Albertyn R*. Rode H*, Thomas J*, Dept of Surgery and Anaesthetics, Red Cross Children's Hospital, and the Univ of Cape Town, Rondebosch 7700, Cape Town, South Africa.

Aim: The assessment of pain and anxiety in South African Paediat-ric bum survivors are complicated by language barriers, cultural differences, socio-economic difficulties and delayed cognitive development. We have developed an observational assessment method based on the CHEOPS (Children's Hospital Of Eastern Ontario's Pain Scale) that differentiate between pain and anxiety and excludes patient involvement in assessments.

Patients and Methods: This method involves video graphing burned children during debridcment procedures, and daily pain assessment by independent trained observers. Respondents (n=100, inclusive of both sexes, admitted for bum injuries >20% TBSA) were divided into the pre-operational cognitive developmental stage (ages 2 -6) and pre-operational developmental stage (ages 7 - 11). Included were parental interviews (previous pain experiences, response, attitudes) and comparative physical assessments (i.e., pulse) done before and after procedures. Pain/ anxiety scores were obtained through allocating a numerical component to each observed behavioural category (i.e., facial expression, body movements, verbal expression and touch).

Results: Complete data was available for 25 patients in each developmental phase. The instrument was found to be applicable to both developmental phases, with high inter rater reliability and repro-ducibility, although some categories (i.e., crying) were more pronounced in the pre-operational stage. Scores differentiated between pain and anxiety with anxiety measured in the absence of direct observable pain stimuli. Cultural differences and parental attitudes dominates patient pain responses.

Conclusion: Pain, anxiety and prescribed analgesia can be effectively assessed without patient involvement by using the video-observational method.

EVALUATION OF THE RELIABILITY AND VALIDITY OF NEONATAL PAIN MEASURES

B. Brady-Fryer*. F. Wamock*, J. Landcr (SPON: C. A. Hoff-meyer), Univ. of Alberta, Edmonton, AB, Canada T6G 2G3.

Aim of Investigation. A number of tools have been developed to quantify neonatal pain for accurate assessment, but their psychometric properties have not been thoroughly investigated. This study was designed to evaluate the validity and reliability of two infant pain measures, the Neonatal Pain Scale (NIPS) and the Neonatal Facial Coding System (NFCS).

Methods: A segment of the data from a study of pain interventions for neonatal circumcision were subjected to secondary analysis. Two independent raters, blinded to group assignment, analysed behavioural data from videotapes of circumcisions of 27 infants in two topical drug application groups (EMLA, n=15: placebo EMLA, n=l 2). The raters used the NFCS and the NIPS to assess infant's responses during 6 circumcision procedure phases (baseline A, drug application, baseline B, preparation, circumcision, and post-circumcision). Infant state was coded for each phase.

Results: Several strategies will be employed to assess the validity and reliability of the NIPS and the NFCS including: analyses of the concurrent validity (correlation- N1PS/NFCS scores and mean phase heart rate), construct validity (within subjects analysis of N1PS/NFCS scores across 6 phases), discriminant validity (EMLA vs. placebo- EMLA scores) and equivalence (intcr-rater reliability and N1PS/NFCS agreement).

Conclusions: The results of this study will offer evidence of the validity and reliability of two existing neonatal pain measures.

HEART RATE VARIABILITY CHANGES IN VERY LOW BIRTH WEIGHT INFANTS FOLLOWING HEEL-STICK

Linda S. Franck. Bonnie Stevens, Cclcste Johnston, Patricia Petryshen*, School of Nursing, Univ of California, San Francisco, San Francisco, CA

Aim of Investigation: To describe changes in heart rate variability (HRV) following heelstick (HS) and to identify factors that influence HRV response to pain in very low birth weight (VLBW) infants.

Methods: HRV was measured by calculating vagal tone index (VTI) from 3 minutes of continuous electrocardiogram recording before and after HS in a subsample of 26 VLBW infants (.97± 0.22 kg birth weight) participating in a randomized controlled trial of non-pharmacologic pain interventions. Severity of illness (SOI) was measured using the Score of Neonatal Acute Physiology (SNAP). Pain was measured using the Premature Infant Pain Profile (PIPP) upon HS incision.

Results: There was no overall difference in mean VTI before and after HS, however, VTI remained below baseline levels 10 minutes following HS in 11 (42%) of the infants. These infants had lower baseline VTI than infants whose VTI returned to baseline values within 10 minutes of the HS (2.09 vs. 3.35, p=.02). Decreased post-HS VTI was associated with increased SNAP scores (r= -.50, p= 0.04) and lower PIPP scores (r= .55, p= 0.02). There was no association between SNAP and PIPP scores (r= -.08, p= 0.47).

