Acta Biomed 2016; Vol. 87, N.

3: 347-352 © Mattioli 1885

Original article - Emergency medicine

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
Rebecca Capponi1, Valentina Loguercio2, Stefania Guerrini3, Giampietro Beltrami2,
Andrea Vesprini4, Fabrizio Giostra2
1
UNIVPM Corso di Laura in Infermieristica - Fermo 2 Medicina e Chirurgia d’Accettazione e d’Urgenza - Area Vasta 4 Fer-
mo, 3 Pronto Soccorso e Medicina d’Urgenza – Azienda Ospedaliero-Universitaria di Bologna Policlinico S. Orsola Malpighi,
4
Qualità e Programmazione Sanitaria - Area Vasta 4 Fermo

Summary. Abstract: Pain evaluation at triage in Emergency Department (ED) is fundamental, as it influences
significantly patients color code determination. Different scales have been proposed to quantify pain but they
are not always reliable. This study aims to determine a) how important is for triage nurses pain measurement
b) reliability of Numeric Rating Scale (NRS), the most used instrument to evaluate pain in Italian EDs,
because it frequently shows higher pain scores than others scales. Methods: End point 1: a questionnaire was
administered to triage nurses in some hospitals of northern Italy. End point 2: 250 patients arriving at the ED
referring pain have been evaluated using, randomly, either the NRS or a fake “30-50” scale. Results: End point
1: Triage nurses acknowledge to modify frequently the referred pain intensity. This for several reasons: nurses
think that patients may exaggerate to obtain a higher priority color code; they may be influenced by specific
patients categories (non EU citizens, drugs-addicted, elderly); the pain score referred by patients is not corre-
spondent to nurse perception. End point 2: Data show that the mean value obtained with NRS is significantly
(p<0.05) higher that the mean obtained with the “30-50” scale. Conclusion: Manipulation on pain evaluation
performed by nurses might result in a dangerous underestimation of this symptom. At the same time, the use
of NRS seems to allow patients to exaggerate pain perception with consequent altered attribution of color
code at triage. (www.actabiomedica.it)

Key words: pain, Numeric Rating Scale, Visual Analog Scale, triage

Introduction pain measurement should be a fundamental part of the

assessment of every single patient in order to correctly
Pain is one of the leading symptoms of presen- lead the therapeutic strategy (3).
tation to the Emergency Department (ED). In Italy, Assessment scales are validated and shared instru-
the spread of pain control and treatment had a great ments, whose purpose is welfare improvement through
pulse due to law 38 published in 2010 (1). According a systematic collection of clinical data. The scales for
to the American Pain Society (2) pain represents the evaluation of pain proposed in literature have the fol-
fifth vital sign; therefore, the evaluation of its intensity lowing features: easiness of use, data recording and
should be part of patient assessment and documenta- processing, comprehension, fulfilling criteria of valid-
tion. As pain perception is extremely patient-related, ity, sensitivity and reliability. Based on different collec-
its objective evaluation remains difficult. Nevertheless, tion criteria, scales can be either subjective or objective.
348 R. Capponi, V. Loguercio, S. Guerrini, et al.

