Blackwell Publishing IncMalden, USAPMEPain Medicine1526-2375American Academy of Pain Medicine?

200783223234
Original ArticlePostoperative Pain Intensity Assessment in ChineseLi et al.

PA I N M E D I C I N E
Volume 8 • Number 3 • 2007

Postoperative Pain Intensity Assessment: A Comparison of Four

Scales in Chinese Adults

Li Li, MSN, RN,* Xueqin Liu, BSN, RN,* and Keela Herr, PhD, RN, FAAN†
*Department of Nursing, Zhujiang Hospital, Southern Medical University, Guangzhou, China; †College of Nursing, The
University of Iowa, Iowa City, Iowa, USA

ABSTRACT

ABSTRACT Objective. To determine the psychometric properties and applicability of four pain scales in Chinese
postoperative adults.
Design. A prospective clinical study.
Setting. A university-affiliated hospital.
Patients. In total, 173 Chinese patients (age range 18–78 years) undergoing scheduled surgery.
Interventions. Recalled pain and anticipated postoperative pain intensity were rated preoperatively
with a visual analog scale (VAS), a numeric rating scale (NRS), a verbal descriptor scale (VDS), and
the Faces Pain Scale Revised (FPS-R). From the day of surgery to the sixth postoperative day,
patients were interviewed for the scores of current operative pain intensity and the worst, least, and
average pain on that day. On the sixth postoperative day, retrospective ratings over the 7 days were
also obtained and tool preferences were investigated.
Outcome Measures. Scale reliability was evaluated using intraclass correlation coefficients (ICCs).
Scale validity was assessed by correlations between scales, analysis of variance with repeated mea-
sures, and the sensitivity of the scales to interventions. Chi-square tests were used to investigate if
error rate and preference rate were related to gender, age, and educational level.
Results. All four pain intensity scales had good reliability and validity when used with Chinese
adults. The ICCs of the four scales across current, worst, least, and average pain on each postop-
erative day were consistently high (0.673–0.825), and all scales at each rating were strongly corre-
lated (r = 0.71–0.99). Analysis of variance with repeated measures revealed significant decreases in
scores associated with postoperative days, and all four scales were sensitive in evaluating analgesic
efficacy. Both the VDS and the FPS-R had low error rates. Nearly half of the participants (48.1%)
preferred the FPS-R, followed by the NRS (24.4%), the VDS (23.1%), and the VAS (4.4%);
however, no significant differences were noted in terms of gender, age, and educational level.
Conclusions. These findings demonstrate that although all four scales can be options for Chinese
adults to report pain intensity, the FPS-R appears to be the best one. Providing tool options to
address individual needs or preferences is suggested.

Key Words. Postoperative Pain; Pain Intensity; Pain Assessment; Visual Analog Scale; Numeric
Rating Scale; Verbal Descriptor Scale; Faces Pain Scale Revised; Chinese Adults

Introduction

Reprint requests to: Xueqin Liu,

Department of Nursing, Zhujiang
BSN, RN, Director,
Hospital, 253 Gongye
P ain is a major health problem and has been
measured and treated as the fifth vital sign [1–
3]. Despite much attention paid to surgical pain,
Road M., Guangzhou 510280, China. Tel: 86-20-
61643040; Fax: 86-20-61643042; E-mail: liuxueqin_1@ evidence suggests that postoperative pain is still
126.com. undertreated [4–7]. Appropriate pain management

© American Academy of Pain Medicine 1526-2375/07/$15.00/223 223–234 doi:10.1111/j.1526-4637.2007.00296.x

