REVIEW ARTICLE

The research gap in chronic paediatric pain: A systematic

review of randomised controlled trials
R. Boulkedid1,2,3, A.Y. Abdou1, E. Desselas4,5, M. Mane
gat1, T.G. de Leeuw6, J. Avez-Couturier7,8,
9 10 11 1,2,3
S. Dugue , C. Mareau , B. Charron , C. Alberti , F. Kaguelidou4,5,12, on behalf of the GAPP
Consortium*
1 AP-HP, Ho ^pital Robert Debre
, Unite

d’Epide

miologie Clinique, Paris, France
2 Universit
e Paris Diderot, Sorbonne Paris Cit
e, UMR-1123, ECEVE, Paris, France
3 Inserm, U1123 and CICEC 1426, Paris, France
4 Inserm, CIC 1426, Paris, France
5 Department of Pediatric Pharmacology and Pharmacogenetics, AP-HP, Ho ^pital Robert Debre

, Paris, France
6 Department of Anesthesia, Sophia Children’s Hospital and Center for Pain Medicine, Erasmus Medical Centre, Rotterdam, The Netherlands
7 Department of Pediatric Neurology, CHU Lille, Children Pain Clinic, Lille, France
8 CHU Lille, Clinical Investigation Center – Innovative Technologies, INSERM CIC-IT 1403, Lille, France
9 Pain Management Unit, Ho ^pital Robert Debre
, APHP, Paris, France
10 Center of Chronic Pain and Migraine Evaluation and Management in Adults and Children, Centre Hospitalier Universitaire de la Timone,
Marseille, France
11 Pain Management Unit, Ho ^pital Necker-Enfants Malades, APHP, Paris, France
12 Universit
e Paris Diderot, Sorbonne Paris Cit
e, EA08, Paris, France

Correspondence Abstract
Florentia Kaguelidou
E-mail: florentia.kaguelidou@aphp.fr, Background and Objective: Chronic pain is associated with
florentia.kaguelidou@rdb.aphp.fr significant functional and social impairment. The objective of this review
was to assess the characteristics and quality of randomized controlled
Funding sources trials (RCTs) evaluating pain management interventions in children and
The study is part of the European GAPP
adolescents with chronic pain.
study that has received funding from the
Methods: We performed a systematic search of PubMed, Embase and the
European Union Seventh Framework Pro-
gramme for research, technological develop- Cochrane Library up to July 2017. We included RCTs that involved
ment and demonstration under Grant children and adolescents (3 months-18 years) and evaluated the use of
Agreement No 602041”. The funders had no pharmacological or non-pharmacological intervention(s) in the context of
role in study design, data collection and pain persisting or re-occurring for more than 3 months. Methodological
analysis, decision to publish, or preparation quality was evaluated using the Cochrane Risk of Bias (ROB) Tool.
of the manuscript
Results: A total of 58 RCTs were identified and numbers steadily
increased over time. The majority were conducted in single hospital
Conflicts of interest
None declared. institutions, with no information on study funding. Median sample size
was 47.5 participants (Q1,Q3: 32, 70). Forty-five percent of RCTs
*The members of the GAPP consortium are included both adults and children and the median of the mean ages at
presented before Reference Section. inclusion was 12.9 years (Q1,Q3: 11, 15). Testing of non-
pharmacological interventions was predominant and only 5 RCTs
evaluated analgesics or co-analgesics. Abdominal pain, headache/
Accepted for publication
migraine and musculoskeletal pain were the most common types of
4 October 2017
chronic pain among participants. Methodological quality was poor with
doi:10.1002/ejp.1137 90% of RCTs presenting a high or unclear ROB.
Conclusions: Evaluation of analgesics targeting chronic pain relief in
children and adolescents through RCTs is marginal. Infants and children
with long-lasting painful conditions are insufficiently represented in
RCTs. We discuss possible research constraints and challenges as well as
methodologies to circumvent them.

© 2017 European Pain Federation - EFICâ Eur J Pain  (2017) – 1

Chronic pain trials in paediatrics R. Boulkedid et al.

Significance: There is a substantial research gap regarding analgesic

interventions for children and adolescents with chronic pain. Most
clinical trials in the field focus on the evaluation of non-pharmacological
interventions and are of low methodological quality. There is also a
specific lack of trials involving infants and children and adolescents with
long-lasting diseases.

