- Email: [email protected]
A Comparison of Three Self-Report Pain Scales in Adults With Acute Pain
The Journal of Emergency Medicine, Vol. -, No. -, pp. 1–9, 2014 Copyright Ó 2014 Elsevier Inc. Printed in the USA. All rights reserved 0736-4679/$ - see front matter
http://dx.doi.org/10.1016/j.jemermed.2014.07.039
Original Contributions
A COMPARISON OF THREE SELF-REPORT PAIN SCALES IN ADULTS WITH ACUTE PAIN Maryam Bahreini, MD,* Mohammad Jalili, MD,* and Maziar Moradi-Lakeh, MD† *Emergency Medicine Department, Tehran University of Medical Sciences, Tehran, Iran and †Iran University of Medical Sciences, Tehran, Iran Reprint Address: Mohammad Jalili, MD, Emergency Medicine Department, Tehran University of Medical Sciences, Keshavarz Boulevard, Tehran, Iran
, Abstract—Background: Several pain rating methods are used to quantify pain. Although these instruments have been extensively studied, their inter-agreement, especially in emergency department (ED) settings, has yet to be determined. Objective: This study was designed to assess the agreement between Visual Analog Scale (VAS), Color Analog Scale (CAS), and verbally administered Numeric Rating Scale (NRS) in the emergency setting. Methods: A sample of 150 adult patients presenting with acute pain to two EDs was recruited. Patients’ pain severity at presentation, 30 and 60 min later was assessed using the three pain scales. The agreement between pain scales was assessed using Bland-Altman method and Spearman correlation. We described a composite measure to serve as the gold standard and to be compared with each score. Factor analysis was also performed to assess the underlying construct. Results: Spearman correlation coefficients between NRS and CAS, NRS and VAS, and CAS and VAS were 0.95, 0.94, and 0.94, respectively (p < 0.001). On a scale of 0 to 10, the 95% limits of agreement between the paired NRS and VAS, VAS and CAS, and CAS and NRS as measured by Bland-Altman method ranged from 2.0 to 2.6, from 2.7 to 2.0, and from 2.1 to 2.0, respectively. The Kaiser Meyer Oklin measure of sampling adequacy was 0.785 and Bartlett’s test for sphericity was significant (p < 0.001). Conclusions: The three pain scales were strongly correlated at all time
periods. The findings suggest that NRS, CAS, and VAS can be interchangeably applied for acute pain measurement in adult patients. Ó 2014 Elsevier Inc. , Keywords—acute pain; pain assessment; emergency department; Visual Analog Scale; Numeric Rating Scale; Color Analog Scale
INTRODUCTION Pain is one of the most frequently encountered symptoms in the emergency department (ED), representing > 60% of all ED complaints. Systematic identification of patients in pain can improve the quality of pain management. That is why Joint Commission on Accreditation of Healthcare Organizations requires hospitals to routinely screen all patients for pain (1). In addition, because pain reduction or relief is an integral part of the emergency physicians’ practice, its accurate measurement has become a major area of focus of the emergency medicine research agenda. Because pain is an inherently subjective symptom, no objective method exists for its assessment. Although several pain measurement tools are currently available for use, pain measurement relies primarily on the patients’ subjective reports (2,3). The National Institutes of Health states that patient self-report is the most reliable indicator of the existence and intensity of pain (2). Emergency medicine literature abounds with
The Institutional Review Board approved the ethics of the study. The study was exempted from obtaining written informed consent but all participants provided their verbal assent.
RECEIVED: 2 February 2014; FINAL SUBMISSION RECEIVED: 6 June 2014; ACCEPTED: 1 July 2014 1
2
M. Bahreini et al.
numerous pain scoring systems for assessment of pain. These include, but is not limited to, the 100-mm Visual Analog Scale (VAS), the verbally administered Numeric Rating Scale (NRS), and the Color Analog Scale (CAS). VAS has been extensively proved to have excellent reliability and validity (3). It is thought to be superior to fixed-interval scales, relative-pain scales, and verbal reports of pain in the setting of chronic pain (4 8). NRS has also been validated as a scale of pain intensity in patients with pain and it is short, easy to administer and score, and provides multiple response options (5). In addition, no age-related difficulties exist (9,10). CAS is also considered to have excellent reliability and validity. It has been validated in children aged older than 5 years for acute, recurrent, or chronic pain, and in the ED setting (3,11–13). It is not clear how the results obtained from different tools are comparable and which one is superior to the others. Several studies have compared different instruments for pain-intensity measurement, mostly in chronic or postoperative pain (2,9,14 17). Despite abundance of the evidence in chronic and postoperative pain settings, the studies comparing these pain scales in acute pain assessment in the ED are scarce. Even when acute pain measurement is the focus of study, special populations, such as pediatric, elderly, and the cognitively impaired patients have been investigated (12,18). Besides agreement between pain scales, patient perception of the convenience of each instrument can also be considered an important aspect and, therefore, is a simple important objective that has been considered in only a few studies (9,17). This study was designed to determine the correlation and agreement between pain scores obtained using three different pain scales (NRS, VAS, and CAS) in adult patients presenting to ED with acute pain. We also tried to find out which pain scale was preferred by patients. METHODS Study Design The Institutional Review Board approved the ethics of the study. The study was exempted from obtaining written informed consent but all participants provided their verbal assent. This was a cross-sectional study designed to compare the results of pain assessment using three different selfreport scales. Participant enrollment took place in the ED of two university-affiliated tertiary care urban centers, each with an annual census of > 40,000 visits. All shifts (day, evening, and night) and all days of the week were included. The EDs are open 24/7 and are staffed by emergency medicine (EM) trained faculty and EM residents.
Study Setting and Population All adult patients (18 years of age or older) who presented to the ED with acute pain of any degree from May 2012 to September 2012 were eligible for the study. All patients exhibiting altered mental status or any degree of cognitive impairment, anyone who could not understand explanations and commands given in Persian language (including deaf patients or those with significant auditory problems making explanations of the pain scales impractical), or patients whose pain was relieved at the onset of study, as well as noncooperative patients and those who were unwilling to give consent, were excluded. Exclusion criteria also included significant visual problems (for the reason of difficulty marking on the paper), and motor abnormalities that precluded scoring of the scales. Study Protocol A convenience sample of patients was recruited when one of the researchers (MB) was in the ED on clinical shift and all data was collected by the same researcher. She evaluated potential participation of the patients and then approached eligible patients and explained the purpose and nature of the study. All participants provided verbal consent, and the study protocol was then started. Patient age, sex, education, pain location (head and neck, torso, limb, or generalized), and type of pain (traumatic or nontraumatic) were documented. Pain scores were recorded regardless of the administration of analgesic. All participants obtained analgesic as needed according to the current ED protocols or at the discretion of the attending physician caring for the patient. The receipt (or failure to receive) of analgesic was not recorded. Patients’ pain severity was assessed using three pain scales: VAS, NRS, and CAS. The VAS is a 100-mm horizontally positioned line with anchor points of ‘‘no pain’’ and ‘‘worst possible pain’’ at each end. The CAS is a 100-mm-long triangular shape. Its width ranges between 10 mm at the lower anchor ‘‘no pain’’ and 30 mm at the upper anchor ‘‘worst possible pain.’’ CAS is also color graduated from white (lower anchor) to dark red (upper anchor). The following questions were asked verbatim from the patient: for VAS: ‘‘Show me the level of your pain on the line: Here is no pain and there is the worst possible pain’’; for verbally administered NRS: ‘‘Tell me on a scale of 0 to 10 what is the level of your pain, 0 representing no pain and 10 is the worst possible pain’’; for CAS: ‘‘Show me on the ruler the intensity of the red color corresponding to the level of your pain: here is no pain and there is the worst possible pain.’’ For VAS, the patient crossed a line to indicate his or her level of pain, and the distance between the ‘‘no pain’’ anchor to the point marked by the patient was
Self-Report Pain Scales in the ED
3
Table 1. Incorrect Responses and the Protocol for Remedial Actions Incorrect Response
Definition/Example
Nonexistent responses
Extended beyond the numbers allowed in the NRS Between two appropriate responses Response falling between two appropriate responses, such as the reproduced score placed between 2 words of CAS Two answers on the same scale Range of responses
Indicated a range of responses rather than providing a single numerical estimate
Remedial Action The responses were omitted and served as missing data Ultimate response would be the higher pain score before which the pain’s mark existed Ultimate response would be the higher pain score before which the patient’s mark existed This incorrect response was managed by further explanation of the pain scale using the same verbatim.
