A systematic review of the prevalence of chronic postsurgical pain in children

Two prior reviews highlight the scarcity and conflicting nature of available data on chronic postsurgical pain in children, reporting a wide prevalence range of 3.2% to 64% (at ≥3 months). This updated systematic review aimed to consolidate information on the prevalence of pediatric chronic postsurgical pain. A thorough literature search of full English‐text publications from April 2014 to August 2021 was conducted using Ovid MEDLINE, PubMed, and Cochrane Database of Systematic Reviews, with search terms: postoperative pain, child, preschool, pediatrics, adolescent, chronic pain. Seventeen relevant studies were identified. Most assessed chronicity once greater than 3 months duration postoperatively (82%), were predominantly prospective (71%) and conducted in inpatient settings (88%). The surgeries examined included orthopedic (scoliosis and limb), urological, laparotomy, inguinal, and cardiothoracic procedures, involving numbers ranging from 36 to 750, totaling 3137 participants/2792 completers. The studies had wide variations in median age at surgery (6 days to 16 years), the percentage of female participants (unspecified or 12.5% to 90%), and follow‐up duration (2.5 months to 9 years). Various pain, functional, psychosocial, and health‐related quality of life outcomes were documented. Chronic postsurgical pain prevalence varied widely from 2% to 100%. Despite increased data, challenges persist due to heterogeneity in definitions, patient demographics, mixed versus single surgical populations, diverse perioperative analgesic interventions, follow‐up durations and reported outcomes. Interpretation is further complicated by limited information on impact, long‐term analgesia and healthcare utilization, and relatively small sample sizes, hindering the assessment of reported associations. In some cases, preoperative pain and deformity may not have been addressed by surgery and persisting pain postoperatively may then be inappropriately termed chronic postsurgical pain. Larger‐scale, procedure‐specific data to better assess current prevalence, impact, and whether modifiable factors link to negative long‐term outcomes, would be more useful and allow targeted perioperative interventions for at‐risk pediatric surgical patients.


| INTRODUC TI ON
Given the number of surgical interventions performed worldwide in the children (estimated in the United States at 3.9 million annually 1 ), there is a need for an enhanced understanding of chronic postsurgical pain (CPSP) in the pediatric demographic.The potential long-term impact of CPSP on the affected child's future wellbeing, educational attainment, employment prospects, and societal engagement necessitates comprehensive research to address existing knowledge gaps.Drawing from the recognized associations between adverse outcomes in children experiencing chronic pain documented in two North American studies, 2,3 it is hypothesized that pediatric patients in the subset experiencing CPSP may face similar challenges.These have included school impairment, 2 diminished rates of higher education and employment, increased reliance on public aid, and compromised quality of relationships compared with pain-free counterparts. 3ilding on the known variation in prevalence of CPSP in adults from 0% up to 85% (dependent upon surgical procedure 4 ) and two previous systematic reviews of pediatric CPSP (each of 12 studies, overlapping by 3, total respondents 1912 and 1835) where prevalence varied from 3.2% to 64% 5 or 20% at a median at 12 months, 6 it is known that pediatric CPSP prevalence also varies with surgical type.The two previous pediatric systematic reviews emphasize the heterogeneity in definitions and methodological variations which lead to challenges in interpreting the associated risk factors in children.Therefore, our hypothesis posits that a systematic review focused on updating and synthesizing the current understanding of CPSP prevalence in the pediatric population may reveal nuanced insights.Through incorporating recent data, this review aimed to update and contribute to a more accurate and comprehensive understanding of the prevalence, associated risk factors, and potential of the long-term implications of CPSP in children undergoing surgical interventions.

| Types of participants
The inclusion criteria for this systematic review encompassed studies focusing on children and adolescents less than 18 years of age.Participants in the selected studies underwent various surgical procedures under general anesthesia at a hospital or day surgery center, ranging from orthopedic, general surgical to cardiothoracic interventions.Surgery for cancer was included when part of a larger series.Noninvasive or diagnostic procedures were excluded.