1. Introduction methodological quality of randomized controlled trials

(RCTs) evaluating pharmacological and/or non-phar-
Chronic pain is a complex, multidimensional experi-
macological intervention(s) for the management of
ence that is generally defined as pain lasting more
persistent or recurrent chronic pain in children and
than 3 months (Merskey and Bogduk, 1994).
adolescents.
Although, the condition is more prevalent in adults,
epidemiological studies have shown that as much as
25% of children and adolescents have experienced
2. Methods
at least once recurrent or persistent pain (Perquin
et al., 2000; King et al., 2011). Migraine or func-
2.1 Electronic search query
tional abdominal pain account for the majority of
painful experiences but chronic long-lasting condi- Relevant RCTs were identified through electronic lit-
tions such as cancer or neurodegenerative conditions erature searches using the following databases:
may also cause significant chronic pain (Hagen et al., MEDLINE, EMBASE and the COCHRANE Library
2008; Palermo, 2009). (CENTRAL and Cochrane Database of Systematic
Experiencing chronic pain clearly has a negative Reviews), all from inception to July 17, 2017. The
impact on patients’ and relatives’ quality of life (Hun- following abbreviated search strategy was used: “per-
feld et al., 2001; Palermo and Eccleston, 2009). In chil- sistent pain”, “recurrent pain”, “continuous pain”,
dren, psychomotor development and social behaviour “chronic pain”, “analgesia”, “analgesics”, “children”,
are severely impaired leading to psychological distress, “paediatric”, “adolescent”, “teenagers”, “clinical tri-
physical disability and school failure (Eccleston et al., als” and other synonyms of these terms combined by
2006; Petersen et al., 2009; World Health Organization the operators Boolean (AND, OR, NOT). The exact
2012; Coffelt et al., 2013). High levels of anxiety and research strategies for all electronic databases are
depression in childhood are major risk factors for devel- given in Supporting Information Appendix S3. The
oping psychological pathologies in adulthood (Fearon search was done without any language restriction or
and Hotopf, 2001; Reinherz et al., 2003). Direct (use of date limits. Lists of references of identified studies,
health care) and indirect (e.g. parents’ work loss) costs systematic reviews and meta-analyses were further
are also particularly increased in the context of paedi- screened for relevant articles.
atric chronic pain (Groenewald et al., 2014). Studies were eligible if (1) they were RCTs defined
All of these reasons pledge for an early and efficient as any prospective study where participants were
treatment of chronically painful conditions in children randomly allocated to study groups, (2) included
and adolescents. Current therapeutic approaches rec- infants, children and adolescents (3 months to less
ognize the value of a multimodal treatment frame- than 18 years of age) and (3) their main objective
work, combining use of analgesics with physical, was to evaluate the effects of pharmacological and/
behavioural and psychological therapies (Odell and or non-pharmacological intervention(s) for the man-
Logan, 2013). However, the analysis of the only sys- agement of chronic pain. Chronic pain was defined
tematic review aiming to evaluate the effects of inten- as pain that persisted or re-occurred in a 3 months
sive interdisciplinary pain treatment in children and time-period (Treede et al., 2015). Pain assessment
adolescents with chronic pain was hampered by the was either the primary or the secondary study out-
non-randomised nature of the studies, their small come. Trials including both children and adults were
number and their methodological weaknesses (Hech- also considered. Abstracts, letters, duplicates, prelimi-
ler et al., 2015). As medical practice should be opti- nary publications and reviews were excluded. RCTs
mally driven by adequate research quality, it is published in languages other than French or English
important to evaluate the research available to sup- were secondarily excluded.
port chronic pain management. The aim of this review Retrieved articles were assessed by two indepen-
was to assess the characteristics and the dent authors (AY, RB), who read the titles and

2 Eur J Pain  (2017) – © 2017 European Pain Federation - EFICâ

R. Boulkedid et al. Chronic pain trials in paediatrics

abstracts to identify the relevant trials. Each author with 18 RCTs identified through manual reference
independently selected the trials to be included in search (Fig. 1; list of selected articles is given in Sup-
this review. Disagreements were resolved by discus- porting Information Appendix S2).
sion with a third researcher (FK). For all abstracts
considered potentially relevant, full texts were 3.1 General trial characteristics and study
retrieved. Full text article selection was indepen- population
dently performed by three authors (AY, RB, ED) and
Table 1 summarizes the main trial characteristics and
disagreements were resolved by consensus.
Fig. 2 displays the number of RCTs per year of publi-
cation and type of intervention evaluated. Most of
2.2 Data extraction
the RCTs were single-centre, hospital-based trials
Data were extracted using a structured data collection from Europe or Northern America and were recently
form (Supporting Information Appendix S1) which published (after year 2005: 36/58 62%). Public fund-
was pre-tested on ten randomly selected articles by ing prevailed although the information was lacking
one researcher (MM) and modified accordingly. The in more than half of the RCTs. Median sample size
form covered the following categories: general charac- was 47.5 participants (Q1,Q3: 32, 70). Only 32
teristics (e.g. study setting, year of publication, fund- (55%) trials were exclusively pediatric (children
ing), study population (e.g. age groups: infants [3– and/or adolescents) and the median of the mean
23 months], children [2–11 years], adolescents [12– ages at inclusion was 12.9 years (Q1,Q3: 11, 15),
17 years], adults [≥18 years]; size), clinical context and none included infants. The majority of RCTs
(e.g. type and source of pain, presence of an underly- (81%) evaluated the impact of a non-pharmacologi-
ing disease), trial design (e.g. nature of intervention cal intervention.
and comparator, outcome measures, pain assessment,
statistical conclusions) and methodological quality. 3.2 Clinical context
Participants presented a chronic pain which was per-
2.3 Methodological quality assessment
sistent in 1 RCTs (2%), recurrent in 26 (45%), both
Methodological quality was assessed using the in 11 (19%) but the type of chronic pain was not
Cochrane Risk of Bias (ROB) Tool implemented based specified in 20 RCTs. Participants presented an
on the guidelines of the Cochrane Collaboration (Hig- underlying disease in 9 RCTs (16%). Only 29% (17/
gins et al., 2011). The tool covers six methodological 58) of RCTs specified the physio-pathological type of
areas: sequence generation, allocation concealment, pain to be treated although most patients presented
blinding, incomplete outcome data, selective outcome with more than one type of pain: nociceptive pain in
reporting and other sources of bias. 10, neuropathic pain in 8, psychogenic pain in 6 and
For each study, ROB is described as low (all six mixed pain in 6 RCTs. Pain locations/causes are
domains are judged to be at low ROB) or high (one given in Table 2 according to type of intervention.
or more domains are judged to be at high ROB) or The majority of studies focused on the management
unclear (one or more domains are judged to be at of abdominal pain (64%) and headache/migraine
unclear ROB and none at high risk). Two authors (47%) however, 23 (40%) trials included patients
(AY, EB) assessed methodological quality indepen- with pain originating from more than one location/
dently and discrepancies were solved by consensus. cause. Of note, RCTs on the management of cancer,
myofascial, eye and psychosomatic pain included
2.4 Data analysis both children and adults.

