Accepted Manuscript

Title: Paediatric nurses’ postoperative pain management

practices in hospital based non-critical care settings: A
narrative review

Author: Alison Twycross Paula Forgeron Anna Williams

PII: S0020-7489(15)00010-3
DOI: http://dx.doi.org/doi:10.1016/j.ijnurstu.2015.01.009
Reference: NS 2500

To appear in:

Received date: 20-9-2013

Revised date: 16-1-2015
Accepted date: 16-1-2015

Please cite this article as: Twycross, A., Forgeron, P., Williams, A.,Paediatric
nurses’ postoperative pain management practices in hospital based non-critical
care settings: A narrative review, International Journal of Nursing Studies (2015),
http://dx.doi.org/10.1016/j.ijnurstu.2015.01.009

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
 Paediatric nurses’ postoperative pain management practices in
hospital settings: A systematic review

Authors:

Alison Twycross, PhD, RN

London South Bank University

t
ip
Email: a.twycross@lsbu.ac.uk

cr
Paula Forgeron, PhD, RN

us
University of Ottawa

Email: paula.forgeron@uottawa.ca
an
M
Anna Williams, PhD

Centre for Nursing and Allied Health Research, Great Ormond Street Hospital for Children
NHS Foundation Trust
d

Email: anna.williams@gosh.nhs.uk
e
pt
ce
Ac

Page 1 of 56
 1
2
3 Paediatric nurses’ postoperative pain management practices in
4
5 hospital based non-critical care settings: A narrative review
6
7
8 Abstract
9
10
Objectives: To investigate paediatric nurses’ postoperative pain management practices
11
12

t
13 with the aim of identifying the factors associated with undermanaged paediatric

ip
14
15 postoperative pain.
16

cr
17
18
19 Design: Systematic search and review

us
20
21
22
23
24 Data sources: PsychInfo, CINAHL, PubMed, EMBASE and hand searching

an
25
26
27
28 Review methods
M
29
30
31 English peer-reviewed quantitative, qualitative, or mixed methods research articles
32
33 published between 1990 and 2012 exploring registered nurses’ paediatric postoperative
d

34
35 pain management practices were included. Articles with a primary focus on nurses’ pain
te

36
37 management practices in the neonatal or paediatric intensive care units, recovery room,
38
p

39
40
and/or focused on children with cognitive impairment were excluded. The search terms
41
ce

42 used were: postoperative pain; nurs*; paediatrics; pediatrics; children; pain assessment;
43
44 non-pharm*; analges*. Titles and abstracts were used for initial screening. Two
45
Ac

46 researchers conducted data extraction and assessment of rigour for each paper.
47
48
49
50
51 Results
52
53 From the initial 248 citations, 27 studies were included. Most studies were descriptive
54
55 and examined relationships between personal factors and nurses' pain management
56
57 practices. Observational data from four papers added insights beyond that provided in
58
59
60
61
1
62
63
64 Page 2 of 56
65
 1
2
3 self-report studies. Two articles used experimental designs with vignettes. Data were
4
5 categorised into four topics: pain assessment; pharmacological practices; non-
6
7 pharmacological practices; and factors affecting practices. Despite improvements in
8
9
10 analgesic administration over the past 20 years, practices remain suboptimal. Children’s
11
12 behaviour appears to influence nurses’ pain assessment more than validated measures.

t
13

ip
14 A significant proportion of children did not have pain scores recorded in the first 24-hours
15
16 postoperatively. Children receive more analgesia when ordered around the clock

cr
17
18
19 compared to as required. However, around the clock analgesia prescription did not

us
20
21 guarantee administration. Nurses reported using several non-pharmacological strategies
22
23 routinely but some are not evidence based.
24

an
25
26
27
28
Conclusions
M
29
30 The results of this review indicate nurses’ assessment and management of children’s
31
32 pain is not consistent with published guidelines. Results of studies exploring nurse and
33
d

34 child related factors are inconclusive. Research needs to examine the impact of
35
te

36
organisational factors on nurses’ pain care practices. Intervention studies are needed to
37
38
determine the most effective strategies to support and improve nurses’ pain care for
p

39
40
41 children.
ce

42
43
44
45 Keywords: paediatric; children; nurses; postoperative; pain, pain assessment, pain
Ac

46
47
48 management.
49
50
51
52
53
54
55
56
57
58
59
60
61
2
62
63
64 Page 3 of 56
65
 1
2
3 What is already known about the topic?
4
5  Despite research about the management of pain and some improvements over time
6
7
to clinical practice, children continue to experience avoidable moderate to severe
8
9
10 pain in hospital postoperatively.
11
12  Gaps remain in nurses’ knowledge about paediatric pain; knowledge alone is

t
13

ip
14 insufficient to change practice.
15
16
 Several reasons have been postulated as contributing to continuing suboptimal pain

cr
17
18
19 management including nurses’ perceptions of children’s pain and pain management

us
20
21 and organisational culture.
22
23
24

an
25
26
27
What this paper adds
28

M
29 Nurses may have preconceived ideas about how children should behave
30
31 postoperatively. Behavioural cues of pain appear to take precedence over a child’s
32
33
d
self-report.
34
35

te

36 Analgesic administration has improved over the past 20 years but practices are still
37
38 not optimised.
p

39
40  Non-pharmacological interventions are not used as often as they could be and are
41
ce

42 sometimes used inappropriately.

43
44
45
Ac

46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
3
62
63
64 Page 4 of 56
65
 1
2
3 1. Introduction
4
5 There have been numerous paediatric postoperative pain management studies over the
6
7
8 past 20 years, yet children continue to experience avoidable postoperative pain
9
10 (Kozlowski et al. 2014, Twycross & Finley 2013). This is despite evidence to guide pain
11
12 management practices being readily available in the form of clinical guidelines. In

t
13

ip
14 England, national acute pain management recommendations are not followed in more
15
16
than a quarter of hospitals (National Confidential Enquiry into Patient Outcome and

cr
17
18
19 Death 2011) and children in low and middle income countries have been found to

us
20
21 experience high rates of pain (Forgeron et al. 2005) suggesting pain management
22
23 remains a global problem. Unrelieved pain has a number of undesirable physiological
24

an
25
26
and psychosocial consequences that can affect the child at the time and later in life, for
27
28 example, through increased sensitivity to later pain events (Grunau et al. 1998, Saxe et
M
29
30 al. 2001, Taddio et al. 2002). It is, therefore, imperative to ensure pain is managed
31
32 effectively.
33
d

34
35
te

36
37 The reasons why children continue to experience avoidable postoperative pain need to
38
be the focus of on-going research if we are to design interventions to bring about
p

39
40
41 significant improvement. Despite the amount of literature available no reviews have been
ce

42
43 carried out in relation to nurses’ paediatric postoperative pain management. One
44
45
Ac

46 literature review focused on the use of pain assessment tools in acute paediatric pain
47
48 and concluded there was no robust evidence to support their use in practice (Franck &
49
50 Bruce 2009). However, this study did not separate postoperative pain assessment
51
52 practices from other forms of acute pain such as cancer pain and, therefore, it is not
53
54
55
known if there are differences in how nurses use pain assessment tools when pain may
56
57 change more substantially like over the course of the postoperative period. One
58
59 Cochrane review looked at the effectiveness of non-pharmacological methods for
60
61
4
62
63
64 Page 5 of 56
65
 1
2
3 procedural pain (Uman et al. 2010) but does not describe how often nurses use these in
4
5 their practice. Another review has focused on the use of relaxation for the treatment of
6
7 paediatric chronic pain (Eccleston et al. 2012) and although evidence supports the use
8
9
10 of relaxation in the treatment of paediatric headache, it does not shed light on nurses’
11
12 use of relaxation or other non-pharmacological strategies in their care of children during

t
13

ip
14 the postoperative period. Other reviews have focused on the adult population. This
15
16 systematic search and review, therefore, set out to investigate paediatric nurses’

cr
17
18
19 postoperative pain management practices. The aim of this review was to identify factors

us
20
21 associated with undermanaged paediatric postoperative pain, whether improvements in
22
23 practice occurred during the time period selected, and the implications for nursing
24

an
25 practice and research.
26
27
28
M
29
30 2. Methods
31
32 A systematic search and review is the suggested approach when the question is broad
33
d

34
35 as in this paper. It incorporates multiple study types to provide a complete picture of the
te

36
37 prevalence of research on a topic (Grant & Booth 2009). This method of review
38
p

39 combines the search strategies and inclusion and exclusion criteria associated with
40
41
ce

systematic reviews as well as the analytical synthesis of a critical review. Although it is

42
43
44 not necessary to subject the articles included to a methodological critique in a systematic
45
Ac

46 search and review, Caldwell et al.’s (2011) framework was used to provide insights into
47
48 the strengths and weaknesses of the studies included in this review. A description of this
49
50 framework is provided in Section 2.2.2.
51
52
53
54
55 2.1 Inclusion/Exclusion Criteria
56
57 Inclusion criteria were:
58
59  Peer-reviewed primary research exploring registered nurses’ postoperative pain
60
61
5
62
63
64 Page 6 of 56
65
 1
2
3 management practices
4
5  Research carried out in the paediatric population (newborn to 18 years of age).
6
7
8  Quantitative, qualitative, or mixed methods studies.
9
10  Published in English between 1990 and October 2012 (to allow any improvements in
11
12 practices over time to be identified).

t
13

ip
14
15 Studies that included other health professionals or other patient populations were
16

cr
17 considered if the data on paediatric postoperative pain management by nurses were
18
19 reported separately.

us
20
21
22
23
24

an
The exclusion criteria were:
25
26
27  Studies relating primarily to pain management in the neonatal or paediatric intensive
28
M
29 care unit and the recovery room as the complexity of patients (i.e. ventilated,
30
31 paralysed, sedated) and the staffing ratios make comparison to ward based pain
32
33
d
care difficult.
34
35

te

36 Studies where the primary research population were children with cognitive
37
38 impairment.
p

39
40  Postoperative pain management after discharge.
41
ce

42
 Books, book chapters, commentaries, and dissertations.
43
44
45
Ac

46
47
48 2.2 Data Sources and Selection
49
50 Electronic database searches (PsychInfo, CINAHL, PubMed and EMBASE) were
51
52 conducted by a medical librarian using the following search terms: postoperative pain;
53
54
55
nurs*; paediatrics; pediatrics; children; pain assessment; non-pharm*; analges*. To
56
57 ensure relevant papers were not missed in the database search the authors hand
58
59 searched the reference list of each included article. The number of articles extracted
60
61
6
62
63
64 Page 7 of 56
65
 1
2
3 from each database can be seen in Table 1. Twenty-seven articles were subsequently
4
5 included in the review. Since the inclusion of dissertation evidence has rarely impacted
6
7 the conclusions of systematic reviews (Vickers & Smith 2000) and due to the difficulties
8
9
10 often encountered in retrieving these sources, dissertations were not included. Two
11
12 researchers (AT, PF) carried out initial screening of all the databases using study titles

t
13

ip
14 and abstracts and recorded the number meeting the criteria, the number excluded,
15
16 reasons for exclusion, and if the article had to be reviewed to determine eligibility.

cr
17
18
19 Disagreements on included and excluded articles were to be reviewed by a third

us
20
21 reviewer but this was not required. Once agreement was reached on eligibility, each
22
23 article was read in full by two of the authors to assess methodological rigour and extract
24

an
25 the data pertaining to postoperative nursing care practices. The selection process is
26
27
28
illustrated in Figure 1.
M
29
30
31
32 2.2.1 Data extraction and synthesis
33
d

34 Data from each study was extracted by two of the authors using customised forms
35
te

36
specifying authors, country, study design, sample characteristics, data collection
37
38
techniques, and postoperative pain care findings linked to nursing. Findings were
p

39
40
41 organised into four main categories: assessment, documentation and communication;
ce

42
43 pharmacological practices; non-pharmacological practices; and factors affecting
44
45 practices. The first three categories relate to best practice as identified in current clinical
Ac

46
47
48 guidelines (Association of Paediatric Anaesthetists 2012) and the fourth allows any
49
50 factors impacting on practice to be highlighted in line with the aim of the review. Most
51
52 studies contained data related to more than one category. Details of the individual
53
54 studies are included in Table 2.
55
56
57
58
59
60
61
7
62
63
64 Page 8 of 56
65
 1
2
3 2.2.2 Assessment of methodological quality
4
5 Unlike systematic reviews and meta-analyses that are conducted to synthesise findings
6
7 of studies to determine effectiveness, systematic search and reviews, like scoping
8
9
10 reviews, are undertaken to summarize the activity in the field, identify gaps, determine
11
12 the need for a systematic review, and to summarize the findings for dissemination

t
13

ip
14 (Arksey & O’Malley 2005). Given their aims, most scoping reviews do not assess the
15
16 quality of bias in the included studies (Davis et al. 2009, Grant & Booth 2009). Whether

cr
17
18
19 or not it is appropriate not to assess the study quality to determine the risk of bias has

us
20
21 been debated (Levac et al. 2010). In the light of this a framework was used to guide the
22
23 assessment of scientific quality (risk of bias) of the included articles (Caldwell et al.
24

an
25 2011). This framework consists of an overall approach to study critique using specific
26
27
28
items based on the methodology and provides a list of criteria for qualitative, quantitative,
M
29
30 and mixed methods research to assist the reader in assessing the fidelity of the study to
31
32 its stated design and determine the dependability of the results. Items include such
33
d

34 things as: is there a hypothesis; are key variables defined; is the selection of participants
35
te

36
described and sampling method defined; and are major concepts defined. It does not
37
38
produce a single numerical score to represent quality but rather guides the reader to
p

39
40
41 examine all aspects of a study to help determine strengths and weaknesses. Insights
ce

42
43 into the major weaknesses and strengths of the included articles are included in Table 2.
44
45 No articles were excluded based an assessment of study quality as biases are inherent
Ac

46
47
48 in the wide range of research designs included in this review. Including all studies that
49
50 meet inclusion criteria, regardless of quality assessment, is in keeping with recent
51
52 practices in systematic reviews (Higgins & Green 2011) but caution is noted in the
53
54 reporting of findings below when the study quality was found to be at a higher risk for
55
56
57 bias above the inherent risk within the type of study. Small sample sizes and single site
58
59 studies limited the generalizability of results in some studies. For instance, from 18
60
61
8
62
63
64 Page 9 of 56
65
 1
2
3 quantitative studies, a small sample size was identified as a key weakness in five papers
4
5 (Asprey 1994, Boughton et al. 1998, Rheiner et al. 1998, Twycross 2007b, 2008). Other
6
7 studies lacked essential information to judge the appropriateness of sample, data
8
9
10 collection method, and analysis. The quantitative studies judged to be most robust
11
12 employed experimental designs (Hamers et al. 1996, Hamers et al. 1997, Hudson-Barr

t
13

ip
14 et al. 1998). Where the design was chart review, stronger studies included an
15
16 appropriate control group (Boughton et al. 1998) or employed appropriate inferential

cr
17
18
19 statistics, whereas weaker quantitative studies overall were purely descriptive.

us
20
21 Qualitative studies evaluated as being of higher quality clearly explained the selection
22
23 and use of appropriate methodological approaches relevant to the phenomena being
24

an
25 investigated (Byrne et al. 2001, Twycross 2007a, Woodgate & Kristjanson 1996) or
26
27
28
addressed issues of rigour and trustworthiness (Woodgate & Kristjanson 1996).
M
29
30
31
32 3. Results
33
d

34
3.1 Study details
35
te

36
37 Of the 28 studies reviewed, 10 included data about nurses’ pain assessment practices,
38
10 included data on nurses’ postoperative analgesic practices, six reported on nurses’
p

39
40
41 use of non-pharmacological strategies and 14 reported on factors impacting on practice.
ce

42
43 There was a mix of quantitative (n=18), qualitative (n=4), and mixed-methods (n=6)
44
45
Ac

46 studies, using a range of data collection methods. Key methodological issues identified
47
48 through critique using Caldwell’s framework included: not enough details provided in the
49
50 article to make a judgement on a framework criteria; limited sampling; use of study
51
52 specific (potentially not valid) tools; limited justification for statistical tests; no
53
54
55
philosophical foundation in qualitative studies; and limited or no details of thematic
56
57 analysis approach. For details of specific methodological issues identified for individual
58
59 studies please see Table 2.
60
61
9
62
63
64 Page 10 of 56
65
 1
2
3
4
5 3.2 Pain assessment and documentation
6
7 3.2.1 Strategies used to assess pain
8
9
10 The strategies nurses used to assess pain in children are discussed in five papers
11
12 (Nethercott 1994, Rheiner et al. 1998, Salantera et al. 1999, Twycross 2007a, Woodgate

t
13

ip
14 & Kristjanson 1996) (Table 2). These papers give us an insight into nurses’ pain
15
16 assessment practices. A variety of data collection tools were used including an interview,

cr
17
18
19 chart audits, observational techniques and mixed methods. Generally there is consensus

us
20
21 in results across all five studies that means we can be fairly confident an accurate
22
23 picture of practice has been obtained. It is worth noting that nurses collected data on
24

an
25 their own practices in Rheiner et al.’s study and this may have had an impact on the
26
27
28
results. The pain assessment strategies used include: pain assessment tools,
M
29
30 behavioural cues and physiological cues.
31
32
33
d