CAS = Color Analog Scale; NRS = Numeric Rating Scale.
the measure of the patient’s pain intensity. For CAS, the patient marked the color associated with his or her level of pain and the line was reproduced by the physician on the ruler and measured later. The order of administration of pain rating scales was randomly varied among patients using a 6 6 array Latin Square design. The three scales were offered to patients separately, with 15 to 20 s between each assessment. The same procedure was repeated and the scores were obtained on two other occasions at 30 and 60 min post baseline in the same order by the same physician. Patients were blinded to their prior assessment(s) and were not reminded of their previous response(s). Incorrect responses and the remedial actions are presented in Table 1. If the pain was relieved in the second or third episode of the questioning, a ‘‘0’’ was documented and the study stopped. Finally, patients were surveyed regarding the pain scale with which they preferred to self-report their pain intensity. Measures We entered 150 patients to the study and some of them were assessed for the second and third times. In total, 404 pain measurements were obtained. The pain scores obtained on different occasions, from the same participant, were considered to be from different patients. Outcome Our study outcome was to determine the equivalence of these methods of pain measurement in adult patients who are in acutely painful conditions. Data Analysis All data were entered into an Excel spreadsheet program (Microsoft, Redmond, WA) and exported to SPSS software (version 11.5, SPSS Inc., Chicago, IL), Stata software (version 9.1, StataCorp, College Station, TX), and MedCalc software (version 12.1.4.0, MedCalc Software,
Mariakerke, Belgium) for analysis. The precision of estimates was expressed with 95% confidence intervals (CIs). Considering that the pain scores did not have a normal distribution and the scales were ordinal, we calculated two-tailed Spearman r correlation coefficients to assess the correlation between them. To assess the agreement between measures, the VAS and CAS scores were regressed on NRS, and the CAS scores were also regressed on VAS scores. The difference between each paired scores, the y-intercepts and slopes were determined. However, it should be noted that a high correlation does not necessarily imply that there is good agreement between methods. To overcome this problem and evaluate the agreement between the three measures and to determine the bias and the limits of agreement between the scales, a series of analyses were conducted using Bland Altman method. The differences between each paired score, and the interval that included 95% of these differences, were calculated. We established a priori the maximum limit of agreement at 6 20 mm (the repeatability coefficient for test-retest of the same scale in an adult) (12). To evaluate the hypothesis that the three instruments are related and measure practically a single dimension, we conducted a factor analysis under the assumption of a common factor underlying all the variables (ie, rating scales). We used Kaiser Meyer Oklin (KMO) measures of sampling adequacy and Bartlett’s test of sphericity to assess suitability of factor analysis with our data. Because none of the scores could be assumed as the reference standard, we analyzed the three scores of each patient on each episode by means of principal component analysis and specified a composite score as ‘‘the most representative score’’ of pain severity for that particular patient on that particular occasion. We then compared each patient’s individual pain score on each scale in each occasion with this composite score and their correlation was assessed. In addition, to assess the association between the measures, these composite scores were then regressed on each of the scores (VAS,
4
M. Bahreini et al.
Figure 1. Flow of participants in the study. ED = emergency department.
NRS, and CAS) separately. If the measures were equivalent, we would expect a y-intercept of 0 and a slope of 1. Finally, to evaluate the effect of extreme scores on the data analysis, an additional subset of data composed of scores (excluding 0 and 10 scores) was also assessed separately. This ensured that the correlation coefficients obtained are not inflated by these extreme scores.
RESULTS Characteristics of Study Subjects Two hundred and forty-two patients were approached and assessed for eligibility. A total of 157 (65%) met all inclusion criteria and were invited to participate in this study. Of those, 7 patients refused to
Self-Report Pain Scales in the ED
5
Table 2. Demographic and Pain-Related Characteristics of Participants (n = 150) Characteristics Age (years), mean 6 SD Sex (%) Female Male Education (%) Illiterate Primary school Secondary school Higher education Pain location (%) Head/neck Torso Abdomen Limb Generalized pain Pain etiology (%) Traumatic Nontraumatic
32.2 6 16.3 51.3 48.7 16.6 13.3 38.0 32.0 17.3 18.7 30.7 32.0 1.3 32.0 68.0
participate and were excluded; therefore, data were collected on 150 eligible participants who consented. All 150 patients rated their pain at baseline, 142 patients at 30 min post baseline, and 112 patients at 60 min post baseline. This drop in number was because some patients were either discharged or left the ED for further workup (eg, sent to the radiology suite) before the next occasion of assessment. Overall, pain assessment was performed on 404 occasions, each time with a set of three randomly administered scales. Of the 150 patients, 53 were asked the NRS first and 51 were asked the VAS first, and the rest answered the CAS first. Flow of participants in the study is depicted in Figure 1. Demographic and Pain-Related Descriptive Statistics Table 2 summarizes descriptive data on the characteristics of the patients included in the study. On the basis of one-sample Kolmogorov-Smirnov Test, the pain scores did not have a normal distribution. The range of pain measured was 0 to 10, but the sample consisted primarily of patients with initially moderate to high pain levels. Mean and standard deviations of the pain scores are summarized in Table 3.