In subjective scales, the assessment method is based on to Marche Region as well as of northern and central It-
verbal or analog pain description and can be limited aly (Appendix 2). The collected data were subsequently
by cognitive and communicative abilities and patient analyzed using Microsoft Office’ Excel, version 2013.
age. The second objective approach implies evalua- A total of 154 nurses answered the questionnaire.
tion of specific behavioral and physiological indices in As for the second aim of this study, a numerical
response to a painful stimulus, so that a score related measurement scale was administered to all patients
to pain intensity can be obtained. One of the preva- older than 14 years who presented in the same period
lent scales used is the Numeric Rating Scale (NRS), to the ED with pain. Two measuring scales were used:
in which patients indicate verbally or graphically the i) the classical NRS numerical scale and ii) a new, non
intensity of the perceived pain assigning a number in- existing scale ranging from 30 to 50, in which 30 cor-
cluded between 0 to 10 (4). Another common scale responded to “absence of pain” and 50 to “worst pain
is the Visual Analog Scale (VAS): in this second case ever experienced”, in order to avoid “school” influence.
patients are asked to analyze a 10 cm line in which Each patient compiled one scale. The administration
at each extremity different pain intensitivies are de- of one or the other scale was randomized. A sheet of
scribed, starting from “no pain” to the left up to “the paper with the inscription “NRS” or “30-50” identi-
worst pain experienced” to the right. fying the type of scale administered was included in
This study had a dual objective. We focused on non-transparent envelopes. The envelopes were then
evaluation of pain at ED triage because, in this area, closed, mixed and numbered from 1 to 250 and opened
the evaluation of this parameter significantly affects in consecutive order. Patients data were recorded on
the subsequent take in charge of patients. Therefore, a proper file. A total of 250 patients lamenting pain
our first purpose was to determine whether or not pain as main symptom participated the study, 125 for each
evaluation performed by triage nurses could influence different scale (“30-50” or NRS). The available data are
code assignment together with the eventual correct both quantitative and qualitative.
use of pain scales for code determination. Secondly, we
applied NRS at triage to determine whether patients,
influenced by numbers, tend to overestimate their level Results
of pain by assigning a higher numeric value not truly
corresponding to their real pain intensity. In previous Aim 1
studies, NRS showed to determine a substantial and
widespread increase of pain values. compared to VAS. The questionnaire that nurses had to fill included
This might be related to an old school legacy accord- eight questions, the first of which (Appendix 1) aimed
ing to which the highest value is more rewarding and to analyze the volume (i.e., number of patients visited
6 represents the minimum threshold. In this study we per year) of the ED in which the involved nurses used
wanted therefore to compare NRS with another ficti- to work. Results showed how 88% of nurses worked in
tious numerical scale that could not be connected to an ED with more than 35.000 accesses, 6% between
school evaluation (0 to 10). 20 and 35.000 and a further 6% less than 20.000. The
second question investigated the kind of pain scale
locally adopted. 126 nurses (81.8%) used NRS, 12
Methods (7.8%) VAS and 16 (10.4%) claimed to use other types
of instruments. Thirdly, nurses were asked to assign a
For a 3 months period (January 1st-March 31 score from 0 to 100 to indicate the usefulness of pain
2015), we administered to triage nurses working in quantification at triage. 37 nurses (24%) affirmed pain
the Emergency Department of “A. Murri” Hospital in measurement is essential (100) in triage; 28 (18.2%)
Fermo the “Questionnaire on the use of pain scales in responded 90, 32 (20.8%) 80, 14 nurses (9.1%) chose
triage” (Appendix 1). In the same period of time, the 70. For the remaining nurses (43-27.9%) the impor-
questionnaire was distributed in other EDs belonging tance had a relevance quantifiable in a value definitely
A correct pain evaluation at triage in ED 349

inferior to 50. As for the fourth question (“when the egories. For 56.7% of nurses the answer was “yes”, thus
patient main symptom is pain, in which percentage confirming the existence of categories. Among these
do you quantify it by using the adopted assessment categories we must mention specific ethnic groups,
scale?”), 16 nurses answered “<30%”, 24 “30-50%”, 55 different social classes (i.e. people using drugs or al-
“51-80%”, while 59 nurses indicated “>80%”. In the cohol, homeless) and age groups. In 43.3% of cases,
fifth question, we asked whether they believed pa- instead the answer was negative (Figure 3) .
tients answer on pain would depend on the attempt Question 8 investigated the existence of internal
to receive a higher priority code. As shown in Figure protocol in which pain intensity is included and con-
1, only 3 nurses were convinced that this was not the tributes to assignment of the priority code and, if so,
case, whereas 11 nurses answered “1-30%”, 22 “31- whether triage nurses follow it or modify patient’s re-
50%”, 61 “51-80%”, and finally 57 responded “>80%”. ferred values in order to assign a more reasonable color
The sixth question, instead, investigated the fre- code. For 8 nurses (5.2%) “There are no protocols that
quency by which the triage nurses tend liberally to in- modify the priority code according to pain”, 70 nurs-
terpret patients’ pain, thus modifying patient’s evalu- es (45.45%) claimed to strictly apply pain protocols,
ation. 50 nurses affirmed to never modify the inten- whereas the remaining 49.35% (76 nurses) admitted
sity of the pain reported by patient, while the others to modify patients’ answer in order to assign a more
declared, in varying proportions, to change the value appropriate (according to the nurse personal evalua-
reported by the patients (Figure 2). Furthermore, for tion) color code. Therefore, we asked these last nurses
those who aknowledged to alter in some cases the in- in which percentage this was done, and the most fre-
tensity of pain reported by patient, we asked whether quent response was 30-50% (Figure 4).
this modification was related to specific patients’ cat-