34 Some, but not all, nurses use pain assessment tools. In an interview study, three out of
35
te

36
10 nurses reported using pain assessment tools but indicated they were not used
37
38
routinely (Nethercott 1994). In another study, 34% of nurses (n=303) completing a
p

39
40
41 questionnaire indicated a pain assessment tool was available in their clinical area, and
ce

42
43 the most commonly used tool was the Wong and Baker Faces scale (Salantera et al.
44
45 1999). Four per cent of nurses reported using a Visual Analogue Scale (Salantera et al.
Ac

46
47
48 1999). Even when nurses used assessment tools they did not use them consistently.
49
50 The Wong and Baker Faces scale was reportedly used regularly by 18% of nurses, but
51
52 only with some children in pain by another 12% of nurses (Salantera et al. 1999). Further
53
54 evidence of the continued inconsistent use of a pain assessment tool was reported in an
55
56
57 observational study (Twycross 2007a). In Twycross’ study, nurses were only seen using
58
59 a pain assessment tool on three occasions despite there being at least 22 other times
60
61
10
62
63
64 Page 11 of 56
65
 1
2
3 when it would have been appropriate to do so.
4
5
6
7 Behavioural cues continue to be an important consideration when assessing pain in
8
9
10 children of all ages. Interviews with nurses revealed behavioural changes were an
11
12 important consideration when assessing a child’s postoperative pain with facial

t
13

ip
14 expression being particularly important (Nethercott 1994). Nurses’ most frequent
15
16 response when asked how they knew a child was in pain were crying and other forms of

cr
17
18
19 verbal behaviour, with overt forms of behaviour such as fighting or restlessness being

us
20
21 second (Woodgate & Kristjanson 1996). In another study, if a child self-reported their
22
23 pain and requested analgesics, nurses considered the child’s non-verbal behaviours
24

an
25 before administering analgesics (Rheiner et al. 1998). In a chart review, the reliance on
26
27
28
behavioural signs was high with 98% of nurses (n=303) stating they assessed pain by
M
29
30 observing changes in a child’s behaviour (Salantera et al. 1999). More recently, in an
31
32 observational study nurses stated they used behavioural cues to assess a child’s pain
33
d

34 but did not always appear to act on them (Twycross 2007a).

35
te

36
37
38
The importance attributed to physiological cues such as vital signs as a way of
p

39
40
41 assessing children’s pain is unclear. In one study, 80% of nurses (n=303) reported using
ce

42
43 changes in physiological indicators to ascertain whether a child was in pain (Salantera et
44
45 al., 1999), as did nurses in two other studies (Nethercott 1994, Woodgate & Kristjanson
Ac

46
47
48 1996). However, Rheiner and colleagues (1998) found physiological signs were rarely
49
50 used. Physiological indicators are now considered an unreliable measure of a child’s
51
52 pain (von Baeyer & Spagrud 2007) and this perhaps explains why their use is not
53
54 explored in studies from 2000 onwards.
55
56
57
58
59
60
61
11
62
63
64 Page 12 of 56
65
 1
2
3 3.2.3 Does using a pain assessment tool make a difference?
4
5 Whether using pain assessment tools results in better pain care (lower pain intensity)
6
7 has been explored in two studies with conflicting results (Boughton et al. 1998, Rheiner
8
9
10 et al. 1998). An intervention study explored the impact of introducing a pain assessment
11
12 tool but found this did not increase the amount of analgesics administered (Boughton et

t
13

ip
14 al. 1998). In another study, a small but significant relationship was found between self-
15
16 reported pain scores using the Oucher [a pain assessment tool comprising a

cr
17
18
19 photographic faces scale and a 0-10 vertical numerical rating scale (Beyer et al. 2009)]

us
20
21 and the amount of analgesic drugs administered (Rheiner et al. 1998). As mentioned
22
23 previously, nurses in Rheiner et al.’s study collected data on their own practices and this
24

an
25 may have had an impact on the results as could the relatively small sample size (n=19).
26
27
28
M
29
30 3.2.4 Documentation
31
32 Four papers contained data about the quality of nursing documentation in relation to pain
33
d

34 management (Gillies et al. 1999, Prot-Labarthe et al. 2008, Shrestha-Ranjit & Manias
35
te

36
2010, Simons & Moseley 2009, Twycross 2007a). Overall the quality of documentation
37
38
was poor. In one chart audit carried out in the UK, 72% of children (n=175) had a pain
p

39
40
41 score recorded, although a quarter of children did not have a score recorded in the first
ce

42
43 24-hours postoperatively (Simons & Moseley 2009). Another chart audit in Australia
44
45 found assessment of pain using a specific tool was not documented in 87.8% of charts
Ac

46
47
48 (n=106); this was significantly less often than would have been expected according to
49
50 hospital protocol (Shrestha-Ranjit & Manias 2010). The findings of a study comparing
51
52 practices with patients in two different Western countries (France [n=100] and Canada
53
54 [n=100]) noted 2,388 numerical pain ratings in the charts, 71.4% (n=1,706) in one
55
56
57 hospital and 28.6% (n=682) in the other during the five days post-surgery (Prot-Labarthe
58
59 et al. 2008).
60
61
12
62
63
64 Page 13 of 56
65
 1
2
3 Inconsistencies were also noted in relation to documentation about the reassessment of
4
5 pain and analgesics given. In one observational study documentation did not include an
6
7 evaluation of the effectiveness of any pain-relieving interventions (Twycross 2007a). In
8
9
10 another study, 30% of children’s charts (n=50) had no comments about pain during the
11
12 first 24-hours postoperatively (Salantera et al. 1999). Documentation relating to pain in

t
13

ip
14 the study by Salantera and colleagues, declined in the second 24-hours postoperatively
15
16 with only 57% of the charts (n=50) having no comment about pain during this period.

cr
17
18
19 Even when analgesics were administered documentation remained problematic. In

us
20
21 another study with adolescents (n=351) nurses only documented that analgesics were
22
23 given in relation to pain assessment 43% of the time, and documentation of
24

an
25 reassessment following analgesic administration occurred 12% of the time (Gillies et al.
26
27
28
1999). In Gillies et al.’s study, pain was not mentioned in the patient’s notes
M
29
30 approximately 35% of the time. In a chart audit, information about non-pharmacological
31
32 interventions was only documented in 23.6% of charts (n=106) (Shrestha-Ranjit &
33
d

34 Manias 2010).
35
te

36
37
38
3.3 Analgesics
p

39
40
41 Ten studies (from Europe, Australia, and North America) explored nurses’ analgesic
ce

42
43 administration practices. These papers explored the association of such things as
44
45 prescription schedule (i.e. whether prescribed around the clock or pro re nata), route,
Ac

46
47
48 and medication type, with the analgesic administration practices of nurses.
49
50
51
52 3.3.1 Pro re nata versus around the clock administration
53
54 Seven studies explored the influence of the prescription regime on nurses’ postoperative
55
56
57 analgesic administrative practices (Higgins et al. 1999, Nethercott 1994, Romsing 1996,
58
59 Shrestha-Ranjit & Manias 2010, Simons & Moseley 2008, Smyth et al. 2011, Twycross
60
61
13
62
63
64 Page 14 of 56
65
 1
2
3 2007a). The studies that used chart review methods found that regardless of medication
4
5 type (opioid or non-opioid) or surgery type (i.e. general, cardiac, orthopaedic, ear, nose
6
7 and throat), if medication was ordered around the clock it was administered to the child
8
9
10 more often than if the medication was ordered pro re nata (Higgins et al. 1999, Shrestha-
11
12 Ranjit & Manias 2010, Simons & Moseley 2008). When the analgesic was codeine and

t
13

ip
14 prescribed around the clock 65% of children (n=175) received doses whereas if the
15
16 analgesic was paracetamol 96% of children received doses (Simons & Moseley 2008).

cr
17
18
19 However, being ordered around the clock did not guarantee that the child would receive

us
20
21 all prescribed doses. Children (n=114) all had paracetamol prescribed around the clock
22
23 in Higgins et al.’s (1999) study; 92.3% (518 of 561) of the scheduled doses were
24

an
25 administered. If the analgesic was prescribed pro re nata only 43% of the total possible
26
27
28
number of doses were administered to children when the analgesic was codeine (Simon
M
29
30 & Moseley, 2008) and 60% to 74% received a dose of paracetamol (Shrestha-Ranjit &
31
32 Manias 2010, Simons & Moseley 2008). In a recent study, 76% (n=72) children had a
33
d

34 pro re nata analgesic ordered but only about half of those (n=46) actually received an
35
te

36
analgesic and these were administered mostly on the day of surgery and the first day
37
38
postoperatively (Smyth et al. 2011). When the children did not receive around the clock
p

39
40
41 analgesics, perhaps unsurprisingly, their usage of pro re nata analgesics increased from
ce

42
43 41.7% to 85.3% (Higgins et al. 1999).
44
45
Ac

46
47
48 There are a couple of qualifiers to these findings. First, the provision of other forms of
49
50 analgesia may contribute to nurses being reluctant to provide available doses of non-
51
52 opioid medication. For example, in one study although only about half of the of children
53
54 (n=46) who were ordered a pro re nata medication received a dose, just over 70% of the
55
56
57 participants (n= 67) had an analgesia prescribed around the clock for the day of and into
58
59 day one of the postoperative period (Smyth et al. 2011). In another study, although
60
61
14
62
63
64 Page 15 of 56
65
 1
2
3 children did not receive all of their regularly prescribed doses of paracetamol the children
4
5 were also prescribed a regular local anaesthetic, although not all the children received
6
7 all the ordered doses of the local anaesthetic (Higgins et al. 1999). Second, the studies
8
9
10 were based on chart reviews, reported descriptive statistics, were of relatively small
11
12 sample sizes, and did not conduct inferential statistical testing making it difficult to

t
13

ip
14 determine the generalizability of these findings (Table 2). Third, it is possible that the
15
16 withholding of analgesic doses in both the around the clock and pro re nata groups were

cr
17
18
19 due to contraindications. However, no data were presented that indicated nurses

us
20
21 withheld either opioid or non-opioid doses due to contraindications. Given the
22
23 withholding of analgesics is consistent across studies and no author included the
24

an
25 possibility that their findings were due to contraindications, suggests the withholding of
26
27
28
analgesics was due to other factors, but this warrants further study.
M
29
30
31
32 In the studies that used observation and/or interview methods (Nethercott 1994, Smyth
33
d

34 et al. 2011, Twycross 2007a) nurses were found to be reluctant to administer analgesics
35
te

36
(opioid and non-opioid). Nurses’ decisions to administer pro re nata analgesics were
37
38
reported to be influenced by such factors as the nurses’ experience, patients’ verbal and
p

39
40
41 behavioural signs, parent report, not wanting to wake the child, and a preference for
ce

42
43 non-pharmacological measures (Smyth et al. 2011). Based on the range of years that
44
45 these studies covered (1994 to 2011) this reluctance appears to have continued over
Ac

46
47
48 time.
49
50
51
52 It may not only be the pro re nata nature of ordered analgesics that negatively affects
53
54 nurses’ decision-making on administration postoperatively. One study found a specific
55
56
57 issue with rectal dosing (Romsing 1996). Nurses were found to round down the rectal
58
59 dose of paracetamol (despite suppositories being supplied in doses that enabled
60
61
15
62
63
64 Page 16 of 56
65
 1
2
3 administration of the correct dose) and lengthen the dosing intervals. Both these actions
4
5 resulted in suboptimal analgesia being administered to children in the first 24-hours post-
6
7 tonsillectomy. However, despite 72 patients being included in this study the analgesic
8
9
10 administration practices of only two nurses were observed. The findings may, therefore,
11
12 be more an indication of the consistency of analgesic administration practices of these

t
13

ip
14 nurses than a consistent finding among nurses more widely (Table 2).
15
16

cr
17
3.3.2 Opioids
18
19
Six articles reported on nurses’ opioid administrative analgesic practices (Asprey 1994,

us
20
21
22 Gillies et al. 1999, Nethercott 1994, Salantera et al. 1999, Twycross 2007a, Woodgate &
23
24 Kristjanson 1996). Based on the publication dates most of these studies may represent

an
25
26 previous opioid administration practices. It is possible that nurses’ practices were
27
28
improving over this period and thus contribute to the inconsistent findings across studies.
M
29
30
31 For example, only 8% (n=2) of children with various types of surgery did not have an
32
33 order for an opioid postoperatively. Out of the 92% (n=23) who did have an opioid order
d

34
35 only 13% (n=3) of these children did not receive a dose of an opioid postoperatively
te

36
37
38 (Asprey 1994). Similarly, 92% (n=46) of children and adolescents received opioids within
p

39
40 the first 24-hours post appendectomy surgery (Salantera et al. 1999). Both of these
41
ce

42 studies suggest nurses may not be reluctant to administer opioids. However, another
43
44 chart review study found that only 55% (n=196) of adolescent patients having elective
45
Ac

46
47
surgeries had an opioid ordered and out of those 196 patients 71% (n=132) did not
48
49 receive any opioids within the first 24 hours of surgery (Gillies et al. 1999). Even the
50
51 dose of opioid administered when a dose range is available to nurses is non-consistent
52
53 in studies. Asprey (1994) found in 63% of the cases nurses administered the higher
54
55
dose of opioid whereas in an observational study nurses (n=13) were observed to
56
57
58 provide the lower dose of an opioid when they had the choice (Twycross 2007a).
59
60
61
16
62
63
64 Page 17 of 56
65
 1
2
3
4
5 Studies that included interviews provide some insights into the inconsistent findings in
6
7 nurses’ administration of opioids to children. It may be that myths and misconceptions
8
9
10 regarding opioids are starting to be corrected. In 1994, nine out of 10 nurses stated older
11
12 nurses and medical staff were overly concerned about opioid addiction and would not

t
13

ip
14 prescribe or administer them postoperatively (Nethercott, 1994), and nurses in two other
15
16 studies expressed fears that children would become addicted to pain medications

cr
17
18
19 (Gillies et al. 1999, Woodgate & Kristjanson 1996). Although in all probability nurses’

us
20
21 knowledge about opioids has improved over time some still do not maximise analgesics
22
23 for paediatric patients postoperatively which may be partly due to continued
24

an
25 misconceptions. For example, Twycross (2007) observed that nurses generally did not
26
27
28
engage in discussions with parents regarding the administration of opioids except when
M
29
30 it came to discharge instructions. In this circumstance they were observed to often
31
32 encourage the use of paracetamol and ibuprofen and discourage the use of codeine.
33
d

34 However, this study, like many in this review, is based on a small sample size in one
35
te

36
hospital. Opioid administration practice of paediatric nurses in the postoperative period,
37
38
therefore, warrants further research.
p

39
40
41
ce

42
43 3.4 Non-pharmacological methods
44
45 Six studies examined nurses’ use of non-pharmacological strategies in managing
Ac

46
47
48 children’s postoperative pain (He et al. 2011, He et al. 2005, Polkki et al. 2003, Polkki et
49
50 al. 2001, Rheiner et al. 1998, Twycross 2007a) (Table 2). Of these, five explore the type
51
52 of strategy used, and frequency with which these strategies are reportedly used (He et al.
53
54 2011, He et al. 2005, Polkki et al. 2001, Rheiner et al. 1998, Twycross 2007a) and are
55
56
57 reported in this section. Other than Twycross (2007a) where observational methods
58
59 were used, all these studies describe nurses’ self-reported use of non-pharmacological
60
61
17
62
63
64 Page 18 of 56
65
 1
2
3 strategies in managing children’s pain.
4
5
6
7 Three of these studies also report on demographic factors associated with the use of
8
9
10 non-pharmacological methods (He et al. 2011, He et al. 2005, Polkki et al. 2001) and
11
12 these findings are reported in Table 3. Polkki et al. (2001) also report on the relationship

t
13

ip
14 between organisational factors and nurses’ reported use of these methods, He et al.
15
16 (2005, 2011) and one further study (Polkki et al. 2003) explores nurses’ own perceptions

cr
17
18
19 of factors promoting and hindering their use of non-pharmacological strategies. Findings

us
20
21 from these studies relating to these factors are reported in Section 3.5.3 below.
22
23
24

an
25 3.4.1 Non-pharmacological pain management: Strategies used and frequency of use
26
27
28
Three studies used the same questionnaire to examine nurses’ reported use of non-
M
29
30 pharmacological strategies for postoperative pain management (He et al. 2011, He et al.
31
32 2005, Polkki et al. 2001). Across all three studies, several cognitive-behavioural
33
d

34 strategies were reported to be used by most nurses, nearly always or always.

