Pain management in hospitalized neonates is a critical concern and has historically been undertreated.1 During their stay, these patients undergo numerous painful procedures, such as blood sampling, vascular access placement, intubation, and lumbar punctures. Studies estimate that NICU patients experience an average of up to 16 daily procedures.2,3

Pain management in preterm infants is particularly important, as inadequate treatment has been found to have a negative impact on neurodevelopment.4 Specifically, greater exposure to pain during the neonatal period in preterm infants has been associated with poorer cognitive, motor, and behavioral outcomes in childhood.5–7 While untreated pain and inadequate sedation can be harmful, the pharmacologic management of pain and stress in neonates is not without risks. Analgesics and sedatives may cause adverse effects such as cardiorespiratory depression, hypotension, delayed gastric motility, tolerance, and withdrawal syndrome, and there is growing concern about their potential long-term impact on neurodevelopment.8

Both pharmacologic and nonpharmacologic approaches are used for neonatal pain management. Some nonpharmacologic interventions have proven effective for certain procedures.9 However, there is no standardized pharmacologic approach to neonatal sedation and pain management.10 Currently, each NICU must develop an in-house treatment protocol based on the available evidence.11 In our hospital, pain management is standardized. However, in Europe, most of the commonly used drugs have no approved indication for use in preterm infants, and only a few have an approved indication for use in term neonates. Furthermore, the summary of product characteristics for certain drugs, such as methadone and midazolam, explicitly contraindicates use in newborns.12

Due to the lack of approved therapeutic options, dosing in neonates is often extrapolated from pharmacokinetic data from pediatric and adult populations. Such extrapolation is frequently not supported by robust studies demonstrating its safety and efficacy, particularly in preterm infants. Additionally, advanced techniques, such as extracorporeal membrane oxygenation and therapeutic active hypothermia, alter drug pharmacokinetics, although their impact on clinical practice has not yet been adequately characterized.13,14 This uncertainty underscores the need for real-world studies to evaluate the use of these treatments and their potential adverse effects.

Although there have been large-scale studies on neonatal pain medication, the cumulative exposure to these treatments during the stay in the NICU has not been analyzed in depth. Much of the previous evidence comes from surveys or registries that do not provide detailed information on aspects like treatment duration or the administered doses.10 This lack of detailed information is a critical limitation for prescribers, affecting clinical decision-making and safety assessments.

In this context, we designed a study with the primary objective of conducting a detailed analysis of the drugs used for neonatal pain management, including their indications, doses and durations. We also evaluated treatment safety in order to contribute key evidence to help optimize prescribing practices and improve pain management in this vulnerable population.

During the study period, 423 patients were admitted to the NICU. Of this total, 357 met the inclusion criteria; 48 patients were excluded because they were outborn and 18 because they stayed less than 24 hours in the NICU. Table 1 presents the demographic characteristics of the patients admitted to the NICU during the study period.

Of the patient admitted to the NICU, 53.22% (n = 190) required treatment with some sedative or analgesic drug. Considering only those requiring a higher level of analgesia beyond paracetamol, the percentage who received analgesics or sedatives decreased to 31.37% (n = 112) in the overall sample and 33.88% (n = 41) in infants born preterm before 32 weeks of gestation or with birth weights of less than 1500 g. Table 2 shows the proportion of patients who received sedation or analgesia based on their baseline characteristics. Among patients who received sedoanalgesia, 42.1% were female, with a median (IQR) gestational age of 36 (29–39) weeks and birth weight of 2365 (1145–3190) grams. The median (IQR) duration of mechanical ventilation was 1 (0–3) day and the median (IQR) length of stay in the NICU was 14 (6–34) days.

We found a statistically significant difference in sedation and analgesia use between patients born before 32 weeks or with birth weights under 1500 g and those born after 32 weeks with weights of greater than 1500 g (61.15% [74/121] vs 49.15% [116/236]; P = .0041). Additionally, 92.7% (101/109) of patients who required mechanical ventilation received sedation/analgesia, compared to 35.9% (89/248) of those who did not require mechanical ventilation (P < .005).

The median (IQR) duration of sedation or analgesia in patients who received one or both was 4.5 (2–11) days.

Table 3 lists the drugs prescribed for pain management and sedation during the study period, in addition to the indications for which they were used. The same drug may have been administered to a patient for more than one indication. It is important to note that our protocol does not contemplate the routine use of opioids or neuromuscular blockers for noninvasive procedures, such as ultrasound scans. The recorded bolus doses administered for imaging tests correspond to critically ill patients who were already unstable and frequently on mechanical ventilation. In these cases, additional boluses were administered to optimize hemodynamic and respiratory stability during the procedure, rather than as an analgesic or sedative requirement for the imaging itself. Table 4 shows the minimum and maximum doses used for each drug and the median cumulative dose during the stay of patients in the sample.

Paracetamol was the drug of choice for non-opioid analgesics. During the study period, more than half the patients received paracetamol at some point. Metamizole was used exclusively as an antipyretic and never as an analgesic.