Main Results Factor analysis. The KMO measure of sampling adequacy was 0.785 and Bartlett’s test for sphericity was significant (p < 0.001). Extraction communalities were 0.949, 0.960, and 0.961 for VAS, CAS, and NRS, respectively. Only one component had Eigen value > 1 (2.87). The variance explained by the first component was 95.68%. Correlations. The Spearman r revealed a strong correlation between NRS, VAS, and CAS with correlation coefficients between NRS and CAS, NRS and VAS, and CAS and VAS being 0.95, 0.94, and 0.94, respectively (p < 0.001). The regression lines derived from these data show that the slopes of the lines are not 1. There is a strong correlation between the NRS and VAS (r = 0.90 and p < 0.001). The y-intercept was 0.12 and the slope was 0.95 (95% CI 0.83–1.07). For the CAS and VAS the y-intercept was 0.13, with a slope of 0.94 (95% CI 0.81–1.07), and a strong correlation (r = 0.89 and p < 0.0001) existed. NRS and CAS were also correlated (r = 0.91 and p < 0.0001) and the y-intercept was 0.48, with a slope of 0.92 (95% CI 0.44–1.40). All of the correlations were statistically significant (p < 0.001). This finding indicates that although the scores of the three scales correlate well, they are not equivalent. Bland-Altman. The mean differences and the limits of agreement between the three methods of pain measurement are presented in Table 4. The Bland-Altman graphs are presented in Figures 2, 3, and 4. Based on the analysis, the following can be inferred: The limits of agreement between the paired NRS and VAS scores extended from 2.0 to 2.6 around the mean difference between the two measures. Also, the interval between VAS and CAS and that for CAS and NRS were 2.7 to 2.0 and 2.0 to 2.1, respectively (Table 4). The correlation coefficients between NRS, VAS, and CAS scores and the composite measure, regarded as gold standard, were 0.98, 0.97, and 0.98, respectively, which indicates strong correlations. The Spearman correlation coefficients were 0.90, 0.88, and 0.87 assessing the correlation between NRS and CAS, NRS and VAS, and CAS and VAS in data set devoid of 0 and 10, respectively. The mean differences,
Table 3. Pain Intensity Scores for Each of the Measures Rating Scales NRS CAS VAS
Baseline (n = 150)
30 min (n = 142)
60 min (n = 112)
Total Sample (N = 404)
5.0 (4.0–8.0) 6.0 (3.7–8.0) 5.1 (2.6–7.6)
4.0 (1.0–6.2) 4.0 (1.0–6.2) 3.1 (1.0–6.8)
1.0 (0.0–5.0) 1.0 (0.0–5.0) 1.0 (0.0–3.7)
4.0 (1.2–7.0) 4.0 (2.0–7.0) 3.3 (1.0–6.8)
CAS = Color Analog Scale; NRS = Numerical Rating Scale; VAS = Visual Analog Scale. Data are median (interquartile range).
6
M. Bahreini et al.
Table 4. Mean Difference and Limits of Agreement According to the Bland-Altman Method Limits of Agreement (95% CI) Scale VAS/CAS VAS/NRS CAS/NRS
Mean Difference (95% CI) 0.4 ( 0.5 to 0.2) 0.3 (0.1 to 0.4) 0.1 ( 0.2 to 0.0)
Lower Limit 2.7 ( 2.9 to 2.0 ( 2.3 to 2.1 ( 2.3 to
Upper Limit 2.4) 1.9) 1.9)
2.0 (1.7 to 2.1) 2.6 (2.5 to 3.0) 2.0 (1.8 to 2.2)
CAS = Color Analog Scale; CI = confidence interval; NRS = Numerical Rating Scale; VAS = Visual Analog Scale.
lower and upper limits of agreement according to BlandAltman method were 0.1 and 2.0 to 2.1, 0.3 and 2.0 to 2.6 and the latter 0.4 and 2.0 to 2.7, between NRS and CAS, NRS and VAS, and CAS and VAS in data set without 0 and 10, respectively.
that, when the tools are applied to adult patients with acute pain in the ED setting, there is a strong correlation between pain scores obtained using any of the three.
Patient satisfaction. Patients were questioned about the preferred pain scale to identify their pain. Overall, 38% of the participants mentioned no difference between the scales. Among the 62% who did notice a difference, CAS was the preferred scale (32% of total) to score pain. In the illiterate group, most patients (60%) stated that there was no difference between the three pain scales. Interestingly, none of the patients in this group chose VAS as the most preferable. In addition, although 32.5% of women preferred VAS, 34.2% of men mentioned no difference between the pain scales.
There are many different instruments for measuring pain intensity in the ED. The degree of agreement among these self-report scales is controversial. In the present study, we compared the performance and comparability of the three pain rating scales commonly used in ED practice (ie, VAS, CAS, and NRS). The findings of this study proved
In our study, a strong correlation was found between NRS and VAS (r = 0.94). This is exactly the same as the finding of the study by Bijur et al., conducted on 297 measurement in adult patients in the ED (r = 0.94) (19). Other investigators concluded that verbally administered NRS can be substituted for VAS in acute pain measurement (20,21). Also, a French study on a sample of 90 ED nontrauma patients and a sample of 200 patients with trauma gave similar results (r = 0.91 and r = 0.80, respectively) (22). Avery conducted the aforementioned study on 100 ED patients and found a correlation coefficient of 0.89 between NRS and VAS (23). Earlier studies in other settings, such as postoperative pain, have also yielded similar results (15,17,24). We also found that the pain scores obtained by any of the three pain rating scales and the composite measure considered as the gold standard were highly correlated. To the best of our knowledge, no previous study has used this method to assess the validity of the measurements. In addition, assuming that patients reporting their pain as maximum
Figure 2. Bland-Altman plot of Visual Analog Scale (VAS) and Color Analog Scale (CAS).
Figure 3. Bland-Altman plot of Numeric Rating Scale (NRS) and Color Analog Scale (CAS).
DISCUSSION
Correlations
Self-Report Pain Scales in the ED
7
Patient Preference
Figure 4. Bland-Altman plot of Numeric Rating Scale (NRS) and Visual Analog Scale (VAS).
or minimum would score similarly with all pain scales and to avoid the possibility of exaggerated correlation between scores at the extremis (0 and 10), we repeated correlation analyses using a data subset, devoid of 0 and 10 scores. Even this conservative analysis did not change the results of the study. Again, we did not find any studies that have adapted this approach. The sole reliance on correlation coefficients is misleading and cannot be used to decide whether one instrument performs superior to another. Agreement The analysis of agreement between the three scales indicated that the three pain scales evaluated are in acceptable agreement to measure pain intensity because the limit of agreement as determined by the Bland-Altman method fell close to the repeatability coefficient determined for adults ( 6 20 mm). Bailey et al. assessed the agreement between VAS, CAS, Verbal Numeric Scale, and Wong-Baker FACES Pain Rating Scale in a pediatric population with acute abdominal pain and concluded that only CAS and VAS were in agreement (12). Performing a regression analysis, we assumed that if y-intercept equals 0, the two scales were equivalent, so that the predicted value of VAS at NRS or CAS score of 0 would be 0, and every unit increase in NRS or CAS would match with a 1cm increase in VAS. The results of our study suggest that there was a small difference, indicating that the measures were closely correlated but not entirely equivalent. The factor analysis of the scoring properties of pain rating instruments suggested a high degree of association between the variables.
