Hrev_master Abstract Since pain perception is highly subjective and culturally medi- ated, its objective evaluation remains difficult. Nevertheless, pain measurement should ideally be a part of the assessment of patients in order to plan adequate pain relief. Several scales have been pro- posed for pain measurement, being the numerical rating scale (NRS) the most widely used, often at triage time. NRS have demonstrated acceptable reliability and validity, in post-operative medicine and in oncologic pain, but data in the Emergency Departments (EDs) are poor. The aim of this study is to evaluate the Emergency Nurses’ (ENs) perception about the reliability of NRS in the triage process. A questionnaire based on 11 items was designed and subsequently administered to a large number of ENs in several EDs in Northern and Central Italy. 301 questionnaires were filled out and returned. The majority declares using NRS scale to measure pain (item 2, mode = 4, mean = 3.8), and attribut- ing priority code based on NRS value (item 3, mode = 4, mean = 3.4). In general, triage nurses believe that NRS is only indicative and that their judgement matters (item 4, mode = 4, mean = 3.2). The vast majority of triage nurses do believe that the patients will indicate a fake higher NRS value with the aim to get a more urgent code (item 5, mode = 5, mean = 4), while only a small minority expects that patients would underestimate their NRS for fear of penalizing more urgent patients. Very few believe that such scale underestimates the patients’ condition, while the majority is ambivalent about whether such scale overestimates it. In conclu- sion, NRS confirms to be a potentially valuable tool for pain eval- uation at triage time, but many nurses express some doubts on its reliability, and will attribute the triage code mainly basing on their own judgement. Introduction The culture of health-parameters measurement appeared only recently in the history of medicine. This effort is not merely an intellectual exercise: the selection of particular health indicators reflects the perceived importance of a problem at a particular his- torical period. Hence, the results of these measurements commonly become tools for the organizations promoting health policies. One classic definition of measurement is “the assignment of numbers to objects or events to represent quantities of attributes according to rules”.1 Hence, the process of measuring health parameters needs standardized questions about a particular condition, and require the assignment of numbers to each answer. These numbers may then be aggregated into a cumulative score, intended to reflect some aspect of health. When measuring any biological variable, the piv- otal questions should be: “Does this instrument measure what it is expected to measure?”, and “Is the parameter I’m measuring dis- tributed in the population according to the Gauss’ law?”. These two questions deal with validity and reliability (namely, the replicabil- ity of an instrument’s results) of the chosen tool. Pain, one of the leading symptoms of presentation to the Emergency Department (ED), should not be considered an exception. Since pain percep- tion is highly subjective and culturally mediated, however, its Emergency Care Journal 2018; volume 14:7787 Correspondence: Gianfranco Cervellin, Emergency Department, University Hospital of Parma, Via Gramsci 14 43126 Parma, Italy. E-mail: gcervellin@ao.pr.it Key words: Pain; Numeric rating scale; Triage; Emergency department. Acknowledgements: the authors acknowledge all the participating nurs- es of the involved EDs, for having kindly responded to the question- naire. They also acknowledge Prof. Alessandra Cassar, Department of Economics, University of San Francisco, for the help with the statistical analysis. Contributions: GC1 and IC conceived the study and made the question- naire; all the Authors contributed to data collection; GC12 coordinated the data collection; GC1 wrote the manuscript; all the Authors reviewed the manuscript. Conflict of interest: the authors declare no potential conflict of interest. Funding: none. Received for publication: 27 August 2018. Revisions received: 2 October 2018. Accepted for publication: 30 October 2018. This work is licensed under a Creative Commons Attribution 4.0 License (by-nc 4.0). ©Copyright G. Cervellin et al., 2018 Licensee PAGEPress, Italy Emergency Care Journal 2018; 14:7787 doi:10.4081/ecj.2018.7787 [Emergency Care Journal 2018; 14:7787] [page 111] Multicenter survey