NIRS offers the advantage of noninvasive continuous oxygen monitoring, allowing for real-time assessments and immediate intervention [18, 19]. Supportive evidence indicates that changes in CrSO2 correlate with changes in brain electrical activity in areas associated with pain perception [20, 21]. This encompasses pain induced experimentally or clinically, such as needle insertion, electrical stimulation, or dental pain, in adolescents and adults post different surgeries. Alterations in CrSO2 linked to pain perception across distinct brain regions have been reliably quantified by NIRS technology. Subsequent investigations evaluated whether the mitigation of pain responses were linked to CrSO2 fluctuations, for example, if the administration of sucrose solutions or tactile stimulation during procedures on infants was efficacious in reducing CrSO2 fluctuations [22]. These findings align with prior research utilizing other parameters to assess infant distress [14].

LPT infants are frequently managed as term infants, because of similarities in activity levels and physical characteristics, and their procedural profiles often mirror those of term infants. However, LPT infants tend to require longer hospital stays and may undergo more frequent procedural interventions than their term counterparts. Therefore, acknowledging the impact of pain responses stemming from common medical procedures in this subgroup of LTP infants is important for devising appropriate care strategies.

The LPT and term cohorts demonstrated comparable baseline CrSO2 values, establishing a benchmark for LPT infants. Previous examinations of the impact of painful procedures on CrSO2 variations in newborns, were predominantly conducted among very preterm infants and unveiled diverse trends. Some studies have reported an increase, or an initial increase followed by subsequent decrease in CrSO2, indicating an increase in cerebral blood flow in response to neuronal activity via neurovascular coupling [7, 23,24,25,26]. However, these observed patterns lacked consistency across patient groups, with an overall prevailing trend of decreased CrSO2 on average [6,7,8]. Our investigation revealed various patterns of changes in CrSO2, including immediate increases followed by decreases or delayed decreases across patient groups. These inconsistent patterns hinder predictability across the studied cohorts. Despite the variability, the overarching trend indicated a decrease in CrSO2. While there were changes in both increases and decreases in CrSO2, we focused primarily on the decrease aspect. This emphasis stems from the fact that lower CrSO2 values are associated with long-term neurodevelopmental outcomes [27].

In response to routine medical procedures, both LPT and term infants exhibited reduced CrSO2 values, consistent with prior findings in both very preterm and term infants [6, 7, 25]. A recent study by Kumar et al. [6] involving very-low birthweight infants noted a decline in CrSO2 estimated to range from 4% to 7%, following procedures such as heelstick or venipuncture. Remarkably, this closely aligns with the 5% reduction in CrSO2 observed in the LPT group in our study. In our investigation, 6.7% and 10% of the LPT and term infants, respectively, exhibited Min PoP CrSO2 values below the normal acceptable threshold of 60%, with no significant difference between the groups. Importantly, none of the neonates experienced CrSO2 levels below 55%, which are considered critical values. Therefore, this study affirms that routine medical procedures do not result in dangerously low CrSO2 levels.

We aimed to explore and compare these responses specifically in LPT infants, which revealed distinct reactions to painful procedures compared to term infants. These were characterized by lower variability, as indicated by a reduced IQR, a lesser decline from baseline CrSO2, and smaller fluctuations in CrSO2, as depicted by differences between the minimum and maximum CrSO2 values compared to term infants. These findings suggest a mature response to maintain cerebral oxygenation during painful stimuli in the LPT group. A previous investigation highlighted an extended latency of peak responses to painful procedures in very premature infants, attributed to slow conduction velocity and delayed synaptic response to nociceptors [25]. Enhanced neural circuit development during the late preterm period potentially influences the cortical processing of pain response [10]. We observed a similar time interval to reach the Min PoP CrSO2 between LPT and term infants. Given the comparable trend in the change of PoP CrSO2 and the time interval to reach the minimum PoP CrSO2, our findings suggest that LPT infants exhibit a response to pain processing comparable to that observed in term infants. Nevertheless, since the overall PoP CrSO2 values decreased in both groups, indicating a cortical response to common procedural pain, it is advisable to limit the number of painful procedures to the essential and avoid unnecessary repetitive interventions.