An important factor to consider when choosing among rating scales is their ease of administration and patients’ comfort reporting pain. In our study, we assessed patients’ preference for different pain rating scales by asking them a direct question. Although the largest group of patients (38%) stated that there was no difference between the pain scales, the CAS was preferred by a relatively large proportion of patients (32%). Surprisingly, none of the uneducated patients sustaining acute pain mentioned VAS as the instrument of choice for quantifying pain. Several studies have also dealt with this question, but all of them have been undertaken in settings other than the ED (14,16,25 29). These studies have shown that the Verbal Rating Scale (VRS) was preferred by the less educated and the elderly, and the NRS was the instrument of choice in an age-mixed population in chronic pain patients and in head and neck cancer pati ents (9,14,16,22,25,26,28,29). None of these studies, however, used CAS, and our study did not include VRS. In summary, the strong correlation between verbally administered NRS, VAS, and CAS and the close interscale agreement obtained supports their interchangeable application in acute adult pain setting. Limitations Our study faces some limitations that should be acknowledged. Firstly, the three pain ratings were obtained almost simultaneously. Although the order was randomized, the degree of agreement between the measures may have been overestimated. Secondly, one fifth of the measurements were repetitions that were obtained from the same patient at different occasions. These were considered independent observations for the purpose of this study. This, however, could have biased the results. In this study, we assessed patients’ perspectives about the three pain scales. It will also be of great value to determine the pain score most easily administered from the physicians’ viewpoint. This was not possible in our study because one researcher obtained all of the data. CONCLUSIONS The strong correlation between verbally administered NRS, VAS, and CAS and the close inter-scale agreement obtained support their interchangeable application in acute adult pain setting. Acknowledgments—The authors appreciate Ahmad Bahreini for data management. We would also like to extend our gratitude to
8
M. Bahreini et al.
Dr. Akbar Fotouhi and Dr. Nourieh Sharifi for their insightful critique and suggestions.
REFERENCES 1. Joint Commission on Accreditation of Healthcare Organizations. 2006 Comprehensive Accreditation Manual for Hospitals: The Official Handbook. Oakbrook Terrace, IL: Joint Commission on Accreditation of Healthcare Organizations; 2006. 2. Ferraz MB, Quaresma MR, Aquino LR, Atra E, Tugwell P, Goldsmith CH. Reliability of pain scales in the assessment of literate and illiterate patients with rheumatoid arthritis. J Rheumatol 1990;17:1022–4. 3. O’ Rourke D. The measurement of pain in infants, children, and adolescents: from policy to practice. Phys Ther 2004;84:560–70. 4. Carlsson AM. Assessment of chronic pain. I. Aspects of the reliability and validity of the visual analog scale. Pain 1983;16:87–101. 5. Jensen MP, Turner JA, Romano JM, Fisher LD. Comparative reliability and validity of chronic pain intensity measures. Pain 1999; 83:157–62. 6. Joyce CRB, Autshi DW, Mason RM. Comparison of fixed interval and visual analog scales for rating chronic pain. Eur J Clin Pharmacol 1975;8:415–20. 7. McCormack HM, Home DJ, Sheather S. Clinical applications of visual analog scales: a critical review. Psychol Med 1988;10:1007–19. 8. Ohnhaus EE, Adler R. Methodological problems in the measurement of pain a comparison between the verbal rating scale and the visual analog scale. Pain 1975;1:379–84. 9. Jensen MP, Karoly P, Braver S. The measurement of clinical pain intensity: a comparison of six methods. Pain 1986;27:117–26. 10. Krebs EE, Carey TS, Weinberger M. Accuracy of the pain numeric rating scale as a screening test in primary care. J Gen Intern Med 2007;22:1453–8. 11. Hain R. Pain scales in children: a review. Palliat Med 1997;11: 341–50. 12. Bailey B, Bergeron S, Gravel J, Daoust R. Comparison of four pain scales in children with acute abdominal pain in a pediatric emergency department. Ann Emerg Med 2007;50:379–83. 383.e1-2. 13. Bulloch B, Tenenbein M. Validation of 2 pain scales for use in the pediatric emergency department. Pediatrics 2002;110:e33. 14. Clark P, Lavielle P, Martinez H. Learning from pain scales: patient perspective. J Rheumatol 2003;30:1584–8. 15. DeLoach LJ, Higgins MS, Caplan AB, Stiff JL. The visual analog scale in the immediate postoperative period: intrasubject vari-
16.
17. 18. 19. 20. 21. 22. 23.
24.
25. 26.
27. 28. 29.
ability and correlation with a numeric scale. Anesth Analg 1998; 86:102–6. Peters ML, Patijn J, Lame I. Pain assessment in younger and older pain patients: psychometric properties and patient preference of five commonly used measures of pain intensity. Pain Med 2007; 8:601–10. Breivik EK, Bjo¨rnsson GA, Skovlund E. A comparison of pain rating scales by sampling from clinical trial data. Clin J Pain 2000;16:22–8. Ahlers SJ, van Gulik L, van der Veen AM, et al. Comparison of different pain scoring systems in critically ill patients in a general ICU. Crit Care 2008;12:R15. 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. Ekblom A, Hansson P. Pain intensity measurements in patients with acute pain receiving afferent stimulation. J Neurol Neurosurg Psychiatry 1988;51:481–6. Daoust R, Beaulieu P, Manzini C, Chauny JM, Lavigne G. Estimation of pain intensity in emergency medicine: a validation study. Pain 2008;138:565–70. Berthier F, Potel G, Leconte P, Touze MD, Baron D. Comparative study of methods of measuring acute pain intensity in an ED. Am J Emerg Med 1998;16:132–6. Avery VR. Comparing the Visual Analog Scale and Verbally Administered Numeric Rating Scale in Traumatic versus NonTraumatic Pain in a Community Hospital Emergency Center. Master’s Project. Toledo, Ohio: University of Toledo; 2012. Hjermstad MJ, Fayers PM, Haugen DF, et al. European Palliative Care Research Collaborative (EPCRC). Studies comparing Numerical Rating Scales, Verbal Rating Scales, and Visual Analogue Scales for assessment of pain intensity in adults: a systematic literature review. J Pain Symptom Manage 2011;41:1073–93. Herr KA, Mobily PR. Comparison of selected pain assessment tools for use with the elderly. Appl Nurs Res 1993;6:39–46. Herr KA, Spratt K, Mobily PR, Richardson G. Pain intensity assessment in older adults: use of experimental pain to compare psychometric properties and usability of selected pain scales with younger adults. Clin J Pain 2004;20:207–19. Kremer E, Atkinson JH, Ignelzi RJ. Measurement of pain: patient preference does not confound pain measurement. Pain 1981;10: 241–8. Rodriguez CS, McMillan S, Yarandi H. Pain measurement in older adults with head and neck cancer and communication impairments. Cancer Nurs 2004;27:425–33. Williams AC, Davies HTO, Chadury Y. Simple pain rating scales hide complex idiosyncratic meanings. Pain 2000;85:457–63.
Self-Report Pain Scales in the ED
ARTICLE SUMMARY 1. Why is this topic important? Pain represents nearly 60% of all emergency patient complaints. Accurate measurement of pain can enhance the quality of patient management. 2. What does this study attempt to show? The goal of this study was to assess correlation and agreement between the Visual Analog Scale (VAS), Color Analog Scale (CAS), and Numeric Rating Scale (NRS) adult patients in the emergency department. 3. What are the key findings? There was strong correlation and close inter-scale agreement between verbally administered NRS, VAS, and CAS. 4. How is patient care impacted? Patient perception of the convenience of each instrument was identified. This study assessed various pain rating scales for better pain reduction or relief.
