The efficacy and safety of negative pressure wound therapy in paediatric burns: a systematic review and meta-analysis of randomized controlled trials Authors: Jiaqi Lou, Xiaoyu Zhu, Ziyi Xiang, Youfen Fan, Jingyao Song, Neng Huang, Jiliang Li, Guoying Jin, Shengyong Cui Published: 2024 ABSTRACT BackgroundAlthough the effective application of negative pressure wound therapy (NPWT) has been exemplified in diverse clinical studies, its potential and safety, specifically regarding paediatric burns, are yet to be fully confirmed. Our most recent systematic review and meta-analysis strive to investigate the impact of NPWT within the realm of paediatric burns.MethodsWe sourced relevant articles from databases including PubMed, Embase, the Cochrane Database, Web of Science, the International Clinical Trials Registry Platform, China National Knowledge Infrastructureris, the VIP Database for Chinese Technical Periodicals, and the Wanfang database. We defined the primary outcome measure as the healing time, while healing rate, numbers of dressing changes, detection rate of positive bacteria, incidence rate of adverse reactions, scar scale scores, and treatment costs were considered as secondary outcome measures. Pooling of data was conducted and the results were articulated as relative risk (RR), mean difference (MD), and standardized mean difference (SMD), all with a 95% confidence interval (CI).ResultsIn this systematic review and meta-analysis, a total of 12 studies involving 1033 individuals were examined, including 559 paediatric burn patients who underwent NPWT (referred to as the treatment group) and 543 patients who received treatments other than NPWT (referred to as the control group). The amalgamated data from these studies exhibited that the treatment group experienced significant reductions in healing time (SMD = -1.60; 95% CI: -2.26 - -0.95; p < 0.001, I2 = 92.8%), the number of required dressing changes (SMD = -4.6; 95% CI: -5.84 - -3.36; p < 0.001, I2 = 92.4%), positive bacteria detection rate (RR = 0.61; 95% CI: 0.26–1.46; p = 0.004, I2 = 81.8%), incidence of adverse reactions (RR = 0.61; 95% CI: 0.33–1.12; p = 0.005, I2 = 68%), scar scale scores (SMD = -1.66; 95% CI: -2.54 - -0.79; p < 0.001, I2 = 89.4%), as well as in treatment costs (SMD = 0.92; 95% CI: -1.66–3.49; p < 0.001, I2 = 98.4%). Additionally, these individuals showed an increased rate of healing (RR = 1.17; 95% CI: 0.99–1.39; p < 0.001, I2 = 78%). Subgroup analysis did not find that the degree of burn was one of the sources of high heterogeneity.ConclusionOur meta-analysis points to the effectiveness of NPWT in treating paediatric burns. Notably, it significantly mitigates healing duration, frequency of dressing alterations, positive bacterial detection rate, adverse reactions incidence, scar scale scores and treatment costs, all while propelling the acceleration of wound healing.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12887-024-05302-z. Background Children are particularly susceptible to external factors that can cause injury and are prone to various accidental injuries, with burns being more prevalent. The thin and delicate nature of children’s skin, coupled with their limited appendages [1], makes them more vulnerable to severe damage from the same injurious factors compared to adults. Furthermore, due to the incomplete development of their organs and immune systems [2], children are at a higher risk of infection and multi-organ failure, which can be life-threatening and exacerbate wound depth while promoting scar formation [3]. This not only impacts appearance and functionality but also has long-term consequences in terms of disfigurement and disability for children; hence, greater attention should be given to the treatment of childhood burns. The depth of children burns often require surgical treatment. Studies [4] have reported that vacuum sealing drainage technology used in the treatment of granulation tissue wounds in children can reduce the pain of children during dressing change, reduce the incidence of sepsis, and shorten the course of disease. In several clinical trials [5, 6], No complications related to NPWT equipment have had been identified. However, another study [7] reported that there were some complications in the application of NPWT in pediatric burns, including bleeding, local infections, and mechanical device issues. In addition, The additional hypothesised caused by NPWT treatment and its subsequent treatment also seem to test the feasibility, acceptability and safety of this approach [8], multiple contextual factors affect negative pressure wound therapy uptake in acute pediatric burn Settings [9]. Currently lack more evidence to determine negative pressure wound therapy in paediatric burns the efficacy and safety, and, evidence available about the method is scarce and low quality [10]. Therefore, this study through the existing literature database search reivew and meta-analysis to clear NPWT therapeutic effect on treating paediatric burns, provide reference for the clinical treatment and care. Methods In this systematic review, we adhered faithfully to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) Statement [11] for reporting our research protocol, outcomes, and other pertinent items. Additionally, the protocol for this meta-analysis has been officially registered with PROSPERO under the identifier CRD42024548748. Data source and search strategy We conducted a comprehensive search for pertinent articles across various databases, including PubMed, Embase, Cochrane Database, Web of Science, International Clinical Trials Registry Platform, China National Knowledge Infrastructureris, VIP Database for Chinese Technical Periodicals, and the Wanfang database. The search strategy incorporated specific subject headings and keywords, such as ‘Burns’, ‘Negative Pressure Wound Therapy’, ‘Topical Negative-Pressure Therapy’, ‘Negative-Pressure Dressings’, ‘Vacuum-Assisted Closure’, ‘vacuum sealing drainage’, ‘VSD’, ‘negative pressure’, ‘vacuum assisted closure’, ‘VAC’, ‘Pediatrics’, ‘pediatric’, ‘child’, and ‘Children’, as well as their respective synonyms. References from the initially included studies and relevant review articles were also screened as additional potential sources of eligible clinical trials not identified during the primary database search. All located articles were imported, securely stored, and managed using EndNote 20. Each search outcome underwent an independent eligibility review by three authors, JQ L, XY Z and ZY X, with any emerging disagreement resolved by the corresponding author, SY C. Eligibility criteria Criteria for Inclusion: (1) Study Type: Prospective Randomized Controlled Trials (RCTs), including before-after randomized controlled trials, were considered; (2) Participants: Focus was on paediatric patients with burns; (3) Intervention: Comparisons made between NPWT application alone or alongside supplementary treatments, and treatments excluding NPWT; and (4) Outcomes: Details on healing duration, healing speed, frequency of dressing changes, detection rate of positive bacteria, incidence of adverse reactions, scar scale scores, and treatment expenses were required. Criteria for Exclusion: (1) Studies that did not expressly state their inclusion criteria; (2) Studies that did not clearly mention their outcome parameters; (3) Studies that were uncontrolled or retrospectively randomized trials; and (4) Preclinical studies conducted using animal