We computed medians (first and third quartiles; Q1,

3.3 Trial design
Q3) for continuous variables and the number (per-
centages) for categorical variables. Analysis was per- All RCTs aimed to evaluate the efficacy of a pain
formed using SAS software version 9.3 (SAS Inc, management intervention and the majority were
Cary, NC, USA). parallel-group superiority trials (n = 52; 90%). Num-
ber of arms was 2 in 47 RCTs (81%), 3 in 9 RCTs
(16%) and more than 3 in 2 RCTs (3%). Median
3. Results
duration of study was 28.5 (Q1, Q3: 21, 50) months
Electronic search yielded a total of 936 articles. Alto- and median individual participation was 6 months
gether, 40 RCTs were selected for analysis together (Q1, Q3: 3, 12), respectively. Table 3 summarizes

© 2017 European Pain Federation - EFICâ Eur J Pain  (2017) – 3

Chronic pain trials in paediatrics R. Boulkedid et al.

Figure 1 Flow chart of RCTs.

the characteristics of the RCTs’ outcome measures. not required in 13 (20%) and not reported in 44
Assessment of pain was the primary outcome in RCTs (78%).
86% (50/58) of RCTs (single assessment [n = 12] or Among the 11 RCTs evaluating pharmacological
part of a composite outcome [n = 38]), and a sec- interventions, five evaluated the use of diclofenac or
ondary outcome in 8% (n = 8). Self-assessment of nefopam, acetaminophen-codeine with or without
pain was privileged (95% of RCTs) and the numeri- doxylamine, amitriptylin with or without pindolol,
cal rating scale (NRS-11, 50%) was the most fre- chlormezanone and drotaverine hydrochloride
quently used pain scale. Other important outcomes respectively; the remaining trials evaluated the
measures such as quality of life and pain-related dis- effects of antibiotics (n = 3) or combinations of vita-
ability were more rarely assessed (Table 3). mins (n = 3) to treat chronic painful symptoms.
A baseline period of pain assessment before ran- Overall, control groups comprised a placebo (n = 5)
domization was required in 83% (48/58) of RCTs. or an active reference treatment (n = 6). Of note,
However, the duration of this assessment period was only 4 of these 11 RCTs were exclusively pediatric.
reported only in 67% (32/48) of these trials and var- For RCTs evaluating non-pharmacological
ied between 1 day and 6 months (median: interventions (n = 47): 39 evaluated the efficacy of a
14.5 days). Also, a treatment ‘wash-out’ period was single intervention, (psychotherapy, n = 13; comple-
required in 1 pharmacological RCT (7 day duration), mentary therapy [e.g. hypnotherapy], n = 5;

4 Eur J Pain  (2017) – © 2017 European Pain Federation - EFICâ

R. Boulkedid et al. Chronic pain trials in paediatrics

Table 1 Characteristics of selected RCTs on paediatric chronic pain (N = 58)

Non-pharmacological intervention
All RCTs (n = 58) Pharmacological intervention (n = 11) (N = 47)
Number (%) or Median [Q1–Q3] Number (%) or Median [Q1–Q3] Number (%) or Median [Q1–Q3]

Continent of corresponding author

Oceania 1 (2) 0 1 (2)
South America 1 (2) 0 1 (2)
Asia 4 (7) 2 (18) 2 (4)
Europe 24 (41) 6 (54) 18 (38)
North America 28 (48) 3 (27) 25 (53)
Funding
Public 29 (50) 1 (9) 28 (60)
Private 4 (7) 2 (18) 2 (4)
Not reported 26 (45) 8 (73) 18 (38)
Number of centres
Single-centre 30 (52) 5 (45) 25 (53)
Multicentre national 10 (17) 1 (9) 9 (19)
Munticentre international 4 (7) 1 (9) 3 (6)
Not reported 14 (24) 4 (36) 10 (21)
Trial settinga
Hospital based 40 (69) 7 (63) 33 (70)
Community 23 (40) 4 (36) 19 (40)
Not reported 2 (3) 2 (18) 0
Study population
Minimal age at inclusion (years)b 9 [7–12] 10 [4–16] 9 [7–11]
Min-max 2–17 2–17 4–15
Maximal age at inclusion (years)c 17.5 [16–18] 20 [16–65] 17 [16–18]
Min-max 12–82 12–80 12–82
Age groupsa
Infants: (28 days to 23 months) 0 0 0
Children (24 months to 11 years) 42 (72) 6 (54) 36 (76)
Adolescents (12–18 years) 58 (100) 11 (100) 47 (100)
Adults (≥18 years) 26 (45) 7 (64) 19 (40)
Sample size 47.5 [32–70] 48 [31–90] 47 [32–70]
Min-Max 15–273 20–110 15–273
Number of patients analysed 39.5 [28–66] 39 [26–62] 40 [28–69]
Min-Max 6–269 19–124 6–269
Study duration (inclusion + 28.5 [21–50] 25 [9–29] 32 [22–50]
follow-up) (months)d
Min-Max 1–129 1–58 2–129
Duration of participation per 6 [3–12] 2.4 [0.69–12.2] 6 [3–12]
individual (months)e
Min-Max 0.0068–36 0.0068–36 1–15
a
The total percentage may exceed 100% because some RCTs include more than one category.
b
Missing data in 3 RCTs.
c
Missing data in 4 RCTs.
d
Missing data in 28 RCTs.
e
Missing data in 5 RCTs.