224
for surgical patients contributes to earlier mobili-
zation, shortened hospital stay, reduced cost, and
increased patient satisfaction, whereas undertreat-
ment of pain may have short- or long-term harm-
ful effects on health [8–10]. It is well established
that accurate pain assessment is essential to effec-
tive postoperative pain relief [11–13].
Pain intensity is thought to be one of the pri-
mary factors that determine the impact of pain on
a person’s overall function and sense of well-being
[14]. Fortunately, using pain scales helps to quan-
tify pain intensity and guide treatment decisions
and monitoring for effectiveness. Although many
pain assessment scales exist, each scale has no stan-
dardized title or definition [15]. It has been shown
that the visual analog scale (VAS), the numeric
rating scale (NRS), the verbal descriptor scale
(VDS) or verbal rating scale (VRS), and the faces
scale are the most commonly used measures
of pain intensity. To facilitate communication
between health care providers and patients, a com-
mon metric scale should be adopted [16,17]. In
clinical practice, both the 0–5 and 0–10 scales are
widely used [15], but the 0–10 scale is preferred
due to the need to standardize clinical pain inten-
sity rating among patients [17].
Compared with the quantity of studies of pain
assessment with cancer pain or other chronic pain,
research on surgical pain in adults with assessment
scales is limited. Carey et al. [18] conducted a
study to identify which of the three pain intensity
measurement scales, the Wong–Baker Faces Scale
[19], the NRS, and the VAS, was the most appro-
priate and effective for hospitalized patients. They
found that the Wong–Baker Faces Scale was the
most selected (48.6%), followed by the NRS
(35.3%), and the VAS (16.1%), with a reliability
coefficient of 0.88. Briggs and Closs found that
the VAS and the VRS correlated well when used
to assess postoperative pain in orthopedic
patients, but the VRS had lower noncompliance
rate [20]. Jensen et al. reported that scales with
relatively few levels, such as the 4-point VRS,
appeared less sensitive than the VAS; however, the
VAS may also be less sensitive than the NRS, pos-
sibly because the VAS is more difficult to compre-
hend for some patients [21]. Studies have shown
these tools are effective in Caucasian older adults
as well. Despite the fact that pain may persist for
much longer, previous research on postoperative
pain was usually assessed from several hours to
3 days, with different scales using different met-
rics (e.g., 0–4, 0–5, 0–6, 0–10, 0–20, and 0–100)
[21–23].
Li et al.
Cultural differences should also be considered
when assessing pain. Culture influences how a
patient responds to and reports pain, and cultural
background has been identified as an important
factor influencing pain behavior and expression
[24–26]. However, data are limited on pain scale
use in individuals from other cultures, specifically
in Chinese people. Aun et al. reported that the
VAS was suitable for Chinese and a vertical version
was easier to understand than a horizontal one
[27]. Zhao et al. found that the NRS and the VRS-
4 correlated well when used in 50 Chinese patients
after cardiac or general surgery and suggested that
a scale incorporating an NRS with a VRS was
appropriate [28]. In these two studies, faces scales
were not included.
It has been suggested that the use of faces scales
avoids language and may cross cultural differences
[29]. The Faces Pain Scale (FPS) [30] and the
Wong–Baker Faces Pain Rating Scale [19] are the
two most commonly used scales, although they
were originally developed for use in children.
Despite that recent findings have established the
validity and reliability of the FPS in adult and
older populations [29,31–34], no evidence is avail-
able for the use of a revised FPS (FPS-R) [16] to
assess postoperative pain in Chinese.
The purpose of this study was to determine the
psychometric properties and applicability of four
selected pain intensity scales (VAS, NRS, VDS,
and FPS-R) scored from 0 to 10 in Chinese adults
for 7 days after scheduled surgery.
Methods
Subjects
The study was conducted in a teaching hospital
that carried out approximately 7,000 surgical
operations annually in Guangzhou, China.
Respondents were selected according to the fol-
lowing inclusion criteria: age over 18 years, no
confusion or cognitive impairment diagnosis from
medical record, admission for scheduled opera-
tion, American Society of Anesthesiology score of
≤ 3, an expected hospital stay of no less than 7 days
after surgery, and willing and able to participate in
the study.
During a 17-week period, 210 patients under-
going scheduled surgery were enrolled, of whom
173 completed the study. The average age was
45.3 years (SD = 15.0 years), with a range from 18
to 78 years. Fifty-five percent (N = 95) were male
and 45% (N = 78) female. Fifty-one percent
(N = 89) had less than a high school education,
Postoperative Pain Intensity Assessment in Chinese
23% (N = 39) had completed high school, and
26% (N = 45) had more than a high school
education.
Twenty-six percent of participants (N = 45)
had operation history. The average hospital stay
was 20 days (SD = 13.3, range 7–81 days). The
operation sites were skeleton and joint (N = 29),
spine (N = 37), gastrointestinal (N = 12), liver,
gallbladder, and pancreas (N = 20), urinary
(N = 26), reproductive (N = 32), and other
(N = 17). The anesthesia types were general anes-
thesia (N = 31), local anesthesia (N = 56), epidu-
ral anesthesia (N = 83), and spinal block (N = 3).
After surgery, 28% (N = 48) were given patient-
controlled analgesia (PCA), 17.3% (N = 30)
received an “on demand” (pro re nata [prn]) dose
of analgesics (including nine patients after discon-
tinuation with PCA). The operation sites of those
with prn analgesics were as follows: skeleton and
joint (N = 6), spine (N = 10), gastrointestinal
(N = 3), liver, gallbladder, and pancreas (N = 5),
urinary (N = 3), reproductive (N = 2), and other
(N = 1).
Pain Measures
Pain intensity was measured with the following
scales: VAS, NRS, VDS, and FPS-R.
Visual Analog Scale
The VAS is a 10-cm line, with the end point 0 for
“no pain” and 10 for “worst pain.” Participants
were asked to make a mark on the line that repre-
sented their pain intensity, and pain intensity level
was scored by measuring the distance from the “no
pain” end to the patient’s mark. The VAS is a tool
commonly used in research and clinical practice,
and its reliability and validity in pain assessment
has been clearly demonstrated [35–37]. Pain
intensity measured by the VAS is significantly
correlated with those measured by verbal and
numeric scales [38,39]. The main strength of the
VAS is its ratio scale properties [40], but its draw-
backs are that it is time-consuming, needs abstract
thinking, which makes it difficult to understand
and complete, and results in high failure response
rates (especially in older people) [20,38].
Numeric Rating Scale
The 0–10 NRS uses 11 numbers (0 through 10)
to measure pain intensity. Participants were
instructed to select the number that best reflected
the intensity of pain, with 0 equaling no pain and
10 the worst pain. This scale was selected because
it is more commonly used in clinical practice and
225
has been found to be a reliable and valid pain
intensity measure [41,42].
Verbal Descriptor Scale
The VDS consists of a list of adjectives describing
different levels of pain intensity: no pain, mild
pain, moderate pain, severe pain, extreme pain,
and worst pain [43]. The scores 0, 2, 4, 6, 8, and
10 were assigned to each of the verbal descriptors,
with “none” scored as 0 to “worst pain” scored 10,
with higher numbers associated with more intense
adjectives. There may be issues with the lack of
congruence between the NRS and the VDS using
numbers. Participants were asked to pick the word
that best described their pain intensity, and their
VDS intensity score was the number associated
with the word they chose. The tool incorporates
numerically ranked descriptors that are organized
with an ordinal approach and describe different
levels of pain [38]. The VDS is easy to administer
and score, but its weaknesses include the use of
words that do not necessarily express what the
patient is experiencing or do not have the same
meaning for each participant, and intervals
between words that do not represent equal seg-
ments or identical steps in the scale [44].
Faces Pain Scale Revised
The FPS, developed by Bieri et al., consists of
seven line-drawn faces presented in a horizontal
format, representing no pain to worst pain [30].
The FPS was selected over other options for sev-
eral reasons: the facial depictions appeared less
childlike, the absence of tears avoids potential cul-
tural bias about pain expression, and the use of a
neutral face to represent no pain instead of a happy
face [32]. Preliminary validation testing of the FPS
was satisfactory in previous studies [29,31–34]. In
2001, the FPS was revised by Hicks et al. to make
it more suitable for use with the widely used
metric scoring 0–10 scale, using visual depictions
of faces to represent increasing levels of pain
intensity along a six-face continuum [16]. The
FPS-R was used in this study. Participants were
instructed to point to the face that best repre-
sented the intensity of their pain; the scores 0, 2,
4, 6, 8, and 10 were assigned to each face consec-
utively, with higher numbers representing more
painful faces.
The VAS, the NRS, and the VDS have hori-
zontal or vertical formats. In this study, the vertical
formats were replicated from the literature, using
Chinese word descriptors and end points to facil-
itate understanding. The word translations in
226 Li et al.