models. In cases where multiple articles reported the same or overlapping data sets, primary attention was given to the article that documented a longer intervention duration or provided a larger sample size. Data extraction and quality assessment Two co-authors, JQ L and XY Z, independently evaluated articles based on their screening of titles and abstracts. Subsequently, full-text articles were examined for their eligibility and completeness of information. For every study meeting the eligibility criteria, we extracted information pertaining to: the source of the study (including author names, year of publication, journal, and country), characteristics of the study population (such as burn status, sample size, study design, type of subjects, gender ratio, average initial age, duration of follow-up, etc.), interventions, and outcomes. Essential outcomes included at least one of the following: healing time, rate of healing, frequency of dressing changes, rate of positive bacteria detection, incidence of adverse reactions, scar scale scores, and cost of treatment. All data extraction was standardised and executed using a template and Microsoft Excel (Version: 16.78.3). Each RCT included in the review underwent a methodological quality assessment utilizing the Cochrane Collaboration’s risk of bias tool, as outlined in version 5.0.1 of the Cochrane Handbook for Systematic Reviews of Interventions [12]. According to the handbook’s criteria, we classified the risk of bias for each item as either low, high, or unclear. Both the task of data extraction and quality assessment were independently handled by two authors, JQ L and ZY X. In cases of discrepancy, a resolution was sought through discussion and agreement. Whenever a consensus could not be reached, an arbitrator (YF F) intervened. The data synthesis was then carried out using Review Manager (RevMan) 5.4 by another author, JY S. Statistical analysis Data extraction was independently carried out by two authors (JQ L and JL L) from complete texts of the studies and consolidated into collective sheets. Extracted details included the first author, year of publication, intervention, and specific outcomes. Data validation was performed by a different author (N H) following a standardized process. The assessment of methodological quality, showcased in Fig. 1, was aligned with the Cochrane Reviewers’ Handbook [13] guidelines. For discrete numerical variables, we present risk ratios (RR) and 95% confidence intervals (CIs). Continuous variables, on the other hand, appear as Standardized Mean Differences (SMD). SMD was used as a summary statistic in our meta-analysis due to the employment of diverse assessment methods for identical outcomes [14]. Fig. 1PRISMA diagram detailing the literature search and the study selection/exclusion process. PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-analyses; RCT: randomized controlled trials PRISMA diagram detailing the literature search and the study selection/exclusion process. PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-analyses; RCT: randomized controlled trials Heterogeneity was quantified using the I2 statistic, and forest plots were generated and double-checked by three authors (JQ L, GY J, and SY C). We deemed pooled outcomes with I2 < 50% as having low statistical heterogeneity, while those with I2 > 50% were considered to have high statistical heterogeneity [15]. Analyses characterized by low heterogeneity employed a fixed-effects model; those showing high heterogeneity used a random-effects model. We scrutinized potential sources of heterogeneity utilizing techniques like funnel plots, Egger’s regression test, trim and fill method, and sensitivity analysis. The trim and fill technique, a nonparametric (rank-based) data augmentation approach, while Egger’s regression test offers a linear regression approach for parametric data, given that it relates the intervention effect estimates to their standard errors - weighted by their inverse variance. If the p-value was less than 0.05, we deemed the results statistically significant [16]. All statistical analyses were executed utilizing Stata 15.0. A p-value of less than 0.05 was deemed to signify statistical relevance for all tests, except for heterogeneity test, where a p-value threshold of less than 0.10 was utilized. Subgroup analysis We also performed subgroup analyses according to the degree of pediatric burns. Children with %TBSA < 5% were considered as mild burns, and children with %TBSA ≥ 5% were considered as moderate or severe burns. Results Characteristics of included studies Figure 1 illustrates the PRISMA statement flowchart outlining the study selection process. The initial search yielded 606 records, which were reduced to 310 after software filtration removed 296 duplicates. These 310 articles underwent manual screening to ensure alignment with the inclusion and exclusion criteria, which the 171 of them did not meet. Additionally, we were unable to procure the full texts of 5 articles. Consequently, we diligently reviewed the remaining 134 records. Our exhaustive review resulted in the selection of 12 studies [17–28] that were consistent with our criteria, and thus, were incorporated into the current systematic review and meta-analysis. Table 1 showcases a summary of the characteristics of the 12 studies. Altogether, we grouped the 1102 patients as follows: 559 in the NPWT group and 543 in the control group. Interestingly, we noted that 49 patients participated in within-subject controlled trials, serving in both experimental and control capacities. Table 1 Detailed information of the included studiesStudy(author/year)Journal CountryPatientsDesignProcessScalesTreatmentNumber of patientsMale (n)Female (n)Mean age (sd, Year)%TBSAOutcomesShen 2013Chinese Journal of BurnsChinaChildren aged 1-7 years with burnsProspective randomized controlled trialWithin 72h after injury treatment begins, the end of the 9 d treatment, don't change dressings, suction cover after debridement, maintaining negative pressure - 16 kpa/NPWT14581643.3±1.210±3(1)(2)(3)(6)(7)Conventional therapy16285773.1±1.810±4Yao 2013-1Infection, Inflammation, RepairChinaBurn children aged 8 months to 5 years oldProspective randomized controlled trialEscharectomy was performed under general anesthesia within 24 hours post injury, and VSD was covered after escharectomy without dressing changeVSSNPWT20//2.6±0.410.3±1.0(1)(4)(5)(6)The wound was covered with self-made ordinary petroleum petroleum gauze after dermabrasion within 24 hours after injury. The outer dressing was replaced on the second day after surgery, and the soggy outer dressing was replaced in timeVaseline gauze dressing change treatment20//2.5±0.49.8±1.3Yao 2013-2Infection, Inflammation, RepairChinaBurn children aged 8 months to 5 years oldProspective randomized controlled trialEscharectomy was performed under general anesthesia within 24 hours post injury, and VSD was covered after escharectomy without dressing changeVSSNPWT20//2.6±0.410.3±1.0(1)(4)(5)(6)The eschar was dermabrasion under general anesthesia within 24 hours post injury. After dermabrasion, the xenogenic acellular dermal matrix was covered, and dressing was changed regularlyXenogenic acellular dermal matrix20//2.6±0.510.1±1.2Bai 2016Journal of Traumatic SurgeryChinaChildren with burnsProspective randomized controlled trialAfter debridement, VSD was performed directly. The negative pressure was -53 to -26kPa, and the sponge and