Fentanyl and morphine were administered as continuous infusion for maintenance pain management in 82 patients. Ninety-six percent (n = 79) of the patients initially received fentanyl, with 18.9% (n = 15) transitioning to morphine. Morphine was initially used in three patients, two of whom were hemodynamically stable. Remifentanil was used only in intermittent doses in 93.3% (n = 14) of patients who required intubation during the procedure.

Regarding sedation, 68 patients received continuous dexmedetomidine infusions. Nineteen patients also received midazolam in critical situations requiring deep sedation. The median (IQR) corrected gestational age was 39 (35–40) weeks in patients who received midazolam and 37 (27–39) weeks in those who received dexmedetomidine.

Cisatracurium and rocuronium were the drugs used for muscle relaxation. Continuous infusions were administered only during ECMO treatment and in the critical and postoperative period of two patients with congenital diaphragmatic hernia, always in patients on mechanical ventilation (n = 5). Single doses were administered for all other procedures in which muscle relaxants were used.

Five patients underwent therapeutic hypothermia, and all received fentanyl for analgesia during cooling. The dose was similar to the doses used in other patients; none received a dose higher than 2 μg/kg/hour, and the median (IQR) maximum dose was 1.5 μ/kg/hour (1-2). All patients received dexmedetomidine as a sedative, and two also received midazolam.

As for the adverse events associated with the use of these drugs, there was one case of paracetamol poisoning during the study period that required treatment with N-acetylcysteine. Due to an administration error, the patient received a 75 mg/kg dose instead of 7.5 mg/kg. The patient was 27+4 weeks of corrected gestational age of weighed 850 g. Four hours after drug administration, the serum paracetamol level in the patient was 49.5 μg/mL. The paracetamol levels subsequently normalized to undetectable levels, and the patient did not exhibit any associated clinical manifestations.

Furthermore, 8.8% (6/68) of patients treated with dexmedetomidine required a dose reduction or temporary discontinuation due to hypotension or bradycardia. In regard to drug withdrawal, 50.7% (34/67) of the patients treated with benzodiazepines or opioids received some form of replacement therapy for prophylaxis or treatment for withdrawal syndrome. Among the patients who received continuous opioid infusion, 53.1% (34/64) received methadone. Additionally, some patients (32.3%, 22/68) required treatment with clonidine or lorazepam as a substitute for sedative therapy (dexmedetomidine or midazolam).

Our study provides a detailed overview of sedation and analgesia use in a high-acuity NICU, highlighting both the frequency of their use and their dosage. In this sample, 53.2% of the patients required sedation. This percentage is higher compared to the 34% reported in the EUROPAIN study cohort, in which fewer patients appeared to require sedation.10 However, comparing data for patients born before 32 weeks of gestation in the Swedish cohort, the percentage of patients requiring intravenous sedation and analgesia was similar (38.5% in the Swedish cohort vs. 33.8% in our study).15 Previous studies have shown that attitudes toward analgesia in neonates can vary depending on the cultural context and national treatment guidelines.16 The high proportion of sedation and analgesia in our cohort suggests a greater sensitivity to neonatal pain management in our setting, as well as awareness/knowledge among our professionals that pain in early life has significant long-term effects on health outcomes persisting into childhood and even adulthood.

The results show that fentanyl was the first-line opioid in most cases, with subsequent switching to morphine in patients with prolonged or more severe conditions. Remifentanil was used almost exclusively for orotracheal intubation, which was consistent with the previous literature.17 Sedation was primarily based on dexmedetomidine, with midazolam reserved for patients in more critical situations requiring deep sedation. This approach was consistent with previous evidence suggesting that dexmedetomidine may be a safer alternative to benzodiazepines, given its potential neuroprotective effect and the adverse effects on neurodevelopment associated with benzodiazepines.18,19 The use of muscle relaxants was exceptional and limited to particular situations, such as ECMO or major surgery.

When we compared our results with those of the EUROPAIN study, we found that 82% of ventilated neonates in that study received some form of sedation and analgesia,10 compared to 92.7% of ventilated neonates in our cohort. With respect to opioids, the EUROPAIN cohort primarily received morphine, whereas fentanyl was the most commonly prescribed opioid in our unit. Furthermore, 25% of neonates in the EUROPAIN cohort received midazolam, whereas, in our cohort, dexmedetomidine was the primary sedative, with less frequent use of midazolam. Notably, the use of neuromuscular blockers in continuous infusion in ventilated patients was significantly more frequent in the EUROPAIN study (7% compared to 4.6% in our study). We also found a critical difference in the use of drugs for treating or preventing withdrawal syndrome. In the EUROPAIN study, 8.1% of patients treated with opioids or benzodiazepines received such drugs, compared to 50.7% in our cohort. This discrepancy may be due to changes in withdrawal management practices in recent years, as the EUROPAIN study was published ten years ago. In our unit, the decision to initiate methadone is not based on predefined cumulative dose thresholds but rather on the duration of opioid exposure, the presence of clinical signs suggestive of withdrawal, and the judgment of the care team.