Figure 1. Attempt to obtain a higher priority code Figure 3. Personal evaluation of pain depending on different
patients categories

Figure 2. Modification of pain perceptions Figure 4. Modification of patients answers in determining co-
lor priority code
350 R. Capponi, V. Loguercio, S. Guerrini, et al.

Aim 2 Table 1. Pain assessment with NRS

Sample size 125
Data analysis comparing pain measurement ob- Males 73 58%
tained with NRS and the “30-50” showed how, with Females 52 42%
NRS, the value assigned by patient was 1.78 higher Average age 51.9
than that awarded by “30-50” scale. Although further Average pain referred 7.36
confirmation is needed, we observed a statistically sig- Standard deviation 2.15
nificant difference (p<0.05) between the pain values​​ Variance 4.62
reported with the NRS scale (which has an average
value of 7.3) and the fictitious “30-50” (average value
41 converted into the 1 to 10 scale is equivalent to 5.5). Table 2. Pain assessment with “30-50” scale
As suggested, NRS might be related to pain overes-
Sample size 125
timation as the patient, due to a scholar “memory”,
Males 57 46%
unconsciously starts from a minimum psychologi-
Females 68 54%
cal threshold (represented by 6) and therefore might
Average age 50.2
assign a higher value. As a consequence, the “30-50”
Average pain referred 5.5*
scale is likely to be more objective scale and compa-
Standard deviation 2.73
rable to VAS, although this should be the subject of
Variance 7.44
another, specifically addressed study. We also must
mention that, as for other assessment tools, this scale *normalized value
has limitations, especially with elderly and foreigners.
Very importantly, the use of the “30-50” scale could the intensity of pain reported by patients, in order to
positively influence color code assignment as, accord- assign the color code considered more appropriate.
ing to triage guidelines, a pain level superior or equal Moreover, many nurses also are influenced by certain
to 7 determines a yellow code. In the present study, categories of “fragile” patients. As for other operating
although performed in a relatively small sample of Units but particularly for Emergency Departments,
patients, pain evaluation measured applying the NRS pain assessment and management remain a crucial
scale obtains an average value of 7.36 resulting in a yel- topic and further investigations are needed. There is
low code whereas using the “30-50” scale the average a sort of “skepticism” that pushes triage nurses to per-
normalized value is 5.5, which corresponds to a green sonally interpret the pain, rather than rely on patients
code. This certainly can significantly modify dynamics judgment. Nurses’ efforts is to reconcile the patients
of triage and Emergency Department. The two sam- signs with their personal intuition and sensitivity, thus
ples under study are described in Tables 1 and 2. leading sometimes to pain underestimation by medical
staff (6). Frequently nurses tend to believe that patients
emphasize the perceived pain in order to be assigned
Conclusions a higher priority code and access more easily to treat-
ment. Therefore, pain underestimation remains an im-
“Oligo-analgesia” is a term coined in 1989 by portant issue and, although some improvements have
Wilson and Pendleton and indicates the inadequate been obtained, it is essential to investigate the problem
recognition and treatment of pain (5). The problem and make pain truly become the “fifth vital sign”.
with oligo-analgesia in patients accessing the ED is In recent years, the use and analysis of pain meas-
widespread and the attention from the medical staff urement instruments have been investigated. The pre-
is still insufficient. In our study, several contradictions sent study has focused on the adequacy of NRS that,
have emerged, although the majority of nurses judges despite being easily and quickly applicable, therefore
pain assessment at triage of crucial relevance. Most being suitable for the ED setting, presents some biases
nurses participating the study acknowledged to change attributable to self-assessment.
A correct pain evaluation at triage in ED 351