9
http://dx.doi.org/10.1016/j.jemermed.2014.07.039
Original Contributions
A COMPARISON OF THREE SELF-REPORT PAIN SCALES IN ADULTS WITH ACUTE PAIN Maryam Bahreini, MD,* Mohammad Jalili, MD,* and Maziar Moradi-Lakeh, MD† *Emergency Medicine Department, Tehran University of Medical Sciences, Tehran, Iran and †Iran University of Medical Sciences, Tehran, Iran Reprint Address: Mohammad Jalili, MD, Emergency Medicine Department, Tehran University of Medical Sciences, Keshavarz Boulevard, Tehran, Iran
, Abstract—Background: Several pain rating methods are used to quantify pain. Although these instruments have been extensively studied, their inter-agreement, especially in emergency department (ED) settings, has yet to be determined. Objective: This study was designed to assess the agreement between Visual Analog Scale (VAS), Color Analog Scale (CAS), and verbally administered Numeric Rating Scale (NRS) in the emergency setting. Methods: A sample of 150 adult patients presenting with acute pain to two EDs was recruited. Patients’ pain severity at presentation, 30 and 60 min later was assessed using the three pain scales. The agreement between pain scales was assessed using Bland-Altman method and Spearman correlation. We described a composite measure to serve as the gold standard and to be compared with each score. Factor analysis was also performed to assess the underlying construct. Results: Spearman correlation coefficients between NRS and CAS, NRS and VAS, and CAS and VAS were 0.95, 0.94, and 0.94, respectively (p < 0.001). On a scale of 0 to 10, the 95% limits of agreement between the paired NRS and VAS, VAS and CAS, and CAS and NRS as measured by Bland-Altman method ranged from 2.0 to 2.6, from 2.7 to 2.0, and from 2.1 to 2.0, respectively. The Kaiser Meyer Oklin measure of sampling adequacy was 0.785 and Bartlett’s test for sphericity was significant (p < 0.001). Conclusions: The three pain scales were strongly correlated at all time
periods. The findings suggest that NRS, CAS, and VAS can be interchangeably applied for acute pain measurement in adult patients. Ó 2014 Elsevier Inc. , Keywords—acute pain; pain assessment; emergency department; Visual Analog Scale; Numeric Rating Scale; Color Analog Scale
INTRODUCTION Pain is one of the most frequently encountered symptoms in the emergency department (ED), representing > 60% of all ED complaints. Systematic identification of patients in pain can improve the quality of pain management. That is why Joint Commission on Accreditation of Healthcare Organizations requires hospitals to routinely screen all patients for pain (1). In addition, because pain reduction or relief is an integral part of the emergency physicians’ practice, its accurate measurement has become a major area of focus of the emergency medicine research agenda. Because pain is an inherently subjective symptom, no objective method exists for its assessment. Although several pain measurement tools are currently available for use, pain measurement relies primarily on the patients’ subjective reports (2,3). The National Institutes of Health states that patient self-report is the most reliable indicator of the existence and intensity of pain (2). Emergency medicine literature abounds with
The Institutional Review Board approved the ethics of the study. The study was exempted from obtaining written informed consent but all participants provided their verbal assent.
RECEIVED: 2 February 2014; FINAL SUBMISSION RECEIVED: 6 June 2014; ACCEPTED: 1 July 2014 1
2
M. Bahreini et al.
numerous pain scoring systems for assessment of pain. These include, but is not limited to, the 100-mm Visual Analog Scale (VAS), the verbally administered Numeric Rating Scale (NRS), and the Color Analog Scale (CAS). VAS has been extensively proved to have excellent reliability and validity (3). It is thought to be superior to fixed-interval scales, relative-pain scales, and verbal reports of pain in the setting of chronic pain (4 8). NRS has also been validated as a scale of pain intensity in patients with pain and it is short, easy to administer and score, and provides multiple response options (5). In addition, no age-related difficulties exist (9,10). CAS is also considered to have excellent reliability and validity. It has been validated in children aged older than 5 years for acute, recurrent, or chronic pain, and in the ED setting (3,11–13). It is not clear how the results obtained from different tools are comparable and which one is superior to the others. Several studies have compared different instruments for pain-intensity measurement, mostly in chronic or postoperative pain (2,9,14 17). Despite abundance of the evidence in chronic and postoperative pain settings, the studies comparing these pain scales in acute pain assessment in the ED are scarce. Even when acute pain measurement is the focus of study, special populations, such as pediatric, elderly, and the cognitively impaired patients have been investigated (12,18). Besides agreement between pain scales, patient perception of the convenience of each instrument can also be considered an important aspect and, therefore, is a simple important objective that has been considered in only a few studies (9,17). This study was designed to determine the correlation and agreement between pain scores obtained using three different pain scales (NRS, VAS, and CAS) in adult patients presenting to ED with acute pain. We also tried to find out which pain scale was preferred by patients. METHODS Study Design The Institutional Review Board approved the ethics of the study. The study was exempted from obtaining written informed consent but all participants provided their verbal assent. This was a cross-sectional study designed to compare the results of pain assessment using three different selfreport scales. Participant enrollment took place in the ED of two university-affiliated tertiary care urban centers, each with an annual census of > 40,000 visits. All shifts (day, evening, and night) and all days of the week were included. The EDs are open 24/7 and are staffed by emergency medicine (EM) trained faculty and EM residents.
Study Setting and Population All adult patients (18 years of age or older) who presented to the ED with acute pain of any degree from May 2012 to September 2012 were eligible for the study. All patients exhibiting altered mental status or any degree of cognitive impairment, anyone who could not understand explanations and commands given in Persian language (including deaf patients or those with significant auditory problems making explanations of the pain scales impractical), or patients whose pain was relieved at the onset of study, as well as noncooperative patients and those who were unwilling to give consent, were excluded. Exclusion criteria also included significant visual problems (for the reason of difficulty marking on the paper), and motor abnormalities that precluded scoring of the scales. Study Protocol A convenience sample of patients was recruited when one of the researchers (MB) was in the ED on clinical shift and all data was collected by the same researcher. She evaluated potential participation of the patients and then approached eligible patients and explained the purpose and nature of the study. All participants provided verbal consent, and the study protocol was then started. Patient age, sex, education, pain location (head and neck, torso, limb, or generalized), and type of pain (traumatic or nontraumatic) were documented. Pain scores were recorded regardless of the administration of analgesic. All participants obtained analgesic as needed according to the current ED protocols or at the discretion of the attending physician caring for the patient. The receipt (or failure to receive) of analgesic was not recorded. Patients’ pain severity was assessed using three pain scales: VAS, NRS, and CAS. The VAS is a 100-mm horizontally positioned line with anchor points of ‘‘no pain’’ and ‘‘worst possible pain’’ at each end. The CAS is a 100-mm-long triangular shape. Its width ranges between 10 mm at the lower anchor ‘‘no pain’’ and 30 mm at the upper anchor ‘‘worst possible pain.’’ CAS is also color graduated from white (lower anchor) to dark red (upper anchor). The following questions were asked verbatim from the patient: for VAS: ‘‘Show me the level of your pain on the line: Here is no pain and there is the worst possible pain’’; for verbally administered NRS: ‘‘Tell me on a scale of 0 to 10 what is the level of your pain, 0 representing no pain and 10 is the worst possible pain’’; for CAS: ‘‘Show me on the ruler the intensity of the red color corresponding to the level of your pain: here is no pain and there is the worst possible pain.’’ For VAS, the patient crossed a line to indicate his or her level of pain, and the distance between the ‘‘no pain’’ anchor to the point marked by the patient was
Self-Report Pain Scales in the ED
3
Table 1. Incorrect Responses and the Protocol for Remedial Actions Incorrect Response
Definition/Example
Nonexistent responses
Extended beyond the numbers allowed in the NRS Between two appropriate responses Response falling between two appropriate responses, such as the reproduced score placed between 2 words of CAS Two answers on the same scale Range of responses
Indicated a range of responses rather than providing a single numerical estimate
Remedial Action The responses were omitted and served as missing data Ultimate response would be the higher pain score before which the pain’s mark existed Ultimate response would be the higher pain score before which the patient’s mark existed This incorrect response was managed by further explanation of the pain scale using the same verbatim.