on emergency nurses’ perception of Numerical Rating Scale reliability at triage time in adult Emergency Department patients Gianfranco Cervellin,1 Ivan Comelli,1 Andrea Bellone,2 Fabrizio Giostra,3 Nicoletta Acierno,4 Michele Mitaritonno,5 Ciro Paolillo,6 Fabrizio Mastroianni,7 Matteo Pistorello,8 Gianni Rastelli,9 Giorgio Ricci,10 Andrea Tenci,11 Guido Caironi12 1Emergency Department, University Hospital of Parma, Parma; 2Emergency Department, Niguarda Hospital, Milan; 3Emergency Department, Hospital of Fermo, Fermo; 4Emergency Department, S. Paolo Hospital, Savona; 5Emergency Department, Hospital of Cosenza, Cosenza; 6Emergency Department, Spedali Riuniti University Hospital, Brescia; 7Emergency Department, Hospital of Aosta, Aosta; 8Emergency Department, Hospital of Treviso, Treviso; 9Emergency Department, Hospital of Vaio, Fidenza; 10Emergency Department, University Hospital of Verona, Verona; 11Emergency Department, Girolamo Fracastoro Hospital, San Bonifacio (VR); 12Emergency Department, Local Health Agency, Como, Italy Non -co mmerc ial us e o nly [page 112] [Emergency Care Journal 2018; 14:7787] objective evaluation remains difficult, and the validity of patient self-report is often questioned.2-4 Nevertheless, pain measurement should ideally be a part of the assessment of patients in order to plan the most appropriate therapeutic management, namely ade- quate pain relief.5 Some Authors have proposed that pain repre- sents the fifth (or sixth, when considering peripheral oxygen satu- ration) vital sign, thus compelling nurses and physicians to consid- er the evaluation of its intensity as part of patient assessment and documentation.6 It is, however, of pivotal importance to under- stand that pain is an intimate experience, and neither blood tests, tissue pathology, diagnostic imaging, physical assessments, nor patient behaviours reliably reflect the true pain experience. Several scales have been proposed in literature for pain mea- surement. For practical purposes, they should be easy to use, have clear data recording and processing, and fulfil criteria of validity, sensitivity and reliability. The four main tools presently used to rate pain intensity are: i) the numerical rating scale (NRS), ii) the visual analog scale (VAS), iii) the graphical rating scale (GRS), and iv) the verbal rating scale (VRS). In the majority of western countries, notably in adult patients, in most EDs various form of NRS are routinely used, often at triage time.5 NRS consists of a range of numbers, usually from 0 to 10, from which one is selected by the patient as most representative of his/her level of pain inten- sity. In this range, zero represents no pain and 10 represents some maximal measure of pain, like pain as intense as you can imagine or worst pain possible. NRS have demonstrated acceptable relia- bility and validity, especially in post-operative medicine and in oncologic pain,7-12 but data in the EDs are poor. Nevertheless, in many Italian EDs NRS contributes to the attribution of the triage code, thus establishing the priority to the first medical contact. Many patients are aware of this, and recently some concerns have been raised about the possibility of them adopting some forms of simulation with the goal to achieve a higher triage code, and, con- sequently, a shorter length of stay in the ED. As such, the aim of this study is to evaluate the Emergency Nurses’ (ENs) perception about the reliability of NRS in the triage process. Materials and Methods Design A questionnaire based on 11 items/statements (Table 1) was designed by a group of members of our scientific society – the Academy of Emergency Medicine and Care (AcEMC) – and sub- sequently administered either directly (by a physician or by the head of nurses) or via e-mail to a large number of ENs in several Italians EDs. The nurses were informed that the questionnaire was anonymous, and they were free to participate or not. Data collection The questionnaire asks about pain measurement at triage time in adult patients only (i.e., aged more than 14 years). Some addi- tional information, such as the respondent’s age, gender, and years of work as triage nurse were obtained. The responses were collect- ed in an anonymous way. For each center, a physician or a nurse was responsible for the distribution and collection of the question- naires, as well as for returning the surveys to our administrative assistant. All the questionnaires were filled out and collected between May 1st and June 15th, 