35
te

36
Preparatory information (including cognitive and sensory information, and different ways
37
38
of giving information) was the most commonly reported cognitive behavioural strategy
p

39
40
41 used by Chinese nurses in He et al.’s (2005) and Finnish nurses in Polkki et al.’s (2001)
ce

42
43 studies, but not by the sample of Singaporean nurses in He et al.’s (2011) study. Of
44
45 physical methods, positioning was reported to be the most frequently used strategy by
Ac

46
47
48 nurses from all three studies. The proportion of nurses saying they nearly always or
49
50 always used various emotional support strategies varied substantially between studies,
51
52 but comforting/reassurance was the most commonly endorsed emotional support
53
54 strategy in all three samples. It is unclear whether variance between studies is
55
56
57 statistically significant and observing or highlighting differences between studies relies
58
59 on percentage comparisons without statistical power. It is also unclear whether between-
60
61
18
62
63
64 Page 19 of 56
65
 1
2
3 study variations can be explained in terms of cultural or institutional differences. This
4
5 warrants further exploration.
6
7
8
9
A North American study that did not use the same questionnaire as the three studies
10
11
12 mentioned above reported the most frequently used non-pharmacological methods in

t
13

ip
14 the first five postoperative days by nurses (n=29) as being positioning (45.1%),
15
16 reassurance (25.4%) and ice (9.1%) (Rheiner et al. 1998). Other strategies (massage,

cr
17
18 heat, relaxation, teaching and distraction/play), accounted for less than 5% of all non-
19

us
20
21 pharmacological strategies used in this study (Rheiner et al. 1998). In an observational
22
23 study carried out in a UK hospital, during 185 hours of data collection non-
24

an
25 pharmacological strategies were only used on four occasions, with distraction being
26
27 used three times for procedural pain during the postoperative period (Twycross 2007a).
28
M
29
30
31
32 From these findings it can be seen that nurses report a wide variability in their use of a
33
d

34 range of non-pharmacological methods, and it is of concern that some of those methods

35
te

36
37
that are promoted are not necessarily evidence based (i.e. reassurance). Reliance on
38
self-report data and limited recent research impacts on the credibility of conclusions
p

39
40
41 reached about nurses’ use of non-pharmacological strategies in practice, and make it
ce

42
43 difficult to establish whether these findings are up to date or practice has improved over
44
45
Ac

time.
46
47
48
49
50 3.5 Factors influencing nurses’ management of children’s postoperative pain
51
52 Factors influencing nurses’ management of children’s pain are discussed in 20 papers
53
54
55 (Asprey 1994, Byrne et al. 2001, Gillies et al. 1999, Hamers et al. 1996, Hamers et al.
56
57 1997, He et al. 2011, He et al. 2005, Higgins et al. 1999, Hudson-Barr et al. 1998,
58
59 Nethercott 1994, Polkki et al. 2001, Prot-Labarthe et al. 2008, Rheiner et al. 1998, Ross
60
61
19
62
63
64 Page 20 of 56
65
 1
2
3 et al. 1991, Simons & Moseley 2009, Smyth et al. 2011, Twycross 2007b, 2008, Vincent
4
5 & Gaddy 2009, Woodgate & Kristjanson 1996). These factors can be divided into those
6
7 relating to the child, nurses, and the organisation.
8
9
10
11
12 3.5.1 Factors relating to the child

t
13

ip
14 The results of studies in this review provide a clear indication that children’s behavioural
15
16 cues impact on nurses’ perceptions of how much pain they are in (Hamers et al. 1996,

cr
17
18
19 Hudson-Barr et al. 1998, Nethercott 1994, Ross et al. 1991, Woodgate & Kristjanson

us
20
21 1996). There is evidence in one study that noisy children who had parents who
22
23 intervened frequently received more pain medications compared to quiet children or
24

an
25 those with passive parents (Woodgate & Kristjanson 1996). Three studies explored the
26
27
28
influence of children’s behaviour on nurses’ administration practices (Hamers et al. 1996,
M
29
30 Hudson-Barr et al. 1998, Vincent & Gaddy 2009). Nurses were significantly more likely
31
32 to administer a non-opioid analgesic to a child who expressed their pain vocally (Hamers
33
d

34 et al. 1996). Nurses were asked about the amount of morphine they would administer to
35
te

36
two patients, both reporting a pain score of eight out of 10; one child was depicted as
37
38
smiling and the other was depicted as grimacing (Vincent & Gaddy 2009). Nurses
p

39
40
41 indicating they would give a higher dose of morphine to the smiling child were more
ce

42
43 likely to make positive or neutral comments about the importance of self-report versus
44
45 behaviour in pain assessment. Those who would not give a full dose of morphine made
Ac

46
47
48 negative comments about the mismatch between smiling and pain ratings perhaps
49
50 indicating that some nurses still gauge their pain management practices on what have
51
52 been traditionally considered to be expressions of pain (i.e. grimacing). In an
53
54 experimental study, nurses (n=55) were shown video snippets of infants recorded after
55
56
57 receiving analgesia or during a time when analgesia was deemed inactive and asked if
58
59 they would administer analgesia (Hudson-Barr et al. 1998). Regression analysis did not
60
61
20
62
63
64 Page 21 of 56
65
 1
2
3 find any consistent group of behaviours (facial grimace, body movement, etc.) that
4
5 predicted nurse accuracy while viewing infant snippets regardless of the grouping of the
6
7 baby (active analgesia or inactive analgesia). Agreement between the coded snippets in
8
9
10 the inactive analgesia group (in need of analgesic) versus the active group (those who
11
12 recently received analgesic) and the nurses’ decision to medicate the infants in this

t
13

ip
14 group was 16.4% versus 82.8% respectively, suggesting nurses were more accurate at
15
16 identifying a baby not in pain than in pain.

cr
17
18
19

us
20
21 There is some evidence, from two observational studies (Byrne et al. 2001, Woodgate &
22
23 Kristjanson 1996), that nurses have a clear perceived idea about how children in pain
24

an
25 should behave, and that not complying with this results in children being labelled as bad
26
27
28
or poorly behaved. In one study, good children were those who did not complain; the
M
29
30 more overt children’s behaviours the more likely nurses were to label them as hysterical,
31
32 whining or miserable (Woodgate & Kristjanson 1996). Nurses appeared to have a
33
d

34 predetermined trajectory of pain behaviour through which children were expected to

35
te

36
move during the postoperative period; children were seen as bad patients if they did not
37
38
comply with this path (Byrne et al. 2001). When pain behaviour occurred outside of this
p

39
40
41 predetermined trajectory, nurses construed pain as unreal, unwarranted or not deserving
ce

42
43 treatment. As these studies are older it remains unknown if or whether this reliance on
44
45 an expected invariable postoperative pain trajectory remains a barrier to postoperative
Ac

46
47
48 pain management.
49
50
51
52 There is equivocal evidence about the impact of children’s age on nurses’ assessment
53
54 and management of pain. Nurses reported children’s age was a factor when assessing
55
56
57 pain (Nethercott 1994). Simons and Moseley (2009) found the frequency of pain
58
59 assessment by nurses varied with child age, with older children having more pain scores
60
61
21
62
63
64 Page 22 of 56
65
 1
2
3 recorded. However, in an experimental vignette study age was not correlated with
4
5 nurses’ postoperative pain assessments, although a trend was found indicating nurses
6
7 may attribute more pain to older children (Hamers et al. 1996). The influence of
8
9
10 children’s age on nurses’ management of their pain (i.e. analgesic administration) was
11
12 explored in two studies. In one study, children’s age was not correlated with nurses’

t
13

ip
14 decision to administer a non-opioid analgesic to the vignette character (Hamers et al.
15
16 1996). In another study, younger children (less than 24-months old) received fewer

cr
17
18
19 doses of around the clock paracetamol compared with older patients regardless of the

us
20
21 type of surgery incision or the day of surgery (Higgins et al. 1999). Patients less than 24-
22
23 months old (n= 38) received 84.2% (64 of the 76 doses) versus 94.7% (142 of the 150
24

an
25 doses) for patients over 24-months (n=76) of ordered paracetamol doses on the day of
26
27
28
surgery. On the first postoperative day 89.3% (100 of the 112 doses) for patients under
M
29
30 24-months were administered versus 95.1% (212 of the 223 doses) for patients over 24-
31
32 months.
33
d

34
35
te

36
How gender impacts on nurses’ perceptions of a child’s postoperative pain has not been
37
38
explored in detail. One study found no difference between the pain scores of boys and
p

39
40
41 girls (Simons & Moseley 2009) and children’s gender made no difference in the
ce

42
43 analgesic administrative practices of nurses in two studies (Gillies et al. 1999, Hudson-
44
45 Barr et al. 1998). When it comes to the administration of analgesics the evidence of an
Ac

46
47
48 association with administration practices and a child’s gender is equivocal. In one study
49
50 no differences were found in the morphine equivalents administered to girls versus boys
51
52 (Rheiner et al. 1998). However, Asprey (1994) reported two of the three children who
53
54 had an opioid prescribed but did not receive any opioid were boys, whereas all the girls
55
56
57 with an opioid order received at least one dose. Further complicating our understanding
58
59 of effect of children’s gender on opioid administration is that when boys did receive
60
61
22
62
63
64 Page 23 of 56
65
 1
2
3 opioids they received more doses than girls (3.5 doses/day versus 3.1 doses/day)
4
5 (Asprey 1994) although statistical tests to determine if this apparent difference in doses
6
7 between boys and girls was significant were not conducted.
8
9
10
11
12 There is some evidence a child’s diagnosis (type of surgery), the seriousness of their

t
13

ip
14 condition, and/or the type of surgery impacts on nurses’ perceptions of pain (Prot-
15
16 Labarthe et al. 2008, Ross et al. 1991, Simons & Moseley 2009, Woodgate &

cr
17
18
19 Kristjanson 1996). One qualitative study described a perceived hierarchy of suffering

us
20
21 based on child’s diagnosis (Woodgate & Kristjanson 1996). In a chart audit, the type of
22
23 surgery was found to affect how often pain scores were recorded; there were frequent
24

an
25 recordings for abdominal surgical cases but very infrequent recording for
26
27
28
appendectomies and tonsillectomies (Simons & Moseley 2009). The influence of surgery
M
29
30 type on pain assessments was reported in another study (Prot-Labarthe et al. 2008). Of
31
32 the 2,388 numerical pain assessments recorded 15% (359) were for appendectomies;
33
d

34 14.6% (348) related to children who had had a cholecystectomy; 14.8% (354) were after
35
te

36
a ureteroneocyctostomy; and 55.6% (1,327) after spinal fusion (Prot-Labarthe et al.
37
38
2008). However, no statistical differences were found between pain assessment and
p

39
40
41 diagnosis in another study (Hamers et al. 1996).
ce

42
43
44
45 Most of the studies include children who had undergone a variety of surgical procedures
Ac

46
47
but did not have the statistical power to conduct any analysis based on surgical
48
49
50 procedures. This means the impact of different surgeries on nurses’ postoperative
51
52 analgesic administration practices remains unclear. The one study that did explicitly
53
54 included surgery type as a variable did not find any correlation between surgery type and
55
56 nurses’ intent to medicate infants postoperatively (Hudson-Barr et al. 1998). However, in
57
58
59 a study by Higgins and colleagues (1999), regardless of age, those who had
60
61
23
62
63
64 Page 24 of 56
65
 1
2
3 thoracotomy incisions were less likely than those with sternotomy incisions to receive
4
5 prescribed around the clock paracetamol doses. This may be due to the fact children
6
7 undergoing a thoracotomy received doses of local anaesthetic administered by
8
9
10 anaesthetists (although the children only received about half of the prescribed local
11
12 anaesthetic doses). Although Ross and colleagues (1991) did not explicitly explore

t
13

ip
14 surgery type, they did examine the effect of serious sequelae from surgery on the pro re
15
16 nata opioid administration practices of nurses. They found that when vignettes describe

cr
17
18
19 the children as having serious permanent sequelae postoperatively, nurses administered

us
20
21 more opioids. Additionally, nurses have reported that knowing the type of surgery
22
23 (seriousness of the illness) helped them determine the amount of pain the child was
24

an
25 experiencing (Smyth et al. 2011).
26
27
28
M
29
30 3.5.2 Factors relating to the nurse
31
32 Nurse factors were explored in eight studies (Hamers et al. 1996, Hamers et al. 1997,
33
d

34 He et al. 2011, He et al. 2005, Polkki et al. 2001, Smyth et al. 2011, Twycross 2007b,
35
te

36
2008). Nurses’ perceptions of the degree of pain intensity a child was experiencing was
37
38
not related to the nurses’ confidence in their practice ability (Hamers et al., 1996). When
p

39
40
41 observational data were compared to the responses on a questionnaire no link was
ce

42
43 found between the perceived importance of a pain management task (Twycross 2008) or
44
45 an individual nurse’s knowledge about pain management (Twycross 2007b), and the
Ac

46
47
48 quality of their practices. None of the studies found nurses’ age, years of experience, or
49
50 education were associated with assessment practices.
51
52
53
54 Hamers et al. (1997) captured data using a series of video vignettes and reported on
55
56
57 factors related to nurses’ intent to administer a non-opioid analgesic. Nurses were
58
59 divided into three groups; experts (registered nurses), intermediate nursing students,
60
61
24
62
63
64 Page 25 of 56
65
 1
2
3 and novice nursing students. Expert nurses expressed intent to administer non-opioid
4
5 analgesia significantly more than novices and intermediates. Nurses’ years of
6
7 experience were also found to make them more inclined to provide analgesia and more
8
9
10 confident in their ability to decide whether a child needed analgesia (Smyth et al. 2011).
11
12 However, nurses with more than 10 years of experience were significantly less inclined

t
13

ip
14 to administer an analgesic compared to nurses with less than 10 years of experience
15
16 (Hamers et al. 1997) making the effect of nurses’ experience on analgesic administration

cr
17
18
19 unclear.

us
20
21
22
23 Three studies explored demographic factors affecting nurses’ use of non-
24

an
25 pharmacological strategies specifically (He et al. 2011, He et al. 2005, Polkki et al. 2001).
26
27
28
These are summarised in Table 3. Not all factors were significantly associated with all
M
29
30 methods and although some demographic factors were associated with the overall use
31
32 of non-pharmacological strategies there were no specific associations between
33
d

34 individual strategies and nurse characteristics.