We further investigated the correlation between clinical pain scores and changes in PoP CrSO2 during painful procedures in both LPT and term infants. PIPP-R and NIPS scores are commonly employed for acute pain assessments [28]. However, a systematic review revealed that several scoring systems lack a statistically significant correlation with neurophysiological indices, such as brainwave patterns, which are currently considered direct indicators of pain perception based on brain activity [29]. A recent study in preterm infants demonstrated strong positive correlations between changes in CrSO2 and video-assessed NIPS and PIPP-R scores (r = 0.71 and 0.78, respectively) [6]. NIPS scores based on muscle tone changes appear to provide a clear indication of pain response. We observed an overall poor to moderate correlation between the decrease in PoP CrSO2 and higher NIPS scores, primarily in term infants, but only poor correlation in LPT infants. The lower correlation in the LPT group may stem from reduced facial expression changes in response to pain with decreasing GA [30], suggesting that facial expressions alone may not be a precise assessment method, especially in preterm infants.

Hence, there has been a shift toward incorporating physiological measures into pain assessment tools, including the PIPP-R score, which utilizes assessments of physiological, behavioral, and contextual parameters [31]. Interestingly, we did not find a substantial correlation between PIPP-R scores and changes in PoP CrSO2. This lack of correlation may be partly attributed to the contextual scores derived from GA and behavioral stage, which remain constant during the procedure and do not vary with the pain response. Therefore, LPT infants with lower PoP CrSO2 values may exhibit lower overall PIPP-R scores due to reduced responses in both physiological and behavioral aspects. The rationale as to why LPT infants demonstrated a moderate correlation between higher PIPP-R scores and a wider difference between the Max and Min PoP CrSO2, unlike term infants, without differences in group pain scores, remains unexplained. Further investigation is warranted to elucidate these intriguing findings. However, our findings indicating that NIPS and PIPP-R scores do not correlate with CrSO2, suggest that the use of pain scores alone may not accurately reflect the cortical response elicited by acute painful procedures, at least in LPT and term infants. The findings align with those of Slater et al., who observed no alteration in facial expression in 13 out of 33 assessments, with 10 of these instances (30%) exhibiting changes in CrSO2 levels [15]. Gestational and postmenstrual age dependent variations among individual components can also be observed in behavioral (posture/tone, cry, facial expression, sleep pattern) and physiological (color, respirations, heart rate, blood pressure and saturation) parameters [32].

There are several strengths of our study. It is the first to prospectively examine the impact of routine painful procedures on LPT infants. We maintained internal validity by implementing stringent controls and precise monitoring of all variables. The selection criteria were meticulously applied to include only infants with minimum confounding factors that could influence CrSO2 values, thereby ensuring an accurate portrayal of the effects of the painful procedure on CrSO2 changes. However, certain limitations should be recognized in the generalizability of our findings. First, the implemented procedures were dissimilar and characterized by their brief and minimally invasive nature, resulting in pain experiences that were neither severe nor prolonged enough to induce significant adverse effects. Second, term infants, serving as the comparative cohort, may have introduced potential bias due to abnormalities affecting cerebral autoregulation, such as variations in partial pressure of carbon dioxide (pCO2) or plasma glucose over time. Nevertheless, both groups exhibited normal baseline CrSO2 values, and no significant differences were noted between them. Furthermore, term infants often exhibited more robust crying compared to their LPT counterparts, a factor that may influence the variability of CrSO2 or contribute to elevated intrathoracic pressure, thereby leading to more pronounced fluctuations in CrSO2 compared to the LPT group. Unexpectedly, only a small number of infants (8%) in both groups received oral sucrose for non-pharmacological pain management, which cannot be justified, since the intervention is part of our routine clinical practice for comfort during medical and nursing procedures. In retrospect the lack of any comfort measure in our study that focused on pain, is a serious oversight and merits timely resolution through a continuous quality improvement initiative. However, the outcomes of our study accurately portray the direct impact of painful procedures with minimal influence from other potential confounders, particularly from analgesic medications. Last, all infants in this study were respiratory and hemodynamically stable, and may have resulted in a selective bias towards only stable infants. It is important to note that these results may manifest differently in very ill infants undergoing more intense procedures that may lead to more pronounced changes in CrSO2. However, our focus was specifically on common procedures in stable infants, which reflects common clinical practice.

In summary, both LPT and term infants demonstrated decreased CrSO2 values in response to painful procedures. Notably, LPT infants displayed milder changes in CrSO2, indicating an intact cortical response to pain comparable to term infants. The correlations between changes in CrSO2 and PIPP-R or NIPS scores were poor to moderate, highlighting the complex nature of these associations relative to gestational age.