| Types of studies
Full-text, peer-reviewed, published English-language manuscripts were considered for inclusion.Studies had to be prospective or cross-sectional cohort studies or randomized controlled trials and specifically address the prevalence of pediatric chronic postsurgical pain (CPSP).Articles were excluded if the study population included adults, lacked an English-language translation, or lacked relevant prevalence data.

| Definition of chronic postsurgical pain (CPSP)
The International Association for the Study of Pain (IASP) 7 and the International Classification of Diseases (ICD-11) 8 chronic pain definitions emphasize a duration of ≥3 months.The ICD-11 defines chronic postsurgical (and posttraumatic) pain as developing after a "surgical procedure (or tissue injury involving any trauma, including burns) and persisting at least 3 months" as a diagnosis of exclusion.
Both definitions incorporate the presence of pain without specifying severity or functional impact.

| Prevalence of CPSP
The primary outcome was the prevalence of CPSP in the pediatric population.The studies selected were required to provide insights into the occurrence and persistence of pain 3 months or more postsurgery.

| Risk factors for CPSP in children
The review also aimed to identify prior or new potential risk factors associated with pediatric CPSP.These risk factors encompass biological, medical, and functional and psychosocial variables that could contribute to the development and persistence of CPSP.

| Search strategy
A systematic literature search, conducted from April 2014 to August 2021, involved comprehensive queries in Ovid MEDLINE, PubMed, and the Cochrane Database of Systematic Reviews.The search terms included "postoperative pain," "child," "preschool," "pediatrics," "adolescent," and "chronic pain."Boolean operators were strategically employed for each database (refer to Appendix S1 for detailed search terms).The selection of April 2014 as the starting point aimed to capture studies published since the 2014 review by Nikolajsen and Brix and align with the 2015 update of the International Classification of Diseases (ICD-11). 8

| Registration
This systematic review was not prospectively registered.A protocol submission to PROSPERO was attempted and declined due to the prioritization of COVID-related articles during that period.

| Study selection
One reviewer (NS) screened the abstracts in the literature search list.The other three authors reviewed the literature search list as a group confirming the articles that should be reviewed as full texts.Two reviewers (NS; GP) then accessed the full-text manuscripts assessing them for eligibility for inclusion in the systematic review.
Only studies providing substantial information on the prevalence of pediatric CPSP were included.Additionally, the two reviewers scrutinized the reference lists of included studies' manuscripts for potential additional papers meeting the inclusion criteria.Disagreements if they arose were discussed with the third and fourth authors.

| Data extraction
Data extraction included key information such as first author name, publication date, country of study, study type, surgery type, patient demographics (age and sex), sample size, follow-up duration, CPSP definition and determination method, pain measures, functional impact, quality of life, psychological measures (child and parent), and additional factors like analgesic and healthcare utilization.The extracted data were categorized according to specific surgical subgroups by two reviewers (NS; GP).

| Risk of bias
The risk of bias of included studies was assessed by two authors (NS; GC) using the Quality in Prognosis Studies (QUIPS tool). 9Where disagreements between the two assessors occurred, a third author (GP) arbitrated.The assessed QUIPS domains included study participation, study attrition, prognostic factor measurement, outcome measurement, study confounding, and statistical analysis and reporting.

| Prevalence of CPSP
We aimed to report the studies prevalence results.Where randomized controlled studies or comparative cohort studies were available, we aimed to report median and interquartile range of prevalence of chronic postsurgical pain between 3 and 12 months across the surgical subgroup.We also aimed to summarize pain frequency, severity, and quality if provided at follow-up.

| Subgroup analyses
Where enough data were available, we aimed to summarize this according to individual surgical cohort (e.g., ≥2 studies).We anticipated that subgroup analysis may be possible for the following surgery types: spinal fusion, laparotomy, pectus surgery, inguinal hernia repair, thoracotomy, and sternotomy as prior reviews included multiple studies for spinal fusion, mixed surgery, laparotomy, and thoracotomy samples.