educational approaches, n = 2; surgery, n = 1 and 3.4 Methodological quality assessment

other interventions, n = 13 [e.g. exercise rehabilita-
Among the 26 RCTs conducted with a blinded
tion programme] and 7 evaluated the effects of
assessment of intervention efficacy, ten were con-
multiple interventions. Control groups were no
ducted in a double-blind and sixteen in a single-
intervention, n = 15; standard medical care, n = 12;
blind approach. Among the 58 RCTs, only 30/58
educational approaches, n = 7; placebo, n = 3; phar-
(52%) defined the exact methods of randomization
macological treatment, n = 1 and other (e.g. surgery
and 19/30 (63%) used computer random number
or dietary therapy), n = 9.

© 2017 European Pain Federation - EFICâ Eur J Pain  (2017) – 5

Chronic pain trials in paediatrics R. Boulkedid et al.

Figure 2 Trends of time of the number of RCTs according to the type of evaluated intervention.

generator. The allocation method was detailed only 4. Discussion

in fourteen RCTs (24%). Sealed envelope techniques
were more often used as method of treatment arm This is the first review to describe the current
allocation (9/14; 64%). research on pharmacological and non-pharmacologi-
Only six RCTs (10%) presented a low ROB (high cal pain management interventions in children and
ROB, n = 14; unclear ROB, n = 38) and they were all adolescents with chronic pain. Overall, few RCTs
published after year 2005 (data not shown). However, have been published, mainly single-institution, pub-
ROB varied with the type of intervention tested. licly funded trials of limited size. Only 55% were
Among RCTs testing pharmacological interventions, exclusively paediatric and none involved children of
more than half applied adequate methods of blinding <2 years of age. The majority focused on the evalua-
of outcome assessment, blinding of participants and tion of non-pharmacological interventions in chil-
personnel and reported complete outcome data dren presenting with headache/migraine or
(Fig. 3). For non-pharmacological interventions, abdominal pain. Methodological quality was poor,
blinding of participants and personnel, reporting of most probably related to the absence of adequate
complete outcome data and blinding of outcome reporting of study features.
assessment were inadequate or not described in more Chronic pain is acknowledged as a growing prob-
than half of the RCTs (Fig. 3). Also, RCTs involving lem with significant individual and societal repercus-
both children and adults presented a lower risk of bias sions that requires adequate and often
(4/26; 15%) than exclusively pediatric RCTs (2/32; multidisciplinary treatment approaches (Hechler
6%). Only 23/58 (40%) RCTs reported a sample size et al., 2015). Still, this complex health problem lacks
calculation and the majority (46/58) presented statis- consensus on clinical definitions, severity scaling and
tically significant results. intervention outcomes of interest even in adult