Pain, widely considered as the “fifth vital sign”, is 2. Morone NE, Weiner DK. Pain as the 5th vital sign: expos-
a symptom often underestimated. The Authors’ per- ing the vital need for pain education. Clin Ther 2013 Nov;
35(11): 1728-32.
sonal perception is that the evaluation of pain at triage 3. Minute M, Massaro M, Barbi E. Trattamento del dolore in
can be compared to a chess game, in which players are pronto soccorso. Pediatria d’urgenza 2012; 42 (167): 143-
represented by patients and nurses. Everyone, more or 150.
less unconsciously, tends to provide an interpretation 4. Knox H. Todd, MD, MPH. Pain Assessment Instruments for
Use in the Emergency Department. Emerg Med Clin N Am
and quantification that goes to their advantage. Ac-
2005; 23: 285-95.
cording to the authors of the present study, this work 5. D ucharme J. The future of Pain Management in Emergency
might suggest some interesting ideas for pain correct Medicine. Emerg Med Clin N Am 2005; 23: 467-475.
management. Firstly, we suggest that a more “aseptic” 6. Amaducci, Bagattoni, Pocaforza, Mecugni. L’assessment del
measurement scale should be proposed; at the same dolore in pronto soccorso: l’impatto sugli infermieri preposti
al triage; IPASVI-ECM-Rivista infermiere 2012; 1: 36-39.
time, training programs should be organized, aimed to
motivate nurses in investigating the fifth vital sign and
teach to avoid personal interpretations of pain.
Received: 28 September 2016
Accepted: 15 December 2016
Correspondence:
References Fabrizio Giostra,
Area Vasta 4 Via Zeppilli 18
1. 
Legge 15 marzo 2010, n. 38 “Disposizioni per garantire 63900 Fermo
l’accesso alle cure palliative e alla terapia del dolore” Tel. 3288642485; 0734 6252278; 0734 625111

Appendix 1.
Questionnaire on use of pain scales at triage
We kindly ask you to answer the following questions by choosing one of the multiple choices available.

1. The volume (i.e., number of patients per year) of your ED is:

- <20.000
- 20.000-35.000
- >35.000
2. In your ED what scale is adopted to measure pain?
- NRS
- VAS
- other (please specify)……
3. According to you, on a scale from 0 to 100 how relevant is pain measurement at triage?
Nr….
4. When the patient main symptom at triage is pain, in what percentage do you quantify it by using the adopted assessment scale?
- 0%
- 1-30%
- 31-50%
- 51-80%
- >80%
5. In what percentage do you believe patient’s answer on pain might depend on the attempt to receive a higher priority code?
- 0%
- 1-30%
- 31-50%
- 51-80%
- >80%
352 R. Capponi, V. Loguercio, S. Guerrini, et al.

6. When you investigate pain at triage, do you tend to modify patient’s evaluation, and if so, in what percentage?
- 0%
- 1-30%
- 31-50%
- 51-80%
- >80%

7. If you answer to the previous question was affirmative, are there specific patients’ categories that are mostly related to your modi-
fication according to your personal interpretation (for instance social classes, ethnic categories, age classes)?
- yes
- no

8. Does pain intensity contribute to color code assignment?

- there are no protocols that allow to modify the priority code according to pain
- I strictly apply pain protocol
- Sometimes I modify patient’s answer in order to assign a more appropriate code according to my personal evaluation

9. If the answer to the previous question was “Sometimes I modify patient’s answer in order to assign a more appropriate code ac-
cording to my personal evaluation”, in what percentage of patients does this happen?
- <30%
- 30-50%
- 51-80%
- >80%

Appendix 2

- Ospedale “A. Murri” di Fermo (A.S U R. Area vasta n. 4)

- Ospedale “S. Maria della Misericordia” di Urbino (A.S U R. Area vasta n. 1)
- Ospedale “Madonna del Soccorso” di San Benedetto del Tronto (A.S.U.R . Area vasta n. 5)
- Ospedale “Engles Profili” di Fabriano (A.S U R. Area Vasta n. 2)
- Ospedale “S. Maria della Scaletta” di Imola
- Ospedale “Luigi Sacco” di Milano
- Azienda Ospedaliero-Universitaria di Parma
- Ospedale Regionale “Umberto Parini” di Aosta
- Azienda ospedaliera universitaria integrata Verona-Borgo Trento
- Azienda ospedaliera “S. Croce e Carle” di Cuneo
- Ospedale degli infermi di Biella
- Azienda Ospedaliera-Universitaria di Bologna