2018. The study was performed in agreement with the ethical stan- dards established by the institution in which the experiments were performed and the Helsinki Declaration of 1975. Due to the intrin- sic nature of the study, an approval by ethic committee was not required. Each nurse freely decided to respond or not to the ques- tionnaire. Statistical analysis The statistical analysis was performed using Stata/SE 15.1®. The mean values ± Standard Deviation (SD) of the respondents’ rating of their agreement or disagreement with each of the eleven statements are reported in Table 2. The linear regressions for each item Y was performed according to the following formula: Yi=β0 + β1Agei + β2Genderi +β3Experiencei+β4Triagei + εi Article Table 1. The questionnaire. 1 2 3 4 5 Totally Partially Neither agree Partially Totally disagree disagree nor disagree agree agree 1 I consider essential to measure, not only to evaluate, patients’ pain at triage. 2 I always measure patients’ pain using NRS. 3 I attribute the triage code based on NRS results. 4 NRS results are only approximate. I decide the triage code based on my personal judgement. 5 Several patients declare a high NRS value with the aim to achieve a higher triage code. 6 Several patients declare a low NRS value due to the worry to penalize patients with worse problems. 7 I believe to give right triage codes, appropriate to the suffering condition of the patients. 8 Sometimes I overrate triage codes to avoid conflicts and problems. 9 Sometimes I underrate triage codes to avoid ED overload. 10 By using NRS there is a risk of underestimate patients’ problems. 11 By using NRS there is a trend toward overestimate patients’ problems. NRS, numerical rating scale; ED, emergency department. Non -co mmerc ial us e o nly where Yi represents respondent i’s answer to how he/she agrees with statement Y, with answers ranging from 1=Totally disagree to 5=Totally agree. Agei represents the respondent’s age categories divided in 5 brackets (20-30, 30-40, 40-50, 50-60, >60). Genderi represents the respondent’s sex, coded as 1 if female, 0 if male. Experiencei indicates years of experience as nurse in general, coded in 5 brackets (<3, 3-5, 5-10, 10-20, >20). Triagei reports years of experience as triage nurse divided into 6 brackets (≤1, 1- 3, 3-5, 5-10, 10-20, >20). Results A total of 301 questionnaires were filled out and returned, with a variable number of responses to each single item (leading to a total of responses varying from 264 to 301 per item). Most of the answers came from female nurses [169 vs 95; 37 not declared]. Age distribution, number of years from degree, number of years working as a triage nurse are shown in Tables 3-5. The overall results to each response are briefly shown in Figure 1. Table 2 reports in detail the mean values ± SD of the respondents’ rating of their agreement or disagreement with each of the eleven statements. Table 6 reports the results of the linear regressions for each item Y, according to the formula described in the Materials and Methods section. Our results indicate a substantial agreement of the triage nurses on the necessity to measure the patient’s pain at triage time (item 1), with the most frequent answer (mode) equal to 5 (completely agree) and mean equal to 4.3. The majority declares using NRS scale to measure pain (item 2, mode = 4, mean = 3.8), and attribut- ing priority code based on NRS value (item 3, mode = 4, mean = 3.4). We didn’t find any significant difference in the responses with respect to age, sex, years of experience or triage for any of these items. In general, triage nurses believe that NRS is only indicative and that their judgement matters (item 4 NRS results are only approximate. I decide the triage code based on my personal judge- ment, mode = 4, mean = 3.2). For this item, we found a significant difference based on gender, with females regarding the NRS relia- bility higher than males who instead value more their personal judgment (females 3.1 vs males 3.4; P=0.021 uncontrolled t-test; coeff.=-0.39, P=0.015 from regression controlling for the other factors), and a trend toward higher confidence as years of triage experience increase (coeff.