| Risk factors for CPSP
Due to the lack and heterogeneity of demographic and outcome data, we were unable to perform meta-analysis on parental or child biopsychosocial risk factors.Therefore, we described the individual study findings in terms of baseline functional impact, quality of life, and psychological risk factors assessed as associated with the presence or severity of CPSP (at 3 to 12 months).Previous reviews have assessed impact of female versus male sex, age, preoperative pain severity, and preoperative child and parent anxiety, depression, and pain catastrophizing.We anticipated scar length as an indicator of operation size, postoperative morphine consumption, and early pain presence (e.g., at 1 month) may be relevant.

| Search results-Article identification and selection process
A total of 174 articles were initially identified (NS).After eliminating duplicates and screening for relevance based on titles and abstracts (NS; GP), 36 articles from the search and an additional two articles identified through peer review making 38 total that underwent a comprehensive assessment (NS; GP).The full PDF articles were accessed from the University of Melbourne Library database and the Australian and New Zealand College of Anaesthetists' Library's journal subscriptions.
An institutional librarian was not involved.Following this, 21 articles were excluded; Table S1 documents the reasons for exclusion (NS; GP).
The selection process is illustrated in Figure 1, adhering to PRISMA guidelines.There were no discrepancies in terms of included versus excluded papers.Two studies had an earlier definition of chronic pain in a spinal fusion population.They were included as they added to the data available for this surgical subpopulation.Of the two further studies suggested at the peer review stage; one was included and the second was excluded.[12][13][14][15][16][17][18][19][20][21][22][23][24][25][26]

| Study characteristics
Most studies 14/17 (82%) adhered to the International Association for the Study of Pain (IASP) definition requiring pain persistence for ≥3 months.Two included studies 21,22 had earlier patient follow-up specified as "2-3 months."A further study 16 employed a 2-month definition but was included as the participants were approached years postoperatively.Sample sizes ranged from 36 to 750, totaling 3137, with 2792 participants completing follow-up.Prospective studies constituted 71%, while cross-sectional cohort studies represented 29%.Notably, no randomized controlled trials were identified.Risk of bias of the included studies is summarized in Table S2, wherein only three of the 17 studies were judged at low risk of bias, 21,22,26 and three further were low risk in all but one category (two not describing attrition 18,19 and the third not describing confounders 14 ), one was moderately biased in two categories, 20 and the remainder were moderately to highly biased in ≥2 categories with some of these studies failing to assess prognostic factors, 16 study confounders 12 or perform 12,16,17 or adequately report data or statistical analysis. 10,11,13,15,24,253 | Surgical populations

| Patient demographics and setting
Among the 17 studies, the majority (88%) were conducted in an inpatient setting, with one study being ambulatory 11 and another unspecified. 12Female sex ratios varied from 12.5% to 90% or were unspecified.The mean patient age at the time of surgery ranged from 6 days to 16 years, while follow-up durations mean/medians spanned from 2.5 months to 9 years.

| Diverse designs and measures
Questionnaires employed in the studies displayed diverse designs, incorporating various measures (see below and text in Tables 1 and   2) administered preoperatively, the day of surgery, and at different postoperative time points of 0-72 h, 2 weeks, and 1, 2, 3, 4, 6, 10-12, and 12 months (see Table 3 for summary of timing of specific measures).

