Cortical pain responses from 112 human neonates were measured using EEG to a single painful procedure (clinically required blood test). Behavioural and physiological responses were also recorded as part of a standard composite pain score, Premature Infant Pain Profile (PIPP). Extensive notes were collected regarding the infant’s medical and pain history up to the day of the procedure. The data includes 4 second epochs from a 20 channel EEG recording that was time-locked to heel lance, control, and auditory stimuli. PIPP scores for each of the 3 stimulus types are also provided. Up to the time of the procedure, we have collected detailed information from the medical notes for each infant. This information includes: the condition at birth, diagnoses, medications, previous painful procedures, and injuries. Written parental consent was obtained prior to each study. The study was approved by the NHS Health Research Authority and the study conformed to the standards set by the Declaration of Helsinki.Many infants are born too young or too small or with serious medical conditions. As a result, these infants will need to spend time in hospital and undergo numerous clinical procedures, many of which are painful. While even the youngest preterm infant can show behavioural reactions to pain, we do not know how much the pain is processed in their brain. This is important because the brain is responsible for the actual feeling of pain: how much it hurts, where it is and how unpleasant it is. Our research aims to discover this so that infant pain can be prevented or adequately treated. Too much injury and stress in early life, when the brain is still forming connections, is thought to alter normal development of pain pathways in the brain, but this process is poorly understood. We intend to clarify this by measuring brain activity, tissue injury and physiological stress independently and analysing how they influence each other in intensive care. Finally, since each infant is an individual, we need to know how pain pathways in the brain differ between infants and what factors are responsible for these differences. Our work will help ensure that each hospitalised infant receives appropriate pain relief.

Experimental design Brain activity in response to 3 stimuli (noxious heel lance, sham control, and auditory control) was monitored using electroencephalography (EEG). All stimuli were performed using a lancet and time-locked to the ongoing EEG recording (Worley et al., 2012). Behavioural and physiological measures were also collected as part of the Premature Infant Pain Profile (PIPP). A more detailed description of these procedures has been published elsewhere (Jones et al., 2017; Slater et al., 2009, 2010). Infant pain history and relevant infant and maternal medical information was then collected by a research nurse. Stimulation The noxious stimulus was a heel lance that was clinically required to collect a blood sample. The heel was cleaned with sterile water using sterile gauze and the site allowed to air dry. The lancet was placed against the heel for at least 30s prior to the release of the blade. This was to obtain a baseline period free from other stimulation in order to identify evoked changes using EEG and the PIPP. The heel was then gently and intermittently compressed 30s after the release of the blade, again to ensure a post-stimulus period free from other stimuli. Babies were soothed as and when required. The control stimulations (sham and auditory) followed the same procedure as the heel lance, however the blade was facing away from the heel upon its release (sham), or the lancet was not touching the foot at any point (auditory – due to the audible click that is produced following the release of the blade). Electroencephalography EEG was recorded from up to 20 electrodes (disposable Ag/AgCl cup electrodes) in addition to the ground and reference electrodes. Recording electrodes were positioned according to the modified international 10/10 electrode placement system, with high density central-parietal and posterior temporal coverage, overlying primary visual (O1, O2,), primary auditory (T7, T8), association (F7, F3, F4, FCz, F8, P7, P8, TP9, TP10, POz), and somatosensory cortices (C3, Cz, C4, CP3, CPz, CP4). Reference and ground electrodes were respectively placed at Fz and FC1/2. ECG was recorded from both shoulders and respiratory movement was monitored with an abdominal transducer. Electrode/skin contact impedances were kept to a minimum by gently rubbing the skin with a prepping gel and then applying the electrodes with a conductive paste. A soft bonnet was then secured over the electrodes. EEG activity, from DC to ≥500 Hz, was recorded using the Neuroscan SynAmps2 EEG/EP recording system. Signals were digitised with a sampling rate of 2 kHz and a resolution of 24 bit. The EEG recordings were converted into EEGLAB data structure and segmented into 4 second epochs (2 seconds pre- and post-stimulus) using MATLAB (MathWorks) and EEGLAB (Swartz Center for Computational Neuroscience). No pre-processing was performed except for one file (253501L01) which was downsampled from 5 kHz to 2 kHz using EEGLAB. Each file contains the data of the EEG electrodes, ECG, and respiration, with a trigger (denoted with “4” or “8”) indicating the onset of the stimulus at 0s. These can be found in the parent folder “EEG” in the related Figshare item. Premature Infant Pain Profile (PIPP) A PIPP score was calculated for each of the stimulations by combining behavioural (3 facial expressions) and physiological measures (heart rate and blood oxygen saturation) (Stevens et al., 1996). Infant facial behaviour was recorded on video with an LED light placed within the frame. When the blade from the lancet was released, the light flashed allowing for accurate timing of stimulus onset. Beat-by-beat blood oxygenation and heart rate were monitored with a pulse oximeter (Nellcor Oximax). Scoring details are provided below. Patient notes Detailed notes were retrieved from numerous sources (1) Nursing charts that logged blood gas results, serum bilirubin levels (SBR), haematology results, microbiology results, observations chart (e.g. vital signs and urine output), feeding times and volume, and medications chart. (2) Cranial ultrasound scan results and any other imaging results (MRI). (3) Echocardiography results. (4) Specialist reports in case of referrals, from within or outside the department. (5) Contemporaneous notes (logging of each event as and when it occurs). All sources were used to record each painful procedure up to the day of the study. As we were testing the responses to a heel lance, additional information regarding previous heel lances was also obtained. No infants in this sample had undergone any surgery prior to the day of the study. Following additional consent from parents we recorded the mother’s medications and conditions during pregnancy. This information was collected using the electronic drug charts, and maternal patient notes if available on the ward. If the notes were not available the midwife was able to provide the information. Maternal information was only collected if the baby was <3 days old or breast feeding and therefore likely to be influenced by maternal factors.