semi-permeable film were replaced after 6-7 days of continuous suction until the patient's wound healedVSSNPWT4024165.4 ±2.412.4±4.0(1)(2)(3)(4)(5)For the treatment of burns, the oil gauze was covered after debridement, and the sterile gauze dressing was bandaged under pressure, and the dressing was changed in time until the wound healedconventional therapy3520155.8 ±2.511.7±4.0Li 2016Clinical Journal of Medical OfficersChinaChildren aged 1-6 years with burnsProspective before-after randomized controlled trialVSD was performed within 24 hours. Cover the VSD after debridement, maintaining pressure in 0.02 ~ 0.06 kPa, continuous treatment 1 week/NPWT242523/2%-15%(1)(6)Type 3 iodine volts disinfection, spray outside skin of recombinant bovine basic fibroblast growth factor, gauze, according to the seepage situation 1 ~ 2 d treatment, until the wounds healedConventional therapy242523/Yin 2016China Medical CosmetologyChinaChildren aged 2-5 years with burnsProspective randomized controlled trialAfter debridement, negative pressure drainage was performed to control the negative pressure at 80-350mmhg, and the pus and blood were treated intermittently/NPWT6032283.1±0.8/(1)(7)Children aged 1-5 years with burnsAfter debridement, the conventional method was carried out on the wound skin orthopedic treatment, iodine gauze bandaging and dressingConventional therapy6031292.4±0.7/Yuan 2016Orthopaedics & Traumatology: Surgery & ResearchChinapediatric burn patientsProspective randomized controlled trialAfter debridement, the wound surface was veiled by a VSD sponge, sealed by a semi-permeable membrane, and fluid removed by a drainage tube made of silica gel under a negative pressure regulated between−50 and−150 mmHg. Then an artificial dermis was then implanted on fresh granulation tissue after debridement following the manufacturer’s instructions.VSSVacuum sealing drainage-artificial dermisimplantation-thin partial thickness skin grafting5335185.5±3.2/(3)Thin partial thickness skin was grafted on the wound surface afterSurgical debridement-change dressings thin partial thickness skin grafting201196.2±2.82.5~60Cai 2017Chinese Journal of BurnsChinaChildren aged 2-13 years with burnsProspective before-after randomized controlled trialScar excision combined with NPWTOSASNPWT2511145.5±2.8/(4)(7)Before scar excision combined with NPWT/2511145.5±2.8/Li 2018Journal of Clinincal Medical LiteratureChinaChildren aged 2-12 years with burnsProspective randomized controlled trialThe wound was disinfected with type 3 iodophor, external sprayed with recombinant bovine basic fibroblast epidermal growth factor, and bandaged with gauze. Dressing change was performed once every 1-2 days according to the exudation, until the wound healed/NPWT151056.9±1.52~15(1)(2)VSD was used within 1 day post injury. After debridement, negative pressure suction was performed, and the pressure was maintained at 0.02-0.06 kPaConventional therapy15966.7±1.32~15Wang 2018Journal of Frontiers of MedicineChinaChildren aged 3-15 years with burnsProspective randomized controlled trialAfter disinfection and debridement of necrotic tissue, vacuum drainage was performed with the negative pressure maintained at 50kPa, and skin grafting was performed after fresh granulation/NPWT4229136.61±3.09/(1)(7)Children aged 4-13 years with burnsAfter escharectomy, the wound was covered with vaseline gauze or antibacterial gauze. Dressing was changed every two days, and skin grafting was performed after fresh granulation grewVaseline gauze or antibiotic gauze dressing change treatment4220225.37±3.19/Zheng 2019Chinese Journal of BurnsChinaChildren aged 0-6 years with burnsProspective randomized controlled trialAfter simple debridement, the wound was covered with micro-negative pressure material and fixed with transparent air sticking film, and the negative pressure material was changed every 3-5 daysVSSMDNPWT3218143.9±1.65.5±2.2(1)(2)(3)(4)(7)After simple debridement, the wound was covered with vaseline gauze with 1% silver sulfadiazine cream, and the dressing was changed once every other dayConventional therapy3220123.8±1.75.8±1.6Frear 2021Scientific ReportsAustralian New Zealand Children under 17 years old with minor burnsDouble-arm parallel-group randomized controlled trialStandard care was combined with adjunctive NPWT/Standard care (silver impregnated dressings)+NPWT475942/0.015(7)Standard care (silver dip dressing)Standard care (silver impregnated dressings)54 1–2Zhang 2021-1Chinese Journal of Injury Repair and Wound Healing (Electronic Edition)ChinaChildren aged 1-14 years with burnsProspective randomized controlled trialAfter debridement, a large piece of skin was implanted, and MDNPWT was cut to the size of the skin graft area. No external negative pressure was needed, and the outer elastic gauze bandage was fixed/MDNPWT292273.8±2.48.3±1.2(2)(5)(6)(7)After debridement, a large piece of skin was implanted, and the skin graft was intermittently sutured to the edge of the wound with 4-0 non-absorbable suture. After knotting, the skin graft was covered with vaseline gauze. Multi-layer medical fat-free cotton or gauze blocks were used to fill the skin graft area, and the suture lines were knotted and bound up under pressurePressure Pack3424103.7±2.68.5±1.4Zhang 2021-2Chinese Journal of Injury Repair and Wound Healing (Electronic Edition)ChinaChildren aged 1-14 years with burnsProspective randomized controlled trialAfter debridement, a large piece of skin was implanted, and the negative pressure of VSD was maintained at -75 to -50 MMHG, and the negative pressure suction was continued/NPWT272164.1±2.28.7±1.3(2)(5)(6)(7)After debridement, a large piece of skin was implanted, and the skin graft was intermittently sutured to the edge of the wound with 4-0 non-absorbable suture. After knotting, the skin graft was covered with vaseline gauze. Multi-layer medical fat-free cotton or gauze blocks were used to fill the skin graft area, and the suture lines were knotted and bound up under pressurePressure Pack3424103.7±2.68.5±1.4Different serial numbers corresponding to different outcomes: (1) Healing time (Day); (2) Healing rate (%); (3) Detection rate of positive bacteria (%); (4) Scar scale scores (Point); (5) treatment costs (Yuan); (6) Numbers of dressing changes; (7) Incidence of adverse reactions (%) Table 1 Detailed information of the included studiesStudy(author/year)Journal CountryPatientsDesignProcessScalesTreatmentNumber of patientsMale (n)Female (n)Mean age (sd, Year)%TBSAOutcomesShen 2013Chinese Journal of BurnsChinaChildren aged 1-7 years with burnsProspective randomized controlled trialWithin 72h after injury treatment begins, the end of the 9 d treatment, don't change dressings, suction cover after debridement, maintaining negative pressure - 16 kpa/NPWT14581643.3±1.210±3(1)(2)(3)(6)(7)Conventional therapy16285773.1±1.810±4Yao 2013-1Infection, Inflammation, RepairChinaBurn children aged 8 months to 5 years oldProspective randomized controlled trialEscharectomy was performed under general anesthesia within 24 hours post injury, and VSD was covered after escharectomy without dressing changeVSSNPWT20//2.6±0.410.3±1.0(1)(4)(5)(6)The wound was covered with self-made ordinary petroleum petroleum gauze after dermabrasion within 24 hours after injury. The outer dressing was replaced on the second day after surgery, and the soggy outer dressing was replaced in timeVaseline gauze dressing change treatment20//2.5±0.49.8±1.3Yao 2013-2Infection, Inflammation, RepairChinaBurn