Chinese were validated by two English teachers
who were native Chinese and by 10 Chinese sur-
gical patients. Each tool was printed on a separate
sheet of paper. To prevent the patients from just
picking the same number across scales, only the
NRS has the numbers on the tool.
Data Collection
The study was approved by the institutional
review board. Operation schedule sheets were
reviewed from Sunday to Thursday (5 days/week).
The exclusion criteria were determined from
patients’ medical records. Those who met the eli-
gibility criteria were approached to determine
interest in participating. Signed informed consent
was obtained on the day before the operation.
Tools were explained and practiced preoperatively
with two different tasks on the four scales (order:
VAS, NRS, VDS, and FPS-R). First, subjects were
asked to rate any vividly recalled pain (such as a
toothache, childbirth, and headache). Second,
they were asked to rate their anticipated postop-
erative pain intensity on each of the four scales.
Then participants were visited by the first author
and two trained staff nurses in the evenings
(18:00–21:00) from the operation day to the sixth
postoperative day. This time was considered so as
not to interrupt the daytime medical routine work.
The patients completed the scales by marking, or
verbally stating or pointing to the rating. Current
operative pain intensity scores and daily retrospec-
tive worst, least, and average pain levels were
obtained by using the above scales every evening.
On the sixth postoperative day, single retrospec-
tive ratings on worst, least, and average pain for
the 7 days for each scale were also obtained.
In order to determine the sensitivity of the four
scales to interventions, subjects with prn analge-
sics were asked to make pain ratings before and
30 minutes after pharmacological interventions.
At the completion of data collection, participants
were also asked to identify the scale they preferred
to use as a pain intensity measure.
Data Analysis
Scale reliability was evaluated using intraclass cor-
relation coefficients (ICCs). Scale validity was
assessed by correlations between scales, analysis of
variance with repeated measures, and the sensitiv-
ity of the scales to interventions. For patients with
analgesics, the difference score was calculated by
subtracting the post-treatment score from the pre-
treatment score, and paired t-test was used.
Error rates were determined by the frequency
with which patients were unable to correctly use
the pain scale (e.g., items left blank; responses
falling between two numbers, words, or facial
expressions; or selecting more than one response
or options outside the scale range) [32]. McNemar
tests were performed to compare the error rates
for each pair of scales, using the Bonferroni cor-
rection to control the overall significance level.
Chi-square tests were used to investigate if error
rate and preference rate were related to gender,
age, and educational level.
Results
Scale Reliability
On the day before the operation, the ICC of
recalled pain across four scales was 0.723
(F = 11.439, P < 0.001), and the ICC of antici-
pated postoperative pain across four scales was
0.667 (F = 8.999, P < 0.001). On each postopera-
tive day, ICCs were calculated for four scales
across ratings of current pain, daily retrospective
ratings of worst, least, and average pain. The ICCs
of the four scales across current, worst, least, and
average pain were consistently high (0.673–0.825)
(P < 0.001) (Table 1). On the sixth postoperative

Table 1 Reliability coefficients for four scales across current, worst, least, and average pain on 7 postoperative days*
VAS NRS VDS FPS-R
Time ICC F ICC F ICC F ICC F
d0 0.825 19.799 0.822 19.477 0.785 15.606 0.754 13.228
d1 0.775 14.752 0.760 13.660 0.710 10.809 0.717 11.121
d2 0.781 15.265 0.771 14.431 0.748 12.900 0.719 11.258
d3 0.735 12.121 0.750 12.975 0.735 12.115 0.717 11.130
d4 0.773 14.639 0.783 15.462 0.723 11.426 0.744 12.645
d5 0.759 13.569 0.727 11.670 0.677 9.400 0.715 11.044
d6 0.703 10.490 0.673 9.221 0.678 9.435 0.674 9.276

* P < 0.001.
d0 = operation day; d1 = 1st postoperative day; d2 = 2nd postoperative day; d3 = 3rd postoperative day; d4 = 4th postoperative day; d5 = 5th postoperative day;
and d6 = 6th postoperative day.
FPS-R = Faces Pain Scale Revised; ICC = intraclass correlation coefficient; NRS = numeric rating scale; VAS = visual analog scale; VDS = verbal descriptor scale.
Postoperative Pain Intensity Assessment in Chinese 227

day, the ICCs of the four scales across single ret-

<0.001

<0.001

<0.001

<0.001
rospective ratings of worst, least, and average pain

P
for the 7 days ranged from 0.383 to 0.420
(P < 0.001).