6 Eur J Pain  (2017) – © 2017 European Pain Federation - EFICâ

R. Boulkedid et al. Chronic pain trials in paediatrics

Table 2 Pain locations/causes persistent pain (Eccleston et al., 2012) although, one
Non- cannot refute the necessity of pharmacological treat-
Pharmacological pharmacological ments and the positive interactions between the two
Source of All RCTs intervention intervention therapeutic approaches. Still, evaluation of analgesics
pain N (%)a (N = 58) (N = 11) (N = 47) for the treatment of chronic pain in children was very
Abdominal painb 37 (64) 9 (82) 28 (60) scarce. In addition, RCTs focused mainly on two
Headache/migraine 27 (47) 4 (36) 23 (50) causes of pain: headache/migraine and abdominal
Musculoskeletal 15 (26) 1 (9) 14 (30) pain, while neglecting children and adolescents with
pain long-lasting conditions causing substantial pain e.g.
Sickle cell disease 2 (3) 0 2 (4)
cancer or sickle cell disease. Both findings underline
Fibromyalgia 2 (3) 0 2 (4)
the fact that for some clinical conditions, pain therapy
Postoperative pain 1 (2) 0 1 (2)
CRPSc 2 (3) 0 2 (4) remains empirical and mainly based on extrapolation
Cancer 1 (2) 1 (9) 0 of therapeutic schemas from adults (Gregoire and Fin-
Myofascial pain 1 (2) 0 1 (2) ley, 2013; Mercadante and Giarratano, 2014).
Eye pain 1 (2) 1 (9) 0 Clinical trials in chronic pain are altogether diffi-
Psychosomatic 1 (2) 0 1 (2) cult to design, conduct and interpret even in adult
pain
practice (Polydefkis and Raja, 2008; Dworkin et al.,
Not specified 3 (5) 1 (9) 2 (4)
2010; Moore et al., 2013a). Several methodological
a
The total percentage exceed 100% because some RCTs include more challenges, e.g. the heightened placebo response and
than one category. the use of subjective outcomes are also encountered
b
Includes epigastric pain, recurrent or functional abdominal pain,
in children (Dworkin et al., 2005, 2010; Birnie et al.,
recurrent abdominal pain of psychosomatic origin, chronic abdominal
pain, functional bowel disorders, irritable bowel syndrome. . .etc.
2012; Weimer et al., 2013). However, paediatric pain
c
Omplex regional pain syndrome. research may be exposed to additional challenges.
First, the number of children and adolescents pre-
medicine (Treede et al., 2008; Bouhassira and Attal, senting certain types of chronic pain, e.g. neuro-
2011; Moore et al., 2013b). The absence of diagnos- pathic pain, is known to be very small compared to
tic tools and the difficulties in classifying chronic adults (Howard et al., 2014). Thus, small sample
pain was reflected by the fact that one-third of RCTs sizes and trial participants with highly variable dis-
did not provide information about the physio-patho- ease profiles preclude treatment evaluation (Moore
logical type of participants’ chronic pain. This may et al., 1998). Second, the choice of an adequate
also be the reason why no RCT involved infants comparator to test therapeutic interventions is often
<2 years of age. Although pain perception and its problematic. Placebo controlled trials, the gold stan-
negative effects have been clearly identified in dard for drug testing, are not ethically acceptable in
infants as young as preterm neonates (Allegaert sometimes severely affected children and adoles-
et al., 2013), recent experimental studies advocate cents. In our review, placebo arms have only been
that neuropathic chronic pain is suppressed in the implemented when testing vitamins’ and antibiotics
youngest and may emerge later in adolescence for the management of abdominal pain. On the
(McKelvey et al., 2015; Fitzgerald and McKelvey, other hand, there are currently no active compara-
2016). But how can we clinically confirm the tors proven to be efficacious and considered as the
absence of persistent pain in infants who are unable standard of care in paediatric chronic pain (Walco
to verbalise pain or discomfort and without adequate et al., 2010; World Health Organization 2012). Third,
tools to recognize it? Yet, the challenge of identifying several study features like the duration of the base-
and quantifying ongoing pain is not specific to the line pain intensity assessment period, washout of
youngest as pain intensity scales used in children prohibited medications before inclusion and accep-
and adolescents have been essentially developed to tance or not of concomitant analgesics during the
evaluate acute or procedural pain (Hummel and van trial were rarely reported in the RCTs reviewed. Yet,
Dijk, 2006; Stinson et al., 2006; von Baeyer and Spa- these are important trial features that may impact
grud, 2007; Palermo, 2009). Though, it is recom- acceptance of the trial by patients/families and treat-
mended to use the same scales for chronic pain, ing physicians and consequently influence trial
there is no evidence on their psychometric properties recruitment. Finally, participation of patients in the
in this clinical context (McGrath et al., 2008). RCTs was found to be rather short (median
A variety of psychological therapies have proven to 6 months) for a condition such as chronic pain,
be beneficial for children and adolescents with although research needed more than 2 years to

© 2017 European Pain Federation - EFICâ Eur J Pain  (2017) – 7

Chronic pain trials in paediatrics R. Boulkedid et al.

Table 3 Description of RCTs’ outcome measures

All RCT (n = 58) Pharmacological Non pharmacological

Number (%) treatment n = 11 treatment N = 47

Pain intensity
Simple primary outcome 12 (21) 4 (36) 8 (17)
Complex/composite primary outcome 38 (66) 5 (45) 33 (70)
Secondary outcome 8 (14) 2 (18) 6 (13)
Pain intensity outcome defined as
Qualitative variable 5 (9) 0 5 (11)
Binary based on threshold 1 (20) 1 (20)
Binary based on % of response 3 (60) 3 (60)
Other (pain unchanged/improvement/pain free) 1 (20) 1 (20)
Quantitative variable 52 (90) 10 (91) 42 (89)
Mean/median pain score 22 (42) 1 (10) 21 (50)
Decrease on initial score in term of points or % 29 (56) 9 (90) 20 (48)
Other 1 (2) 0 1 (2)
Not reported 1 (2) 1 (9) 0
Tools used to assess pain intensitya
Simple verbal scale 2 (3) 1 (9) 1 (2)
Visual analogue scale 22 (38) 7 (64) 15 (32)
Numerical rating scale (NRS) 29 (50) 5 (45) 24 (51)
Wong Baker Faces Pain Rating Scale 1 (2) 1 (9) 0
Faces Pain Scale – Revised (FPS-R) 7 (12) 2 (18) 5 (11)
Other 13 (22) 5 (45) 8 (17)
Not reported 5 (9) 3 (27) 2 (4)
Pain intensity assessora
Patient 55 (95) 10 (91) 45 (96)
Medical doctor 1 (2) 0 1 (2)
Nurse 1 (2) 1 (9) 0
Parents 16 (28) 2 (18) 14 (30)
Research assistant 1 (2) 0 1 (2)
Not reported 2 (3) 1 (9) 1 (2)
Other outcome measuresa
QOL 15 (26) 2 (18) 13 (28)
Primary outcome 3 0 3
Secondary outcome 12 2 10
Depression 16 (28) 1 (9) 15 (32)
Primary outcome 2 0 2
Secondary outcome 14 1 13
Anxiety 9 (16) 1 (9) 8 (17)
Pr Primary outcome 2 0 2
Secondary outcome 7 1 6
Disability 12 (21) 0 12 (25)
Primary outcome 7 0 7
Secondary outcome 5 0 5
Safety 7 (12) 4 (36) 3 (6)
Primary outcome 1 0 1
Secondary outcome 6 4 2
Quality of sleep 5 (9) 0 5 (11)
Primary outcome 1 0 1
Secondary outcome 4 0 4
Satisfaction 6 (10) 1 (9) 5 (11)
Primary outcome 0 0 0
Secondary outcome 6 1 5
Other 44 (76) 8 (73) 36 (77)

QOL, Quality of life.

a
The total percentage may exceed 100% because some studies used more than one criterion.