=0.111, P=0.069). The vast majority of triage nurses do believe that the patients will indicate a fake higher NRS value with the aim to get a more urgent code (item 5, mode = 5, mean = 4). Only a small minority, on the contrary, expects that patients would underestimate their NRS for fear of penalizing Article Figure 1. The figure shows the mean values of responses to each statement. Table 2. Mean values ± standard deviation (SD) of rating attrib- uted from responders to the eleven statements. Question N. Number of responsesMean value ± SD 1 300 4.29 ± 1.00 2 301 3.76 ± 1.10 3 297 3.36 ± 1.22 4 297 3.21 ± 1.24 5 298 3.99 ± 1.05 6 301 1.51 ± 0.94 7 300 3.94 ± 0.82 8 299 2.34 ± 1.09 9 301 1.87 ± 1.02 10 300 2.10 ± 1.13 11 300 2.76 ± 1.29 Table 3. Age classes of the responders. Age class N % 20-30 43 14.3 30-40 104 34.6 40-50 112 37.2 50-60 36 11.9 >60 4 1.3 Not declared 2 0.7 Table 4. Subdivision of the sample according to the number of years from degree. Years from degree N % <3 15 4.9 3-5 67 22.3 5-10 56 18.6 10-20 76 25.2 20 81 26.9 Not declared 6 2 Table 5. Subdivision of the sample according to the number of years working as a triage nurse. Years working as a triage nurse N % <1 46 15.3 1-3 43 14.3 5-5 31 10.3 5-10 82 27.4 10-20 80 26.6 >20 17 5.6 Not declared 2 0.7 [Emergency Care Journal 2018; 14:7787] [page 113] Non -co mmerc ial us e o nly [page 114] [Emergency Care Journal 2018; 14:7787] more urgent patients (item 6, mode = 1, mean =1.5). For this last item (Several patients declare a low NRS value due to the worry to penalize patients with worse problems), the data show a positive correlation between such belief and age categories (coeff.=0.15; P=0.046) and a negative one with increasing categories of triage experience (coeff.=-0,13; P=0.002). Triage nurses exhibit confidence in their ability to assess the appropriate color code based on patient suffering (item 7, mode = 4, mean = 3.9) without much overestimation to avoid patient’s complaints (item 8, mode = 2, mean = 2.3) and even less underes- timation for fear of overburdening the ED (item 9, mode = 1, mean = 1.9). For item 7 (I believe to give correct triage codes, appropri- ate to the suffering condition of the patients) we found a significant positive correlation between agreement with this statement and increasing categories of years of triage experience (coeff.=0.09; P=0.024). With respect to what triage nurses think about whether NRS provides a biased estimate of a patient’s condition, very few believe that such scale underestimates it (item 10, mode = 1, mean = 2.1), while the majority is ambivalent about whether such scale overestimates it (item 11, mode = 3, mean = 2.8). For item 10 (By using NRS there is a risk of underestimate patients’ problems) we found a significant positive correlation with increasing age cate- gories (coeff.=0.26; P=0.008), and a significant negative one with increasing years of working experience as a nurse (coeff.=-0.16; P=0.022). Finally, for item 11 (By using NRS there is a trend toward overestimate patients’ problems), we found a significant negative correlation between agreement with this statement and increasing years of working experience as a nurse (coeff.=-0.17; P=0.029). Discussion The results of this study seem to suggest that a group of Italian ENs are confident with NRS scale for pain measurement at triage time, and that the vast majority of them always use this tool. Nevertheless, there is a high percentage of nurses that consider NRS score as only suggestive, thus basing their judgement on per- sonal experience. This seems to mainly originate from the belief that several patients will indicate an artificially higher NRS value to get a more urgent code. In fact, responses somewhat conflicting have been obtained from statements 3 to 6. It should be considered that patients population with a painful condition in the ED is extremely heterogeneous, and obviously dif- ferent from selected patients managed in hospital pain clinics. Several cognitive, emotional, psychological, cultural and social factors, including anxiety and haste to be visited, could jeopardize the reliability of NRS in the ED.3,4,13-15 The studies on this topic have reached a kaleidoscope of con- clusions, sometimes non concordant or even conflicting. For example, an interesting study has shown that a visual-analogic scale, namely the Wong-Baker scale (usually used in the pediatric setting), performed better than NRS in an overcrowded ED.16 Another interesting and intriguing study, performed in Italians EDs, comparing NRS (rated 1-10) with a fake NRS (rated 30-50), has shown that the fake scale underrated pain intensity when com- paring with official NRS (mean 5.5 vs 7.4). The Authors argue that, since in Italy there is an old school legacy according to which the highest value is more rewarding