45.075

49.496

53.283

55.102
Scale Validity

F
Spearman correlations included 42 pairs (6 for

Average Pain
each day × 7 postoperative days) for current,

2.54 ± 2.73
0.91 ± 1.34
0.59 ± 1.24
2.68 ± 2.74
0.97 ± 1.34
0.59 ± 1.22
2.40 ± 2.33
0.93 ± 1.21
0.48 ± 1.03
2.63 ± 2.31
1.00 ± 1.30
0.55 ± 1.17
worst, least, and average pain, respectively. Spear-

X±S
man correlations between the four scales ranged
from 0.74 to 0.95 for ratings of current pain, 0.80
to 0.99 for worst pain, 0.71 to 0.97 for least pain,

<0.001

<0.001

<0.001

<0.001
and 0.72 to 0.95 for average pain. All scales at each
rating were strongly correlated (P < 0.01), but the

P
weakest correlations appeared to be between the
FPS-R and the VDS when assessing current,

34.617

38.491

39.693

46.615
worst, and least pain, and between the VDS and

Analysis of variance with repeated measures of the four scales on selected postoperative day pain intensity
the NRS when assessing average pain. For the

F
same scale, the correlations between current and
average pain were stronger (r = 0.89–0.96), while

1.97 ± 2.54
0.54 ± 1.11
0.37 ± 1.09
2.06 ± 2.53
0.58 ± 1.11
0.35 ± 1.03
1.87 ± 2.21
0.63 ± 1.03
0.37 ± 0.94
2.10 ± 2.12
0.70 ± 1.16
0.37 ± 1.05
Least Pain
the correlations between least and worst pain were

X±S
weaker (r = 0.49–0.69).
The retrospective ratings on worst, least, and
average pain between scales across the postopera-
tive period were strongly correlated (P < 0.01),

FPS-R = Faces Pain Scale Revised; NRS = numeric rating scale; VAS = visual analog scale; VDS = verbal descriptor scale.
<0.001

<0.001

<0.001

<0.001
with Spearman correlations ranging from 0.88 to
P

0.95 for ratings of worst pain, 0.80 to 0.93 for least

pain, and 0.80 to 0.93 for average pain.
59.102

57.550

60.454

59.666
Analysis of variance with repeated measures
revealed significant decreases in scores associated
F

with postoperative days. Specifically, significant

differences in scores were found within the first
4.45 ± 3.33
2.12 ± 2.20
1.13 ± 1.86
4.49 ± 3.33
2.26 ± 2.21
1.22 ± 1.93
4.21 ± 3.06
2.27 ± 2.00
1.06 ± 1.57
4.55 ± 3.11
2.32 ± 1.97
1.31 ± 1.92
Worst Pain

three postoperative days, between the operative

X±S

day to the second postoperative day and the fourth

to the sixth postoperative day (P < 0.05, Table 2
d0 = operation day; d3 = third postoperative day; d6 = sixth postoperative day.
and Figures 1–4; detailed results are available from
the authors).
<0.001

<0.001

<0.001

<0.001

During the period from the operation day to

P

the sixth postoperative day, 28 patients with prn

analgesics were able to make pain ratings preanal-
44.128

46.676

50.412

55.560

gesics and 30 minutes postanalgesics, with pain

intensity scores of 7.5–7.9 on average prior to
F

analgesics, and 3.2–3.7 after treatment. Results

showed that all four scales were sensitive for eval-
Current Pain

2.81 ± 2.87
0.92 ± 1.40
0.58 ± 1.34
2.89 ± 2.87
0.98 ± 1.41
0.56 ± 1.29
2.66 ± 2.57
0.94 ± 1.27
0.54 ± 1.10
2.88 ± 2.48
1.08 ± 1.37
0.56 ± 1.20

uating the efficacy of analgesic interventions

(Table 3).
X±S

Error Rate
Errors in tool use were considered a response to a
Time

tool that could not be interpreted as a single score

d0
d3
d6
d0
d3
d6
d0
d3
d6
d0
d3
d6

(e.g., circling two numbers, words, or faces; or

marking between two numbers, words, or faces;
Table 2

FPS-R

using a circle or check mark across the VAS) or

NRS

VDS
VAS

simply not attempting to use the scale. Although

228 Li et al.

3.5

3.0
mean score of current pain intensity

2.5

2.0
scales
1.5 Figure 1 Mean score of postopera-
VAS tive current pain on four scales.
d0 = operation day; d1 = 1st postoper-
1.0 NRS ative day; d2 = 2nd postoperative day;
d3 = 3rd postoperative day; d4 = 4th
VDS postoperative day; d5 = 5th postoper-
.5
ative day; and d6 = 6th postopera-
FPS-R tive day. FPS-R = Faces Pain Scale
0.0 Revised; NRS = numeric rating scale;
d0 d1 d2 d3 d4 d5 d6 VAS = visual analog scale; VDS =
time verbal descriptor scale.