8 Eur J Pain  (2017) – © 2017 European Pain Federation - EFICâ

R. Boulkedid et al. Chronic pain trials in paediatrics

Figure 3 Risk of bias assessment.

complete (median 28.5 months). Long-lasting RCTs developed and applied in adults should be adapted
mainly due to recruitment difficulties tend to and validated in paediatrics, e.g. in neuropathic pain,
increase research costs while losing their scientific the DN4 questionnaire or the quantitative sensory
interest. testing (QST), whose value in clinical practice should
In a significant number of RCTs included in our be further explored (Howard et al., 2014; Mainka
review the risk of bias was unclear probably because et al., 2015). Moreover, pain intensity is only one
authors do not follow guidelines for reporting of dimension of the chronic pain experience (Birnie
RCTs and this is consistent with conclusions from et al., 2012). In our review, assessment of pain
previous reviews in adults (Turner et al., 2012a,b). intensity was the primary outcome for all studies but
Our review intended to explore the potential quality of life or satisfaction with treatment was
research gap in pediatric chronic pain management rarely assessed. Confining evaluation to merely pain
and to discuss the underlying reasons for this gap. In intensity does not accurately reflect anticipated ben-
any case, the methodological weaknesses of the efits in pain-related disability and may potentially
RCTs included and the heterogeneity of interven- impair testing of promising therapies (McGrath et al.,
tions tested prevent from drawing any conclusions 2008; Lynch-Jordan et al., 2014). Finally, interna-
on the effectiveness of the latter. Our review is also tional expert initiatives to define adequate method-
based on published RCTs and may not comprise ologies and study features when performing chronic
negative studies. Therefore, our results may not pain trials in paediatrics are greatly needed. The
completely reflect research efforts to improve man- IMMPACT initiative (Grol et al., 2008) whose mis-
agement of paediatric chronic pain but they shed sion was to develop consensus reviews and recommen-
light into the dearth and challenges of research in dations for improving the conduct of clinical trials of
the field. Although there are many ways to tackle treatments for pain comprised a paediatric component
these challenges, three should be further high- for outcome measures in trials. However, as opposed to
lighted. Properly identifying ongoing painful condi- adults (Dworkin et al., 2010, 2011), IMMPACT did not
tions in children and adolescents is the first step to issue specific recommendations for the design of confir-
adequate treatment. Some diagnostic tools initially matory chronic pain clinical trials in children and

© 2017 European Pain Federation - EFICâ Eur J Pain  (2017) – 9

Chronic pain trials in paediatrics R. Boulkedid et al.

adolescents considering specific methodological chal- Marcello Allegretti, Dr Ornella Bellagamba, Prof.
lenges. Currently, only one US expert group has pro- Franca Benini, Dr Donato Bonifazi, Dr Daniela
posed guidance on how and when to perform RCTs in Caprino, Prof. Adriana Ceci, Dr Sabrina Congedi,
children and adolescents but merely in the context of Dr Francesco Craig, Dr Sandro Dallorso, Dr Antuan
acute pain (Berde et al., 2012). In addition, alternative Divisic, Dr Mariagrazia Felisi, Dr Marco Gentile,
and innovative approaches to clinical trial design such Dr Andrea De Giacomo, Dr Rebecca Lundin, Dr Luca
as the randomised withdrawal or adaptive designs may Manfredini, Dr Laura Mangiarini, Dr Emilia Matera,
represent more feasible and reliable options to perform Prof. Lucia Margari, Dr Alessandro Mazza, Dr Virgilio
clinical research in children (McQuay et al., 2008; Pace, Dr Chiara Di Pede, Dr Maria Giuseppina Pet-
Baiardi et al., 2011; Moore et al., 2013a). International ruzzelli, Dr Pieradelchi Ruffini, Dr Luigina Tagli-
consensus on these methods would certainly urge reg- avacca, Dr Maria Traverso; the Netherlands:
ulatory acceptance and contribute in developing effec- Dr Maarten O. Mensink, Prof. Dick Tibboel, Prof Sas-
tive treatments in children and adolescents with kia N de Wildt, Dr Tjitske van der Zanden; United
chronic pain. Kingdom: Dr Oscar Della Pasqua, Dr Paul Healy.