Conclusions: Previous studies have not clearly demonstrated a relationship between pain and SOI in VLBW infants. The findings from this preliminary study indicate that stressed infants have prolonged recovery of HRV following painful procedures and suggest that HRV may be a useful measure of the pain-related stress in VLBW infants.

Acknowledgments: Funded by the National Inst for Nursing Research (NR03916).

IDENTIFICATION OF PAIN CUES OF CHILDREN WITH SEVERE NEUROLOGICAL IMPAIRMENT

AM Hunt. A Goldman, K Mastroyannopoulou³ , K Seers. Inst of Child Health, London WC1N 1 EH, UK,'Royal College of Nursing Inst, Oxford 0X2 6HE, UK; Univ of East Anglia, Norwich, UK

Aims: To identify pain sources and cues in children with severe neurological impairment. To explore issues that may affect assessment of pain in this group and guide the development of a behavioural pain rating scale.

Methods: Description of pain sources and behavioural cues was sought through a parent questionnaire (n=265) and in-depth interviews of parents (n--20) and professionals (n=26). Qualitative data was analysed using Grounded Theory. Results: 120 (45%) parents completed the questionnaire. Children were aged 1-25 years (median 1 1 years) and suffered both static (70%) and progressive (30%) encephalopathies. Eighty-three percent of parents reported their child suffered pain. Most common categories of pain were gastro-intestinal (56%), musculo-skcletal (54%), seizures (21%) and headaches (16%). Parents primarily identified pain through changes from their child's usual posture and movements (96%), cry (96%), facial expression (88%), mood (65%) and physiological changes (e.g. sweating) (58%o). Parents described difficulties in convincing professionals of their child's pain and obtaining pain relief. Knowledge of the child, familiarity with the population and an ability to sustain a long term relationship with the family were important factors in assessing and managing pain in this group

Conclusions Pain can be a significant problem for these children and needs to be addressed The assessment tool should be designed so that it recognises parents as experts in their child, is capable of reflecting the degree of variation in pain from one day to another and from one pain to another, is credible and valued by professionals across different specialties and disciplines Acknowledgments The study is funded by Action Research

NON-VERBAL EXPRESSIONS THAT NURSES USE TO DETERMINE PAIN IN INDIVIDUALS WITH SEVERE OR PROFOUND COGNITIVE IMPAIRMENT

Clemens Kctels*, Katinka A J. van Dongen. Jan P H Hamers, Huda Huljer Abu-Saad, Dept of Nursing Science, Maastricht Univ, P 0 Box 616, 6200 MD Maastricht, The Netherlands

Aim of Investigation To investigate the type and relative importance of non-verbal expressions nurses in the field of developmental disabilities use to assess pain in people with a severe or profound cognitive impairment

Methods Subjects were 112 nurses (14 male and 98 female) specialized (9 years experience) in the daily care of people with severe or profound cognitive impairment from 9 institutions throughout the Netherlands Nurses were presented with 158 indicators of pain collected from existing scales on measuring pain in non-verbal populations and based on the preliminary results of an ongoing research project in this area Nurses were asked to give ratings on a scale of 1 to 10 for each of the indicators with 1 representing 'not at all important' and 10 'extremely important to indicate pain' Furthermore nurses were asked to indicate the combination of nonverbal expressions they consider important in diagnosing pain in this population

Results Nurses chose all 158 indicators to indicate pain The highest mean scores were given to the indicators 'painful expression during manipulation' (9 12, SD 1 78), and 'moaning during manipulation' (9.02, SD 1 75) The lowest mean score (5 38, SD 2 46) was given to the indicator 'chewing' Nurses referring to pain in the severely cognitively impaired gave different scores The highest mean scores in this group were given to the indicators 'does not want to eat' (8 86, SD 1.56), and 'does not want to drink' (8 86, SD 1.66) All, except for one indicator had a range of 9 An additional number of 14 non-verbal expressions to indicate pain were mentioned. Of all subjects, 96% answered that combinations of indicators are important to assess pain, 'facial' expression was mentioned by 30 nurses as being important in association with other indicators

Conclusions: A multitude of indicators was used by nurses to assess pain in this patient population Attempts are currently underway to reduce the number of indicators using suitable statistical methodologies.