children aged 8 months to 5 years oldProspective randomized controlled trialEscharectomy was performed under general anesthesia within 24 hours post injury, and VSD was covered after escharectomy without dressing changeVSSNPWT20//2.6±0.410.3±1.0(1)(4)(5)(6)The eschar was dermabrasion under general anesthesia within 24 hours post injury. After dermabrasion, the xenogenic acellular dermal matrix was covered, and dressing was changed regularlyXenogenic acellular dermal matrix20//2.6±0.510.1±1.2Bai 2016Journal of Traumatic SurgeryChinaChildren with burnsProspective randomized controlled trialAfter debridement, VSD was performed directly. The negative pressure was -53 to -26kPa, and the sponge and semi-permeable film were replaced after 6-7 days of continuous suction until the patient's wound healedVSSNPWT4024165.4 ±2.412.4±4.0(1)(2)(3)(4)(5)For the treatment of burns, the oil gauze was covered after debridement, and the sterile gauze dressing was bandaged under pressure, and the dressing was changed in time until the wound healedconventional therapy3520155.8 ±2.511.7±4.0Li 2016Clinical Journal of Medical OfficersChinaChildren aged 1-6 years with burnsProspective before-after randomized controlled trialVSD was performed within 24 hours. Cover the VSD after debridement, maintaining pressure in 0.02 ~ 0.06 kPa, continuous treatment 1 week/NPWT242523/2%-15%(1)(6)Type 3 iodine volts disinfection, spray outside skin of recombinant bovine basic fibroblast growth factor, gauze, according to the seepage situation 1 ~ 2 d treatment, until the wounds healedConventional therapy242523/Yin 2016China Medical CosmetologyChinaChildren aged 2-5 years with burnsProspective randomized controlled trialAfter debridement, negative pressure drainage was performed to control the negative pressure at 80-350mmhg, and the pus and blood were treated intermittently/NPWT6032283.1±0.8/(1)(7)Children aged 1-5 years with burnsAfter debridement, the conventional method was carried out on the wound skin orthopedic treatment, iodine gauze bandaging and dressingConventional therapy6031292.4±0.7/Yuan 2016Orthopaedics & Traumatology: Surgery & ResearchChinapediatric burn patientsProspective randomized controlled trialAfter debridement, the wound surface was veiled by a VSD sponge, sealed by a semi-permeable membrane, and fluid removed by a drainage tube made of silica gel under a negative pressure regulated between−50 and−150 mmHg. Then an artificial dermis was then implanted on fresh granulation tissue after debridement following the manufacturer’s instructions.VSSVacuum sealing drainage-artificial dermisimplantation-thin partial thickness skin grafting5335185.5±3.2/(3)Thin partial thickness skin was grafted on the wound surface afterSurgical debridement-change dressings thin partial thickness skin grafting201196.2±2.82.5~60Cai 2017Chinese Journal of BurnsChinaChildren aged 2-13 years with burnsProspective before-after randomized controlled trialScar excision combined with NPWTOSASNPWT2511145.5±2.8/(4)(7)Before scar excision combined with NPWT/2511145.5±2.8/Li 2018Journal of Clinincal Medical LiteratureChinaChildren aged 2-12 years with burnsProspective randomized controlled trialThe wound was disinfected with type 3 iodophor, external sprayed with recombinant bovine basic fibroblast epidermal growth factor, and bandaged with gauze. Dressing change was performed once every 1-2 days according to the exudation, until the wound healed/NPWT151056.9±1.52~15(1)(2)VSD was used within 1 day post injury. After debridement, negative pressure suction was performed, and the pressure was maintained at 0.02-0.06 kPaConventional therapy15966.7±1.32~15Wang 2018Journal of Frontiers of MedicineChinaChildren aged 3-15 years with burnsProspective randomized controlled trialAfter disinfection and debridement of necrotic tissue, vacuum drainage was performed with the negative pressure maintained at 50kPa, and skin grafting was performed after fresh granulation/NPWT4229136.61±3.09/(1)(7)Children aged 4-13 years with burnsAfter escharectomy, the wound was covered with vaseline gauze or antibacterial gauze. Dressing was changed every two days, and skin grafting was performed after fresh granulation grewVaseline gauze or antibiotic gauze dressing change treatment4220225.37±3.19/Zheng 2019Chinese Journal of BurnsChinaChildren aged 0-6 years with burnsProspective randomized controlled trialAfter simple debridement, the wound was covered with micro-negative pressure material and fixed with transparent air sticking film, and the negative pressure material was changed every 3-5 daysVSSMDNPWT3218143.9±1.65.5±2.2(1)(2)(3)(4)(7)After simple debridement, the wound was covered with vaseline gauze with 1% silver sulfadiazine cream, and the dressing was changed once every other dayConventional therapy3220123.8±1.75.8±1.6Frear 2021Scientific ReportsAustralian New Zealand Children under 17 years old with minor burnsDouble-arm parallel-group randomized controlled trialStandard care was combined with adjunctive NPWT/Standard care (silver impregnated dressings)+NPWT475942/0.015(7)Standard care (silver dip dressing)Standard care (silver impregnated dressings)54 1–2Zhang 2021-1Chinese Journal of Injury Repair and Wound Healing (Electronic Edition)ChinaChildren aged 1-14 years with burnsProspective randomized controlled trialAfter debridement, a large piece of skin was implanted, and MDNPWT was cut to the size of the skin graft area. No external negative pressure was needed, and the outer elastic gauze bandage was fixed/MDNPWT292273.8±2.48.3±1.2(2)(5)(6)(7)After debridement, a large piece of skin was implanted, and the skin graft was intermittently sutured to the edge of the wound with 4-0 non-absorbable suture. After knotting, the skin graft was covered with vaseline gauze. Multi-layer medical fat-free cotton or gauze blocks were used to fill the skin graft area, and the suture lines were knotted and bound up under pressurePressure Pack3424103.7±2.68.5±1.4Zhang 2021-2Chinese Journal of Injury Repair and Wound Healing (Electronic Edition)ChinaChildren aged 1-14 years with burnsProspective randomized controlled trialAfter debridement, a large piece of skin was implanted, and the negative pressure of VSD was maintained at -75 to -50 MMHG, and the negative pressure suction was continued/NPWT272164.1±2.28.7±1.3(2)(5)(6)(7)After debridement, a large piece of skin was implanted, and the skin graft was intermittently sutured to the edge of the wound with 4-0 non-absorbable suture. After knotting, the skin graft was covered with vaseline gauze. Multi-layer medical fat-free cotton or gauze blocks were used to fill the skin graft area, and the suture lines were knotted and bound up under pressurePressure Pack3424103.7±2.68.5±1.4Different serial numbers corresponding to different outcomes: (1) Healing time (Day); (2) Healing rate (%); (3) Detection rate of positive bacteria (%); (4) Scar scale scores (Point); (5) treatment costs (Yuan); (6) Numbers of dressing changes; (7) Incidence of adverse reactions (%) Detailed information of the included studies Different serial numbers corresponding to different outcomes: (1) Healing time (Day); (2) Healing rate (%); (3) Detection rate of positive bacteria (%); (4) Scar scale scores (Point); (5) treatment costs (Yuan); (6) Numbers of dressing changes; (7) Incidence of adverse reactions (%) These studies, published between 2013 and 2021, exhibited a range in individual trial sample sizes from 24 to 307. Two trials were published in English, while ten trials were available in Chinese. The trials were predominantly conducted in China, with one study carried out in Australia. Almost all patients were pediatric burn attendees. Each patient was admitted due to various degrees of burn injuries. The intervention approaches in the selected trials primarily consisted of diverse forms of NPWT, with the majority being NPWT alone. Newer technologies were also adopted, like in Yuan et al.