5.0

4.0
mean score of worst pain intensity

3.0

scales
Figure 2 Mean score of postopera-
2.0 VAS
tive worst pain on four scales.
d0 = operation day; d1 = 1st postoper-
NRS ative day; d2 = 2nd postoperative day;
1.0 d3 = 3rd postoperative day; d4 = 4th
VDS postoperative day; d5 = 5th postoper-
ative day; and d6 = 6th postoperative
FPS-R day. FPS-R = Faces Pain Scale
0.0
Revised; NRS = numeric rating scale;
d0 d1 d2 d3 d4 d5 d6 VAS = visual analog scale; VDS =
time verbal descriptor scale.

none of the patients had used any pain scales report a recalled pain, including different types of
before, a majority of them completed the scales pain, such as back pain, leg pain, toothache, and
successfully. On preoperative assessment, as part childbirth. Fifty-one patients did not have a
of the tool practice, 122 patients were able to recalled pain that could be used for this task. Error
rates for the first-time use to assess recalled pain
demonstrated significant differences as follows:
Table 3 The change of pain intensity pre- and postanal- VAS (12.3%), NRS (4.9%), VDS (0.8%), and
gesics on four scales (N = 28)
FPS-R (2.5%) (χ2 = 19.351, P < 0.001). Significant
VAS NRS VDS FPS-R differences of error rates were found between the
Premedication 7.9 ± 1.6 7.9 ± 1.6 7.8 ± 1.6 7.5 ± 1.8 VAS and the VDS, and the VAS and the FPS-R
Postmedication 3.7 ± 2.8 3.7 ± 2.7 3.3 ± 2.6 3.2 ± 2.4 (P < 0.05) (Table 4). There was a difference in the
Mean difference 4.3 ± 2.4 4.2 ± 2.3 4.5 ± 2.1 4.3 ± 1.9
t 9.495 9.776 10.632 11.684
error rate between genders, with more errors for
P <0.001 <0.001 <0.001 <0.001 the VAS among female participants (χ2 = 5.514,
FPS-R = Faces Pain Scale Revised; NRS = numeric rating scale; VAS = visual
P = 0.026), and a difference in the error rate
analog scale; VDS = verbal descriptor scale. between education levels, with more errors for the
Postoperative Pain Intensity Assessment in Chinese 229

2.5

2.0

mean score of least pain intensity

1.5

scales
Figure 3 Mean score of postopera-
tive least pain on four scales. 1.0 VAS
d0 = operation day; d1 = 1st postop-
erative day; d2 = 2nd postoperative NRS
day; d3 = 3rd postoperative day; .5
d4 = 4th postoperative day; d5 = 5th VDS
postoperative day; and d6 = 6th post-
operative day. FPS-R = Faces Pain FPS-R
Scale Revised; NRS = numeric rating 0.0
scale; VAS = visual analog scale; d0 d1 d2 d3 d4 d5 d6
VDS = verbal descriptor scale. time

3.0

2.5
mean score of average pain intensity

2.0

1.5 scales
Figure 4 Mean score of postopera-
VAS
tive average pain on four scales.
1.0
d0 = operation day; d1 = 1st postop-
erative day; d2 = 2nd postoperative NRS
day; d3 = 3rd postoperative day;
d4 = 4th postoperative day; d5 = 5th .5 VDS
postoperative day; and d6 = 6th post-
operative day. FPS-R = Faces Pain FPS-R
0.0
Scale Revised; NRS = numeric rating
scale; VAS = visual analog scale; d0 d1 d2 d3 d4 d5 d6
VDS = verbal descriptor scale. time

Table 4 Scale error rates for the first-time use to assess recalled pain (N = 122)
VAS NRS VDS
Right Error Right Error Right Error
NRS Right 103 13
Error 4 2
P 0.049
VDS Right 107 14 115 6
Error 0 1 1 0
P <0.001 0.125
FPS-R Right 106 13 113 6 118 1
Error 1 2 3 0 3 0
P 0.002 0.508 0.625

Significance level at 0.05 with Bonferroni correction is 0.0083 (= 0.05/6).

FPS-R = Faces Pain Scale Revised; NRS = numeric rating scale; VAS = visual analog scale; VDS = verbal descriptor scale.
230 Li et al.

NRS among those with more than high school

0.089

1.000

0.248
education (χ2 = 7.672, P = 0.022) (Table 5). The

P
results showed that those who failed to respond to
the VAS said it was too difficult to understand, so

3.747

0.715

2.787
they just left them blank. Those who unsuccess-

χ2
fully completed the NRS, the VDS, or the FPS-R
mostly had responses falling between two num-

Error
bers, words, or facial expressions, respectively.

0
3

3
0

3
0
0
When rating the anticipated postoperative pain,

FPS-R
29 patients were not able to anticipate a pain that

Right
could be used for this task. A total of 144 patients

67
52

96
23

61
26
32
could do so, and the error rate of each scale
dropped as follows: the VAS (0.7%), the NRS

0.451

1.000

0.633
(2.8%), the VDS (0.7%), and the FPS-R (0.7%),

P
and no significant difference was found (χ2 =
3.905, P = 0.272). When the actual postoperative

1.228

0.234

0.914
pain for 7 postoperative days was assessed, the

Scale error rates for the first-time use to assess recalled pain by gender, age, and education level (N = 122)

χ2

FPS-R = Faces Pain Scale Revised; HS = high school education; NRS = numeric rating scale; VAS = visual analog scale; VDS = verbal descriptor scale.
number of patients able to report pain intensity
gradually increased from 140 to 173. The error

Error
rates dropped further, except for the VDS, which

0
1

1
0

1
0
0
had a slight increase for several time points. Error
rates for the NRS and the VDS were relatively

Right
high when assessing actual postoperative pain, VDS

67
54

98
23

63
26
32
with responses more frequently falling between
two words or numbers.
1.000