EMERGENCY MEDICINE: PATIENT'S ACUTE PAIN IS UNDERESTIMATED BY RESCUE TEAM

T.J. Luger. M. Gassner*, M. Baubin*, I H. Lorenz, Dept of Anaesthesia and Intensive Care Medicine, Univ of Innsbruck, A-6020 Innsbruck, Austria

Aim of Investigation: In emergency medicine the correct estimation of a patient's acute pain is complicated by the specific situation outdoors. Less time is thus available for correct evaluation Previous in-hospital data show a tendency to underestimate or ignore the patient's acute pain situation (1). The goal of this study was to examine the patient's pain as estimated by the rescue team. Methods. The study was approved by the Ethics Committee of the Univ of Innsbruck and included 70 patients who needed a rescue doctor. Of these, 19 were excluded because of logistics problems, leaving 51 patients to be studied (female:male: 30-21, medical-trauma 36.15; age 59.9±23.5 years ). Measurements were performed using a visual analog scale (VAS ranging from "no pain" to "maximum possible pain") and a verbal pain scale (VPS ranging from "no pam-0" to "maximum pam=5") The acute pain situation was estimated in the same order (rescue doctor, paramedic, driver, patient) at the emergency location, during transportation to the hospital and on arrival at the hospital Statistics ANOVA for repeated measures with Bonferroni corrections, p<0 05-statistical significance

Results The patients had an acute pain of 6 0 ± 2 7 cm on the VAS scale that was significantly underestimated by the rescue doctor (4 8±2 4, p=0 001), the paramedic (4 7±2 8, p=01) and the driver (4 4±2 4, p-0 0001), only 31%-33% correct estimations were evaluated Analysis ofcovanates demonstrated a significant influence of the emergency situation, medical or trauma (p=0 009), but no significance of age (p=0 81) or sex (p=0 08) During transporta tion to the hospital the patient's diminished pain was better estimated by all groups (ANOVA time F= 135 6, p-0 0001, group F=2 2, p=NS, group x time F=1 86, p-NS)

Conclusion The main finding of this study is that (I) a patient's acute pain is significantly underestimated by the rescue team independent of age or sex and (2) during transportation the patient's diminished pain is better estimated We therefore recommend the use of these pain scales (VAS or VPS) in emergency medicine outdoors

References Boisaubm EV et al Clin J Pain, 1989

ASSESSING PAIN IN INTENSIVE CARE UNIT (ICU) PATIENTS: THE STATE OF PRACTITIONERS' KNOWLEDGE

Kathleen Puntillo. Hildy Schell*, Neal Cohen* Univ of California, San Francisco School of Nursing and Dept of Anesthesia, UCSF Medical Center, San Francisco, CA 94143, USA

Aim of Investigation Optimal treatment choices for pain and anxiety are predicated on accurate assessments Our aim was to identify and examine assessment parameters used by critical care practitioners to determine the presence of pain and anxiety in their patients We also examined their knowledge of the effects of certain drugs administered to patients that can make pain assessment extremely difficult, e g , neuromuscular blocking agents (NMBAs) and propo-fol, an anesthetic

Methods Using a detailed, pretested questionnaire, we queried critical care RNs (N=49) and MDs (N-34) in four ICUs about methods they use to assess pain and anxiety in verbal and nonverbal critical care patients

Results: Most RNs (84%) and 50% of MDs in this study had spent an average of less than 10 hours studying pain assessment techniques in the year prior to this study Practitioners were more apt to use patient behaviors and physiological signs of pain than patients' self-reports They also emphasized the difficulty in assessing patients who are receiving NMBAs Some noted (N-28) that practitioners should assume patient has pain and routinely medicate if the patient is receiving NMBAs.

Conclusions We found that RNs and MDs at this institution frequently used indirect methods of pain assessment These methods are often, but not always, the best choices in ICU settings, especially if the patient is able to self-report Findings from this study will provide direction to future educational efforts that stress the importance of individualizing pain assessment according to critical care patients' medication regimes and clinical status

PRELIMINARY VALIDATION OF A NEW BEHAVIORAL INFANT PAIN SCALE (BIPS)

Sandberg B*. Larsson B*, Hansson L*, Palmgren K*, Olsson AK*, Pappila K*, Beskow A*, Westerlmg D. Depts ofThoracic Surgery, Anesthesiology and Intensive Care, Univ Hospital, S-221 85 Lund, Dept ofPaediatnc, Anaesthesia and Intensive Care, Karolmska Hospital, S-116 78 Stockholm, Sweden

Aim of Investigation To investigate a new behavioral infant pain scale (Bips) in infants aged 1-11 months (mean 5 months) after thoracic surgery.