References
5. Conclusions
Allegaert, K., Tibboel, D., van den Anker, J. (2013). Pharmacological
This is the first review to illustrate the substantial treatment of neonatal pain: In search of a new equipoise. Semin Fetal
research gap regarding analgesic interventions for Neonatal Med 18, 42–47.
children and adolescents with chronic pain. There is von Baeyer, C.L., Spagrud, L.J. (2007). Systematic review of
observational (behavioral) measures of pain for children and
a lack of clinical trials to evaluate pharmacological adolescents aged 3 to 18 years. Pain 127, 140–150.
interventions particularly in infants and in children Baiardi, P., Giaquinto, C., Girotto, S., Manfredi, C., Ceci, A. (2011).
and adolescents with long-lasting diseases. Our Innovative study design for paediatric clinical trials. Eur J Clin
Pharmacol 67(Suppl 1), 109–115.
results underline the difficulties in conducting such Berde, C.B., Walco, G.A., Krane, E.J., Anand, K.J., Aranda, J.V. et al.
trials and point out the absence of methodological (2012). Pediatric analgesic clinical trial designs, measures, and
guidance and implementation of innovative method- extrapolation: Report of an FDA scientific workshop. Pediatrics 129,
354–364.
ologies in this specific field. Birnie, K.A., McGrath, P.J., Chambers, C.T. (2012). When does pain
matter? Acknowledging the subjectivity of clinical significance. Pain
153, 2311–2314.
Acknowledgements Bouhassira, D., Attal, N. (2011). Diagnosis and assessment of
neuropathic pain: The saga of clinical tools. Pain 152, S74–S83.
All the collaborators were involved in the context of the Coffelt, T.A., Bauer, B.D., Carroll, A.E. (2013). Inpatient characteristics
FP7 GAPP Project (GA no. 602962) and under the of the child admitted with chronic pain. Pediatrics 132, e422–e429.
Dworkin, R.H., Turk, D.C., Farrar, J.T., Haythornthwaite, J.A., Jensen,
umbrella of TEDDY – European Network of Excellence for
M.P. et al. (2005). Core outcome measures for chronic pain clinical
Paediatric Clinical Research. trials: IMMPACT recommendations. Pain 113, 9–19.
Dworkin, R.H., Turk, D.C., Peirce-Sandner, S., Baron, R., Bellamy, N.
et al. (2010). Research design considerations for confirmatory chronic
Author contributions pain clinical trials: IMMPACT recommendations. Pain 149, 177–193.
Dworkin, R.H., Turk, D.C., Katz, N.P., Rowbotham, M.C., Peirce-
F.K., C.A and R.B. designed the study. R.B., A.YA., E.D. Sandner, S. et al. (2011). Evidence-based clinical trial design for
and M.M. conducted data analysis. F.K., R.B. and A.YA. chronic pain pharmacotherapy: A blueprint for ACTION. Pain 152,
S107–S115.
interpreted study results. The manuscript was initially
Eccleston, C., Jordan, A.L., Crombez, G. (2006). The impact of chronic
drafted by R.B. and F.K. All authors discussed the results pain on adolescents: A review of previously used measures. J Pediatr
and commented on the manuscript. Psychol 31, 684–697.
Eccleston, C., Palermo, T.M., Fisher, E., Law, E. (2012). Psychological
interventions for parents of children and adolescents with chronic
GAPP Consortium members illness. Cochrane Database Syst Rev CD009660.
Fearon, P., Hotopf, M. (2001). Relation between headache in childhood
Albania: Dr Donjeta Bali, Dr Alketa Hoxha(Qosja), and physical and psychiatric symptoms in adulthood: National birth
cohort study. BMJ 322, 1145.
Prof Ermira Kola; Estonia: Dr Inga Talvik; France: Fitzgerald, M., McKelvey, R. (2016). Nerve injury and neuropathic pain
Dr Juliette Andrieu-Galien, Prof Daniel Annequin, – A question of age. Exp Neurol 275(Pt 2), 296–302.
Dr Romy Blanchet, Prof. Isabelle Desguerre, Dr Elisa- Gregoire, M.C., Finley, G.A. (2013). Drugs for chronic pain in children:
A commentary on clinical practice and the absence of evidence. Pain
beth Fournier-Charri
ere, Dr Barbara Tourniaire, Dr Res Manage 18, 47–50.
Chantal Wood; Germany: Dr Antje Neubert, Prof. Groenewald, C.B., Essner, B.S., Wright, D., Fesinmeyer, M.D., Palermo,
Dr. Regina Trollmann, Dr Stefan Wimmer; Greece: T.M. (2014). The economic costs of chronic pain among a cohort of
treatment-seeking adolescents in the United States. J Pain 15, 925–
Dr Eleana Garini, Dr Panagoula Mammi; Italy: Dr 933.