1.000

0.022
P

Subject Scale Preference

On the sixth postoperative day, 92% (N = 160) of
0.062

0.020

7.672
the patients identified a preferred pain intensity
χ2

scale. Nearly half of the participants (48.1%) pre-

ferred the FPS-R, followed by the NRS (24.4%),
Error

the VDS (23.1%), and the VAS (4.4%). It

3
3

5
1

0
2
appeared that elderly Chinese patients preferred 4
the NRS although no significant difference was
Right

found. However, no significant differences were

NRS

64
52

94
22

64
24
28

noted in terms of gender, age, and educational

level (P > 0.05) (Table 6).
0.026

0.734

0.183
P

Discussion

New findings from this study demonstrated that

5.514

0.341

3.395

all four pain intensity scales had good reliability

χ2

and validity when used with Chinese adults. When

assessing acute postoperative pain, all the correla-
Error

4
11

13
2

11
1
3

tions between the four scales were significant,

which is similar to previous reports [20,33,38,39].
The results showed that all four scales led to the
Right
VAS

same conclusions regarding a patient’s experience

63
44

86
21

53
25
29

of pain following surgery. However, the correla-

tions between the VDS and the FPS-R on current,
Education level

worst, and least pain, and the correlations between

Age (years)
Female

the VDS and the NRS on average pain, appeared

Table 5

Gender
Male

<HS

>HS
<60
≥60

weaker. The reason may be that the faces of the

HS

FPS-R could express other unpleasant feelings of

Postoperative Pain Intensity Assessment in Chinese
Table 6 Postoperative patients’ scale preference by gender, age, and education level (N = 160)
Gender
Male Female
VAS 4 3
NRS 27 12
VDS 16 21
FPS-R 41 36
χ2 5.366
P 0.147
FPS-R = Faces Pain Scale Revised; HS = high school education; NRS = numeric rating scale; VAS = visual analog scale; VDS = verbal descriptor scale.
pain more than intensity, and the numbers of the
NRS could better represent the average level of
pain, while the words of the VDS were too rough
to describe that sensation or express the average
pain intensity.
Prior studies have shown that the method of
retrospective pain scoring can produce reliable
ratings [45,46], and the most important aspect of
a pain measure’s validity is perhaps its sensitivity
or the ability to detect changes in pain over time
or due to treatment [21]. The 7-day postopera-
tive pain assessment in this study indicated that
each scale could discriminate different levels of
pain and display the changing trend of pain
intensity, regardless of whether current or retro-
spective pain was being assessed or not. That the
correlations between current and average pain
were stronger suggests that average pain intensity
could be a better indicator of retrospective pain.
Moreover, the four scales could also monitor
pain changes before and after analgesics, further
indicating that each scale could be used to assess
clinical pain in Chinese adults.
As using scales to assess pain is a learning and
participating process for the patients, Chinese
patients should be taught to correctly use pain
intensity scales. Pain assessment scales have not
been widely used in China, thus all the participants
in this study did not have any experience in using
the scales. The recalled pain and anticipated pain
were to teach the patients how to use the tools and
validate their understanding. Through learning,
the majority of them could understand and use the
selected four scales to rate their recalled pain or
anticipated postoperative pain. This suggests that
Chinese patients do not require any more practice
or training with assessment scales. When the
scales were used again to assess the actual postop-
erative pain, the number of subjects able to give
ratings increased day by day and the error rates of
the four scales dropped further. The reason for
improved ability to use the scales over postopera-
tive days may include improvement of physical and
231
Age (years) Education Level
<60 ≥60 <HS HS >HS
7 0 3 0 4
29 10 18 9 12
30 7 16 10 11
69 8 44 18 15
6.169 7.510
0.104 0.276
cognitive status during recovery, as well as famil-
iarity with the tools or experience in using the
tool.
The present results showed that the VAS was
the scale with the highest error rate (12.3%) at
first-time use among the four scales, consistent
with those reported in the literature (5.3% to
20%) [20,38]. As soon as patients learned how to
use the VAS, the error rate for assessing actual
postoperative pain declined, suggesting that the
greater number of levels of pain rated on the VAS
can accurately express a patient’s current pain
intensity. The NRS and the VDS (on a 0–10 scale)
are two simple and easy scales to describe pain
[31,32,47]. This study also found that these two
scales were easy to understand when assessing
recalled pain with low failure rates; however, error
rates were relatively high when assessing actual
postoperative pain, with responses falling between
two words or numbers, perhaps because the ability
to use the pain scales for postoperative pain report-
ing and for retrospective report of worst, least, and
average pain depends on the recovery process of
patients. In this study, the revised FPS was found
to have low error rates in rating both recalled pain
and actual surgical pain in Chinese adults, further
supporting its applicability and feasibility.
In the meantime, preferences as well as factors
affecting the successful completion of a scale
should also be considered for the selection of a
right tool. Nearly half of the patients preferred the
FPS-R across gender, age, and educational level.
This may be because the tool was simple and
reflected more than pain sensation. The prefer-