CAS = Color Analog Scale; NRS = Numeric Rating Scale.
the measure of the patient’s pain intensity. For CAS, the patient marked the color associated with his or her level of pain and the line was reproduced by the physician on the ruler and measured later. The order of administration of pain rating scales was randomly varied among patients using a 6 6 array Latin Square design. The three scales were offered to patients separately, with 15 to 20 s between each assessment. The same procedure was repeated and the scores were obtained on two other occasions at 30 and 60 min post baseline in the same order by the same physician. Patients were blinded to their prior assessment(s) and were not reminded of their previous response(s). Incorrect responses and the remedial actions are presented in Table 1. If the pain was relieved in the second or third episode of the questioning, a ‘‘0’’ was documented and the study stopped. Finally, patients were surveyed regarding the pain scale with which they preferred to self-report their pain intensity. Measures We entered 150 patients to the study and some of them were assessed for the second and third times. In total, 404 pain measurements were obtained. The pain scores obtained on different occasions, from the same participant, were considered to be from different patients. Outcome Our study outcome was to determine the equivalence of these methods of pain measurement in adult patients who are in acutely painful conditions. Data Analysis All data were entered into an Excel spreadsheet program (Microsoft, Redmond, WA) and exported to SPSS software (version 11.5, SPSS Inc., Chicago, IL), Stata software (version 9.1, StataCorp, College Station, TX), and MedCalc software (version 12.1.4.0, MedCalc Software,
Mariakerke, Belgium) for analysis. The precision of estimates was expressed with 95% confidence intervals (CIs). Considering that the pain scores did not have a normal distribution and the scales were ordinal, we calculated two-tailed Spearman r correlation coefficients to assess the correlation between them. To assess the agreement between measures, the VAS and CAS scores were regressed on NRS, and the CAS scores were also regressed on VAS scores. The difference between each paired scores, the y-intercepts and slopes were determined. However, it should be noted that a high correlation does not necessarily imply that there is good agreement between methods. To overcome this problem and evaluate the agreement between the three measures and to determine the bias and the limits of agreement between the scales, a series of analyses were conducted using Bland Altman method. The differences between each paired score, and the interval that included 95% of these differences, were calculated. We established a priori the maximum limit of agreement at 6 20 mm (the repeatability coefficient for test-retest of the same scale in an adult) (12). To evaluate the hypothesis that the three instruments are related and measure practically a single dimension, we conducted a factor analysis under the assumption of a common factor underlying all the variables (ie, rating scales). We used Kaiser Meyer Oklin (KMO) measures of sampling adequacy and Bartlett’s test of sphericity to assess suitability of factor analysis with our data. Because none of the scores could be assumed as the reference standard, we analyzed the three scores of each patient on each episode by means of principal component analysis and specified a composite score as ‘‘the most representative score’’ of pain severity for that particular patient on that particular occasion. We then compared each patient’s individual pain score on each scale in each occasion with this composite score and their correlation was assessed. In addition, to assess the association between the measures, these composite scores were then regressed on each of the scores (VAS,
4
M. Bahreini et al.
Figure 1. Flow of participants in the study. ED = emergency department.
NRS, and CAS) separately. If the measures were equivalent, we would expect a y-intercept of 0 and a slope of 1. Finally, to evaluate the effect of extreme scores on the data analysis, an additional subset of data composed of scores (excluding 0 and 10 scores) was also assessed separately. This ensured that the correlation coefficients obtained are not inflated by these extreme scores.
RESULTS Characteristics of Study Subjects Two hundred and forty-two patients were approached and assessed for eligibility. A total of 157 (65%) met all inclusion criteria and were invited to participate in this study. Of those, 7 patients refused to
Self-Report Pain Scales in the ED
5
Table 2. Demographic and Pain-Related Characteristics of Participants (n = 150) Characteristics Age (years), mean 6 SD Sex (%) Female Male Education (%) Illiterate Primary school Secondary school Higher education Pain location (%) Head/neck Torso Abdomen Limb Generalized pain Pain etiology (%) Traumatic Nontraumatic
32.2 6 16.3 51.3 48.7 16.6 13.3 38.0 32.0 17.3 18.7 30.7 32.0 1.3 32.0 68.0
participate and were excluded; therefore, data were collected on 150 eligible participants who consented. All 150 patients rated their pain at baseline, 142 patients at 30 min post baseline, and 112 patients at 60 min post baseline. This drop in number was because some patients were either discharged or left the ED for further workup (eg, sent to the radiology suite) before the next occasion of assessment. Overall, pain assessment was performed on 404 occasions, each time with a set of three randomly administered scales. Of the 150 patients, 53 were asked the NRS first and 51 were asked the VAS first, and the rest answered the CAS first. Flow of participants in the study is depicted in Figure 1. Demographic and Pain-Related Descriptive Statistics Table 2 summarizes descriptive data on the characteristics of the patients included in the study. On the basis of one-sample Kolmogorov-Smirnov Test, the pain scores did not have a normal distribution. The range of pain measured was 0 to 10, but the sample consisted primarily of patients with initially moderate to high pain levels. Mean and standard deviations of the pain scores are summarized in Table 3.