| Pain assessment tools
Pain assessment tools ranged from simple yes-no responses 12 to ageappropriate use of unidimensional scales like the Visual Analogue Scale (VAS), 10,18 Numeric Rating Scale (NRS), 16,19,21,26 17 One study used the Wong-Baker Scale for all age groups pain report. 15Additionally, two studies used multidimensional measures including the Parent's Postoperative Pain Measure (PPPM) 11 or a Pain Management Index (pain scores combined with analgesic use 17 ).Some studies enquired about pain as present, or as a percentage or Likert scale over a period before questionnaire administration, commonly in the 1-4 weeks prior, but others up to 3 months prior.Some studies specified "any pain" as reportable, while others incorporated severity (e.g., moderate with VAS or NRS ≥3 or ≥4/10) or physical or functional impact as part of the definition (Table 1).
Pain quality was described either with simple descriptors such as burning, 17 tingling or shooting, 14,15,21 or questionnaires including Self-report of the Leeds Assessment of Neuropathic Symptoms and Signs (S-LANSS), 16 painDETECT™ 24 validated in adolescents, or Douleur Neuropathique 4 (DN4) validated in adults. 10,206 | Prevalence of pediatric chronic postsurgical pain (CPSP)

| Overall
The prevalence of CPSP in mostly inpatient pediatric populations, as reported in the 17 selected studies, displayed a wide range from 2% to 100%.Table 1 summarizes the definitions used and prevalence rates including totals according to surgical subtype.There is a wide variation in the surgical populations, duration of follow-up, ages of the patients at surgery and when interviewed, study type and design, and study quality or bias, with sample size and recall being of issue, as well as descriptive presentation lacking in detail of data presentation, attrition or statistical analysis.

| Prevalence within surgical cohorts
1][22][23][24][25] The mixed orthopedic surgery (two studies with similarly high scoliosis representations, n = 120 vs. pectus surgery, n = 40) also in teenagers by the same author group have differing prevalence of 20% at 4 months 18 and 72% at 12 months. 19e mixed surgery group (three studies with different compositions of orthopedic, n = 413 vs general surgery, n = 397; day-case 11 vs. inpatient 10,26 ) in younger children 11 vs. teenagers 10,26 had CPSP prevalences of 2%, 10 4% 11 and 36% at 6 months, 26 without subgroup information.Teenagers post infant laparotomy (n = 750) had CPSP prevalence of 24% several years later 13 and of teenagers (n = 158) post open appendicectomy, 18% had CPSP 6 months later. 14The prevalence following inguinal hernia repair was low in mostly male teenagers (1 study n = 210) years later at 7%, 12 while in a subgroup of younger children (n = 82 of 327) was slightly higher at 9.2% at 6 months over other surgery types in a mixed general surgery group. 11 The CPSP prevalence following thoracotomy in younger children more than 1 year postoperatively was less than 5%. 16,17Several years after cardiac surgery, mostly via sternotomy, CPSP was moderate at 17%. 15 3.6.*Where multiple pain measurement tools were used for intensity, the authors provided appropriate detail of validity and employed these in the appropriate patient age group.

TA B L E 3 (Continued)
1 to 4 weeks, 10,19 and one the week prior using the painDETECT questionnaire, 24 but none reported pain frequency data.One study requested diary completion using frequency >50% as part of CPSP definition. 18Three studies incorporated severity in their definition of CPSP: >3/10 following mixed surgery 11 and spinal fusion. 21,23ree studies transformed VAS scores to subcategories of mild, moderate, and severe. 10,11,26Eight studies focused on pain quality either through neuropathic descriptor enquiry 14,17,21 or through questionnaires reported neuropathic 10,15,20 versus nociceptive predominance. 16,24llowing mixed surgery, two studies report different patient pain trajectories over 12 months.The first incorporated a 2-week postdischarge profile to subsequent recovery defining early and late recovery trajectories. 19The second included preoperative pain report with mild versus moderate pain (and separate unpleasantness) trajectories. 26The moderate pain trajectory included 48% of patients with mean pain scores of 3.4 to 4.3/10 and moderate unpleasantness trajectory included 28% of patients with mean scores of 5.6 to 5.8/10 but overlap was not described. 26One spinal fusion study reported four acute postoperative pain trajectories (one mild and three moderate). 25

Patient age and sex
Two mixed surgical studies with low prevalence both identified association with older age in the univariate, but not multivariate analysis (general 10 and orthopedic 18 ) and three mixed studies did not (general day-case, 11 general, 26 and orthopedic 19 ).Three studies identified association with female sex, following mixed surgery for pain intensity (but not unpleasantness), 26 postlaparotomy, 13 and postappendicectomy, 14 while a mixed orthopedic 19 and spinal fusion study 20 found no association with sex.