Methods Forty-two babies were included in the study after informed consent from parents During mechanical ventilation in 21 babies, a continuous intravenous (iv) infusion of morphine (20-40 (ig kg 'h ') was given Intermittent iv injections of the opioid keto-bemidon 25 ug kg ' were given as required to treat pain in the thoracic pediatnc intensive care unit Paracetamol was given rectally 60-100 mg kg 'day ' After extubation, during spontaneous respiration, analgesia was rated before and after ketobemidon injections by experienced nurses with the use of Bips Bips encompasses 5 infant pain behaviors, each given a rating of 0-2, with a maximal level of pain defined as 10 and a pain-free state as 0 Results A total of 182 assessments by Bips were made in the study period. The ratings of Bips up to one hour after ketobemidon (0 7 ± 1 6, n=84) were significantly (p<0.001) lower than the pre-mjec-tion pain ratings (5.8 ± 2 5) The injection was repeated within an hour at 16 occasions, where analgesia was regarded as inadequate, as judged by persistent high ratings of Bips

Conclusion. Bips was well accepted, easy to leam and use Following an analgesic dose of ketobemidon, an opioid with a potency of 1 1-2 compared to morphine, ratings of pain by Bips were significantly lower than before injections, indicating that ratings of Bips were related to sensations of pain in the studied babies After due validation, Bips may be a useful tool for assessment of pain in infants and young children

DETERMINING THE CLINICAL UTILITY OF TWO MEASURES OF NEONATAL PAIN

C Schiller*. B. Stevens. S Sidani*, M Ballantyne*, C McNair*, Faculty of Nursing, Univ of Toronto, 50 St George St, Toronto, M5S3H4

Aim of Investigation To evaluate the clinical utility of two measures of neonatal pain, the CRIES' and the Premature Infant Pain Profile (PIPP)² for assessing post-operative pain in neonates in the Neonatal Intensive Care Unit (NICU)

Methods A measure of clinical utility including indicators of time, cost, instruction, format and acceptability was developed The clinical utility measure was completed by 61 nurses from the NICU who cared for 51 term and preterm neonates who had non-emergency surgery Participating nurses were randomly assigned to use either the CRIES or PIPP to assess pain The scores of all indicators were compared using the Kruskal-Wallis ANOVA or the t-test Results Both measures attained high scores on the clinical utility measure but there were differences on several individual indicators The CRIES rated higher on the average clarification time and overall format The PIPP rated higher on the potential for future use within the NICU

Conclusions: Both the CRIES and the PIPP can be considered clinically useful measures of neonatal pain

Acknowledgments Supported in part by a grant from The Grace Evelyn Simpson Reeves Nursing Research Fund, The Hospital for Sick Children, Toronto, Canada 'Krechel & Bildner, 1995 ² Stevens, Johnston, Petryshen & Tad-dio, 1996

VALIDATION OF THE PREMATURE INFANT PAIN PROFILE (PIPP) WITH VERY LOW BIRTH WEIGHT (VLBW) NEONATES

B. Stevens. C Johnston, L Franck, P Petryshen*, Faculty of Nursing, Univ of Toronto, 50 St George St, Toronto, Canada M5S 3H4

Aim of Investigation To establish the sensitivity and validity of the PIPP for assessing procedural pain in VLBW neonates Methods' PIPP scores of 19 neonates, <28 weeks Gestational Age (GA) were compared with 103 neonates >28 weeks GA who each received 4 randomly ordered non-pharmacological interventions (swaddling, positioning prone, pacifier + water, pacifier •+• sucrose) for 4 consecutive heel lances

Results There were no significant differences in PIPP scores between the 2 GA groups of neonates for any of the 4 interventions (> 05) Differences in the efficacy of the interventions were consistent across both GA groups with significantly lower PIPP scores in the pacifier + water and pacifier + sucrose interventions as compared to the swaddling or prone positioning interventions

Conclusions The PIPP is a sensitive measure of pain in neonates <28 weeks GA as well as in neonates >28 weeks GA There is support for the construct validity of the PIPP with VLBW neonates However, as the sample size was small, construct validity will be further examined in future studies with larger numbers of VLBW neonates

