Newborns in the neonatal intensive care unit (NICU) often undergo invasive procedures or surgeries that cause prolonged pain. Cumulative pain exposures during early brain development alter brain growth, decrease brain volume, and lead to neurodevelopmental sequelae. Therefore, monitoring and treating neonatal pain is a central goal of NICU care, and opioid-based pain regimens are regularly used to treat pain after surgery. However, prolonged opioid exposures can also lead to delayed feeding, prolonged mechanical ventilation, impaired brain growth, and neurodevelopmental sequelae. Optimally, neonatal pain should be managed to minimize pain, stress, and opioid exposures. If bedside nurses can detect the infant’s pain while it is still mild, then non-opioid drugs and non-pharmacologic therapies can be used to prevent it from becoming severe. Thus, optimized post-surgical analgesia will protect neonates from the trauma of either prolonged acute pain vs. prolonged opioid exposures, and the long-term effects of both these on the newborn brain. However, measuring neonatal pain accurately and consistently is difficult, as the current approaches are subjective and unreliable when practiced by different NICU nurses. Nursing workload in the NICU also does not allow for continuous pain assessment and monitoring. The objective of the proposed research is to develop an automated pain monitoring system that objectively measures pain continuously based on several biomarkers. The proposed Neonatal Pain Monitoring System (NPMS) will prompt the nurse with pain alarms, coupled with a pain management protocol that minimizes pain, stress, and opioid dosing during post-operative neonatal care. This project in the UG3 phase aims to: 1) develop and validate AI-driven neonatal pain biomarker signatures and an automated NPMS that can constantly assess neonatal pain to identify the onset and offset of postoperative pain, 2) create a pain management protocol that utilizes the information provided by NPMS to minimize postoperative pain and opioid use. The UH3 phase aims to complete a randomized controlled trial using postoperative pain scores, clinical outcomes, pain-related stress, acute pain events, pain trajectories, and provider surveys to compare the two randomized groups with and without AI-driven pain biomarker monitoring. Expected outcomes include neonatal pain biomarker signature models, an automated NPMS that uses these models, and a pain management protocol that utilizes NPMS to minimize postoperative pain and opioid use. These novel tools will equip the bedside nurses to manage surgical newborn pain pre-emptively and more effectively, which will reduce central sensitization, thus reducing pain scores (primary outcome), the need for opioids or other analgesics/sedatives, and improving their post-surgical clinical outcomes (secondary outcomes). This collaborative, multi-disciplinary team has the track-records, advanced skills, and access to sufficient patient populations to successfully execute the proposed aims and significantly advance the field of post-surgical care.

Newborns in the neonatal intensive care unit (NICU) undergoing invasive procedures or surgeries are frequently exposed to prolonged pain, postoperative complications, and human suffering because of the major difficulties in measuring neonatal pain, inconsistent and ineffective pain management, and delayed pain treatments due to nursing workload. Our proposed research aims to develop an AI-driven automated system to objectively and continuously measure newborn pain and pain sensitization based on several biomarkers, and when detected, to send their pain alarms to bedside nurses in real-time. The proposed neonatal pain monitoring system allows us to create therapeutic strategies that minimize pain, stress, and total opioid doses given during the post-operative care of newborns, thereby improving their clinical and post-surgical outcomes.