35
te

36
37
38
3.5.3 Factors relating to the organisation
p

39
40
41 In one of the studies exploring nurses’ use of non-pharmacological strategies in
ce

42
43 managing pain, the specific hospital and ward were significantly associated with use of
44
45 some strategies, including transcutaneous electrical nerve stimulation (TENS), touch,
Ac

46
47
48 and imagery (Polkki et al. 2001). Hospital ward only was associated with the reported
49
50 use of thermal regulation, positioning, and presence. These findings suggest resources
51
52 as well as knowledge and skills, and unit culture may influence nurses’ use of different
53
54 non-pharmacological strategies.
55
56
57
58
59 Polkki et al. (2003) used factor analysis to examine the individual and contextual factors
60
61
25
62
63
64 Page 26 of 56
65
 1
2
3 perceived by Finnish nurses as promoting and hindering the use of non-pharmacological
4
5 pain management techniques. Five promoting factors emerged, including nurses’ own
6
7 perceived competence (e.g. their knowledge, education and experience), versatile use of
8
9
10 pain-relieving strategies (relating to beliefs about the effectiveness of non-
11
12 pharmacological strategies in relation to medication, the organisation of work, and a

t
13

ip
14 desire to learn effective methods of pain-relief), children’s age and ability to cooperate,
15
16 and parental participation. Identified barriers to effective pain management included

cr
17
18
19 nurses’ confidence (or lack thereof) in using non-pharmacological strategies, a perceived

us
20
21 need for education and negative early experiences in using these strategies, beliefs
22
23 regarding parental roles (the importance of parental participation in pain relief, and that
24

an
25 non-pharmacological strategies are parental roles) and about children’s ability to express
26
27
28
pain (a quiet and passive child not requiring pain relief) and the organisational work
M
29
30 model and rate of patient turnover. Workload and time emerged as both promoting and
31
32 hindering factors. Similarly, a sample of Chinese nurses in He et al.’s study (2005)
33
d

34 reported a shortage of nurses or heavy workloads as a limiting factor in the use of non-
35
te

36
pharmacological methods, as well as nurses’ lack of knowledge about pain management,
37
38
which by some nurses was attributed to the impact of traditional cultural beliefs on pain
p

39
40
41 management practice. Further, nurses reported that the use of non-pharmacological
ce

42
43 methods was not part of conventional postoperative practice and that this hindered their
44
45 use of such methods. Across both studies therefore, time/workload, nurses’ knowledge,
Ac

46
47
48 and beliefs surrounding pain management emerge as important perceived barriers to the
49
50 use of non-pharmacological methods in managing children’s pain.
51
52
53
54 4. Discussion
55
56
57 A summary of the practice reported in the papers reviewed compared to current clinical
58
59 guidelines can be seen in Table 4. Nurses do not always adhere to evidence-based
60
61
26
62
63
64 Page 27 of 56
65
 1
2
3 guidelines despite efforts to improve care over time. Several factors were identified that
4
5 may impact on nurses’ postoperative pain management practices. Much of what we
6
7 know about nursing practices is based on the results of chart reviews, meaning a lot of
8
9
10 what we know may be based on incomplete data. Many other studies collected data
11
12 using self-report tools that provide information about what nurses say they do which is

t
13

ip
14 not always supported by observational studies. The results of this review will now be
15
16 considered and the implications for practice and future research identified. Where

cr
17
18
19 appropriate, research from outside the review will be drawn upon where it illuminates our

us
20
21 findings further.
22
23
24

an
25
26
27
28
M
29
30
31
32
33
d

34
35
te

36
37
38
p

39
40
41
ce

42
43
44
45
Ac

46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
27
62
63
64 Page 28 of 56
65
 1
2
3 4.1 What do we know about nurses’ postoperative pain management practices?
4 4.1.1 Pain assessment and documentation
5
6 Nurses’ use of self-report pain assessment tools appears to be increasing over time but
7
8
9 evidence in this review and a recent study suggest that nurses continue to use them
10
11 inconsistently (Twycross et al. 2013b). The use of pain assessment tools varies between
12

t
13 different wards suggesting organisational factors may influence practice. Pain scores,

ip
14
15 even when recorded, were not found to guide choices about nurses’ postoperative pain-
16

cr
17
relieving interventions. This seeming disconnect between pain assessment and nurses’
18
19
intervention is consistent with the results of two other studies (Johnston et al. 2007,

us
20
21
22 Twycross et al. 2013b). A review of the literature on the use of pain assessment tools in
23
24 all paediatric contexts (not solely postoperative pain management) concluded the

an
25
26 evidence suggesting possible benefits of nurses using standardised pain assessment
27
28
tools needed interpreting with caution due to methodological issues in many of the
M
29
30
31 studies (Franck & Bruce 2009). However, one of the findings of the current review was
32
33 that the effectiveness of using self-report pain assessment tools has not been
d

34
35 adequately explored in the postoperative period. Moreover, a prospective comparative
te

36
37
38 study focusing on acute paediatric pain management (not just postoperative) found the
p

39
40 use of a self-report pain assessment tool in conjunction with an analgesic algorithm
41
ce

42 made a difference as children in the intervention group had significantly lower pain
43
44 scores (Falanga et al. 2006). Research is needed to explore whether the use of a self-
45
Ac

46
47
report pain assessment tool in the postoperative period results in children experiencing
48
49 less pain.
50
51
52
53 This review found that many nurses use behavioural signs of pain (crying, grimacing,
54
55
etc.) as a way of assessing children’s pain. However, there were no papers that
56
57
58 assessed the use of behavioural pain assessment tool and their impact on children’s
59
60
61
28
62
63
64 Page 29 of 56
65
 1
2
3 pain experience. Where behavioural assessment of pain was discussed (i.e. Hudson-
4
5 Barr et al. 1998) this focused on the children’s behaviours rather than using a pain
6
7 assessment tool. This is an area where further research is needed
8
9
10
11
12 Documentation relating to pain remains a challenge. Pain assessments were not always

t
13

ip
14 recorded with documentation tending to focus on the medications given. The pain-
15
16 relieving interventions used were rarely recorded in children’s notes, and there was little

cr
17
18
19 evidence of their effectiveness being evaluated. Similar findings have been noted in

us
20
21 other acute pain studies (Johnston et al. 2007, Taylor et al. 2008, Twycross et al. 2013b).
22
23 This is concerning as a lack of documentation may contribute to inadequate pain care
24

an
25 (Scott 1994). We need to know more about why nurses find it challenging to chart
26
27
28
effectively as opposed to continuing to find that they are not charting about postoperative
M
29
30 pain management consistently.
31
32
33
d

34 Communication with children and parents was only explored in one study. Nurses did
35
te

36
discuss children’s postoperative pain care with parents although this was often initiated
37
38
by parents and tended to focus on analgesic drugs (Twycross 2007a). Communication
p

39
40
41 between nurses and children also tended to focus on pain medications. Similar findings
ce

42
43 have been seen in other acute pain studies (Twycross & Collis 2013, Twycross et al.
44
45 2013b). This is of concern as the United Nations convention on the rights of the child
Ac

46
47
48 makes it clear that children should be involved in decisions relating to their care (United
49
50 Nations 1989). Studies that explore barriers to this involvement and how nurses do
51
52 engage children and families in postoperative pain management are needed.
53
54
55
56
57 4.1.2 Analgesics
58
59 Improvements in nurses’ analgesic administration practices in the care of children
60
61
29
62
63
64 Page 30 of 56
65
 1
2
3 postoperatively have occurred since Eland’s seminal study in 1974 as children now
4
5 receive analgesics more often in the postoperative period. However, results from this
6
7 review suggest nurses may still not be optimising the use of analgesics in the
8
9
10 postoperative period, particularly multi-modal analgesia. Only two studies discussed the
11
12 concurrent use of different classes of analgesics (Higgins et al. 1999, Twycross 2007a).

t
13

ip
14 The study by Smyth and colleagues (2011) noted that paracetamol, ibuprofen, and
15
16 various opioids were ordered but did not differentiate between these medications in their

cr
17
18
19 findings. Thus, how extensive the administration of a single class of analgesic by nurses

us
20
21 is when prescriptions either allow for or actually order multimodal analgesia warrants
22
23 further study. The reluctance to administer opioids, which was apparent in this review
24

an
25 (either alone or in combination with other analgesics) mirrors the finding of other studies
26
27
28
looking at postoperative pain management (Twycross et al. 2013b, Vincent & Denyes
M
29
30 2004) as well as in paediatric cancer-related pain (Finley et al. 2008, Forgeron et al.
31
32 2009). In recent years reports about the non-medical use of opioids have increased in
33
d

34 adults (Hernandez & Nelson 2010, Phillips 2013) which may increase nurses’ unfounded
35
te

36
fears of addiction in the paediatric postoperative context.
37
38
p

39
40
41 Perhaps unsurprisingly, children received more doses of analgesics (both opioid and
ce

42
43 non-opioid) when around the clock analgesic was prescribed compared to pro re nata
44
45 and required fewer rescue doses of analgesic (Higgins et al. 1999, Ross et al. 1991).
Ac

46
47
48 However, even when analgesics are prescribed around the clock children are not
49
50 guaranteed to receive them suggesting that factors relating to nurses or the structure of
51
52 nurses’ work also plays a part. The reasons for withholding analgesics are not reported,
53
54 and it is unclear if the rates with which prescribed analgesics are withheld are
55
56
57 comparable to other types of medications or for physiological reasons. In clinical practice
58
59 not administering an antibiotic would be seen as an adverse event whereas omitting a
60
61
30
62
63
64 Page 31 of 56
65
 1
2
3 prescribed dose of analgesics would not be considered in this light. This is despite
4
5 recent calls for mismanaged pain to be considered an adverse event (Chorney et al.
6
7 2010, Twycross et al. 2013a). Research is needed that specifies the reasons why nurses
8
9
10 withhold analgesics (both opioid and non-opioid) in the postoperative period, as this was
11
12 not specifically detailed in the included studies. Additionally, the impact of analgesic

t
13

ip
14 administration practices on children’s pain scores is worthy of investigation.
15
16

cr
17
18
19 4.1.3 Non-pharmacological strategies

us
20
21 Knowledge about the frequency and pattern of use of non-pharmacological strategies in
22
23 managing children’s postoperative pain remains limited. The majority of studies explored
24

an
25 the use of these strategies through self-report surveys or chart audits. This in itself is
26
27
28
useful to an extent, because it may reflect the degree to which nurses feel certain
M
29
30 strategies are appropriate and effective for use with their patients. However, this data
31
32 must be interpreted with caution, as it may be subject to reporting biases.
33
d

34
35
te

36
Several studies employed the same or adapted questionnaire, which allows for closer
37
38
comparison between study findings (He et al. 2011, He et al. 2005, Polkki et al. 2001)
p

39
40
41 and found some consistency in the patterning of reported non-pharmacological strategy
ce

42
43 use. However, some of the methods incorporated into the questionnaire raise concerns
44
45 about whether or not these are evidence-based and represent the best possible
Ac

46
47
48 approach to managing children’s pain. It is unclear whether the use of
49
50 comforting/reassurance, which was both reportedly used by nurses and suggested as a
51
52 possible technique in guidance for parents, is an effective method of pain relief. It has
53
54 been shown elsewhere that parental reassurance (e.g. comforting the child by saying it’s
55
56
57 ok) increases child distress in comparison to distraction (McMurtry et al. 2010).
58
59
60
61
31
62
63
64 Page 32 of 56
65
 1
2
3 The possibility of designing interventions to improve the use of non-pharmacological
4
5 strategies in managing pain in this context relies on a deeper understanding of influential
6
7 and modifiable factors and how they may interact. Some of the findings presented above
8
9
10 highlight nurses’ perceptions on the barriers and facilitators in the use of these strategies
11
12 (He et al. 2005, Polkki et al. 2003). This suggests contextual factors and perceptions

t
13

ip
14 about whose responsibility non-pharmacological strategies are, may play a role in how
15
16 nurses manage children’s postoperative pain. There is consistency in the direction of

cr
17
18
19 relationships between nurse demographic factors (i.e. older age, higher level of

us
20
21 experience) and the use of non-pharmacological strategies overall, but not in relation to
22
23 specific strategies themselves.
24

an
25
26
27
28
4.1.4 Factors impacting on nursing practices
M
29
30 The only factor found to be associated consistently with nurses’ pain assessment and
31
32 analgesic administration practices was children’s behaviour. Children whose behavioural
33
d

34 cues were more overt received more doses of analgesics or had their pain perceived as
35
te

36
valid in many of the included studies and this is consistent over time (i.e. Nethercott
37
38
1994; Smyth et al. 2011). There is evidence some nurses believe in a normal
p

39
40
41 postoperative pain trajectory that negatively affects analgesic administration practices
ce

42
43 (Byrne et al. 2001, Woodgate & Kristjanson 1996). Pain is a subjective phenomenon and
44
45 a child’s expression of pain can be affected by several factors including age (Kleiber et
Ac

46
47
48 al. 2007) and culture (Kristjansdottir et al. 2012) as well as previous experiences of pain
49
50 (von Baeyer et al. 2004). There is a need to explore strategies that help nurses
51
52 understand the individuality of a child’s pain experience and adjust the care plan
53
54 accordingly. Setting a pain goal has been used successfully for children with cancer-
55
56
57 related pain (Anghelescu & Oakes 2002, Oakes et al. 2008). The use of this intervention
58
59 for postoperative pain needs exploring. This may also facilitate communication between
60
61
32
62
63
64 Page 33 of 56
65
 1
2
3 the nurse and the child and parent about pain management.
4
5
6
7 Variability in nurses’ aims when managing postoperative pain have been found
8
9
10 (Twycross & Finley 2014) further suggesting there are individual differences in practices,
11
12 although in another study no association was found between nurse specific factors and

t
13

ip
14 pain care (Latimer et al. 2009). In this review, there was equivocal evidence about the
15
16 factors relating to individual nurses and factors relating to the child. This brings into

cr
17
18
19 question if these inconsistent findings are meditated by a different unmeasured factor

us
20
21 such as organisational context.
22
23
24

an
25 4.2 Other considerations
26
27
28
Limited theoretical knowledge about managing pain in children has been suggested as
M
29
30 one reason paediatric nurses do not manage pain effectively. Gaps remain in nurses’
31
32 knowledge about pain, particularly in relation to pain assessment, analgesic
33
d

34 administration and the use of psychological and physical methods (Ekim & Ocakci 2013,
35
te

36
Rieman & Gordon 2007, Twycross 2004, Van Hulle 2005). Gaps in knowledge about the
37
38
pharmacology of analgesics and the anatomy and physiology of pain may mean the
p

39
40
41 rationale for implementing pain-relieving interventions is not well understood. If nurses
ce

42
43 lack knowledge in these areas they may lack confidence or skill in assessing children’s
44
45 pain or using non-pharmacological methods of pain-relief. However, even when nurses
Ac

46
47
48 have a good level of knowledge, this may not be reflected in their pain management
49
50 practices (Twycross 2007b). Other factors may be responsible for knowledge not being
51
52 used in practice.
53
54
55
56
57 Over the past decade there has been a growing awareness that a unit (ward) has a set
58
59 of informal rules that determine how pain is managed. Organisational culture was found
60
61
33
62
63
64 Page 34 of 56
65
 1
2
3 to influence the care provided in eight paediatric hospitals in Canada (Estabrooks et al.
4
5 2011). The impact of unit culture was demonstrated in an ethnographic study on two
6
7 (adult) units in one hospital in the USA (Lauzon Clabo 2008). Participants described a
8
9
10 clear, but different, pattern of pain assessment on each ward. Further, in one Canadian
11
12 study, paediatric nurses described the unit’s pain management culture as giving pain

t
13

ip
14 medications regularly even if they are prescribed PRN; this cultural norm appeared to be
15
16 the factor that impacted most on practice (Twycross et al. 2013b). Context

cr
17
18
19 (organisational culture) clearly has a place in ensuring knowledge is used in practice.

us
20
21 Indeed, it has been postulated that pain management practices remain poor because
22
23 contextual factors are not taken into account (Bucknall et al. 2001, Craig 2009). There is
24

an
25 some evidence that interventions including contextual factors improve paediatric cancer
26
27
28
pain management (Finley et al. 2008). However, no studies have examined contextual
M
29
30 factors in relation to paediatric postoperative pain management. Studies are needed to
31
32 explore which organisational factors and strategies facilitate sustainable changes in unit
33
d

34 culture.
35
te

36
37
38 4.3 Limitations
p

39
40 This review has several limitations. First, many of the studies included had some
41
ce

42 methodological weaknesses and inherent bias that could influence the generalisability of
43
44 the findings. However, the findings of studies from diverse methodologies (qualitative,
45
Ac

46
47
quantitative, mixed) often supported each other (Table 2). In the instances of conflicting
48
49 studies (i.e. nursing and child factors) these studies shared similar methodology.
50
51 Second, 11 of the studies were from five datasets, therefore, some similar findings
52
53 across studies may relate to this issue. Third, the age range in the studies was large,
54
55
though this is not atypical in paediatric studies. Studies were not rejected based on age
56
57
58 as long as the study was conducted with age ranges typically seen in non-critical care
59
60
61
34
62
63
64 Page 35 of 56
65
 1
2
3 areas, as the aim was to have a broad understanding of ward based practices.
4
5 Additionally, most studies to date include the wide age range of paediatric patients but
6
7 due to relatively small sample sizes subgroup analysis based on age were not
8
9
10 conducted. It may, therefore, be that some of these findings relate more to some age
11
12 groups (i.e. infants, preverbal patients) compared to others (i.e. adolescents).

t
13

ip
14 Nevertheless studies did find that adolescents did not receive analgesic drugs as often
15
16 as they could and this was also found for younger children. Clearly more research is

cr
17
18
19 warranted to understand how developmental stages influence nurses postoperative care

us
20
21 practices.
22
23
24

an
25 Fourth, no definition of postoperative period was offered in any of the studies. Some
26
27
28
studies examined practices within the first 24-hours postoperatively, some three days
M
29
30 postoperatively, whereas others did not define the period. Thus it may be that certain
31
32 practices adhere more to recommended guidelines immediately postoperatively but
33
d

34 wane over time (although this did not hold for multimodal analgesic administration or
35
te

36
documentation of pain assessment). Fifth, the majority of the articles reported on studies
37
38
conducted in Western countries and, therefore, the postoperative paediatric pain care
p

39
40
41 practices of nurses in non-Western countries remains unclear. Sixth, many of the studies
ce

42
43 included a chart audit method of data collection thus the reported findings may be more
44
45 reflective of nurses’ documentation than their practices. However, there were a small
Ac

46
47
48 number of observational studies that support some of the chart audit findings. Despite
49
50 these limitations this review’s strength is in its focus on paediatric nurses’ postoperative
51
52 pain care practices over time allowing science in the area to build on what is already
53
54 known.
55
56
57
58
59
60
61
35
62
63
64 Page 36 of 56
65
 1
2
3 Conclusion
4
5 Findings of this systematic search and review suggest that in the postoperative period
6
7
8 paediatric nurses do not use self-report pain assessment tools consistently. Visible
9
10 distress or changes in behaviour may continue to have the biggest impact on some
11
12 nurses’ practice. Nurses do not always maximise analgesics as effectively as possible,

t
13

ip
14 particularly multimodal analgesics and the use of non-pharmacological methods is
15
16
limited and not necessarily implemented into routine care or based on strong evidence.

cr
17
18
19 Few intervention studies have been conducted to determine which modifiable variables

us
20
21 support improvements in nurses’ pain management practices. There is a need to move
22
23 past child specific and nurse specific factors to determine other barriers (i.e.
24

an
25
26
organisational structures) and strategies to evidence-informed care.
27
28
M
29 References
30
31 Anghelescu D & Oakes L (2002): Working toward better cancer pain management for children.
32 Cancer Practice 10, S52-S57.
33 Arksey H & O’Malley L (2005): Scoping Studies: Towards a Methodological Framework.
d

34 International Journal of Social Research Methodology 8, 19-32.

35 Asprey JR (1994): Postoperative analgesic prescription and administration in a pediatric
te

36 population. Journal of Pediatric Nursing 9, 150-157.