Baseline analgesic use
In one spinal fusion study, baseline analgesic use was associated with moderate trajectories over mild acute pain trajectory. 25

Surgery type, duration, and scar length
A mixed surgical study with low prevalence identified associations with thoracic and orthopedic over other surgery subtypes. 10rgical duration was associated with CPSP following long segment posterior spinal fusion in only one 21 of three studies. 20,24nger scar length was not associated with CPSP following spinal fusion 24 nor following mixed surgeries, 10 but was following open appendicectomy. 14

Pain
Preoperative pain presence, intensity, and occurrence at rest.One mixed surgical study identified association of CPSP with moderate and severe recent preoperative pain intensity, after adjustment for regional anesthesia and early postoperative pain within 0-24 h. 10 Four spinal fusion studies identified preoperative pain presence, 21,22 neuropathic in type, 20 and severity and presence of back pain at rest 25 as associated with CPSP.Two further mixed orthopedic studies found no association with preoperative pain. 18,19rly postoperative pain presence and intensity.One spinal fusion study identified moderate over mild acute postoperative pain trajectory as associated with CPSP severity at 6 months. 25A second spinal fusion study found genetic association with both immediate postoperative pain and CPSP. 22The mixed day-case study identified postoperative pain presence at 1 month as associated with CPSP presence at 6 months. 11eoperative or early postoperative pain quality.No studies assessed the association of CPSP with preoperative or early postoperative pain quality, that is, with neuropathic versus non-neuropathic features.

Postoperative morphine use
Three spinal fusion studies reported early postoperative morphine use Day 1 but not beyond 20 and Day 1 and Day 2 21,22 as associated with CPSP, but in a mixed surgical population (with no subgroup analysis of the spinal fusions within this) cumulative 5-day opioid consumption was not.Pain Inventory Short Form Interference Scale, 16 and Child Activity Limitations Interview (CALI21). 11,19eep assessment involved open-ended questions 10,11,21 or standardized scales including CALI21, 11,19 the Adolescent Sleep-Wake Scale and Insomnia Severity Index.18 Mood was evaluated through various scales including the Centre for Epidemiological Study Depression Scale-Children, 26 the Revised Child Anxiety and Depression Scale, 18 the Child Pain Anxiety Symptom Scale and Multidimensional Anxiety Scale, 26 the Child Anxiety Sensitivity Index, 21,23,26 or VAS scale report of anxiety. 10 Pain atastrophizing scales (child or parent report) 19,26 and pediatric health-related quality of life (QoL) scores, specifically Scoliosis Research Society-30 (SRS-30) quality of life, 25 Pediatric Quality of Life Inventory (PedsQL™) subscales or totals, were also employed.14,15,18,19 Of note, the PedsQL™ measures patient's or carer's perceptions of impact on the patient's physical, psychological, and social functioning. A higer score reflects better quality of life, with normal >80/100. 27e functional impact of CPSP is summarized in Table 1, while Table 2 details the explored associations of CPSP with mood and QoL and Table 3 provides an overview of the time points and different measures used, demonstrating the heterogeneity.

Spinal fusion only studies
Of the six spinal fusion studies, three described the presence of functional disability.One determined four postoperative trajectories where a small percent of patients were physically impacted at 6 months (5% in the mild pain group increasing to 9%-12% in the moderate pain groups). 25One study included sleep and daily impact in the definition of chronic pain. 21One reported measuring functional disability without reporting the data. 22The other three did not measure physical status.
In terms of psychological associations with CPSP, this was measured in three of six studies.One found positive association of the moderate trajectories with baseline child catastrophizing status and number of school days missed, 25 one found positive association with early postoperative measure of child anxiety (CASI), 21 and one did not. 22The extended study 23 explored a shared pathway in the epigenetic regulation of CPSP and anxiety.