Main Results Factor analysis. The KMO measure of sampling adequacy was 0.785 and Bartlett’s test for sphericity was significant (p < 0.001). Extraction communalities were 0.949, 0.960, and 0.961 for VAS, CAS, and NRS, respectively. Only one component had Eigen value > 1 (2.87). The variance explained by the first component was 95.68%. Correlations. The Spearman r revealed a strong correlation between NRS, VAS, and CAS with correlation coefficients between NRS and CAS, NRS and VAS, and CAS and VAS being 0.95, 0.94, and 0.94, respectively (p < 0.001). The regression lines derived from these data show that the slopes of the lines are not 1. There is a strong correlation between the NRS and VAS (r = 0.90 and p < 0.001). The y-intercept was 0.12 and the slope was 0.95 (95% CI 0.83–1.07). For the CAS and VAS the y-intercept was 0.13, with a slope of 0.94 (95% CI 0.81–1.07), and a strong correlation (r = 0.89 and p < 0.0001) existed. NRS and CAS were also correlated (r = 0.91 and p < 0.0001) and the y-intercept was 0.48, with a slope of 0.92 (95% CI 0.44–1.40). All of the correlations were statistically significant (p < 0.001). This finding indicates that although the scores of the three scales correlate well, they are not equivalent. Bland-Altman. The mean differences and the limits of agreement between the three methods of pain measurement are presented in Table 4. The Bland-Altman graphs are presented in Figures 2, 3, and 4. Based on the analysis, the following can be inferred: The limits of agreement between the paired NRS and VAS scores extended from 2.0 to 2.6 around the mean difference between the two measures. Also, the interval between VAS and CAS and that for CAS and NRS were 2.7 to 2.0 and 2.0 to 2.1, respectively (Table 4). The correlation coefficients between NRS, VAS, and CAS scores and the composite measure, regarded as gold standard, were 0.98, 0.97, and 0.98, respectively, which indicates strong correlations. The Spearman correlation coefficients were 0.90, 0.88, and 0.87 assessing the correlation between NRS and CAS, NRS and VAS, and CAS and VAS in data set devoid of 0 and 10, respectively. The mean differences,
Table 3. Pain Intensity Scores for Each of the Measures Rating Scales NRS CAS VAS
Baseline (n = 150)
30 min (n = 142)
60 min (n = 112)
Total Sample (N = 404)
5.0 (4.0–8.0) 6.0 (3.7–8.0) 5.1 (2.6–7.6)
4.0 (1.0–6.2) 4.0 (1.0–6.2) 3.1 (1.0–6.8)
1.0 (0.0–5.0) 1.0 (0.0–5.0) 1.0 (0.0–3.7)
4.0 (1.2–7.0) 4.0 (2.0–7.0) 3.3 (1.0–6.8)
CAS = Color Analog Scale; NRS = Numerical Rating Scale; VAS = Visual Analog Scale. Data are median (interquartile range).
6
M. Bahreini et al.
Table 4. Mean Difference and Limits of Agreement According to the Bland-Altman Method Limits of Agreement (95% CI) Scale VAS/CAS VAS/NRS CAS/NRS
Mean Difference (95% CI) 0.4 ( 0.5 to 0.2) 0.3 (0.1 to 0.4) 0.1 ( 0.2 to 0.0)
Lower Limit 2.7 ( 2.9 to 2.0 ( 2.3 to 2.1 ( 2.3 to
Upper Limit 2.4) 1.9) 1.9)
2.0 (1.7 to 2.1) 2.6 (2.5 to 3.0) 2.0 (1.8 to 2.2)
CAS = Color Analog Scale; CI = confidence interval; NRS = Numerical Rating Scale; VAS = Visual Analog Scale.
lower and upper limits of agreement according to BlandAltman method were 0.1 and 2.0 to 2.1, 0.3 and 2.0 to 2.6 and the latter 0.4 and 2.0 to 2.7, between NRS and CAS, NRS and VAS, and CAS and VAS in data set without 0 and 10, respectively.
that, when the tools are applied to adult patients with acute pain in the ED setting, there is a strong correlation between pain scores obtained using any of the three.
Patient satisfaction. Patients were questioned about the preferred pain scale to identify their pain. Overall, 38% of the participants mentioned no difference between the scales. Among the 62% who did notice a difference, CAS was the preferred scale (32% of total) to score pain. In the illiterate group, most patients (60%) stated that there was no difference between the three pain scales. Interestingly, none of the patients in this group chose VAS as the most preferable. In addition, although 32.5% of women preferred VAS, 34.2% of men mentioned no difference between the pain scales.
There are many different instruments for measuring pain intensity in the ED. The degree of agreement among these self-report scales is controversial. In the present study, we compared the performance and comparability of the three pain rating scales commonly used in ED practice (ie, VAS, CAS, and NRS). The findings of this study proved
In our study, a strong correlation was found between NRS and VAS (r = 0.94). This is exactly the same as the finding of the study by Bijur et al., conducted on 297 measurement in adult patients in the ED (r = 0.94) (19). Other investigators concluded that verbally administered NRS can be substituted for VAS in acute pain measurement (20,21). Also, a French study on a sample of 90 ED nontrauma patients and a sample of 200 patients with trauma gave similar results (r = 0.91 and r = 0.80, respectively) (22). Avery conducted the aforementioned study on 100 ED patients and found a correlation coefficient of 0.89 between NRS and VAS (23). Earlier studies in other settings, such as postoperative pain, have also yielded similar results (15,17,24). We also found that the pain scores obtained by any of the three pain rating scales and the composite measure considered as the gold standard were highly correlated. To the best of our knowledge, no previous study has used this method to assess the validity of the measurements. In addition, assuming that patients reporting their pain as maximum
Figure 2. Bland-Altman plot of Visual Analog Scale (VAS) and Color Analog Scale (CAS).
Figure 3. Bland-Altman plot of Numeric Rating Scale (NRS) and Color Analog Scale (CAS).
DISCUSSION
Correlations
Self-Report Pain Scales in the ED
7
Patient Preference
Figure 4. Bland-Altman plot of Numeric Rating Scale (NRS) and Visual Analog Scale (VAS).
or minimum would score similarly with all pain scales and to avoid the possibility of exaggerated correlation between scores at the extremis (0 and 10), we repeated correlation analyses using a data subset, devoid of 0 and 10 scores. Even this conservative analysis did not change the results of the study. Again, we did not find any studies that have adapted this approach. The sole reliance on correlation coefficients is misleading and cannot be used to decide whether one instrument performs superior to another. Agreement The analysis of agreement between the three scales indicated that the three pain scales evaluated are in acceptable agreement to measure pain intensity because the limit of agreement as determined by the Bland-Altman method fell close to the repeatability coefficient determined for adults ( 6 20 mm). Bailey et al. assessed the agreement between VAS, CAS, Verbal Numeric Scale, and Wong-Baker FACES Pain Rating Scale in a pediatric population with acute abdominal pain and concluded that only CAS and VAS were in agreement (12). Performing a regression analysis, we assumed that if y-intercept equals 0, the two scales were equivalent, so that the predicted value of VAS at NRS or CAS score of 0 would be 0, and every unit increase in NRS or CAS would match with a 1cm increase in VAS. The results of our study suggest that there was a small difference, indicating that the measures were closely correlated but not entirely equivalent. The factor analysis of the scoring properties of pain rating instruments suggested a high degree of association between the variables.