Acknowledgments Supported by NIH NINR Grant NR 03916 R01

VALIDITY CHALLENGES IN PAIN ASSESSMENT OF ELDERLY LONG-TERM CARE RESIDENTS WITH COMMUNICATION IMPAIRMENTS

Norma Stewart. College of Nursing, Univ of Saskatchewan, Terry Hartley*, Regma District Health, Joan Middleton*, Christie Ife*, Sandy Knezacek*, Parkndge Centre, Saskatoon, SK, S7M 5N9 CANADA

Aim of Investigation To develop a valid, reliable clinical tool to assess pain in institutionalized elderly persons with impaired communication

Methods A combination of qualitative and quantitative methods will be used in tool development and psychometric evaluation The focus here will be on the qualitative convergent validity process of separating painful vs nonpainful episodes on the videotapes of 25 residents in potentially painful situations Inclusion criteria were age 65 and over, pain diagnosis, and impaired communication due to Alzheimer disease or stroke Informed proxy consent was obtained after Univ ethics approval was received

Results The convergent validity results will be presented from an iterative process oftnangulation by research team members from the disciplines of nursing, speech therapy, and pharmacy Independent validity data were obtained from the physiotherapist who conducted range of motion maneuvers A series of validity challenges were met with research team decisions in the process of meeting the goals of the investigation

Conclusion Clinical judgement of a cluster of dimensions (facial, verbal, vocal, behavioral, and physiological) is important for valid assessment oftherapeutically significant pain in this population The dimensions of primary focus will shift depending on the type and degree of impaired communication Acknowledgments This project is supported by the Health Services Utilization and Research Commission of Saskatchewan

ARE CHEMICAL SOMATOSENSORY EVOKED POTENTIALS (CSSEPS) AN USEFUL METHOD FOR PAIN MONITORING OF PATIENTS IN THE INTENSIVE CARE UNIT(ICU)?

Remhild Strauss*. Rudolf Morawetz, Alexander Kaske*, Dons Schreithofer*, Dept ofAnaesthesiology, Univ Clinic Innsbruck, A-6020 Innsbruck, Austria

Aim of Investigation Most pain assessment methods depend on communication with the patient The standard physiological parameters used to indicate pain are not pain specific CSSEPs proved to be a valid physiologic tool to evaluate the analgesic level in healthy volunteers The aim of our study was therefore to assess if CSSEPs are a sensitive indicator for analgesic treatment at the ICU

Methods. A feasibility-study on ICU-patients was done (n = 18) CSSEPs were recorded after painful stimulation of the nasal mu-cosa with CO; (stimulus intensity 90 vol%, mterstimulus interval 10 sec, stimulus duration: 200 millisec) Baseline and repeated measurements were done to monitor changes of amplitudes and latencies when analgo-sedative medication was reduced Results. CSSEPs appeared in 89% of the study sample In 38% (n=7) repeated measurements were done Length of latencies corresponded to analgesic potency of the drug and decreased in 85% when medication was reduced Number of averaged stimuli greatly influenced amplitudes but not latencies Due to low sample size and mhomogeneous sample no statistical significances could be shown

Conclusion. We showed that CSSEPs can be used for evaluation of the analgesic level in ICU- patients We recommend latency changes in relation to analgesic treatment as indicator rather amplitude changes because the latter are grossly influenced by number of averaged stimuli In further studies the effects of confounders such as number of averaged stimuli, stimulus intensity, and stimulus duration have to be evaluated by using bigger sample sizes

ON THE DEVELOPMENT OF AN OBSERVATIONAL SCALE TO MEASURE PAIN IN NON-VERBAL CHILDREN WITH SEVERE OR PROFOUND COGNITIVE IMPAIRMENT

Katinka A.J. van Dongen. Huda Huljer Abu-Saad, Jan P H Hamers, Dept of Nursing Science, Maastricht Univ, P O Box 616, 6200 MD Maastricht, The Netherlands

Aim of Investigation To develop an instrument to assess pain in the non-verbal child with a severe or profound cognitive impairment The purpose of the present study was to gain insight into non-verbal pain expressions, used by caregivers and observed in practice, to indicate pain in this population Methods Qualitative research methods were used Data were collected using both observations and unstructured interviews Children, adolescents, and young adults (n=25, mean age 16), were observed during possible painful situations (e g vaccination against influenza, physiotherapy, dental procedures) Furthermore parents (n=5), nurses (n=10), physicians (n=4), physiotherapists (n=5), a dentist, mouth-hygienist, psychologist, neuro-psychologist and an orthopedagogue involved, were interviewed Collected indicators were compared with indicators (n=122) from the literature and included the categories facial expression (n=20), vocal expression (n=16), painful part (n=8), motor activity (n=27), physiological (n=14), activities of daily life (n=10) and social/mood (n=27)