Mixed orthopedic studies
The mixed orthopedic studies (mainly scoliosis but including pectus surgery), by the same author group, both documented the PedsQL™ and different psychological measures (Table 3).One determined statistical association with activity limitation (CALI21) and lower global, physical, and psychological QoL in the late versus early recovery group. 19The second included impaired global QoL in the definition of CPSP, provided severity data and determined positive association between depression and poorer sleep quality. 18Parent pain catastrophizing was statistically associated with late recovery in the earlier 19 but not later study. 18Child pain catastrophizing was not associated with CPSP in either mixed study including spinal surgery (in contrast to the spinal fusion study above).

Mixed surgery studies
Of the two mixed surgery inpatient studies, one assessed functional, sleep, and social outcomes but did not provide the data, 10 while the second well-designed study specifically assessed functional disability, pain interference, and multiple psychological variables. 26ltiple statistical associations were found for CPSP of minimal clinical import as the odds ratios and confidence intervals were only just over 1.0.
The mixed day-case study reported functional impact on sleep and daily activity as affecting 11 participants without providing data specifying duration, surgical subtype, or reporting assessment of association. 11

Single surgery subtypes
The postinfant laparotomy study did not report functional outcomes in the surveyed adults. 13Neither did the postinguinal hernia study. 12th focused on surgical outcomes.
The postopen appendicectomy study reported lower PedsQL™ scores with physical, emotional.and school but not social impact in association with CPSP. 14e first postthoracotomy study conducted several years following lateral thoracotomies reported minimal pain interference in one patient who experienced late-onset pain. 16The second reported no functional outcomes. 17 The study >1 year postcongenital heart surgery, mostly through sternotomy, found statistical association with lower total PedsQL™ scores, but clinically, the median values were similarly high normal, with difference only in the physical domain. 15nalgesic use varied with percentages ranging from 0% to 38.3% for simple analgesia, 10,[16][17][18]21,25 0% to 17% for opioids, 21,25 and 1.4% for an unspecified analgesic type, 15 with one study documenting up to 3 analgesic classes but stating no anti-neuropathic medication use. 10 Pain specialist consultation was reported in three studies as recommended for 1.9% overall 10 or sought for 4.3% overall 15 versus for 22% of CPSP patients.21

| Detailed analyses for specific surgical cohorts
Detailed analyses were not conducted for specific surgical subpopulations as the data as evident above was too heterogeneous (See also Table 3).

| DISCUSS ION
In this updated review on pediatric chronic postsurgical pain (CPSP),

| Definitions of CPSP and heterogeneity of CPSP studies
The definitions of CPSP varied across studies, encompassing pain of "any" to "frequent" occurrence, with some definitions specify-

| Comparison with previous reviews of pediatric CPSP
Comparing this review's findings with previous reviews 5,6 proved challenging due to similar heterogeneity in designs of the new studies.The 2017 review reported a median prevalence of 20% at 12 months, drawn from scoliosis and mixed surgery studies with differing pain definitions. 6The 2014 review reported similar prevalence for mixed surgery and higher prevalence for scoliosis studies at 10 months or greater. 5The current review indicated high prevalence also in scoliosis cohorts but lower prevalence in mixed cohorts, while prevalences for single specific surgical subtypes, for example, thoracotomy, sternotomy, and inguinal hernia repair, were similar to those summarized previously. 5

