactive inotropic score (VIS) based on the highest doses of inotropes in the first 48 h was similar between both groups (Table 4) (See Additional file 2, Table 1: Inotropic support in the first 48 h). 

Three patients in IV paracetamol group were re-intubated versus two patients in the continuous morphine group (p = 0.68).

Overall median NRS pain scores were 0 (IQR 0–3) vs 0 (IQR 0–5) in the IV paracetamol and continuous morphine group, respectively. The number of patients with one or more NRS pain scores of 4 or higher was similar in both groups (Table 4).

The estimate (two-sided 95% CI) for the difference in proportions between the paracetamol group and the continuous morphine group, adjusted for center, was − 3.1% (95% CI − 16.6%–10.3%). Non-inferiority of IV paracetamol administration was proven because the upper bound of the 95% CI was below the predefined non-inferiority margin of 20%.

The concomitant use of sedatives was similar in both groups (Table 4). An overview of the doses given of the used sedatives during the study period are displayed in the Additional file 1 (Table 3: Sedative use in the first 48 h).

Nine patients had hepatic dysfunction (ALAT n = 3 (morphine n = 2, paracetamol n = 1); ASAT n = 8 (morphine n = 4, paracetamol n = 4). Renal dysfunction ranging from mild to severe ( GFR under 60 ml/min) was seen in 48 patients (Table 5). Severe renal dysfunction defined as a GFR of 29 ml/min or less was seen in 8 patients; 6 in the morphine group and 2 in the IV Paracetamol group. Most of these patients (31 vs 17) received study morphine, this was not significantly different between groups (p = 0.71 (2-sided)). There seems to be a trend toward increased renal dysfunction in the morphine group which might support reports of protective aspects of IV paracetamol in these patients.

To the best of our knowledge, this is the first large randomized controlled trial in children under the age of 4 years undergoing cardiac surgery with CPB, and the first to compare IV paracetamol as a primary analgesic postoperatively with the common practice of administering opioids.

About 30% of patients in either group were switched to open label paracetamol and morphine during the study timeframe. In approximately half of these cases the reason was sustained fever after surgery and the need for IV paracetamol to decrease body temperature. Interestingly, fever episodes were similar in both groups. Changing to open label might have actually increased the morphine consumption in patients in the IV paracetamol group who were switched from placebo to open label morphine within 48 h, thereby reducing the observed difference between the groups. The need for open label paracetamol due to fever does not reflect inadequate pain management and would not be an issue in open label paracetamol treatment.

To conclude, administration of intermittent intravenous paracetamol as primary analgesic in children under 3 years of age after cardiac surgery with the use of cardiopulmonary bypass resulted in a substantial reduction of the median weight-adjusted cumulative morphine consumption in the first 48 h postoperatively and reduced the need for continuous morphine infusions in almost 60% of all patients treated with IV paracetamol. Considering the similar need for rescue morphine doses and the same median NRS and COMFORT-B scores in both groups, an analgesic treatment protocol incorporating a loading dose of 100 mcg/kg morphine followed by IV paracetamol maintenance doses and bolus rescue morphine achieves equally effective postoperative pain relief in these patients independent of diagnosis or type of cardiac surgery negating the need for continuous morphine infusions in 40% of all patients.