10 Eur J Pain  (2017) – © 2017 European Pain Federation - EFICâ

R. Boulkedid et al. Chronic pain trials in paediatrics

Grol, R., Berwick, D.M., Wensing, M. (2008). On the trail of quality Palermo, T.M. (2009). Assessment of chronic pain in children: Current
and safety in health care. BMJ 336, 74–76. status and emerging topics. Pain Res Manage 14, 21–26.
Hagen, N.A., Biondo, P., Stiles, C. (2008). Assessment and management Palermo, T.M., Eccleston, C. (2009). Parents of children and adolescents
of breakthrough pain in cancer patients: Current approaches and with chronic pain. Pain 146, 15–17.
emerging research. Curr Pain Headache Rep 12, 241–248. Perquin, C.W., Hazebroek-Kampschreur, A.A., Hunfeld, J.A., Bohnen,
Hechler, T., Kanstrup, M., Holley, A.L., Simons, L.E., Wicksell, R., A.M., van Suijlekom-Smit, L.W., Passchier, J., van der Wouden, J.C.
Hirschfeld, G., Zernikow, B. (2015). Systematic review on intensive (2000). Pain in children and adolescents: A common experience. Pain
interdisciplinary pain treatment of children with chronic pain. 87, 51–58.
Pediatrics 136, 115–127. Petersen, S., Hagglof, B.L., Bergstrom, E.I. (2009). Impaired health-
Higgins, J.P., Altman, D.G., Gotzsche, P.C., Juni, P., Moher, D. et al. related quality of life in children with recurrent pain. Pediatrics 124,
(2011). The Cochrane Collaboration’s tool for assessing risk of bias in e759–e767.
randomised trials. BMJ 343, d5928. Polydefkis, M., Raja, S.N. (2008). What can we learn from failed
Howard, R.F., Wiener, S., Walker, S.M. (2014). Neuropathic pain in neuropathic pain trials? Neurology 70, 250–251.
children. Arch Dis Child 99, 84–89. Reinherz, H.Z., Paradis, A.D., Giaconia, R.M., Stashwick, C.K., Fitzmaurice,
Hummel, P., van Dijk, M. (2006). Pain assessment: Current status and G. (2003). Childhood and adolescent predictors of major depression in
challenges. Semin Fetal Neonatal Med 11, 237–245. the transition to adulthood. Am J Psychiatry 160, 2141–2147.
Hunfeld, J.A., Perquin, C.W., Duivenvoorden, H.J., Hazebroek- Stinson, J.N., Kavanagh, T., Yamada, J., Gill, N., Stevens, B. (2006).
Kampschreur, A.A., Passchier, J., van Suijlekom-Smit, L.W., van der Systematic review of the psychometric properties, interpretability and
Wouden, J.C. (2001). Chronic pain and its impact on quality of life feasibility of self-report pain intensity measures for use in clinical
in adolescents and their families. J Pediatr Psychol 26, 145–153. trials in children and adolescents. Pain 125, 143–157.
King, S., Chambers, C.T., Huguet, A., MacNevin, R.C., McGrath, P.J., Treede, R.D., Jensen, T.S., Campbell, J.N., Cruccu, G., Dostrovsky, J.O.
Parker, L., MacDonald, A.J. (2011). The epidemiology of chronic pain et al. (2008). Neuropathic pain: Redefinition and a grading system
in children and adolescents revisited: A systematic review. Pain 152, for clinical and research purposes. Neurology 70, 1630–1635.
2729–2738. Treede, R.D., Rief, W., Barke, A., Aziz, Q., Bennett, M.I. et al. (2015).
Lynch-Jordan, A.M., Sil, S., Peugh, J., Cunningham, N., Kashikar-Zuck, A classification of chronic pain for ICD-11. Pain 156, 1003–1007.
S., Goldschneider, K.R. (2014). Differential changes in functional Turner, L., Shamseer, L., Altman, D.G., Schulz, K.F., Moher, D.
disability and pain intensity over the course of psychological (2012a). Does use of the CONSORT Statement impact the
treatment for children with chronic pain. Pain 155, 1955–1961. completeness of reporting of randomised controlled trials published in
Mainka, T., Maier, C., Enax-Krumova, E.K. (2015). Neuropathic pain medical journals? A Cochrane review. Syst Rev 1, 60.
assessment: Update on laboratory diagnostic tools. Curr Opin Turner, L., Shamseer, L., Altman, D.G., Weeks, L., Peters, J. et al. (2012b).
Anaesthesiol 28, 537–545. Consolidated standards of reporting trials (CONSORT) and the
McGrath, P.J., Walco, G.A., Turk, D.C., Dworkin, R.H., Brown, M.T. completeness of reporting of randomised controlled trials (RCTs)
et al. (2008). Core outcome domains and measures for pediatric published in medical journals. Cochrane Database Syst Rev 11, MR000030.
acute and chronic/recurrent pain clinical trials: PedIMMPACT Walco, G.A., Dworkin, R.H., Krane, E.J., LeBel, A.A., Treede, R.D.
recommendations. J Pain 9, 771–783. (2010). Neuropathic pain in children: Special considerations. Mayo
McKelvey, R., Berta, T., Old, E., Ji, R.R., Fitzgerald, M. (2015). Clin Proc 85, S33–S41.
Neuropathic pain is constitutively suppressed in early life by anti- Weimer, K., Gulewitsch, M.D., Schlarb, A.A., Schwille-Kiuntke, J.,
inflammatory neuroimmune regulation. J Neurosci 35, 457–466. Klosterhalfen, S., Enck, P. (2013). Placebo effects in children: A
McQuay, H.J., Derry, S., Moore, R.A., Poulain, P., Legout, V. (2008). review. Pediatr Res 74, 96–102.
Enriched enrolment with randomised withdrawal (EERW): Time for World Health Organization (2012). Persisting pain in children package:
a new look at clinical trial design in chronic pain. Pain 135, 217–220. WHO guidelines on the pharmacological treatment of persisting pain
Mercadante, S., Giarratano, A. (2014). Pharmacological management of in children with medical illnesses. Available at: http://whqlibdoc.
cancer pain in children. Crit Rev Oncol Hematol 91, 93–97. who.int/publications/2012/9789241548120_Guidelines.pdf (accessed
Merskey, H., Bogduk, N. (1994). International Association for the Study of 27 July 2017).
Pain. Classification of Chronic Pain: Descriptions of Chronic Pain Syndromes
and Definitions of Pain Terms (Seattle: IASP Press).
Moore, R.A., Gavaghan, D., Tramer, M.R., Collins, S.L., McQuay, H.J. Supporting Information
(1998). Size is everything–large amounts of information are needed
to overcome random effects in estimating direction and magnitude of Additional Supporting Information may be found online in
treatment effects. Pain 78, 209–216.
the supporting information tab for this article:
Moore, R.A., Derry, S., Wiffen, P.J. (2013a). Challenges in design and
interpretation of chronic pain trials. Br J Anaesth 111, 38–45.
Appendix S1. Data extraction form.
Moore, R.A., Straube, S., Aldington, D. (2013b). Pain measures and
cut-offs – ‘no worse than mild pain’ as a simple, universal outcome.
Appendix S2. References of RCTs included in the review.
Anaesthesia 68, 400–412.
Odell, S., Logan, D.E. (2013). Pediatric pain management: The
Appendix S3. Search strategy.
multidisciplinary approach. J Pain Res 6, 785–790.

© 2017 European Pain Federation - EFICâ Eur J Pain  (2017) – 11