An important factor to consider when choosing among rating scales is their ease of administration and patients’ comfort reporting pain. In our study, we assessed patients’ preference for different pain rating scales by asking them a direct question. Although the largest group of patients (38%) stated that there was no difference between the pain scales, the CAS was preferred by a relatively large proportion of patients (32%). Surprisingly, none of the uneducated patients sustaining acute pain mentioned VAS as the instrument of choice for quantifying pain. Several studies have also dealt with this question, but all of them have been undertaken in settings other than the ED (14,16,25 29). These studies have shown that the Verbal Rating Scale (VRS) was preferred by the less educated and the elderly, and the NRS was the instrument of choice in an age-mixed population in chronic pain patients and in head and neck cancer pati ents (9,14,16,22,25,26,28,29). None of these studies, however, used CAS, and our study did not include VRS. In summary, the strong correlation between verbally administered NRS, VAS, and CAS and the close interscale agreement obtained supports their interchangeable application in acute adult pain setting. Limitations Our study faces some limitations that should be acknowledged. Firstly, the three pain ratings were obtained almost simultaneously. Although the order was randomized, the degree of agreement between the measures may have been overestimated. Secondly, one fifth of the measurements were repetitions that were obtained from the same patient at different occasions. These were considered independent observations for the purpose of this study. This, however, could have biased the results. In this study, we assessed patients’ perspectives about the three pain scales. It will also be of great value to determine the pain score most easily administered from the physicians’ viewpoint. This was not possible in our study because one researcher obtained all of the data. CONCLUSIONS The strong correlation between verbally administered NRS, VAS, and CAS and the close inter-scale agreement obtained support their interchangeable application in acute adult pain setting. Acknowledgments—The authors appreciate Ahmad Bahreini for data management. We would also like to extend our gratitude to
8
M. Bahreini et al.
Dr. Akbar Fotouhi and Dr. Nourieh Sharifi for their insightful critique and suggestions.
REFERENCES 1. Joint Commission on Accreditation of Healthcare Organizations. 2006 Comprehensive Accreditation Manual for Hospitals: The Official Handbook. Oakbrook Terrace, IL: Joint Commission on Accreditation of Healthcare Organizations; 2006. 2. Ferraz MB, Quaresma MR, Aquino LR, Atra E, Tugwell P, Goldsmith CH. Reliability of pain scales in the assessment of literate and illiterate patients with rheumatoid arthritis. J Rheumatol 1990;17:1022–4. 3. O’ Rourke D. The measurement of pain in infants, children, and adolescents: from policy to practice. Phys Ther 2004;84:560–70. 4. Carlsson AM. Assessment of chronic pain. I. Aspects of the reliability and validity of the visual analog scale. Pain 1983;16:87–101. 5. Jensen MP, Turner JA, Romano JM, Fisher LD. Comparative reliability and validity of chronic pain intensity measures. Pain 1999; 83:157–62. 6. Joyce CRB, Autshi DW, Mason RM. Comparison of fixed interval and visual analog scales for rating chronic pain. Eur J Clin Pharmacol 1975;8:415–20. 7. McCormack HM, Home DJ, Sheather S. Clinical applications of visual analog scales: a critical review. Psychol Med 1988;10:1007–19. 8. Ohnhaus EE, Adler R. Methodological problems in the measurement of pain a comparison between the verbal rating scale and the visual analog scale. Pain 1975;1:379–84. 9. Jensen MP, Karoly P, Braver S. The measurement of clinical pain intensity: a comparison of six methods. Pain 1986;27:117–26. 10. Krebs EE, Carey TS, Weinberger M. Accuracy of the pain numeric rating scale as a screening test in primary care. J Gen Intern Med 2007;22:1453–8. 11. Hain R. Pain scales in children: a review. Palliat Med 1997;11: 341–50. 12. Bailey B, Bergeron S, Gravel J, Daoust R. Comparison of four pain scales in children with acute abdominal pain in a pediatric emergency department. Ann Emerg Med 2007;50:379–83. 383.e1-2. 13. Bulloch B, Tenenbein M. Validation of 2 pain scales for use in the pediatric emergency department. Pediatrics 2002;110:e33. 14. Clark P, Lavielle P, Martinez H. Learning from pain scales: patient perspective. J Rheumatol 2003;30:1584–8. 15. DeLoach LJ, Higgins MS, Caplan AB, Stiff JL. The visual analog scale in the immediate postoperative period: intrasubject vari-
16.
17. 18. 19. 20. 21. 22. 23.
24.
25. 26.
27. 28. 29.
ability and correlation with a numeric scale. Anesth Analg 1998; 86:102–6. Peters ML, Patijn J, Lame I. Pain assessment in younger and older pain patients: psychometric properties and patient preference of five commonly used measures of pain intensity. Pain Med 2007; 8:601–10. Breivik EK, Bjo¨rnsson GA, Skovlund E. A comparison of pain rating scales by sampling from clinical trial data. Clin J Pain 2000;16:22–8. Ahlers SJ, van Gulik L, van der Veen AM, et al. Comparison of different pain scoring systems in critically ill patients in a general ICU. Crit Care 2008;12:R15. 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. Ekblom A, Hansson P. Pain intensity measurements in patients with acute pain receiving afferent stimulation. J Neurol Neurosurg Psychiatry 1988;51:481–6. Daoust R, Beaulieu P, Manzini C, Chauny JM, Lavigne G. Estimation of pain intensity in emergency medicine: a validation study. Pain 2008;138:565–70. Berthier F, Potel G, Leconte P, Touze MD, Baron D. Comparative study of methods of measuring acute pain intensity in an ED. Am J Emerg Med 1998;16:132–6. Avery VR. Comparing the Visual Analog Scale and Verbally Administered Numeric Rating Scale in Traumatic versus NonTraumatic Pain in a Community Hospital Emergency Center. Master’s Project. Toledo, Ohio: University of Toledo; 2012. Hjermstad MJ, Fayers PM, Haugen DF, et al. European Palliative Care Research Collaborative (EPCRC). Studies comparing Numerical Rating Scales, Verbal Rating Scales, and Visual Analogue Scales for assessment of pain intensity in adults: a systematic literature review. J Pain Symptom Manage 2011;41:1073–93. Herr KA, Mobily PR. Comparison of selected pain assessment tools for use with the elderly. Appl Nurs Res 1993;6:39–46. Herr KA, Spratt K, Mobily PR, Richardson G. Pain intensity assessment in older adults: use of experimental pain to compare psychometric properties and usability of selected pain scales with younger adults. Clin J Pain 2004;20:207–19. Kremer E, Atkinson JH, Ignelzi RJ. Measurement of pain: patient preference does not confound pain measurement. Pain 1981;10: 241–8. Rodriguez CS, McMillan S, Yarandi H. Pain measurement in older adults with head and neck cancer and communication impairments. Cancer Nurs 2004;27:425–33. Williams AC, Davies HTO, Chadury Y. Simple pain rating scales hide complex idiosyncratic meanings. Pain 2000;85:457–63.
Self-Report Pain Scales in the ED
ARTICLE SUMMARY 1. Why is this topic important? Pain represents nearly 60% of all emergency patient complaints. Accurate measurement of pain can enhance the quality of patient management. 2. What does this study attempt to show? The goal of this study was to assess correlation and agreement between the Visual Analog Scale (VAS), Color Analog Scale (CAS), and Numeric Rating Scale (NRS) adult patients in the emergency department. 3. What are the key findings? There was strong correlation and close inter-scale agreement between verbally administered NRS, VAS, and CAS. 4. How is patient care impacted? Patient perception of the convenience of each instrument was identified. This study assessed various pain rating scales for better pain reduction or relief.
9