ence for the NRS and for the VDS were equal, as
numbers or words are simple to administer and
can be given either in written or verbal form.
However, few patients chose the VAS, saying it
was difficult for them to understand. The result is
similar to that of Stuppy [33], who reported that
53% of older patients preferred the FPS, followed
by the NRS (30%). The findings are also similar
to previous research conducted by Herr et al. [31]
232
and Taylor et al. [29], who supported the validity,
reliability as well as the preference of the FPS in
Caucasian or African American older adults. So
there may be an important cultural aspect to this
tool that explains why it is preferred by non-
Caucasian samples.
Overall, it can be drawn from the present study
that the FPS-R was the first choice for assessing
pain intensity, followed by the NRS or the VDS,
and the VAS was the last choice. As a satisfactory
pain rating scale must be reliable, valid, sufficiently
graded for capturing changes in pain intensity, eas-
ily understood, easy to score, and preferred by
patients and staff [15], many institutions in the
United States have incorporated the NRS, the
VDS, and the faces scale in clinical settings [48].
The present study also suggested that the use of
the FPS-R, the NRS, and the VDS as pain inten-
sity tool options may be appropriate to meet the
needs of individual Chinese adult patients.
It is also notable that, on average, the
patients’ postoperative pain was fairly well con-
trolled, with mean pain scores of 3 or less. How-
ever, the patients receiving prn analgesics had
poor pain control. We also found that 43.9% of
the patients (N = 173) would ask for analgesics
only when they could not tolerate the pain [7]. If
the nurses were somewhat reluctant to adminis-
ter prn analgesics, it would remain doubtful
whether there would be pain relief. This cer-
tainly points out the disadvantages of nurse-
administered prn analgesics for pain. So the
education of nurses regarding the use of pain
tools, as well as the importance of regular
administration of pain medications in postopera-
tive pain, is strongly recommended.
There are several limitations of the study that
should be considered. First, cognitive function was
not measured before participation, so some of the
patients, especially older adults, may have experi-
enced undetected cognitive impairment. However,
the low failure rates across age groups do not sup-
port an increase in cognitive impairment or inabil-
ity to use the tools with age. Nevertheless, the
exclusion of those who could not complete the pain
scales may have minimized this possibility. Second,
the order of pain scale completion (VAS, NRS,
VDS, and then FPS-R) was designed according
to the degree of conceptual difficulty reported
[20,31,33,38]. Thus, further study using pain scales
presented in random order would control for the
order effect. Third, the patients who participated
in this study were a selected population, operation
types were diverse, and analgesic choices were not
Li et al.
controlled. Further studies are needed to deter-
mine whether the results can be generalized to
other postoperative Chinese patients, including
those with cognitive impairment.
Conclusion
This study demonstrates that all four scales (VAS,
NRS, VDS, and FPS-R) are reliable and valid
tools for assessing postoperative pain in Chinese
adults. However, the FPS-R emerges as the best
scale with respect to reliability, validity, and error
and preference rates. Providing tool options to
address individual needs or preferences is
suggested.
Acknowledgments
The authors are indebted to Dr. Pingyan Chen for his
invaluable consultation and statistical analyses. The
authors would also like to thank Dr. Shiyuan Xu, Yajie Li,
Xiulan Wang, and Erwei Sun for their critical review and
advice. The authors are very grateful to all the nursing staff
on the Surgical Wards and Operating Room of Zhujiang
Hospital, in particular Head Nurse Huijuan Song and
Chengfan Gu, Staff Nurse Hongchao Ma, and Fang Tang,
for their kind assistance with data collection. To those
patients who participated in the study, the authors extend
special thanks.
References
1 American Pain Society. Pain: The fifth vital sign.
http://www.ampainsoc.org/advocacy/fifth.
htm1995-11-11 (accessed July 4, 2002).
2 McCaffery M, Pasero CL. Pain ratings: The fifth
vital sign. Am J Nurs 1997;97:15–6.
3 Merboth MK, Barnason S. Managing pain: The
fifth vital sign. Nurs Clin North Am 2000;35:375–
83.
4 Zalon ML. Nurses’ assessment of postoperative
patients’ pain. Pain 1993;54:329–34.
5 Carroll KC, Atkins PJ, Herold GR, et al. Pain
assessment and management in critically ill postop-
erative and trauma patients: A multisite study. Am J
Crit Care 1999;8:105–17.
6 Herr K, Titler MG, Schilling ML, et al. Evidence-
based assessment of acute pain in older adults: Cur-
rent nursing practices and perceived barriers. Clin J
Pain 2004;20:331–40.
7 Li L, Liu XQ. [A survey and analysis of postopera-
tive pain in surgical patients]. Zhonghua Hu Li Za
Zhi 2004;39:632–4.
8 Acute Pain Management Guideline Panel. Acute
pain management in adults: Operative or medical pro-
cedures and trauma. AHCPR Pub. No. 92-0019,
Rockville, MD: Agency for Health Care Policy and
Postoperative Pain Intensity Assessment in Chinese
Research, Public Health Service, U.S. Department
of Health and Human Services, 1992.
9 Clark FM, Szokol J, Kyu Koh E, Vender J. Pain
control and postoperative outcome. Anesthesiology
2002;97:1026.
10 Jamison RN, Ross MJ, Hoopman P, et al. Assess-
ment of postoperative pain management: Patient
satisfaction and perceived helpfulness. Clin J Pain
1997;13:229–36.
11 Gillies ML, Smith LN, Parry-Jones WL. Postoper-