Another relevant finding of our study was the extensive use of paracetamol in various clinical settings. Its role as a first-line agent for management of postoperative pain is well established and supported by robust evidence.20 However, in critically ill neonates, paracetamol was also used alongside opioid therapy to achieve an opioid-sparing effect and improve safety.21 In cases of mild pain, such as skin injuries, diaper rash, or minor procedures, paracetamol was often used as the primary systemic analgesic. While this approach is consistent with paracetamol’s profile as a safe and well-tolerated drug, nonpharmacologic measures should be reinforced in these situations to minimize unnecessary drug exposure. These findings emphasize the need to optimize protocols, consolidate the role of paracetamol in postoperative pain and promote its use as an adjuvant in selected critically ill patients. They also encourage rational prescribing in cases of mild pain.

Regarding the cumulative doses administered during the stay, to our knowledge, our study is the first one to collect and analyze this information, so we were unable to compare it these with data from previous studies.

In respect to therapeutic hypothermia, all patients who underwent cooling received analgesia with fentanyl and sedation with dexmedetomidine or midazolam. This contrasts with the results of a survey on the medication used in these patients in Italian NICUs, where 71.4% of patients were treated with monotherapy, for which fentanyl (58.6%) was the most commonly used drug.22 A systematic review comparing sedation and analgesia strategies during therapeutic hypothermia revealed that morphine was the most widely used drug, often as monotherapy, with a dose range of 8 to 60 μg/kg/hour for morphine and 1 to 5 μg/kg/hour for fentanyl.23 Adding dexmedetomidine to opioids may reduce the required opioid dose and the time to extubation or resumption of feeding.24 The combination with dexmedetomidine in our cohort could have allowed for a reduction in opioid doses, and the maximum dose administered was 2 μg/kg/hour.

Sedation and analgesia in neonates with hypoxic-ischemic encephalopathy (HIE) undergoing therapeutic hypothermia is an area of growing interest and ongoing debate. In our center, as in many clinical protocols, routine sedation/analgesia is administered to these patients, based on the ethical principle of preventing pain and stress in this highly vulnerable population. Preclinical evidence suggests that pain and stress may counteract the neuroprotective benefits of hypothermia.25 Clinical data are also emerging: in a retrospective study of neonates with HIE, opioid treatment was associated with reduced brain injury on neuroimaging and improved long-term neurological outcomes.26 In a European multicenter randomized trial of therapeutic hypothermia, a greater treatment efficacy was observed compared to earlier studies (number needed to treat 4 vs 6–9), a finding attributed to the fact that all cooled infants also received morphine or an equivalent fentanyl dose.27 However, other studies have reported neutral findings: in a secondary analysis of the National Institute of Child Health and Human Development (NICHD) hypothermia randomized trial, use of sedation/analgesia was not associated with either better or poorer long-term neurodevelopmental outcomes.28 These results highlight that, although sedation/analgesia can mitigate pain and stress, the potential risks should also be considered. These include interference with neurological and neurophysiological monitoring, prolonged mechanical ventilation, hemodynamic instability, and altered drug pharmacokinetics during hypothermia, potentially leading to increased drug exposure. In line with this, Róka et al reported higher serum morphine concentrations between 24 and 72 hours of life in cooled neonates compared with normothermic infants receiving equivalent doses.29 Although our observational study included a limited number of patients with HIE, it contributes valuable real-world information regarding the routine use of sedation/analgesia and the absence of observed adverse effects in this subset of infants.

A relevant finding in our study was the use of the intranasal route for administration of sedative or analgesic drugs. Fentanyl and dexmedetomidine were administered intranasally in specific procedures, such as venous catheter insertion and intraocular injection, offering a noninvasive alternative that facilitates analgesia in situations where the intravenous route is unavailable or less convenient. The intranasal route appears to be an effective and safe option in neonates, although its use has not yet been standardized across neonatal units. Including this route in analgesia protocols could improve neonatal care.30

Our study, which is based on real-world clinical practice data, contributes to the evidence on the use of sedation and analgesia in critically ill neonates. Real-world data and evidence are crucial tools for improving drug development and safety in neonatal care.31 In this context, the implementation of electronic prescription systems, like the one used in our study, offers an opportunity to improve the safety of neonatal pharmacotherapy.32 Electronic prescriptions facilitate the structured collection of clinical and therapeutic data and enable detailed recording of administered doses.33

This study has several limitations. Due to its retrospective nature, the identification of adverse events may have been limited by the quality of the health records. Furthermore, the efficacy of the treatments was not assessed because the analysis did not include the scores in the pain assessment scales applied during treatment. However, the study also has strengths, including the large sample size, the follow-up of patients through hospital discharge, and the exhaustive recording of administered doses. This stands in contrast with other large-scale studies on pain management in neonates, such as the EUROPAIN study, in which there were no data on the indications and doses.10

Our study contributes to the knowledge of sedation and analgesia practices in the NICU, emphasizing the role of fentanyl as the first-line opioid, the widespread use of dexmedetomidine as a first-line sedative, and the need for strategies to reduce opioid withdrawal syndrome. We provide information on the doses used for all drugs, as well as their cumulative doses, which have been poorly described in previous studies. Incorporating electronic prescription systems and analyzing real-world data could be key tools for improving the safety and efficacy of sedation and analgesia in these patients.