and 6 represents the minimum threshold, the 0-10 NRS scale could have unconsciously oriented patients to set themselves starting from a minimum psychological threshold, represented by 6.17 Moreover, since the triage nurse is usually the first health care provider to assess pain among ED patients, the time of pain relief is often dependent on the efficiency and reliability of the triage system.18 In some facilities, especially in North Europe, patient- reported pain is validated by a nurse to ensure that it is neither over nor underreported.19 A recent Danish study has documented a 70% agreement on pain category between the nurses, being the pain assessment unaffected by age, gender, or environmental issues. The Authors conclude that their findings justify that nurse-per- ceived pain assessment is used for triage in the emergency depart- ment.20 One major concern has been driven by the fact that, despite declaring a high NRS score, several patients refuse analgesia deliv- ered by the triage nurse. In an interesting study conducted in France, 38% of patients refused analgesia at triage, mainly (53%) because Pain is bearable without analgesic. Interestingly, none of the patients were reluctant to acetaminophen delivery by a nurse.21 Article Table 6. Results of the linear regressions for each item Y, according to the following formula: Yi=β0 + β1Agei + β2Genderi +β3Experiencei+β4Triagei + εi. Where Yi represents respondent i’s answer to how he/she agrees with statement Y, with answers ranging from 1=Totally disagree to 5=Totally agree. Variables 1 2 3 4 5 6 7 8 9 10 11 Age -0,01 0,04 0,07 0,04 -0,05 0.15** 0 0,06 -0,06 0.26*** 0 (0,08) (0,10) (0,11) (0,11) (0,09) (0,08) (0,07) (0,10) (0,09) (0,10) (0,12) Female gender -0,09 -0,13 0,02 -0.39** -0,05 -0,12 -0,03 -0,03 -0,1 -0,01 -0,11 (0,12) (0,14) (0,16) (0,16) (0,13) (0,11) (0,10) (0,14) (0,13) (0,14) (0,17) Years from degree 0 0,04 0,02 0,01 0,05 -0,03 -0,01 -0,07 -0,02 -0.16** -0.17** (0,06) (0,07) (0,08) (0,08) (0,06) (0,05) (0,05) (0,07) (0,06) (0,07) (0,08) Years working as a triage nurse -0,02 0 -0,1 0.11* 0,06 -0.13*** 0.09** 0 -0,03 -0,02 0,08 (0,05) (0,05) (0,06) (0,06) (0,05) (0,04) (0,04) (0,05) (0,05) (0,05) (0,06) Constant 4.53*** 3.61*** 3.41*** 2.91*** 3.81*** 1.71*** 3.76*** 2.43*** 2.22*** 2.01*** 3.09*** (0,20) (0,24) (0,26) (0,26) (0,22) (0,18) (0,16) (0,23) (0,22) (0,24) (0,28) Observations 259 260 257 257 257 260 259 260 260 259 259 R-squared 0 0,01 0,01 0,05 0,01 0,04 0,03 0,01 0,01 0,03 0,02 Standard errors in brackets. ***P<0.01, **P<0.05, *P<0.1. Non -co mmerc ial us e o nly [Emergency Care Journal 2018; 14:7787] [page 115] It seems, thus, that the expressed need for analgesia could be more reliable than NRS in initial management of acute pain in the ED. Moreover, it has been documented that after implementation of mandatory documentation regarding pain assessment in the ED, despite a significant increase of analgesics delivery, patients’ satis- faction with pain management did not significantly increase.22 We think that these data could at least in part rely on the differ- ent utilizations of NRS. In post-operative setting, for example, the tool is used to drive analgesic dosing and administration, while at ED’s triage NRS is mainly used to attribute priority code. This rep- resents, obviously, a substantial and biasing difference. The afore- mentioned skepticism has led to a widespread tendency to abandon NRS as a reliable tool in the US EDs.23 As a consequence, some Authors have recently argued that a pain protocol, based on admin- istration of analgesic agents, driven solely by patient response to a simple standardized question repeated at intervals, resulted in achievement of satisfactory analgesia in the vast majority of trau- matic as well as non-traumatic painful conditions, and could be preferable to the use of pain rating scales.24-27 Main limitations of this study are the relatively small sample and the fact that questionnaire deals with subjective perceptions. Otherwise, the concept of perception is declared yet in the title. Conclusions In conclusion, NRS confirms to be a potentially valuable tool for pain evaluation at triage time, but many nurses express some doubts on its reliability, and will attribute the triage code mainly basing on their own judgement. Further studies are needed to assess and validate new tools for pain measurement at triage time in the EDs. References 1. Chapman CR. Measurement of pain: problems and issues. In: Bonica JJ, Albe-Fessard D, eds. Advances in pain research and therapy, Vol. 1. New York: Raven Press, 1976. pp 345-353. 