37 Association of Paediatric Anaesthetists (2012): Good Practice in Postoperative and Procedural
38 Pain Management, 2nd edition. Pediatric Anesthesia 22, 1-79.
p

39 Beyer JE, Villarruel AM & Denyes MJ (2009) The Oucher: A User’s Manual and Technical Report.
40 Boughton K, Blower C, Chartrand C, Dircks P, Stone T, Youwe G & Hagen B (1998): Impact of
41
ce

research on pediatric pain assessment and outcomes. Pediatric Nursing 24, 31-36.
42 Bucknall T, Manias E & Botti M (2001): Acute pain management: Implications of scientific
43 evidence of nursing practice in the postoperative context. International Journal of Nursing
44 Practice 7, 266-273.
45
Ac

Byrne A, Morton J & Salmon P (2001): Defending against patients' pain: A qualitative analysis if
46 nurses' responses to children's postoperative pain. Journal of Psychosomatic Research
47
50, 69-76.
48
Caldwell K, Henshaw L & Taylor G (2011): Developing a framework for critquing health research:
49
An early evaluation. Nurse Education Today 31, e1-7.
50
Chorney JM, McGrath PJ & Finley GA (2010): Pain as the neglected adverse event. Canadian
51
Medical Association Journal 182, 732.
52
Craig KD (2009): The Social Communication Model of Pain. Canadian Psychology 20, 22-32.
53
Davis K, Drey N & Gould D (2009): What are scoping studies? A review of the nursing literature.
54
International Journal of Nursing Studies 46, 1386–1400.
55
56 Eccleston C, Palermo TM, Williams ACDC, Lewandowski A & Morley S (2012): Psychological
57 therapies for the management of chronic and recurrent pain in children and adolescents.
58 Cochrane Database of Systematic Reviews Issue 12.
59
60
61
36
62
63
64 Page 37 of 56
65
 1
2
3 Ekim A & Ocakci AF (2013): Knowledge and attitudes regarding pain management of pediatric
4 nurses in Turkey. Pain Management Nursing 14, e262-e267.
5 Estabrooks CA, Squires JE, Hutchinson AM, Scott S, Cummings GG, Kang SH, Middizi WK &
6 Stevens B (2011): Assessment of variation in the alberta context tool: The contribution of
7 unit level contextual factors and speciality in Canadian pediatric acute care settings. BMC
8 Health Services Research 11, 1-17.
9 Falanga IJ, Lafrenaye S, Mayer SK & Tetrault J-P (2006): Management of acute pain in children:
10 safety and efficacy of a nurse controlled algorithim for pain relief. Acute Pain 8, 45-54.
11 Finley GA, Forgeron PA & Arnaout M (2008): Action research: Developing a pediatric cancer
12 program in Jordan. Journal of Pain and Symptom Management 35, 44-454.

t
13 Forgeron PA, Finley GA & Arnaout M (2005): Pediatric pain prevalence and parents' attitudes at a

ip
14 cancer hospital in Jordan. Journal of Pain and Symptom Management 31, 440-448.
15 Forgeron PA, Jongudomkarn D, Evans J, Finley GA, Thienthong S, Siripul P, Pairojkul S, Sriraj W
16 & Boonyawatanangkool K (2009): Children's pain assessment in northeastern Thailand:

cr
17 Perspectives of health professionals. Qualitative Health Research 19, 71-81.
18 Franck LS & Bruce EA (2009): Putting pain assessment into practice: Why is it so painful? Pain
19 Research and Management 14, 13-20.

us
20 Gillies ML, Smith LN & Parry-Jones WLI (1999): Postoperative pain assessment and
21 management in adolescents. Pain 79, 207-215.
22 Grant MJ & Booth A (2009): A typology of reviews: an analysis of 14 review types and associated
23 methodologies. Health Information and Libraries Journal 26, 91-108.
24

an
Grunau RE, Whitfield MF & Petrie J (1998): Children's judgements about pain at aged 8-10 years:
25 Do extremely low birthweight children differ from their full birthweight peers. Journal of
26 Child Psychology and Psychiatry 39, 587-594.
27 Hamers JPH, Abu-Saad HH, van den Hout MA, Halfens RJG & Kester ADM (1996): The
28 influence of children's vocal expressions, age, medical diagnosis and information
M
29 obtained from parents on nurses' pain assessments and decisisons regarding
30 interventions. Pain 65, 53-61.
31 Hamers JPH, van den Hout MA, Halfens RJG, Abu-Saad HH & Heijtjes AEG (1997): Differences
32 in pain assessment and decisions regarding the administration of analgesics between
33
d
novices, intermediates and experts in pediatric nursing. International Journal of Nursing
34 Studies 34, 325-334.
35 He H-G, Lee T-L, Jahja , Sinnappan , Vehvila inen-Julkunen K, Polkki T & Ang ENK (2011):
te

36 The use of nonpharmacological methods for children's postoperative pain relief:

37 Singapore nurses' perspectives. Journal for Specialists in Pediatric Nursing 16, 27-38.
38 He H-G, Polkki T, Vehvilainen-Julkunen K & Pietila A-M (2005): Chinese nurses' use of non-
p

39 pharmacological methods in children's postoperative pain relief. Journal of Advanced

40
Nursing 51, 335-342.
41
ce

Hernandez SH & Nelson LS (2010): Prescription drug abuse: Insight into the epidemic. Clinical
42
Pharmacology and Therapeutics 88, 307-317.
43
Higgins JPT & Green SuMTCC, 2011. Available from (2011) Assessing risk of bias in included
44
studies. In Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0
45
Ac

(Higgins JPT, Altman DG, Sterne JAC & on behalf of the Cochrane Statistical Methods
46
Group and the Cochrane Bias Methods Group eds.). Cochrane Library.
47
48 Higgins SS, Turley KM & Harr J (1999): Prescription and administration of around the clock
49 analgesics in postoperatvie cardiovascular surgery patients. Progress in Cardiovascular
50 Nursing 14, 19-24.
51 Hudson-Barr D, C,, Duffey MA, Holditch-Davis D, Funk S & Frauman A (1998): Pediatric nurses'
52 use of behaviors to make medication decisions in infants recovering from surgery.
53 Research in Nursing and Health 21, 3-13.
54 Johnston CC, Gagnon AJ, Rennick J, Rosmus C, Patenaude H, Ellis J, Shapiro C, Filion F,
55 Ritchie J & Byron J (2007): One-on-one coaching to improve pain assessment and
56 management practices of pediatric nurses. Journal of Pediatric Nursing 22, 467-478.
57 Kleiber C, Suwanraj M, Dolan LA, Berg M & Kleese A (2007): Pain-sensitive temperament and
58 post-operative pain. Journal of the Society of Pediatric Nursing 12, 149-158.
59
60
61
37
62
63
64 Page 38 of 56
65
 1
2
3 Kozlowski LJ, Kost-Byerley S, Colantuoni E, Thompson CB, Vasquenza KJ, Rothman SK, Billett
4 C, White ED, Yaster M & Monitto CL (2014): Pain prevalence, intensity, assessment and
5 management in a hospitalized pediatric population. Pain Management Nursing 15, 22-35.
6 Kristjansdottir O, Unruh AM, McAlpine L & McGrath PJ (2012): A systematic review of cross-
7 cultural comparison studies of child, parent, and health professional outcomes associated
8 with pediatric medical procedures. The Journal of Pain 12, 207-219.
9 Latimer MA, Johnston C.C., Ritchie JA, Clarke SP & Gillin D (2009): Factors affecting delivery of
10 evidence-based procedural pain care in hospitalized neonates. Journal of Obstetric,
11 Gynecologic and Neonatal Nursing 38, 182-194.
12 Lauzon Clabo LM (2008): An ethnography of pain assessment and the role of social context on

t
13 two postoperative units. Journal of Advanced Nursing 61, 531-539.

ip
14 Levac D, Colquhoun H & O’Brien KK (2010): Scoping studies: advancing the methodology.
15 Implementation Science 5, 1-9.
16 McMurtry CM, Chambers CT, McGrath PJ & Asp E (2010): When “don’t worry” communicates

cr
17 fear: Children’s perceptions of parental reassurance and distraction during a painful
18 medical procedure. . Pain 150, 52-58.
19 National Confidential Enquiry into Patient Outcome and Death (2011) Are we there yet? National

us
20 Confidential Enquiry into Patient Outcome and Death, London.
21 Nethercott S (1994): The assessment and management of post-operative pain in children by
22 registered sick children's nurses: an exploratory study. Journal of Clinical Nursing 3, 109-
23 114.
24

an
Oakes LL, Anghelescu DL, Windsor KB & Barnhill PD (2008): An institutional quality improvement
25 initiative for pain management for pediatric cancer inpatients. Journal of Pain and
26 Symptom Management 35, 656-669.
27 Phillips J (2013): Prescription drug abuse: Problem, policies, and implications. Nursing Outlook
28 61, 78-84.
M
29 Polkki T, Laukkala H, Vehvilainen K & Pietila A-M (2003): Factors influencing nurses' use of
30 nonpharmacological pain alleviation methods in paediatric pain patients. Scandinavian
31 Journal of Caring Science 17, 373-383.
32 Polkki T, Vehvilamen-Julkunen K & Pietila A-M (2001): Non-pharmacological methods in relieving
33
d
children's postoperative pain: a survey on hospital nurses in Finland. Journal of
34 Advanced Nursing 34, 483-492.
35 Prot-Labarthe S, Pelletier E, Winterfield U, Villeneuve E, Wood C, Bussieres J-F, Brion F &
te

36 Bourdon O (2008): Comparision of pain management in paediatric surgical patients in

37 two hospitals in France and Canada. Pharmacy World & Science 30, 251-257.
38 Rheiner JG, Megel ME, Haitt M, Halbach R, Cyronek DA & Quinn J (1998): Nurses' assessments
p

39 and management of pain in children having orthopedic surgery. Issues in Comprehensive

40
Pediatric Nursing 21, 1-18.
41
ce

Rieman MT & Gordon M (2007): Pain management competency evidenced by a survey of

42
pediatric nurses' knowledge and attitudes. Pediatric Nursing 33, 307-312.
43
Romsing J (1996): Assessment of nurses' judgements for analgesic requirements in children.
44
Journal of Clinical Pharmacy and Therapeutics 21, 159-163.
45
Ac

Ross RS, Bush JP & Crummette BD (1991): Factors affecting nurses' decisions to adminster
46
PRN analgesic medication to children after surgery: An analog investigation. Journal of
47
Pediatric Psychology 16, 151-167.
48
49 Salantera S, Lauri S, Salmi T & Aantaa R (1999): Nursing activities and outcomes of care in the
50 assessment, management, and documentation of children's pain. Journal of Pediatric
51 Nursing 14, 408-415.
52 Saxe G, Stoddard F, Courtney D, Cunningham K, Chawla N, Sheridan R, King D & King L (2001):
53 Relationship between acute morphine and the course of PTSD in children with burns.
54 Journal of the American Academy of Child and Adolescent Psychiatry 40, 915-921.
55 Scott IE (1994): Effectiveness of documented assessment of postoperative pain. British Journal
56 of Nursing 3, 494-501.
57 Shrestha-Ranjit JM & Manias E (2010): Pain assessment and management practices in children
58 following surgery of the lower limb. Journal of Clinical Nursing 19, 118-128.
59
60
61
38
62
63
64 Page 39 of 56
65
 1
2
3 Simons J & Moseley L (2008): Post-operative pain: The impact of prescribing patterns on nurses'
4 administration of analgesia. Paediatric Nursing 20, 14-19.
5 Simons J & Moseley L (2009): Influences on nurses' scoring of children's post-operative pain.
6 Journal of Child Health Care 13, 101-115.
7 Smyth W, Toombs J & Usher K (2011): Children's postoperative pro re nata (PRN) analgesia:
8 Nurses' administration practices. Contemporary Nurse 37, 160-172.
9 Taddio A, Shah V, Gilbert-MacLeod C & Katz J (2002): Conditioning and hyperalgesia in
10 newborns exposed to repeated heel lances. JAMA 288, 857-861.
11 Taylor EM, Boyer K & Campbell FA (2008): Pain in hospitalized children: A prospective cross-
12 sectional survey of pain prevalence, intensity, assessment and management in a

t
13 Canadian pediatric teaching hospital. Pain Research and Management 13, 25-32.

ip
14 Twycross A (2004) Children’s Nurses’ Pain Management Practices: Theoretical Knowledge,
15 Perceived Importance and Decision-Making. In Nursing Department. University of Central
16 Lancashire.

cr
17 Twycross A (2007a): Children's nurses' post-operative pain management practices: An
18 observational study. International Journal of Nursing Studies 44, 869-881.
19 Twycross A (2007b): What is the Impact of Theoretical Knowledge on Children's Nurses'

us
20 Postoperative Pain Management Practices? An Exploratory Study. Nurse Education
21 Today 27, 607-707.
22 Twycross A (2008): Does the perceived importance of a pain management task affect the quality
23 of children’s nurses’ post-operative pain management practices? . Journal of Clinical
24

an
Nursing 17, 3205-3216.
25 Twycross A, Chorney J, McGrath PJ, Finley GA, Bolliver D & Mifflin K (2013a): A Delphi Study to
26 Identify Adverse Event Indicators for Pediatric Post-operative and Procedural Pain. Pain
27 Research and Management 18, e68-e74.
28 Twycross A & Collis S (2013): How Well Is Acute Pain In Children Managed? A Snapshot In One
M
29 English Hospital. Pain Management Nursing 14, e204-e215.
30 Twycross A & Finley GA (2013): Children’s and parents’ perceptions of postoperative pain
31 management: A mixed methods study. Journal of Clinical Nursing 22, 3095–3108.
32 Twycross A & Finley GA (2014): Nurses’ aims when managing paediatric post-operative pain: Is
33
d
what they say the same as what they do? Journal for Specialists in Pediatric Nursing 19,
34 17-27.
35 Twycross A, Finley GA & Latimer M (2013b): Pediatric Nurses’ Postoperative Pain Management
te

36 Practices: An Observational Study. Journal for Specialists in Pediatric Nursing 18, 189-
37 201.
38 Uman LS, Chambers CT, McGrath PJ & Kisely S (2010): Psychological interventions for needle-
p

39 related procedural pain and distress in children and adolescents. Cochrane Database of
40
Systematic Reviews.
41
ce

United Nations (1989): Convention on the Rights of the Child. United Nations, New York.
42
Van Hulle CV (2005): Nurses' knowledge, attitudes and practices regarding children's pain. MCN
43
30, 177-183.
44
Vickers AJ & Smith C (2000): Incorporating data from dissertations in systematic reviews.
45
Ac

International Journal of Technology Assessment in Health Care 16, 711-713.