‘s [22] study that combined vacuum sealed drainage (VSD), artificial dermis implantation (ADI), and thin section thickness skin transplantation (TSG) in the treatment group. Moreover, Microdynamic Negative Pressure Wound Therapy (MDNPWT), an extension of NPWT technique, was employed by Zheng et al. [26] and Zhang et al. [28] in their treatment groups. In the control group, most treatment plans consisted of regular dressing changes under non-surgical treatment, although different dressings were utilized. Zhang’s study, for example, used large skin grafts post-debridement in the control group and intermittently sutured the skin at the wound edge using 4 − 0 non-absorbable sutures. The skin was covered with Vaseline gauze after tying the sutures, with leftover threads. The skin graft area was filled with multiple layers of medical defatted cotton or gauze blocks, and the suture was knotted and compressed. This method was used to compare against MDNPWT and NPWT in the other two intervention groups. Nine of the studies conducted follow-ups only until the patients’ wounds were completely healed during hospitalization, with no subsequent follow-up. Exceptions include the respective studies of Bai et al. [19], Cai et al. [23], and Zheng et al.‘s [26] which documented scar development in the surgical region for six months and one year after surgery. In studies that recorded post-operative scar scores, all except Cai et al.‘s [23] study, which utilized the observer scar assessment scale (OSAS) for evaluation, used the Vancouver scar assessment scale (VSS). Table 1 summarizes detailed information from the included studies. It is worth noting that in Table 1, three groups of patients appeared in the studies of Yao et al. and Zhang et al., three groups of patients were composed of one or two groups of NPWT treatment and other group of non-NPWT, which is equivalent to the data of each study can be extracted into two separate randomized controlled trials containing two groups of patients, so in Table 1. In this study, we split the data of these two studies and recorded them as “Yao2013-1”, “Yao2013-2”, “Zhang2021-1” and “Zhang2021-2”. In these four RCTs, which were artificially separated by us, NPWT treatment groups were compared with non-NPWT treatment groups. The reason for this is that the number of RCTs that could be included was too small, and proper splitting of data from the three groups in a given study would increase the amount of data that could be included and allow us to see differences in outcomes with or without NPWT, although this would come with additional heterogeneity. Therefore, a total of 12 studies were included in this study, including 14 RCTs. Risk of bias and quality assessment of individual studies We assessed 12 studies using the Cochrane Collaboration risk of bias tool. Figure 2; Table 1 demonstrate the characteristics of the included studies and their designs, along with the summary of the risk of bias in the included trials across different quality domains of the risk of bias tool. Of the 12 studies, only two studies [25, 27] employed double-blinded outcome assessors. Six studies reported adequate randomized sequence generation, while the method of randomization (which included odd and even sampling by hospitalization number and clinician involvement in allocation) remained unclear in the other six studies. Among all studies, two were double-blinded [25, 27], two were prospective before-after randomized controlled design [20, 23], and one was a double-arm parallel-group randomized design [27]. All studies, except for two, reported an unclear risk of bias in allocation concealment.Though most studies noted an unclear risk of bias in the blinding of participants and key researchers, two studies bore a lower risk of measurement bias. Furthermore, four studies presented a lower risk of follow-up bias and reporting bias, and there was a significant attrition bias in Yao et al.‘s study in 2013 and Zhang et al.‘s study in 2021. During the follow-up observation process, several patients discontinued the trial due to different reasons, such as hospital transfer or family requests to stop treatment. Fig. 2Risk of bias charts. A Risk of bias in the included studies; B Risk of bias summary for all included studies. With the exception of two studies, all others demonstrated a low or uncertain overall risk of bias. Yao et al. (2013) and Zhang et al. (2021) exhibited notable attrition bias in their studies. Throughout the follow-up observation period, several participants withdrew from the trials for various reasons, including hospital transfers and family requests to discontinue treatment Risk of bias charts. A Risk of bias in the included studies; B Risk of bias summary for all included studies. With the exception of two studies, all others demonstrated a low or uncertain overall risk of bias. Yao et al. (2013) and Zhang et al. (2021) exhibited notable attrition bias in their studies. Throughout the follow-up observation period, several participants withdrew from the trials for various reasons, including hospital transfers and family requests to discontinue treatment Following the bias evaluation, all studies maintained a moderate risk of bias in their original literature (Fig. 2). Meta-analysis findings Healing time Data on the healing time was reported in 8 studies, comprising 9 RCTs. The pooled quantitative data revealed a significantly shorter healing time in the NPWT group compared to the control group, with a SMD of −1.60 (95% CI: −2.26 - −0.95; p < 0.001, I2 = 92.8%) (Fig. 3A). However, the funnel plot indicated no significant disparity between the two groups (Fig. 3B). Fig. 3Forest plot, funnel plot and the results of Egger’s regression test, trim and fill method, sensitivity analysis of the meta-analysis illustrating the overall weighted effect size of negative pressure wound therapy versus control on the healing time in paediatric burns. A Forest plot. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. Different colors (green, red, yellow) and symbols (“+”, “-“, “?” “) to denote “low risk of bias”, “high risk of bias” and “unclear risk of bias “, respectively. B Funnel plot. The effect size “SMD” is shown on the abscissa, and the inverse of the standard error of the value of the effect size, SE (SMD), is shown on the ordinate. The dots in the figure are the individual studies included. C Result of Egger’s regression test. D Result of the trim and fill method. E Result of sensitivity analysis Forest plot, funnel plot and the results of Egger’s regression test, trim and fill method, sensitivity analysis of the meta-analysis illustrating the overall weighted effect size of negative pressure wound therapy versus control on the healing time in paediatric burns. A Forest plot. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. Different colors (green, red, yellow) and symbols (“+”, “-“, “?” “) to denote “low risk of bias”, “high risk of bias” and “unclear risk of bias “, respectively. B Funnel plot. The effect size “SMD” is shown on the abscissa, and the inverse of the standard error of the value of the effect size, SE (SMD), is shown on the ordinate. The dots in the figure are the individual studies included. C Result of Egger’s regression test. D Result of the trim and fill method. E Result of sensitivity analysis We also ascertained the presence of publication bias using both Egger’s regression test and the trim and fill method as it is crucial for researchers to inspect potential biases through various methods. The outcomes of Egger’s regression test, which scrutinized the zero intercepts, implied no publication bias. Similarly, the results revealed no publication bias using the trim and fill method (Fig. 3C and D). Moreover, a sensitivity analysis showed no significant bias in the results (Fig. 3E). Healing rate Complete data on the effective rate were accessible in 5 studies. The aggregated data from these 6 RCTs showed a statistically significant increase in the healing rate of the experimental groups relative to the control (RR = 1.17; 95% CI: 0.99–1.39; p < 0.001, I2 = 78%) (Fig. 4A). The constructed funnel plot illustrated no evidence of publication bias (See Fig. 4B). Further verifications for bias using Egger’s regression test, the trim and fill method, and sensitivity analysis didn’t display any significant bias either (Fig. 4C, D and E). Fig. 4Forest plot, funnel plot and the results of Egger’s regression test, trim and fill method, sensitivity analysis of the meta-analysis illustrating the overall weighted effect size of negative pressure wound therapy versus control on the healing rate in paediatric burns. A Forest plot. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. Different colors (green, red, yellow) and symbols (“+”, “-“, “?” “) to denote “low risk of bias”, “high risk of bias” and “unclear risk of bias “, respectively. B Funnel plot. The effect size “SMD” is shown on the abscissa, and the inverse of the standard error of the value of the effect size, SE (SMD), is shown on the ordinate. The dots in the figure are the individual studies included. C Result of Egger’s regression test. D Result of the trim and fill method. E Result of sensitivity analysis Forest plot, funnel plot and the results of Egger’s regression test, trim and fill method, sensitivity analysis of the meta-analysis illustrating the overall weighted effect size of negative pressure wound therapy versus control on the healing rate in paediatric burns. A Forest plot. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. Different colors (green, red, yellow) and symbols (“+”, “-“, “?” “) to denote “low risk of bias”, “high risk of bias” and “unclear risk of bias “, respectively. B Funnel plot. The effect size “SMD” is shown on the abscissa, and the inverse of the standard error of the value of the effect size, SE (SMD), is shown on the ordinate. The dots in the figure are the individual studies included. C Result of Egger’s regression test. D Result of the trim and fill method. E Result of sensitivity analysis Numbers of dressing changes The numbers of dressing changes showed a significant disparity between the treatment and control groups, with a SMD of −4.6 (95% CI: −5.84 - −3.36; p < 0.001, I2 = 92.4%) (Fig. 5A). The funnel plots, Egger’s regression test, trim and fill method, and sensitivity analysis depicted no evidence of publication bias in these analyses (Fig. 5B, C, D and E). Fig. 5Forest plot, funnel plot and the results of Egger’s regression test, trim and fill method, sensitivity analysis of the meta-analysis illustrating the overall weighted effect size of negative pressure wound therapy versus control on the numbers of dressing changes in paediatric burns. A Forest plot. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. Different colors (green, red, yellow) and symbols (“+”, “-“, “?” “) to denote “low risk of bias”, “high risk of bias” and “unclear risk of bias “, respectively. B Funnel plot. The effect size “SMD” is shown on the abscissa, and the inverse of the standard error of the value of the effect size, SE (SMD), is shown on the ordinate. The dots in the figure are the individual studies included. C Result of Egger’s regression test. D Result of the trim and fill method. E Result of sensitivity analysis Forest plot, funnel plot and the results of Egger’s regression test, trim and fill method, sensitivity analysis of the meta-analysis illustrating the overall weighted effect size of negative pressure wound therapy versus control on the numbers of dressing changes in paediatric burns. A Forest plot. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. Different colors (green, red, yellow) and symbols (“+”, “-“, “?” “) to denote “low risk of bias”, “high risk of bias” and “unclear risk of bias “, respectively. B Funnel plot. The effect size “SMD” is shown on the abscissa, and the inverse of the standard error of the value of the effect size, SE (SMD), is shown on the ordinate. The dots in the figure are the individual studies included. C Result of Egger’s regression test. D Result of the trim and fill method. E Result of sensitivity analysis Detection rate of positive bacteria A meta-analysis of 3 RCTs revealed that the detection rate of positive bacteria was higher in the treatment group than in the control group (RR = 0.61; 95% CI: 0.26–1.46; p = 0.004, I2 = 81.8%) (Fig. 6A). The I2 statistic suggested high heterogeneity across these three studies. The funnel plots, Egger’s regression test, trim and fill method, and sensitivity analysis all indicated the absence of discernible publication bias (Fig. 6B, C, D and E). Fig. 6Forest plot, funnel plot and the results of Egger’s regression test, trim and fill method, sensitivity analysis of the meta-analysis illustrating the overall weighted effect size of negative pressure wound therapy versus control on the detection rate of positive bacteria in paediatric burns. A Forest plot. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. Different colors (green, red, yellow) and symbols (“+”, “-“, “?” “) to denote “low risk of bias”, “high risk of bias” and “unclear risk of bias “, respectively. B Funnel plot. The effect size “SMD” is shown on the abscissa, and the inverse of the standard error of the value of the effect size, SE (SMD), is shown on the ordinate. The dots in the figure are the individual studies included. C Result of Egger’s regression test. D Result of the trim and fill method. E Result of sensitivity analysis Forest plot, funnel plot and the results of Egger’s regression test, trim and fill method, sensitivity analysis of the meta-analysis illustrating the overall weighted effect size of negative pressure wound therapy versus control on the detection rate of positive bacteria in paediatric burns. A Forest plot. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. Different colors (green, red, yellow) and symbols (“+”, “-“, “?” “) to denote “low risk of bias”, “high risk of bias” and “unclear risk of bias “, respectively. B Funnel plot. The effect size “SMD” is shown on the abscissa, and the inverse of the standard error of the value of the effect size, SE (SMD), is shown on the ordinate. The dots in the figure are the individual studies included. C Result of Egger’s regression test. D Result of the trim and fill method. E Result of sensitivity analysis Incidence of adverse reactions Data on adverse reactions was reported in 6 RCTs. A significant difference in adverse reactions was observed between the treatment and control groups, with a RR of 0.61 (95% confidence interval (CI): 0.33–1.12; p = 0.005, I2 = 68%). These findings are illustrated in Fig. 7A. No significant publication bias was detected across the selected studies, according to the symmetry of the funnel plot and the outcomes from the Egger’s regression test, trim and fill method, and sensitivity analysis (Fig. 7B, C and D, and 7E). Fig. 7Forest plot, funnel plot and the results of Egger’s regression test, trim and fill method, sensitivity analysis of the meta-analysis illustrating the overall weighted effect size of negative pressure wound therapy versus control on the incidence of adverse reactions in paediatric burns. A Forest plot. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. Different colors (green, red, yellow) and symbols (“+”, “-“, “?” “) to denote “low risk of bias”, “high risk of bias” and “unclear risk of bias “, respectively. B Funnel plot. The effect size “SMD” is shown on the abscissa, and the inverse of the standard error of the value of the effect size, SE (SMD), is shown on the ordinate. The dots in the figure are the individual studies included. C Result of Egger’s regression test. D Result of the trim and fill method. E Result of sensitivity analysis Forest plot, funnel plot and the results of Egger’s regression test, trim and fill method, sensitivity analysis of the meta-analysis illustrating the overall weighted effect size of negative pressure wound therapy versus control on the incidence of adverse reactions in paediatric burns. A Forest plot. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. Different colors (green, red, yellow) and symbols (“+”, “-“, “?” “) to denote “low risk of bias”, “high risk of bias” and “unclear risk of bias “, respectively. B Funnel plot. The effect size “SMD” is shown on the abscissa, and the inverse of the standard error of the value of the effect size, SE (SMD), is shown on the ordinate. The dots in the figure are the individual studies included. C Result of Egger’s regression test. D Result of the trim and fill method. E Result of sensitivity analysis Scar scale scores The scar scale scores showed no significant difference between the treatment and control groups, with a SMD of −1.66 (95% CI: −2.54 - −0.79; p < 0.001, I2 = 89.4%) (Fig. 8A). The funnel plot, which was partially symmetrical, indicated no risk of publication bias. The outcomes from the Egger’s regression test, the trim and fill method, and the sensitivity analysis aligned with these results, revealing no significant bias (Fig. 8B, C and D, and 8E). Fig. 8Forest plot, funnel plot and the results of Egger’s regression test, trim and fill method, sensitivity analysis of the meta-analysis illustrating the overall weighted effect size of negative pressure wound therapy versus control on the scar scale scores in paediatric burns. A Forest plot. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. Different colors (green, red, yellow) and symbols (“+”, “-“, “?” “) to denote “low risk of bias”, “high risk of bias” and “unclear risk of bias “, respectively. B Funnel plot. The effect size “SMD” is shown on the abscissa, and the inverse of the standard error of the value of the effect size, SE (SMD), is shown on the ordinate. The dots in the figure are the individual studies included. C Result of Egger’s regression test. D Result of the trim and fill method. E Result of sensitivity analysis Forest plot, funnel plot and the results of Egger’s regression test, trim and fill method, sensitivity analysis of the meta-analysis illustrating the overall weighted effect size of negative pressure wound therapy versus control on the scar scale scores in paediatric burns. A Forest plot. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. Different colors (green, red, yellow) and symbols (“+”, “-“, “?” “) to denote “low risk of bias”, “high risk of bias” and “unclear risk of bias “, respectively. B Funnel plot. The effect size “SMD” is shown on the abscissa, and the inverse of the standard error of the value of the effect size, SE (SMD), is shown on the ordinate. The dots in the figure are the individual studies included. C Result of Egger’s regression test. D Result of the trim and fill method. E Result of sensitivity analysis Treatment costs The NPWT reportedly influenced the treatment costs among patients. This effect is demonstrated in the forest plots, which depicted a SMD of 0.92 (95% CI: −1.66–3.49; p < 0.001, I2 = 98.4%) (Fig. 9A). NPWT applied to pediatric burn patients during hospitalization significantly reduced costs compared to the control group. Despite the high statistical heterogeneity represented by the large I2 value, no biased pattern was observed through the funnel plot, Egger’s regression test, trim and fill method, and sensitivity analysis. (Figure 9B, C, D and E). Fig. 9Forest plot, funnel plot and the results of Egger’s regression test, trim and fill method, sensitivity analysis of the meta-analysis illustrating the overall weighted effect size of negative pressure wound therapy versus control on the treatment costs in paediatric burns. A Forest plot. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. Different colors (green, red, yellow) and symbols (“+”, “-“, “?” “) to denote “low risk of bias”, “high risk of bias” and “unclear risk of bias “, respectively. B Funnel plot. The effect size “SMD” is shown on the abscissa, and the inverse of the standard error of the value of the effect size, SE (SMD), is shown on the ordinate. The dots in the figure are the individual studies included. C Result of Egger’s regression test. D Result of the trim and fill method. E Result of sensitivity analysis Forest plot, funnel plot and the results of Egger’s regression test, trim and fill method, sensitivity analysis of the meta-analysis illustrating the overall weighted effect size of negative pressure wound therapy versus control on the treatment costs in paediatric burns. A Forest plot. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. Different colors (green, red, yellow) and symbols (“+”, “-“, “?” “) to denote “low risk of bias”, “high risk of bias” and “unclear risk of bias “, respectively. B Funnel plot. The effect size “SMD” is shown on the abscissa, and the inverse of the standard error of the value of the effect size, SE (SMD), is shown on the ordinate. The dots in the figure are the individual studies included. C Result of Egger’s regression test. D Result of the trim and fill method. E Result of sensitivity analysis Subgroup analysis findings To investigate the source of the high heterogeneity in the above results, we performed a subgroup analysis according to whether the burn degree of the child was mild (Labeled as “1”) or moderate-severe (Labeled as “2”). Except for the meta-analysis investigating the numbers of dressing change, the studies included in the meta-analysis were all children with moderate or severe burns, so subgroup analysis could not be performed, the other analyses were performed with subgroup analysis. First, we found that when subgroup analysis was performed according to the degree of burn, there was no significant difference in wound healing time between children with mild burn treated with NPWT and those without NPWT (SMD = −2.01, 95%CI: −2.54 - − 1.47; p = 0.608, I2 = 0.0%), while the healing time of NPWT treatment group was shorter than that of control group in children with moderate or severe burns (SMD = −1.48, 95%CI: −2.24 - − 0.73; p < 0.001, I2 = 