2. Jensen MP, Karoly P, O’Riordan EF, et al. The subjective expe- rience of acute pain. An assessment of the utility of 10 indices. Clin J Pain 1989;5:153-159. 3. Morley S. Psychology of pain. Br J Anaesth 2008;101:25-31. 4. Marco CA, Nagel J, Klink E, et al. Factors associated with self-reported pain scores among ED patients. Am J Emerg Med 2012;30:331-7. 5. Todd KH. Pain assessment instruments for use in the emergen- cy department. Emerg Med Clin N Am 2005;23:285-95. 6. Morone NE, Weiner DK. Pain as the 5th vital sign: exposing the vital need for pain education. Clin Ther 2013;35:1728-32. 7. Berthier F, Potel G, Leconte P, et al. Comparative study of methods of measuring acute pain intensity in an ED. Am J Emerg Med 1998;16:132-6. 8. Bijur PE, Latimer CT, Gallagher EJ. Validation of a verbally administered numerical rating scale of acute pain for use in the emergency department. Acad Emerg Med 2003;10:390-2. 9. Seymour RA. The use of pain scales in assessing the efficacy of analgesics in post-operative dental pain. Eur J Clin Pharmacol 1982;23:441-4. 10. Ekblom A, Hansson P. Pain intensity measurements in patients with acute pain receiving afferent stimulation. J Neurol Neurosurg Psychiatr 1988;51:481-6. 11. Wilkie D, Lovejoy N, Dodd M, et al. Cancer pain intensity measurement: concurrent validity of three tools – finger dynamometer, painintensity number scale, visual analogue scale. Hosp J 1990;6:1-13. 12. Paice J, Cohen F. Validity of a verbally administered numeric rating scale to measure cancer pain intensity. Cancer Nurs 1997;20:88-93. 13. Motov SM, Khan AN. Problems and barriers of pain manage- ment in the emergency department: Are we ever going to get better? J Pain Res 2008;2:5-11. 14. Todd KH, Ducharme J, Choiniere M, et al; PEMI Study Group. Pain in the emergency department: results of the pain and emergency medicine initiative (PEMI) multicenter study. J Pain 2007;8:460-6. 15. Craven P, Cinar O, Madsen T. Patient anxiety may influence the efficacy of ED pain management. Amer J Emerg Med 2013;31:313-8. 16. Riccardi A, De Iaco F, Guiddo G, et al. The use of a Wong baker scale in the assessment of acute pain in adults in the overcrowded environment of the ED. Ital J Emerg Med 2015;4:8-12. 17. Capponi R, Loguercio V, Guerrini S, et al. Does the numeric rating scale (NRS) represent the optimal tool for evaluating pain in the triage process of patients presenting to the ED? Results of a multicenter study. Acta Biomed 2016;87:347-52. 18. Ducharme J1, Tanabe P, Homel P, et al. The influence of triage systems and triage scores on timeliness of ED analgesic administration. Am J Emerg Med 2008;26:867-73. 19. Lindberg SO, Lerche la Cour J, Folkestad L, et al. The use of triage in Danish emergency departments. Dan Med Bull 2011;58:A4301. 20. Hangaard MH, Malling B, Mogensen CB. High inter-observer agreement of observer-perceived pain assessment in the emer- gency department. BMC Emerg Med 2018;18:8. 21. Lecomte F, Audureau E, Guyerdet V, et al. Patients in pain that refuse acetaminophen at triage. Am J Emerg Med 2014;32:388-9. 22. Sturesson L, Falk AC, Castrén M, et al. Mandatory documen- tation of pain in the emergency department increases analgesic administration but does not improve patients' satisfaction of pain management. Scand J Pain 2016;13:32-5. 23. Levy N, Sturgess J, Mills P. “Pain as the fifth vital sign” and dependence on the “numerical pain scale” is being abandoned in the US: Why? Brit J Anaesth 2018;120:435-8. 24. Chang AK, Bijur PE, Holden L, et al. Efficacy of an Acute pain titration protocol driven by patient response to a simple query: do you want more pain medication? Ann Emerg Med 2016;67:565-72. 25. Smith JE, Rockett M, Creanor S, et al. PAin SoluTions In the Emergency Setting (PASTIES) – patient controlled analgesia versus routine care in emergency department patients with non-traumatic abdominal pain: randomised trial. BMJ 2015;350:h3147. 26. Smith JE, Rockett M, Squire R, et al. PAin SoluTions In the Emergency Setting (PASTIES)-patient controlled analgesia versus routine care in emergency department patients with pain from traumatic injuries: randomised trial. BMJ 2015;350:h2988. 27. Rahman NH, DeSilva T. A randomized controlled trial of patient-controlled analgesia compared with boluses of analge- sia for the control of acute traumatic pain in the emergency department. J Emerg Med 2012;43:951-7. Article Non -co mmerc ial us e o nly