| Functional impact and quality of life in pediatric CPSP versus pain clinic cohorts
Functional impact was variably reported in the 2014 review, 5 with two studies (17%) indicating negative impact and five others (42%) specifying "little or no" impact.Pain interference was measured by only one quarter of the new studies.Functional impact and disability were measured by several studies but frequently the data was not provided.One well-designed study focused on independent analysis of associations at 12 months for each of pain intensity and unpleasantness trajectories and functional disability (without providing the data delineating severity). 26A comparison cannot then be made with pediatric chronic pain clinic cohorts where moderate-to-severe functional impact is reported (FDI means of 26-27/60 SD 12). 29alth-related QoL was measured by some included studies but was not the primary focus of the 2017 review. 6While in other new studies, reported CPSP impact varied from none (stated significant but this was statistical only 15 ) to mild-to-moderate impact in PedsQL™ subcategories means of 64-75/100.Sleep impact suggested as an optional core outcome 30 has been assessed in a few of the new studies 10,11,16,21 with only one providing an indicator of severity and establishing an association with CPSP.
Social functioning has been explored in several of the newer studies with a few studies enquiring about this with simple yes-no response regarding impact upon play or as three to five items in a specific measurement tool.Only two explored and found no impairment in the social domain of the PedsQL™ in patients with CPSP. 14,15ychological factors have been previously identified as contributors to pain development and perpetuation. 5,6,26,28,30The timing of the questionnaires specific to psychological measures has varied with completion preoperatively, 18 on the day of surgery (acknowledgeably a period where anxiety may be heightened), early postoperatively, 26 and later, for example, at 6 months. 5In past and current studies, only association with and not the severity of psychological impairment has been considered.The newer CPSP studies have found conflicting results for child anxiety in small scoliosis 21,22 and larger mixed studies, 18 statistical significance for adolescent depression in mixed surgical, 26 and possibly clinical significance in mixed orthopedic populations, 18 with conflicting results for child 18,19 and parent catastrophizing traits. 6,19,26No studies to date have explored the psychological factor of patient and parent satisfaction although this is an outcome in the SRS-30 questionnaire. 25Satisfaction with the cosmetic result of pectus and scoliosis correction can positively influence pain, while dissatisfaction can exacerbate the acute and CPSP experience, for example, with orthopedic or general surgery where inadvertent nerve stretch or injury and new neuropathic pain has occurred.The interplay of identified biological, medical, functional, social, and psychological factors continues to remain complex.

| Pediatric versus adult CPSP
Differences between pediatric and adult CPSP were again evident, with lower prevalence and severity in pediatric surgical subgroups.
Despite one pediatric study identifying thoracotomy as highly associated with CPSP, 10 one previous review reports CPSP prevalence postthoracotomy of 16% 5 while two new thoracotomy studies report very low prevalence of 2% and 5% with low severity and pain interference 16 and requirement only for paracetamol 17 compared with adult prevalence of 5%-65%, severe for 10%. 4 The pediatric inguinal hernia repair data is noteworthy with similar prevalence across the three prior 5 and two newer studies months to years later of 3.2%-13.5% 11,12compared with adult prevalence of 5%-63%, severe for 6%-29%.

| Limitations
There are many limitations of the new studies and hence the review, including the variable risk of bias of the various studies with influences of small sample size (particularly in those studies with mixed surgical populations where subgroup numbers are even smaller), recall bias, the questionnaires' chosen outcomes and design and timing of questionnaire administration and the type of analysis performed.
Some studies report significant findings, but this is only statistical with minor clinical importance as the odds ratios and 95% confidence As highlighted, most studies lack any detail of functional impact of CPSP.Some studies may be well designed with low risk of bias, but their differences further negate the capacity to interpret results across studies.Interpretation is also impaired when grouping multiple surgery types with likely differences in CPSP risk, for example, laparotomy with laparoscopy or posterior spinal fusion with pectus surgery, or where studies have assessed a single surgery, but each have a different focus, for example, neuropathic or nociceptive pain versus establishing associations.A further example of the heterogeneity is exploring association for CPSP following spinal fusion with higher POD1 morphine use at 6-12 months versus early acute with later postoperative pain at an earlier time period of 2-3 months.Although recent studies have become increasingly comprehensive and their

K
E Y W O R D S child, chronic pain, pain, postoperative, prevalence, surgery | 703 SIM et al.