46
Vincent CVH & Denyes MJ (2004): Relieving children's pain; Nurses' abilities and analgesic
47
adminstration practices. Journal of Pediatric Nursing 19, 40-50.
48
49 Vincent CVH & Gaddy EJ (2009): Pediatric nurses' thinking in response to vignettes on
50 administering analgesics. Research in Nursing & Health 32, 530-539.
51 von Baeyer CL, Marche TA, Rocha EM & Salmon K (2004): Children's memory for pain: overview
52 and implications for practice. The Journal of Pain 5, 241-249.
53 von Baeyer CL & Spagrud LJ (2007): Systematic review of observational (behavioral) measures
54 of pain for children and adolescents aged 3 to18 years. Pain 127, 140-150.
55 Woodgate R & Kristjanson L (1996): A young child's pain: How parents and nurses "take care".
56 International Journal of Nursing Studies 33, 271-284.
57
58
59
60
61
39
62
63
64 Page 40 of 56
65
Figure(s)

Potential articles identified from Articles excluded after evaluation of

search strategy (n=328) titles and abstracts (n=287)
Duplicates: 11

Full articles retrieved for detailed Full articles excluded after review of
examination (n=51) full article (n=27)

t
ip
cr
Articles identified by reviewing
reference list of included articles

us
references (n=3)

Articles included in systematic review

(n=27)
an
M
Figure 1: Flow diagram for retrieved, excluded and included studies
e d
pt
ce
Ac

Page 41 of 56
Table(s)

Table 1: Database search results

Database Number of articles Number of full articles

retrieved from literature retrieved
search
Psychinfo 30 6
CINAHL 169 27
(2 duplicates of included
articles from other
databases)
Pubmed 131 8

t
ip
EMBASE 95 10
(9 duplicates of included
articles from other

cr
databases)

us
an
M
e d
pt
ce
Ac

Page 42 of 56
Table 3: Factors impacting on nurses’ reported use of non-pharmacological strategies
(Key:  indicates significant correlation found, + and – signal direction of the relationship. Specific
strategies significantly associated with each factor are indicated in italics. )
Study Polkki et al. 2001 He et al. 2005 He et al. 2011
Factor Finland China Singapore

  
Older age (+) (+) (+)
Type of preparatory Positive Positioning, Presence,
information. Reinforcement, Comforting/Reassurance,
Positioning Touch

t
 

ip
Higher level of education (+) (+)
Type of preparatory Imagery, Breathing

cr
information. Techniques, Positioning,
Massage, Presence,
Comforting/Reassurance,

us
Touch
Nurse’s grade/position  
(+) (+)
Positioning, Massage. Comforting/reassurance,

Nurse’s years of 
an 
Touch


experience (+) Variable relationship (+)
M
Type of preparatory with use of specific Imagery, Presence,
information. strategies dependent Massage, Touch,
on years of Distraction
experience.
d

  
e

Nurse having their own (+) (+) (+)

children
pt

Comforting/Reassura Positioning, Positive Distraction, Breathing

nce, Breathing Reinforcement. Techniques, Positioning,
Techniques, Presence,
Presence, Type of Comforting/Reassurance,
ce

preparatory Touch, Helping with ADL

information, Use of
music as distraction
Ac

Page 43 of 56

Table 4: Reported practice compared to current guidelines
Best practice
Assess pain using a validated
pain assessment tool
Reassess pain following the
implementation of pain-relieving
interventions
Administer analgesic drugs are
administered as prescribed and if
the child complains of pain
Use multi-modal analgesia
e
Use non-pharmacological
methods of pain-relief
pt
ce
Ac
Document pain assessments in
the child’s notes
1
(American Academy of Pediatrics and American Pain Society, 2001, Association of Paediatric Anaesthetists, 2012,
Royal College of Nursing, 2009)
1
Reported practice
 Some increase in the use of pain assessment tools over time
but nurses do not use pain assessment tools consistently.
 Decision-making appears to be guided more by the child’s
behavioural cues than their self-report of pain despite the
implementation of pain assessment tools.
 Little is known about the nurses’ use of non-self report pain
assessment tools in the postoperative period.
 Reassessments continue to be a challenge. They are not
always carried out and rarely charted.
t
ip
 Children received analgesics more frequently when ordered
ATC vs. PRN.
cr
 Even ATC prescriptions did not ensure the child received the
analgesic.
 Some nurses are reluctant to administer opioids
us
postoperatively.
 Nurse experience, surgery type, child gender, and child age
show inconsistent association with analgesic administration.
 Nurses are more likely to administer medication when
•
an
behaviour is perceived to match self-reported pain.
Improvement in the administration of an analgesic has been
noted overtime. Yes the practice of administering multimodal
analgesics is not widely adopted.
M
• An observational study noted that nurses administer
paracetamol and ibuprofen.
• Nurses did not always administer paracetamol when a child
was receiving local aesthetic agents despite paracetamol.
d
being prescribed ATC and no contraindication offered.
 Very little is known about nurses’ use of non-pharmacological
methods; data is mostly self-report.
 Limited recent research makes it difficult to identify whether
practice has improved over time.
 Nurses report a wide variability in their use of a range of non-
pharmacological methods.
 Not all non-pharmacological methods that are promoted are
evidenced based (i.e. reassurance).

Documentation continues to be a challenge over time.
 Children do not always have pain assessments scores
documented even in the immediate postoperative period.
Page 44 of 56
Table(s)

t
ip
Table 2: Summary of studies included in the review
Key: PRN = Pro re nata/as needed; ATC = around the clock; yrs. = years; mos. = months; hrs. = hours; M = mean; postop = postoperative; ADL = activities of daily living; IM = intramuscular; non-pharm = non-pharmacological

cr
Study,
Country, Purpose/Aims Sample Measures/ Data Collection Methods Main findings Quality
(Study Design)
Asprey (1994) Replication of Eland N = 25, matched with Eland Charts reviewed, information extracted by All children had an analgesic prescribed. 23 had an opioid prescribed Strengths: Used same

us
United States (1974) study (1974) sample by age & one investigator: (i.e. pethidine (demerol) prescribed more often than morphine). methodology as
examining surgical procedure  Age, sex, race, weight Paracetamol (acetaminophen) given most often followed by comparison study.
Matched case-control prescription and  Surgical procedure morphine; 63% of the doses prescribed were therapeutic level, 26% of Population similar in both
retrospective chart administration of 4-8 yrs. (M = 5.4 yrs.)  Length of stay cases were sub-therapeutic. studies.
audit analgesics. 11 girls 

Aim to determine if Surgery types: orthopaedic
pain management had (4), neurosurgery (2),
improved. genitourinary (6), general
(3), eye (3), otolaryngologic
(3), cardiac (1), and burns
an
Analgesics prescribed
 Frequency of analgesic Some children did not receive any analgesics. Not all who had opioids Weaknesses: Descriptive
administration. prescribed received opioids 23 had both morphine and paracetamol study with small sample
prescribed but only 19 received both. size thus unable to stratify
the sample on gender or
Gender differences: 2 boys did not receive any analgesics, 5 boys did surgery type.

M
requiring skin grafts (3). not receive any opioids, but boys received on average 3.5 doses of
Length of stay 3 – 55 days (M analgesia/day whereas girls received 3.1 doses/day.
= 11.7 yrs.)
Boughton et al. To examine whether N = 86; study group n=36; Study group: had their pain assessed using No significant effect from using a standardised pain assessment was Strengths: Used a control
(1998) the introduction of n=50 controls (data collected Wong-Baker Faces Scale 4-hourly and 1- found: no significant differences between case and control groups on group design for the

d
Canada standardised pain via a chart audit) hourly post-analgesic administration. either descriptive characteristics or outcome measures. intervention comparison.
assessment had an 5 – 17 yrs. (M = 11 yrs.) Variables:
Case-control study & effect on postop pain 38% girls Descriptive Characteristics: Analgesics prescribed & administered (across both groups): Weaknesses:
chart audit management.

te
Inclusion criteria:
Minimum hospital stay 48-


Age and gender
Length of stay
Outcome measures:

All patients had PRN opioid medication prescribed 2-4 hourly, route
some oral & some IM. 25% children received no analgesics. Average
frequency of analgesic administration was 13.5 times (during hospital
Retrospective control
group.
Pain only assessed q4h
ep
hrs. post-op. Number of times analgesics stay). Twice as many oral medications as IM given. 75.6% of children in whatever child’s pain score
administered study group given analgesia reported effective pain control, 23.3% was.
 Nurses’ evaluations of effectiveness reported partially effective pain control. Descriptive statistics used
of pain medications (from patients’ to compare groups (no
verbal & behavioural responses) Other methods of pain management recorded (across both groups): inferential statistics). Study
 positioning (10.7%), distraction (6.75%), cast adjustment (1%) results not generalizable.
c

Other methods of pain control

 Time of first ambulation combination of methods (33.9%), none (22.75%). Multiple outcomes
 Progress of ambulation (time from examined for small
Ac

first ambulation to walking with number of participants.

comfort)

Byrne et al. (2001) To examine emotional Children: Observations conducted by one Two themes emerged: Strengths: Participant
United Kingdom influences on N = 16 consecutive cases of researcher for each child daily from 2nd to 1) Nurses’ requirements of their patients: observation to provide a
communication in the children undergoing major 6th day postop. - the importance of not displaying distress realistic picture of actual
Qualitative study clinical context, using orthopaedic surgery 8-16 yrs. Nurse looking after the child was asked: - the importance of meeting schedules of recovery practices. Included parent,
using ethnographic paediatric pain as a of age “How is [patient] doing today?” & - ways to deliver these requirements (including coercion, nurse, and child
methods model, to determine: 13 girls prompted to clarify answer (Interviews challenging parents, de-individualising) perspectives.
i) how paediatric Surgical procedures: spinal lasted 2-10 min). Child was asked: “How 2) Nurses explanations for patients’ displays of pain
nurses construe their (n=10), hip osteotomy (n=3), are you doing to day?” (interview length - the trauma of surgery and failure of analgesia Weaknesses: Philosophical
patients’ pain; ii) how lower limbs (n=3) unspecified). - wilfulness stance not stated.
these constructions Parents were asked “how is [patient] - parents’ influence Ethnography used but
relate to the Nurses: doing today? “How do the nurses think - emotional state and personality, unpleasantness and limited discussion of the
emotional challenge N = 13, Nurses interviewed [patient] is doing today? (Interviews possession. organization of work.

Page 45 of 56
 t
ip
of pain; iii) how these or observed; (12 RNs; 1 lasted 5-35 min.) Discuss the approach to
constructions untrained care assistant). Nurses’ accounts of challenges were also explored: the difficulty of data analysis but did not

cr
influenced nurses’ 13 female meeting patient/family expectations, conflicts with parents, children’s state if it was a thematic or
communication with 25 – 52 yrs. rejection of pain relief and nursing care. content analysis.
patients, specifically No discussion of ethic
their pain Parents Nurses tried to prevent children from displaying pain behaviour. When approval or ethical

us
management N = 20 Parents of children pain behaviour occurred they construed pain as unreal, unwarranted considerations.
included in study. or not deserving help.

Author concludes that nurses’ constructions are behavioural and

cognitive strategies to provide a means to protect themselves against

an
the emotional threat that children’s pain poses.
Gillies et al. (1999) To examine the Adolescents, N = 351; Adolescents: Pain intensity using Almost half of adolescents felt that nurses did not know when they Strengths: Discussed
United Kingdom experience & Inpatients, n = 287; Day adolescent paediatric pain tool (APPT) & were in pain. power for the adolescents
management of surgery n = 64 coloured visual analogue scale (VAS), sample. Reported
Cross sectional study postop pain in 12-18 yrs. (M = 15.3) Semi-structured interview about pre- and 99% of adolescents were prescribed postop analgesics (opioids and inferential statistics.

M
combining survey, adolescent patients 63% girls postop experiences. non-opioids). Often prescribed in combination.
interview and chart following elective Weaknesses: No
review. surgery. Healthcare professionals, Parents: questionnaire with similar Nurses aimed: to relieve pain completely (61%) or largely (39%) with justification for
N=157 - 76 RNs; 77 doctors wording to adolescent interview. analgesics. 89% stated that they gave analgesics regularly (at least 4-6 recruitment differination
(43 surgeons & 33 hourly). Many adolescents were not given prescribed (PRN) analgesics, between surgery types.

d
anaesthetists). Healthcare professionals: semi-structured seven (2%) received no analgesics. Review of charts showed 75% of Data related to nurses
interviews to explore attitudes towards adolescents received of at least 1 dose but very small proportion practices were descriptive
Most anaesthetists adolescent patients and their pain received analgesics regularly in the first 24-hrs. post-op. Inpatients only.

te
(74%) and surgeons (90%)
were male; 99% nurses
were female.
assessment and management practices
with adolescents (in general).

Nursing pain documentation reviewed

received analgesics both NSAIDs and other, which included opioids,
more than outpatients.

Nurses documented analgesics given 43% of the time

ep
Documentation of reassessment post analgesic 12% of the time.
Pain was not mentioned in any notes in 35% of the notes.
Hamers et al. (1996) To examine the N = 207 Nurses (11 Narrative & vignettes used to describe Vocal expression was the only variable that significantly influenced Strengths: Experimental
The Netherlands influence of the hospitals), 180 female range of medical diagnosis, child age & nurses’ assessment of pain in children & decision to administer non- design; large sample size
medical diagnosis, 22 – 56 yrs. (M = 32.1 yrs.) information obtained from parents, video opioid analgesics. (but power or not
c

Experimental: 2x2 child’s age and vignettes used to illustrate child discussed).
factorial design expressions, and Stratified by years of expression (less vocal vs. vocal). Across all 3 cases:
Ac

parents’ input on experience, knowledge & Significant main effect of child verbal pain expression: nurses Weaknesses: Only
nurses’ pain hospital, & randomised into Responses to 3 questions using a 100-mm attributed more pain to the child & more likely to administer descriptive statistics
assessment and 4 groups. visual analogue scale: analgesics. A trend for interaction between severity of medical available for group
decisions to 1. Rate the pain experienced by the diagnosis & child’s vocal expression of pain. comparison.
administer analgesic child in the case (0=no pain at all, 10
drugs. = extreme pain) No significant main effects for medical diagnosis on nurses’ Findings may not
2. How sure are you that your pain assessment of pain or decision to administer analgesics. generalize to actual nurse
assessment is the correct one? behaviours in practice
(0=completely unsure, No significant main effect of child age on nurses’ assessment of pain
10=completely sure) or decision to administer analgesics, trend for nurses to attribute
3. Would you administer an analgesic more pain to the older child.
to the child in this case? (0=surely
would not administer, 10= would No significant main effect of child age on nurses’ assessment of pain
surely administer) or decision to administer analgesics, trend for nurses to attribute
more pain to the older child.

Page 46 of 56
 t
ip
No significant between-group differences for perceived confidence in

cr
their assessment.

Strong correlation between pain assessment scores by nurse &

decision to administer analgesics.

us
Hamers et al. (1997) To examine the N = 695; novice (n = 271); Same as Hamers et al. (1996) Experts were more likely to give analgesic drugs but not more likely to Strengths: Experimental
The Netherlands influence of expertise intermediate (n=222); expert do a pain assessment. design using vignettes pilot
on nurses’ assessment (n=202). tested for reliability and
Vigentte of pain and decisions No significant effect of expertise level on assessment. Main effect of validity (via a pilot study).
Experimental to administer Novices: 1st year nursing group, differences between characteristics of the cases presented Hypothesis clearly stated.

an
analgesic drugs. students from 6 institutions, influenced pain assessment. Large sample size.
239 female, 17 – 35 yrs.
A significant main effect of expertise on confidence meaning expert Weaknesses: Group of
Intermediate: 4th year nurses were more confident in their assessment than novice & experts include a wide
nursing students from 5 intermediate but novices more confident than intermediate. range of experience (0-15

M
institutions, 194 female, 20 – years of experience).
43 yrs. Significant main effect of expertise on analgesic administration.

Expert: paediatric nurses Significant main effect of group on analgesic administration.

from 11 hospitals, 180 Interaction (group/expertise) revealed novice nurses’ decisions to

d
female, 22 – 56 yrs. administer analgesics did not differ from other groups.
He et al. (2005) To describe nurses’ N =178 (5 hospitals, 12  Questionnaire designed by Polkki et Cognitive behavioural methods reported as most frequently used. Strengths: Appropriately
use of non-pharm surgical wards), all female - al. (2001) Preparatory information, distraction, breathing techniques; relaxation, conducted questionnaire
China
Cross-sectional survey
methods in managing
paediatric postop
pain. te
20 - 54 yrs.
Range of demographics
(educational backgrounds,
position, professional
 Questionnaire elicited reports of: 1)
preparatory information provided
and 2) information on non-pharm
methods provided to parents.
imagery, positive reinforcement. Comforting/reassurance was most
frequently used emotional support strategy (69% reported always
using it).
study. Used a measure
with reliability and validity
(although validity of
including reassurance is
ep
To describe the experience, children of own,  Plus open ended questions asking + correlation between ↑ age & staff position with use of certain questionable)
relationship between and if their children had ever about nurses’ perceptions of strategies. Experience had variable association with different methods
use of non-pharm been hospitalised. barriers to use of non-pharm postop but positive overall with better preparation in some areas. Having Weaknesses: Self-report
methods, and pain management strategies. children of own was associated with provision of some preparatory and due to inherent bias
individual and information (i.e. type of procedure) & some methods (positioning, difficult to generalize and
c

institutional factors. verbal rewarding). state that results are

reflective of practice. Also
Ac

Most cited barriers/limitation: lack of nurses, lack of knowledge about reassurance highlighted as
pain management, non-pharm methods not part of conventional a non-pharm strategy but
postop practice. Influence of traditional Chinese culture on pain reassurance in pain is
management knowledge cited by half of nurses as barrier. correlated with increase
Open ended questions: most mentioned ‘other’ barriers: lack of pain and pain expression.
efficacy/slow effect of non- pharmacological methods; lack of parental
support or cooperation; issues with hospital environment.