Results A total number of 130 indicators were observed or reported by the interviewees Preliminary results show that 56 indicators do not correspond with the indicators in the literature Only 4 of all indicators are mentioned by over 50% of the interviewed subjects, 'facial expression', 'crying', 'screaming' and 'self injurious behavior'. 40% (n=49) of the indicators from existing scales were not mentioned nor observed

Conclusions Non-verbal expressions that caregivers use to indicate pain in non-verbal children with severe or profound cognitive impairment are diverse There is a discrepancy in the indicators found in the literature and indicators commonly observed and used in practice Research is currently undertaken to further test the validity and reliability of these results

CHANGES OVER SIX WEEKS IN PHYSIOLOGIC AND BEHAVIORAL RESPONSES OF PRETERM NEO-NATES TO A PAINFUL STIMULUS

Marl en e Walden. Nell Hodgson Woodruff School of Nursing, Emory Univ, 531Asbury Circle, Atlanta, GA, 30322, USA

Aim of Investigation To investigate physiologic and behavioral responses ofpreterm infants to a heelstick procedure and to determine how postconceptional age (PCA) may influence pain responses.

Methods: Neonatal Individualized Developmental Care Assessment Program (NIDCAP) to assess physiologic (heart rate, oxygen saturation, and respiratory rate) and behavioral (brow bulge, eye squeeze, 45 NIDCAP variables) responses to a heelstick procedure performed weekly between 27-32 weeks PCA Premature Infant Pain Profile (P1PP) scores were also measured

Results Heart rate increased while oxygen saturation and respiratory rate decreased during the heel lance Increased PIPP scores and occurrence of brow bulge and eye squeeze were observed Trends were noted towards increased behavioral stress between baseline and heelstick values Physiologic and behavioral measures were not sensitive to differences in PCA Conclusions. Former 24-26 week preterm infants who are between 27 and 32 weeks PCA are capable of expressing pain through physiological measures and facial actions in a manner similar to more mature preterm and healthy term infants While NIDCAP behaviors failed to reach statistical significance, this method may be helpful in providing information about the physiologic and behavioral cost of the heelstick event on preterm infants The PIPP shows promise as a valid instrument for assessing pain in infants as young as 27-32 weeks PCA

Acknowledgments Supported in part by research grants from the Foundation for Neonatal Research and Education and Texas Children's Hospital, equipment donation from Children's Medical Ventures, and in-kind consultation from Dr Bonnie Stevens, Faculty of Nursing, The Univ of Toronto

NEONATAL DISTRESS (PAIN) BEHAVIOR

Fay F Wamock*. Jamce Landcr, Univ of Alberta Faculty of Nursing, Edmonton, Alberta Canada, T6G 2B5

Aim of Investigation Our understanding of neonatal distress behaviors is limited despite knowledge about their sensory and perceptual abilities and 15 years of pain research Indeed some continue to question whether young infants even feel pain and most clinicians find the assessment of neonatal pain difficult The aim of this two phased observational study is to describe the behavioral response repertoire of male neonates to a painful procedure We describe the findings of the first phase of this pilot research here

Methods. Ethology, a systematic and precise observational method was used to describe the details (type, frequency and duration) of the neonates responses Videotapes of four full-term healthy new-bom males, ages 24 - 36 hrs, were analyzed to identify comparative behavioral patterns in response to varying stimuli associated with circumcision Observations were linked in real time to the video and the simultaneous collection of physiological data (heart rate, GSR, oxygen saturation) Inter-observer reliability was achieved using standardized agreement scores The aim of the second phase of the research is to develop a taxonomy of neonatal distress behaviors through the grouping and labeling of common behaviors into clusters

Results Observations indicate a wide range in the neonates' repertoire of distress responses and a preliminary understanding of distress trajectories This will be described with photographs and case studies.

Conclusions The study will provide basic information about neonatal pain behaviors that are required for the construction of valid pain assessment tools

Acknowledgment Supported in part by Medical Research Council of Canada PhD Fellowship to first author

9th WORLD CONGRESS ON PAIN, 1999, Vienna, Austria, p. 84 - 87