3. 7 . 3 |
Functional, social, and psychological factors-Physical, sleep, mood, and quality of life impact The studies explored various factors potentially associated with CPSP including impairments or interference with physical function, school, sleep, eating, or play, utilizing yes-no questions 10 or specific outcome measures.Function was assessed using various instruments including the Functional Disability Index 21,22 alone or with the Tampa scale for Kinesiophobia 26 and measures of interference including open-ended questions 10 or PROMIS® Pediatric Pain Interference Scale, 26 Brief 17 additional studies were identified since the 2014 review by Nikolajsen and Brix.5 Notably, only one of the five studies published between 2014 and 2017 was included in the 2017 review by Rabbitts, Fisher et al.,6 while the remaining 12 studies were published subsequently.The prevalence of CPSP in pediatric populations exhibited a wide range from 2% to 100%, spanning various surgical categories such as orthopedic, mixed, general surgical, and cardiothoracic surgeries.Spinal fusion surgery for scoliosis demonstrated the highest prevalence among the surgical subtypes.The studies displayed significant heterogeneity concerning design, definitions, patient demographics, sample sizes, and follow-up durations.

4
Importantly, description of whether this is postelective surgery, or an emergency operation postbowel incarceration is relevant, and, although sharing the same title, the surgical operation differs completely where children receive herniorrhaphy and repair and adults hernioplasty with mesh insertion.The postoperative recovery trajectories also frequently differ, for example, for inguinal hernia repair, young children have mild initial postoperative pain with return to full functioning by POD434 versus adults' 6 weeks of restrictions regarding weightlifting, exercise, and delayed return to full work duties.Data regarding CPSP's functional impact is similarly poorly documented for adults.Differences between CPSP in adult and children should be considered in the context of severity, interference, and objectively measured functional outcomes.30

intervals approximate 1 . 0 .
Most have assessed associations of continuous variables, rather than assessing for clinically meaningful results without defining severity of the functional or psychological impact nor looking for a clinically significant change.Studies pose challenges with recall bias when they, although prospective, enquire about pain in the prior month or more or are retrospective and report that in the distant past.Questionnaire design is highly pertinent and whether questions are open versus closed, include measures of severity, validated tools, and outcomes of patient, family and clinical importance.

Country Surgical procedure types Design Participants (% female [F]) Mean or median age at surgery (range IQR or SD) Definition of CPSP & duration of follow-up Prevalence of CPSP (%) and functional impact
3 | Frequency, severity, quality of CPSP, and trajectoriesMost studies reported pain scores.Four studies collected frequency information, three enquired about frequency in the prior TA B L E 1 Studies included in the systematic review of chronic postsurgical pain (CPSP).Authors (year)in Trajectory 1, 5% tolerated no activity;in Trajectories 2-4 this was ~double at 9%-12% Summary 518 participants,

TABLE 1 (
Continued) TA B L E 2 Analgesic use, pain interventions, and additional outcomes of included CPSP studies.Authors (
Abbreviations: ASWS-SF, Adolescent Sleep-Wake Scale-Short Form; BPI-SF, Brief Pain Inventory Short Form; CALI, Child Activity Limitations Index (21); CASI, Child Anxiety Sensitivity Index; CPSP, chronic postsurgical pain; DN4, Douleur Neuropathique 4; FLACC, Faces Legs Activity Crying Consolability Pain Scale; IHR, inguinal hernia repair; NRS, Numeric Rating Scale; OR, odds ratio; PPPM, Parent's Postoperative Pain Measure; PedsQL, Pediatric Quality of Life Inventory™; POD, postoperative day; R-CADS, Revised Child Anxiety and Depression Scale; SBO, small bowel obstruction; VAS or VAS-C/P, Visual Analogue Scale Child or Parent/10.a Typographic error in original article-no email response from enquiry to first author.TA B L E 2 (Continued) TA B L E 3 Time points of outcome measurements and various measures used in CPSP studies demonstrating heterogeneity.

Table 2
also summarizes analgesic use and pain interventions.Only seven studies detailed analgesic use and pain specialist consultation.
moderate severity.Many of the studies were prospective, al-