He et al. (2011) To examine nurses’ N= 134 (>3 months Questionnaire designed by Polkki et al, Nurses reported that they always or nearly always used: relaxation Strengths: As above:
Singapore use of non-pharm pain experience), 7 paediatric 2001 (see above). (89%); breathing techniques (88%); distraction (75%); positioning Appropriately conducted
management methods wards), 134 female, (61%); emotional support - comforting/reassurance (79%); touch questionnaire study. Used
Cross-sectional survey for children postop. 20-61 yrs. of age. (73%); helping with ADL (82%); creating comfortable environment a measure with reliability
Range of experience in (76%); giving preparatory information (54%). and validity (although
Specifically: 1) what paediatric surgery (4 mos. – validity of including
methods do they use, 20 + yrs.), educational Overall nurses reported using range of non-pharm pain relief reassurance is

Page 47 of 56
 t
ip
2) what background background and position. strategies routinely. Only half provided reported providing questionable)
factors are related to preparatory information routinely. Weaknesses: As above:

cr
methods used. Self-report and due to
inherent bias difficult to
generalize and state that
results are reflective of

us
practice. Also reassurance
highlighted as a non-
pharm strategy but
reassurance in pain is
correlated with increase

an
pain and pain expression.

Higgins et al. (1999) To evaluate the N = 114, 55 girls Wong-Baker Faces Scale used for children All patients had paracetamol prescribed. Not all children received all Strengths: Descriptive
United States administration of ATC 2 mos. to 18 yrs. (M = 5 yrs. where appropriate for age (n=71). prescribed doses. Younger children were less likely to receive study using a prospective
analgesia and 3 mos). paracetamol as prescribed. Thoracotomy group received less data sample.

M
Retrospective chart documentation of prescribed paracetamol overall & thoracotomy & < 24 mos. received
audit pain assessment with 4 groups: Weaknesses: Only
only 76.5% of prescribed doses. Patients receiving ATC local
percentages reported.
pro re nata (prn) 1) sternotomy & <24 mos.; 2) anaesthetic received only 52.3% of prescribed doses.
analgesics in children ANOVA or other inferential
sternotomy & >24 mos.; 3)
post cardiac surgery analysis would have been
thoracotomy & <24 mos.; 4) 71 patients old enough to use the Wong Baker faces scale: 38% had

d
helpful. No reason for non-
thoracotomy & >24 mos. pre PRN analgesic pain scores documented & 15.5% had re-evaluation
administration of ATC
pain scores documented.
ordered analgesic

te Overall 12.3% of patients had no pain assessment documented

despite receiving PRN medication.
provided other than the
use of infiltrate and
therefore unclear if RN
held paracetamol for
ep
appropriate or
misunderstanding of the
role of multimodal
analgesia
c
Ac

Page 48 of 56
 t
ip
Hudson-Barr et al. To determine the N = 55 (all female) Videos of 8 post-op infants (2 girls, ≤ 6 Signal detection theory used to examine nurses’ sensitivity to Strengths: Robust study,
(1998) importance of infant M age: 35.3 yrs. mos. of age). behavioural cues: nurses were not sensitive to differences between strong analysis.

cr
behaviours for 8 videotapes classified as analgesic active the two pain states (analgesic active & inactive), infant behaviour did
United States paediatric nurse (infant in video received analgesic within not prompt nurses to administer an analgesic when the infant was Weaknesses: Vignette
Experimental: within decision-making the previous 4 hrs) & analgesic inactive classified as medication inactive. Nurses more likely to recognise when studies may not match
related to analgesic segments (infant in video had received infants did not need analgesics than when they did (Percent actual behaviours in

us
subjects design
administration. last analgesic over 4 hrs. previously)...
Segments from videos used to capture if
nurses would (or would not) administer
an analgesic to the infant, which
agreement between when nurses stated they would administer practice. Regression
analgesic and when infant required analgesic in the snippet was only analysis not significant so
18.8% whereas the percent agreement was 80% when the nurse results need treating with
stated she would not provide the infant with analgesia and the infant caution.
had just received it prior to video snippet).

an
individual infant behaviours helped them
make their decision, & to describe the
infant’s facial expression, body
movements, sleep-wake state,
vocalisations and self-comfort behaviours,
Behaviours used to make decisions of when to administer medication
when analgesics inactive: a) facial expression (42.8%); b) body
movement (60.7%) c) vocalization (25%); d) sleep/wake state (0); e)
self-comfort measures (42.8%); f) respirations (10.7%); g) general

M
and indicate if these individual behaviours
helped their decision regarding analgesic
administration.
appearance (64.3%).
Behaviours used to make decisions of when to administer medication
when analgesics active: a) facial expression (40.6%); b) body
movement (50%) c) vocalization (3.1%); d) sleep/wake state (9.4); e)

d
self-comfort measures (12.5%); f) respirations (62.5%); g) general
appearance (25%).

Nethercott (1994) To describe the

perceptions of
te
N = 22
RNs (2 different wards 1
Interviews: Structured, semi-structured
and focused interview.
Regression analysis: no consistent set of behaviours to account for
either proportion of agreement or nurse accuracy in this study.
9 of 10 nurses voiced that children’s experience of pain was influenced
by the child’s cognitive development.
Strengths: Included nurses,
parents and children.
ep
United Kingdom paediatric nurses, paediatric hospital; n=8
Interviews parents and children parents whose child had Child demographic and medical data Nurses reported using a range of criteria to assess pain (behavioural Weaknesses: No rationale
about postop pain major surgery; n=4 children collected. changes, nursing judgement, parental assessment, verbal for reason for qualitative
assessment and post surgery. communication, physiological signs, oral expression, body language, approach. No discussion of
management. facial expression, analgesia given, age of child, time since surgery). the methodology that
c

Nurses - 3 mos.– 13 yrs. informed the study or

paediatric nursing All nurses emphasised RN responsibility for pain assessment, most where the researcher is
Ac

experience. referred to a medical tendency to underestimate pain, & some situated within the study.
No other participant expressed anxiety about the use of PRN medication, stating they No discussion of ethical
description. wouldn’t give it as regularly as ATC prescriptions. issues or approval to
conduct the study apart
Almost no nurses used pain assessment tools. from a sentence stating
participants were
informed that they could
withdraw at anytime.
Conclusions beyond
findings.
Polkki et al. (2001) To describe RNs’ use N = 162 (convenience sample Study specific questionnaire. Non-pharm strategies reported as being used frequently Strengths: Robust study
of non-pharm pain of RNs, 8 surgical paediatric 31 questions about RN’s views of factors (always/nearly always): comforting/reassurance (98%), touch (83%), and included limitations
Finland management wards across 5 hospitals), facilitating or impeding their use of non- presence (77%), helping with ADL (99%), creating a comfortable related to study design.
Cross-sectional survey strategies with postop 158 female. pharm pain management techniques. environment (69%), positive reinforcement (verbal rewarding 82% & Inferential statistics used.
patients aged 8 – 12 Aged 22-59 yrs. Range of Scores on 5 point Likert scale (1 = totally concreted action 52%), preparatory information (97%), positioning

Page 49 of 56
 t
ip
yrs. experience (3 mos.- 37 yrs.), agree, 5 totally disagree). (98%) Weaknesses: Related to
64% had children of own, inherent issues in study

cr
1) What strategies are 50% had children previously design (self-report
used & 2) Factors hospitalised. questionnaire) and
associated with their concern re: reassurance as
use. a non-pharm strategy.

us
Polkki et al. (2003) To describe factors See above Questionnaire as per Polkki et al. (2001) & Factor analysis conducted to establish factor categories related to use Strengths: Captured data
promoting and questions to capture demographic of non-pharm techniques: on organizational factors
Finland hindering RN’s use of information & perceptions of in addition to individual
Cross-sectional survey non-pharm pain organisational factors impacting their Promoting factors: nurses’ competence; versatile use of pain nurse factors. Factor
management work. alleviation methods; workload/time; child’s age/ability to cooperative; analysis used to provide a

an
strategies in postop parental participation. more sophisticated
pain with children 8- Hindering factors: nurses’ insecurity; beliefs regarding parental analysis compared to
12 yrs. of age). roles/child’s ability to express pain; heavy workload/lack of time; many of the purely
limited use of pain alleviation methods; work model/patient turnover descriptive studies.
rate.

M
Weaknesses: Simple
Promoting factors: highest number of totally agree responses for correlation used to
nurses’ desire to progress in career, learn different non-pharm determine the relationship
methods, & RN’s experience in use of techniques. between factors and
demographics. Although

d
Hindering factors: highest number of totally agree responses for rapid statistically significant the
patient turnover, lack of education on techniques, & lack of time. correlations were small
therefore may not be

Prot-Labarthe et al.
(2008)
To describe the
evaluation and te
N= 200 with uncomplicated
appendectomy, spinal
Charts were analysed, with special
attention to nurses’ entries, prescriptions,
929 prescriptions and 2,388 numerical pain scores were recorded.
clinically relevant.

Strengths: Included data

from two countries. Clear
ep
treatment of pain in fusion, and medication administration sheets. Pain was recorded at 70.8% of vital-sign evaluations at the Robert- details on rationale,
France and Canada paediatric patients in ureteroneocystostomy Debre Hospital compared to 30.9% at the Sainte-Justine hospital. methods, analysis.
Retrospective chart 2 hospitals or laparoscopic Main outcome measures:
audit cholecystectomy Patient characteristics and variables A validated age-appropriate pain evaluation tool was used in 97.4 and Weaknesses: Did not
pertaining to pain evaluation (tool and 94.1% of evaluations in these two hospitals, respectively. capture details about
c

result) and treatment (date, prescription nursing staff beyond part

and administration details) Analgesic dosage was appropriate in 92.5% of prescriptions at Robert- of the pain team at each
Debre (RD) and 86.0% of those at Sainte-Justine (SJ). site yet chart review
Ac

captured frequency of
Proportion of analgesic actually administered by nurses (excluding PCA assessment and analgesic
and epidural) was 69.9% at RD and 79.7% at SJ for fixed scheduled administration.
analgesics. No information on the appropriateness of withholding Differences noted in pain
fixed scheduled analgesics and 31.3% and 29.4% indicating there was assessment characteristics
no difference between prn dosing at the 2 hospitals by RNs. across the 2 hospitals. No
details on policy
differences or
organizational structure.
Rheiner et al. (1998) To describe selected N = 29 RNs (n= 20 paediatric Demographic information sheet Most common reason reported for intervening was child’s self-report Strengths: Used a
factors associated units, n=9 PICU). developed for the study. Data collected or verbal pain statement; second most common reason was the validated tool to assess
United States with RNs pain Primary focus ward RN pain by staff nurses for first 5 post-op days. Oucher score. Nonverbal behaviours were considered more when a children’s pain.
Retrospective chart assessment and management thus included. child requested pain medication.
audit management of Pain levels reported using the Oucher In some instances a combination of verbal + nonverbal indicators Weaknesses: Purposeful

Page 50 of 56
 t
ip
postop pain for Charts reviewed n= 19 (children >3 yrs.). prompted the nurse to intervene. Physiological cues rarely used sample with no
children undergoing 9 Female justification. Sample size of

cr
orthopaedic surgery. 5-17 yrs. (mean =13.1 yrs.) Study designed data collection form:
systematic recording on pain indicators,
(Oucher scores, physiological, verbal &
non-verbal parameters), analgesics
Significant relationship between the amount of analgesia (converted 19. Confounder of nurses
to morphine equivalent) and children’s self-reports of pain. being taught how to use
the Oucher scale at the
No significant differences found between pain scores reported by boys outset of the study may

us
prescribed & administered, and non- and girls. have influenced their pain
pharm methods used. assessment practices.
Most frequent non-pharm strategies documented were positioning Given the small sample
(45.1%); reassurance (25.4%) and ice (9.1%). Massage, heat, size of 19 correlations
conducted between

an
relaxation, teaching and distraction/play each accounted for less than
5% of these interventions. factors not generalizable.
Romsing (1996) To determine RN N=72 Data on paracetamol dosage & interval All children received at least 1 dose. Most administered doses were < Strengths: Used a reliable
administration of Children within the first 24 between doses collected for each child. optimal rectal paracetamol despite appropriate prescription & and valid measure for
Denmark, paracetamol post hrs. postop tonsillectomy. children’s pain (poker chip
available dose supply.
tonsillectomy and n=35 girls Pain assessment using poker chip tool tool). Based comparison of

M
No design clearly
stated. From reported adenoectomy 3-12 yrs. old. Average paracetamol given rectally was 68.4 mg/kg/day +/- 24.6 mg acetaminophen
data likely to be chart (amount & frequency) and prescribed was 90 mg/kg/day & average dose interval was 8.8 hrs. administered to published
review. +/- 4.4 hrs. guidelines

Pre-post rectal paracetamol change in pain score went from 2.1 to 1.7 Weaknesses: Method

d
(19% change). discussion lacking many
details (how study

te
Great variation was noted; some children received only 21% of variables were captured,
therapeutic dose while another received 141% of therapeutic dose. No ethical approval, states
indication for variation (different nurse or child specific factors). had 72 children enrolled
but only 2 nurses and thus
ep
the study showed the
practice of 2 nurses over
72 children and therefore
difficult to generalize to
other nurses practice.
c

Ross et al. (1991) Investigated postop N = 113 Study specific questionnaire: Nurses chose to provide more analgesic medication: Strengths: Details in
PRN medications RNs from 4 hospitals caring Clinical vignettes (adapted from Burokas  To children depicted (displaying behavioural cues) in greater pain method section clear.
Ac

United States, decisions for 2-13 yrs. old. 1985) describing various situations
Mixed methods: following heart surgery in 5 yr. old boys.
Cross-sectional survey RNs respond to each vignette indicating
how they would respond in practice and
design with quasi -
indicating their choice of PRN-prescribed
experimental phase. analgesics from list provided. Also asked
to estimate the child’s pain intensity level.
15 questions concerning RNs beliefs,
knowledge and comfort regarding opioid
analgesics, knowledge about children’s
experience and expression of pain, and
RNs perceived responsibilities.
Part 3: demographic data
Acknowledged limitations
(Mean 2.30 versus 1.69 on Day 1 post operatively)
 On the first post-op day overall compared to the third day (Mean of the study (i.e. vignette
2.33 versus 1.67). may not represent
behaviour).
On first post-day RNs would provide less medication to children
depicted as having a likelihood of permanent heart problems. The Weaknesses: Those
opposite was true on the third post-op day. Perhaps indicating inherent in study design
concern of side effects from analgesia in the immediate post op (self-report bias),
period. representativeness due to
response rate.
By the third day postop nurses provided the same morphine
equivalent units regardless of the amount of behavioural cues the
child depicted =1.06 which may indicate that children’s pain course
was treated based on days since surgery rather than displays of pain.