ative pain assessment and management in adoles-
cents. Pain 1999;79:207–15.
12 Allcock N. Factors affecting the assessment of post-
operative pain: A literature review. J Adv Nurs
1996;24:1144–51.
13 Sjostrom B, Dahlgren LO, Haljamae H. Strategies
in postoperative pain assessment: Validation study.
Intensive Crit Care Nurs 1999;15:247–58.
14 Dahl JL. Effective pain management in terminal
care. Clin Geriatr Med 1996;12:279–300.
15 McCaffery M, Pasero C. Assessment: Underlying
complexities, misconceptions, and practical tools.
In: McCaffery M, Pasero C, eds. Pain: Clinical
Manual, 2nd edition. St. Louis: Mosby; 1999:35–
102.
16 Hicks CL, von Baeyer CL, Spafford PA, van Korlaar
I, Goodenough B. The Faces Pain Scale-Revised:
Toward a common metric in pediatric pain measure-
ment. Pain 2001;93:173–83.
17 Dalton JA, McNaull F. A call for standardizing the
clinical rating of pain intensity using a 0 to 10 rating
scale. Cancer Nurs 1998;21:46–9.
18 Carey SJ, Turpin C, Smith J, Whatley J, Haddox D.
Improving pain management in an acute care set-
ting. The Crawford Long Hospital of Emory Uni-
versity experience. Orthop Nurs 1997;16:29–36.
19 Wong DL, Baker CM. Pain in children: Compari-
son of assessment scales. Pediatr Nurs 1988;14:9–
17.
20 Briggs M, Closs JS. A descriptive study of the use
of visual analogue scales and verbal rating scales
for the assessment of postoperative pain in ortho-
pedic patients. J Pain Symptom Manage 1999;
18:438–46.
21 Jensen MP, Chen C, Brugger AM. Postsurgical pain
outcome assessment. Pain 2002;99:101–9.
22 Mom T, Bazin JE, Commun F, et al. Assessment
of postoperative pain after laryngeal surgery for
cancer. Arch Otolaryngol Head Neck Surg
1998;124:794–8.
23 Gagliese L, Katz J. Age differences in postoperative
pain are scale dependent: A comparison of measures
of pain intensity and quality in younger and older
surgical patients. Pain 2003;103:11–20.
24 Martinelli AM. Pain and ethnicity. How people
of different cultures experience pain. AORN J
1987;46:273–4.
25 Keefe FJ, Jacobs M, Underwood-Gordon L. Biobe-
havioral pain research: A multi-institute assessment
233
of cross-cutting issues and research needs. Clin J
Pain 1997;13:91–103.
26 McGuire DB. Comprehensive and multidimen-
sional assessment and measurement of pain. J Pain
Symptom Manage 1992;7:312–9.
27 Aun C, Lam YM, Collett B. Evaluation of the use
of visual analogue scale in Chinese patients. Pain
1986;25:215–21.
28 Zhao JJ, Lu XY, Zhao CF, Tuo Y. [The research
on the correlation between the numerical rating
scale and verbal rating scale and the improvement
on pain assessment]. Xian Dai Hu Li 2002;9:660–
1.
29 Taylor LJ, Herr K. Pain intensity assessment: A
comparison of selected pain intensity scales for use
in cognitively intact and cognitively impaired
African American older adults. Pain Manag Nurs
2003;4:87–95.
30 Bieri D, Reeve RA, Champion GD, Addicoat L,
Ziegler JB. The Faces Pain Scale for the self-assess-
ment of the severity of pain experienced by children:
Development, initial validation, and preliminary
investigation for ratio scale properties. Pain
1990;41:139–50.
31 Herr KA, Mobily PR, Kohout FJ, Wagenaar D.
Evaluation of the Faces Pain Scale for use with the
elderly. Clin J Pain 1998;14:29–38.
32 Herr KA, Spratt K, Mobily PR, Richardson G. Pain
intensity assessment in older adults: Use of experi-
mental pain to compare psychometric properties
and usability of selected pain scales with younger
adults. Clin J Pain 2004;20:207–19.
33 Stuppy DJ. The Faces Pain Scale: Reliability and
validity with mature adults. Appl Nurs Res
1998;11:84–9.
34 Taylor LJ, Harris J, Epps CD, Herr K. Psychomet-
ric evaluation of selected pain intensity scales for use
with cognitively impaired and cognitively intact
older adults. Rehabil Nurs 2005;30:55–61.
35 Huskisson EC. Measurement of pain. Lancet
1974;2:1127–31.
36 Carlsson AM. Assessment of chronic pain. I. Aspects
of the reliability and validity of the visual analogue
scale. Pain 1983;16:87–101.
37 Hawksley H. Pain assessment using a visual ana-
logue scale. Prof Nurse 2000;15:593–7.
38 Jensen MP, Karoly P, Braver S. The measurement
of clinical pain intensity: A comparison of six meth-
ods. Pain 1986;27:117–26.
39 Hartrick CT. A four-category verbal rating
scale (VRS-4), an 11-point numeric rating scale
(NRS-11), and a 100-mm visual analog scale
(VAS) were compared in the assessment of acute
pain after oral surgery. Clin J Pain 2001;17:
104–5.
40 Price DD, McGrath PA, Rafii A, Buckingham B.
The validation of visual analogue scales as ratio scale
measures for chronic and experimental pain. Pain
1983;17:45–56.
234
41 Jensen MP, Miller L, Fisher LD. Assessment of pain
during medical procedures: A comparison of three
scales. Clin J Pain 1998;14:343–9.
42 Paice JA, Cohen FL. Validity of a verbally adminis-
tered numeric rating scale to measure cancer pain
intensity. Cancer Nurs 1997;20:88–93.
43 Herr KA, Mobily PR. Comparison of selected pain
assessment tools for use with the elderly. Appl Nurs
Res 1993;6:39–46.
44 Ohnhaus EE, Adler R. Methodological problems in
the measurement of pain: A comparison between
the verbal rating scale and the visual analogue scale.
Pain 1975;1:379–84.
Li et al.
45 Liu WH, Aitkenhead AR. Comparison of contem-
poraneous and retrospective assessment of postop-
erative pain using the visual analogue scale. Br J
Anaesth 1991;67:768–71.
46 Babul N, Darke AC, Johnson DH, Charron-
Vincent K. Using memory for pain in analge-
sic research. Ann Pharmacother 1993;27:9–12.
47 McCaffery M. Using the 0-to-10 pain rating scale.
Am J Nurs 2001;101:81–2.
48 Mayer DM, Torma L, Byock I, Norris K. Speak-
ing the language of pain. Am J Nurs 2001;101:44–
9.