Page 51 of 56
 t
ip
Salantera et al. To explore nurses’ n = 303 (20-59 yrs. old), RNs completed questionnaire about their Pain assessment: Strengths: Included both
(1999) assessment and working in university hospital pain assessment and documentation 34% of RNs indicated a pain assessment tool was available. Younger chart audit and self-report

cr
documentation of caring for 50 paediatric practices, and opinions of their own nurses were more aware of pain tools. 18% used Faces Pain Scale questionnaire. Thus,
Finland pain in children patients post appendectomy readiness to assess and evaluate the Revised regularly, 12% used it sometimes, 4% used a VAS, 20% of used attempted to provide a
Cross-sectional survey process of taking care of children in pain. verbal indications pain words. 98% of nurses assessed pain by clearer understanding of
design combined with observing changes in a child’s behaviour. 80% used changes in reported practice and

us
retrospective chart physiological indicators. knowledge and
audit. behaviours.
Around 90% of nurses felt their ability to assess a child’s pain was
good or fairly good. Weaknesses:
Questionnaire results from

an
Documentation: nurses at several hospitals
35% of nurses reported documenting the pain intensity and duration but chart audit only at 1.
Chart review = 92% documented medication given; 50% general Thus difficult to know if
comment about pain care. the chart audits reflect the
30% of charts had no comments about pain during the first 24 hrs., practice of the nurses who

M
70% had a least one comment in first 24 hrs. 14% of charts had no answered the
comments re pain in second 24 hrs., 43% had some comment. questionnaire. Differences
20% of children did not have pain according to the charts between the questionnaire
Most common notes: “The child is in pain: or “The child is aching”. No and the chart audit may be
comments about assessment technique. due to different nursing

d
populations being studied.
Shrestha-Ranjit & To examine paediatric N=106 charts (5-15 yr. olds) Study specific tool of 25 items. Pain Assessment: Strengths: Clear details in
Manias (2010) nurses’ pain over a 3 day period (day of  Child demographics Assessment of pain using a specific tool was not normally documented methods section.
Australia
Retrospective chart
assessment practices
in children post
surgery for fractured
lower limb and to
te
surgery, day 1 and day 2
postoperatively)

Exclusion criteria: <24 hours




Documented RN’s pain assessment
practices (i.e. use of pain tool).
Pharmacological and non-pharm
interventions
in 87.8% of charts.
RNs assessed (documented) pain less often than expected to in the
postop period according to hospital protocol
For patients who had a pain score documented 75.6% had an elevated
Conducted content validity
survey of study specific
audit tool.
ep
audit
compare these with hospital stay, multiple pain score postop. Weaknesses: Inherent in
evidence-based injuries, admitted to PICU, study design. (i.e. chart
guidelines other fractures. 60% of the ordered PRN paracetamol doses were administered on the audit so reflective of
day of surgery, 70% on day two and three. documentation practices
69.8% had some form of ATC analgesia for the day of and into day one and perhaps not clinical
c

of the postoperative period. practice, postoperative

80% of ordered prn codeine given throughout the postoperative orthopaedic surgery for
Ac

period but significantly (no stats provided) fewer doses of oxycodone fracture repair only).
& tramadol were administered than ordered.

Information about non-pharmacological interventions only

documented in 23.6% of charts

Simons & Moseley To examine influences N = 175 children’s charts Analgesia prescriptions and pain related Analgesic prescriptions: Strengths: Conducted study at
(2008) on RNs’ pain (n=80 paediatric hospital documentation from the charts. Paracetamol was the most frequently prescribed medication, 2 hospitals increasing
management of patients, n= 95 paediatric diclofenac second. Only 74 % of the full amount of paracetamol generalizability. Compared ATC
United Kingdom children in the first 24- ward in general hospital was given. Children had a greater chance of receiving paracetamol with prn and therefore able to
Retrospective chart hours postop. patients). and diclofenac if prescribed ATC. PRN paracetamol only given half show ATC does not guarantee
audit 71 girls as often as was possible in the first day post op. administration of all doses of
analgesia.

Page 52 of 56
 t
ip
Codeine mostly prescribed PRN and least administered analgesic.
Of the children who had codeine prescribed 86% received at least 1 Weaknesses: Codeine listed as

cr
dose within the first 24 hrs. but not all possible doses were one of the oral medications, no
administered. ATC vs. PRN: 84% of the children who had codeine indicate if the child was also on
prescribed regularly received all their doses, only 70% of those for IV morphine and thus making
whom it was prescribed PRN had all possible doses. Reasons for codeine administration

us
not providing the does were not captured. inappropriate. Also did not
examine the interaction
Ibuprofen was the 3rd most often prescribed medication. ATC and between oral medication
PRN data recorded together; only 55% of the possible doses administration and child
administered. Reasons for not providing the doses were not receiving morphine infusion.

an
captured. For example, it is unknown if RNs held this analgesic if
the child also had paracetamol and or codeine.
Simons & Moseley Explored the Same as for Simons & Data collected from prescription and pain- Documentation: Strengths: Conducted study at
(2009) influences on nurses’ Moseley (2008) related observation charts 72% of all children had pain scores recorded in the first 24 hrs. 2 hospitals increasing
pain management of post-op. 28% of all children had no pain score recorded. Of those generalizability. Methods

M
United Kingdom children in the first 24- with a pain score 31% had no pain; 30.1% had mild pain; 30% had clearly detailed.
Retrospective chart hrs. postop. moderate pain; 8.8% had severe pain.
audit Weaknesses: No description of
Paediatric unit nurses recorded pain scores more often than nursing staff at the 2 hospitals
paediatric hospital nurses (22% more often). Pain levels were and thus unsure if there are

d
significantly higher at the paediatric hospital (where they had less systematic differences in the
documentation of assessed pain scores). There were more pain types of nurses (RN, nursing
related documentation forms at paediatric hospital were there was assistant, advanced practice

te a well established pain team as opposed to the paediatric unit

where pain scores were recorded on the vital sheet.

Older children significantly more likely to have pain than younger

nurses, staffing ratios). Wide
variety of surgery types and
little information of the
surgeries within larger
ep
children, Nurses assessed and recorded pain on 75% of children categories (i.e. orthopaedic
under 5 yrs. of age, 66% for children aged 5-12 yrs. of age; 71% for surgery but not if this is
children over 12 yrs. of age. No difference between pain scores of fracture repair or spinal
boys and girls (no statistical test presented). implementation); stated that
surgery type did not appear to
c

Assessment: differ significantly between

Children’s level of pain did not differ between different surgeries types of procedure (no
Ac

(no statistical test presented) but type of surgery affected nurses’ statistics presented yet stated
assessment practices – frequent recordings for abdominal cases an important finding). Aim of
but very infrequent pain assessment for appendectomies and the study was to determine
tonsillectomies. influences on pain
management but only
examined assessment.
Smyth et al. (2011) Explore nursing Chart review: Chart review: age, sex, type of surgery, 67 children had a regular scheduled analgesic with paracetamol Strengths: Combined
practices and decision N= 95 children’s chart type dose and route of prescribed being the most frequently prescribed regular analgesic. 72 (76%) chart audit with
Australia making associated patient. analgesia, prn analgesia administered, had a PRN analgesia prescription and 46 (43.8%) received at least observation follow up.
Mixed methods with prn postop 43 girls; 6 weeks-15 yrs. of who initiated the analgesia, assessment one dose of PRN analgesia; 182 PRN analgesic doses were provided Methods clear.
design (retrospective analgesia in children age. related to the administration of the prn (most PRN given on Day 0 or Day 1 post op). less than 50% of the Weaknesses: No
chart review, Observation/Interviews: analgesia, documentation 182 cases of PRN analgesia had the reasons for administration inferential statistics
participant N=18 RN (range of documented (documentation suggested nurses & child used to examine chart
observation & experience from student communicated about the need for PRN analgesia, 2 occasions audit data (i.e. did
nurses to unit manager). Qualitative: family requested dose). Only in 12 cases (7%) was a pain score young children receive

Page 53 of 56
 t
ip
interviews) Observation 2-5 hrs. per day. documented with the administration of PRN analgesia. less analgesics than
Interview questions asked to consenting older).

cr
nurses during the observation period.
How did you choose the type of
medication to administer for pain?; What
made you decide to give an analgesic to
Nurses reported using clinical judgement to determine analgesia Conclusions beyond the
needs (varied based on experience of the nurse but included age, findings. Authors
verbal & behavioural signs, self-report, parent report, experiences comment that the
with own children & own pain experience). Less experienced complex decision

us
the child? nurses less confident in their assessment of pain and need for making to administer
analgesia. Some used the FLACC or the Wong-Baker Faces scale to or not administer prn
assist with pain assessment of need for analgesia. Pain assessment analgesics was based
chart existed on the ward but most nurses did not know that this on child age and other
existed and thus did not use it. factors. Results

an
provided were only
Nurses were observed carrying out varied practices: not giving a comments about
bolus with low respiratory rates but instituting other non-pharm babies crying and new
strategies and re-evaluating effectiveness. Others noted to not give nurses perhaps lacking
prn in favour of non-pharm but no re-evaluation of effectiveness; experience in

M
did not wake children for ATC analgesia; some children did not behavioural signs of
receive PRN analgesia in a timely manner despite obvious signs of pain to administer
pain; re-evaluation observed to be seldom practiced post analgesics. No other
administering PRN analgesia; some nurses observed to call age related data or
medical staff when pain management inadequate, parents decision-making about

d
involvement varied. analgesic
administration was
offered.
Twycross (2007a)
United Kingdom
To determine if
paediatric nurses pain
management
practices adhere to
te
N= 13 female RNs working
on children’s surgical ward.
20 – 49 yrs. of age
Range of education (D – G).
Study specific structured observational
tool based on the practice guidelines.

Field Notes: Researcher recorded pain

Analgesic use:
Pre-emptive analgesia was not routine (even for procedural pain);
most analgesic administration based on verbalisations of pain.
Strengths: Methods
clear, structured and
unstructured data
collected. Long
ep
Mixed-Methods
Design (this article Royal College of Observations conducted management related nurse behaviours Nurses provided multimodal analgesic (paracetamol & ibuprofen). observation time 185
reports on Nursing (1999) & over 3-4 months. Each nurse (verbal & non-verbal) during the Analgesics not always given even if prescribed ATC; if opioids hrs.)
observational data) Agency for Health shadowed for 2-4 shifts. observation period (5 hours over 2-4 prescribed with a dose range nurses generally administered lower
Care Policy Research shifts for each nurse).Total of 185 hrs. of end of the range. Opioid administration was the type of pain Weaknesses: Data
(1992) best practice observation. management most nurses asked others for advice. No statistical collected on one unit
c

guidelines counts provided thus unsure if nursing

practice would differ
Ac

on a different unit or
Engagement of parents or child in the pain management of pain hospital. Small sample
was minimal and inconsistent. Home pain management upon size.
discharge was explained—encouraged paracetamol & ibuprofen,
discouraged codeine use.

Non-pharm strategies: Nurses seen to use distraction on 3

occasions for procedural pain. Most post op pain management,
even simple methods such as distraction were not used regularly.
Non-pharm strategies observed being used on 4 occasions (by 4
different nurses).

Documentation: Not always consistent; analgesia documented but

not other pain relieving strategies; re-evaluation post analgesia
not: documented even when done.

Page 54 of 56
 t
ip
Twycross (2007b) To examine whether See above Revised Pain Management Knowledge No relationship found between theoretical knowledge and use of Strengths: Methods
there is a relationship Test - originally developed by Salantera, pain assessment tools or the evaluation of effectiveness of clear, observation and

cr
United Kingdom between individual 1999. interventions. valid questionnaire
Mixed methods nurse’s theoretical used to capture nurses
design (this article knowledge and their Observations: See above No association between theoretical knowledge and documentation pain management
reports on RN Pain pain management practices. knowledge. Long

us
Management practices. observation time 185
Knowledge Test No association between theoretical knowledge and use of non- hrs.) Detailed results to
results & pharm methods. illustrate links between
observation). pain knowledge and
practice.

an
Weaknesses: Same
sample as above (see
above). Due to small
sample size no

M
inferential statistics
conducted to
determine the
relationship between
knowledge and

d
practice.
Twycross (2008) To establish whether See above Paediatric Pain Training Needs Taking a pain history: Rated as moderately-highly critical but was Strengths: Methods
there is a relationship Questionnaire. not observed in practice. clear, structured and
United Kingdom
Mixed methods
design
between perceived
importance of a pain
management task and
individual nurses’
te Observations: see above. Using a pain assessment tool: Rated as highly critical by many of
the nurses (8) but only observed being used in practice by 3 nurses.
unstructured data
collected. Long
observation time 185
hrs.). Used a validated
ep
(questionnaire and
observational practices. Using behavioural indicators: Some rated it highly important, measure to assess
methods) others moderate-high importance, unclear whether behavioural nurses’ attitudes
cues were actually being used in practice or not as RNs not towards pain
questioned during the observation period. management tasks.
Provides data to
c

Using physiological indicators: Split responses (re: criticality) across illustrate differences
age groups. No participants observed using physiological between attitudes and
Ac

indicators. practice

Reassessing pain: Rated as highly critical but only 3 RNs observed Weaknesses: Same
doing this practice. sample as previous 2
studies (See above).
Documentation: Rated as moderately to highly critical by all
participants but not substantiated in practice.

Using non-pharm strategies: perceived as highly critical by 7 RNs, 5

moderate-high, 1 moderate. 3 participants used non-pharm
methods on 1 occasion each.
van Hulle Vincent & To determine the n= 30 RNs, 97% Female Vignettes using 2 cases depicting 10 yr. Mismatch: some nurses identified things such as disconnect Strengths: Clearly
Gaddy (2009) factors that influence Age (M= 28.8 yrs.) old boys post abdominal surgery. between smiling and pain level while other nurses stated that presented study.
nurses decision Years of paediatric Difference between the 2 vignettes, 1 boy smiling behaviour does not mean the child is not in pain. This
United States making in experience (M = 4.7 yrs.); smiled and one boy was grimacing both impacted some of the nurses choice around morphine: some did Weaknesses: States

Page 55 of 56

Qualitative administering education (90% with
descriptive design analgesics by undergraduate degree)
controlling for child
factors
Woodgate & To explore parents N = 11 children from 2
Kristjanson (1996) and nurses’ responses different paediatric units
to hospitalised under-going surgery (cardiac,
Canada children’s postop pain. plastic or urological)
Grounded theory. 2 - 6 yrs. of age
d
Average length of stay = 6
days.
te
n=22 parents.
n= 24 RNs & nurses
assistants (all female) with >
5 yrs. experience.
ep
Observation periods were 2-
8 hrs. Total of 250 hrs.
c
Ac
had 8/10 pain.
cr
2 multiple-choice questions (how much
pain do you think this child has & how
much morphine would you administer).
Think aloud (open-ended questions) to
us
capture nurses’ decision-making.
Content analysis (based on deductive, a
priori codes) but also allowed for new
codes (not stated if new codes were
an
needed).
Compared the codes for RNs who chose
to administer↑ dose of morphine & RNs
who chose  dose of morphine.
M
Children observed in the immediate
postop period until their discharge day.
Their responses to pain and how
caregivers (parents and nurses) reacted to
them were noted.
Informal interviews with children, parents
and nurses during observational data
collection
Detailed field notes recorded.
t
ip
not give a full dose because the smiling child did not seem qualitative study using
uncomfortable while others would give a full dose even if smiling content analysis. No
as the child rated their pain a 8/10 and the child may be trying to discussion of
hide the pain. There were also differences for the grimacing child, philosophical
some nurses were concerned about over medicating as he had had grounding of
2 mg two hours ago, and felt that if a lower dose did not work this qualitative work. Codes
time around you could always give more. were a priori not
inductive. Results used
RNs who would administer ↑ dose of morphine more likely to descriptive statistics to
make positive or neutral remarks about the mismatch between represent answers to
pain rating and facial expression. RNs who would administer  questions with minimal
dose of morphine made negative comments about the mismatch exemplars and
between pain rating and facial expression. Some nurses therefore
demonstrated inaccurate information about the duration of action, transferability of
half-life, and amount of morphine needed to cause respiratory findings may prove
depression. difficult.
Theme: How parents and nurses take care. Includes monitoring Strengths: Identified
(looking for signs of pain), comforting, and carrying out technical the type of qualitative
care activities. approached used
(grounded theory) and
Strategies used to help care for children in pain: discussed issues of
Knowing: For nurses knowing the child in general as opposed to rigour
knowing when the child was in pain. (trustworthiness).
Clearly described
Nurses usually had difficulty articulating how they could tell a child
was in pain. Most frequent answer: verbal behaviour (i.e. crying,
stated they had pain), second most frequent response overt non-
verbal behaviour (i.e. fighting or restlessness). Nurses rarely relied
methods.
Weaknesses: Minimal
discussion of
philosophical
on children’s facial expression or in changes in their mood. foundation and
grounded theory. No
The quality and quantity of time a nurse spent with a child theory offered as one
impacted their pain assessment. Nurses would often spend a few would expect from
minutes with the child when they looked comfortable and miss grounded theory.
pain signs as they left the room. Analysis more in
keeping with a
Factors affecting nurses’ caring: thematic analysis.
Value judgements:
Good children were those who did not complain. The more overt
children’s behaviours were the more likely they were to be labelled
as hysterical, whining or miserable. Noisy children who had parents
who intervened usually received more pain medications compared
to quiet children or children with passive parents.
There was a perceived hierarchy of suffering based on child’s
diagnosis
Duration of time in pain rather than intensity often seen as more
important
Misconceptions
Nurses had overwhelming fear of addiction to analgesics.
Page 56 of 56