92.8%), indicating that the degree of burn injury was not one of the sources of heterogeneity. At the same time, the combined results of the two subgroups were statistically significant, which was consistent with the total combined results, indicating that NPWT treatment could reduce the wound healing time of children with different burn degrees (Supplementary Fig. 1). Similarly, there was no significant difference in healing time between the mild burn children treated with NPWT and the control group (RR = 1.33, 95%CI: 1.02–1.74; p = 0.739, I2 = 0.0%), while in children with moderate or severe burns, the healing rate of NPWT treatment group was shorter than that of control group (RR = 1.14, 95%CI: 0.95–1.37; p < 0.001, I2 = 82.9%)(Supplementary Fig. 2. All these results indicated that the degree of burn injury was not one of the sources of heterogeneity in the above results. Subgroup analysis also found that the detection rate of positive bacteria in the NPWT treatment group was lower in children with mild burns and children with moderate or severe burns. High heterogeneity was observed in the subgroup of children with mild burns (I2 = 85.5%) (Supplementary Fig. 3). Then, as for adverse reactions, we also observed that in children with mild burns, NPWT treatment group had fewer adverse reactions (RR = 1.27, 95%CI: 0.19–8.39; p < 0.001, I2 = 91.6%), but no significant difference was found between the two groups in children with moderate or severe burns (RR = 0.42, 95%CI: 0.32–0.55; p = 0.966, I2 = 0.0%)(Supplementary Fig. 4). In addition, the results of subgroup analysis suggested that the scar scores of burn children in the NPWT-treated group were lower than those in the control group regardless of the degree of burn, but high heterogeneity was observed in the subgroups of children with moderate or severe burns (I2 = 91.8%) (Supplementary Fig. 5). Similarly, in terms of treatment cost, subgroup analysis also found that the treatment cost of NPWT treatment group was lower than that of control group regardless of the degree of burn, and high heterogeneity was found in children with moderate or severe burns (I2 = 98.4%) (Supplementary Fig. 6). Finally, among the above subgroup analysis results, the combined results of the two subgroups were statistically significant, which was consistent with the total combined results, indicating that burn degree was not the source of heterogeneity in the results of meta-analysis. NPWT treatment can reduce the wound healing time, improve the wound healing rate, reduce the infection rate of positive bacteria, and reduce adverse reactions in children with different burn degrees. At the same time, it can improve the scar hyperplasia after healing and reduce the treatment cost. Discussion It is crucial to exercise stringent control over surgical indications when addressing pediatric burns. Any process of wound repair surgery has potential adverse effects on the subsequent growth and development of burned children, which determines the complexity of managing pediatric burns. Negative pressure closure drainage represents a novel technique for skin wound repair that has been extensively employed in adult burn treatments. However, there remains insufficient evidence regarding its efficacy and safety specifically within the context of pediatric burns. This systematic review and meta-analysis aims to assess the efficacy and safety of NPWT for pediatric burns. The primary finding of this study is that NPWT can reduce the healing time and improve the wound healing rate in pediatric burns. Most of the intervention groups in the studies [17–20, 24, 26, 27] covered NPWT directly after debridement until the wound was completely healed. It is worth noting that Zheng et al. [26] and Zhang et al. [28] applied micro-powered negative pressure devices. Yin et al. [21] and Yuan et al. [22] covered the negative pressure suction device after plastic surgery, and the intervention group adopted autologous skin transplantation after surgical scar resection and fixed the transplanted skin with negative pressure suction device. However, Wang et al. [25] and Frear et al. [28] performed NPWT after debridement on the children in the intervention group, and then performed routine skin grafting after fresh granulation tissue grew out of the wound. Although the meta-analysis of healing time and healing rate showed high heterogeneity, this heterogeneity was not found in the egger test, trim and fill method and sensitivity analysis. Subgroup analyses of children with different degrees of burns also failed to address this issue. Despite our best efforts such as conducting meta regression and sensitivity analyses, the source of this heterogeneity could not be pinpointed convincingly. A systematic review and meta-analysis of randomized controlled trials by Jiang et al. [29] found that NPWT integrates skin grafts more effectively than non NPWT, increasing the improved percentage of graft take. However, sufficient data on adverse events has not been obtained, and the safety of NPWT in skin grafts has not yet been evaluated. NPWT has the advantages of simple operation, shorter operation time, more conducive to postoperative rehabilitation of children, less postoperative complications and shorter hospital stay [30]. The traditional pressure dressing method is difficult to control the pressure on the skin graft wound, and the force on the skin graft is uneven. The continuous limb movement of the child is easy to lead to the relaxation of the dressing [31], which causes the skin graft to shift, curl or even overlap. It can make the skin graft adhere to the burn wound with uniform force, make the skin graft adhere to the wound more closely, attach the skin graft firmly, reduce the probability of displacement, reduce the hematoma under the skin graft [32], promote the growth of blood vessels in the wound to the skin graft, prevent the transverse traction of new blood vessels between the skin graft and the base [33], make the skin graft get nutrition from the wound in time and fully [34], and improve the survival rate of the skin graft. In the burn wound without skin grafting, NPWT also shows the advantages of accelerating wound healing compared with conventional dressing change. It has the advantages of promoting blood circulation in the treatment of superficial burns in children. Some studies [35, 36] show that NPWT can significantly improve the blood perfusion of the wound, increase the number of capillaries at the edge of the wound, speed up the blood flow, help the capillaries and endothelial cells to restore the normal morphology and structure [37], expand the microvessels, increase the blood supply of the wound, and adjust and improve the microcirculation. Our study also found that NPWT could also reduce the number of dressing changes in pediatric burn patients. Due to children’s activity and pain and discomfort after burns, dressings in the wound area are easy to become loose and fall off. The movement of dressings will also affect the microenvironment of the wound bed, resulting in edema and infection, and eventually cause delayed wound healing [38]. NPWT treatment of superficial partial-thickness burn wounds in children can maintain continuous negative pressure suction for a week without replacement, and most wounds can achieve healing. Compared with the traditional routine dressing treatment, the frequency of dressing change is significantly reduced. Studies [39, 40] have confirmed that dry wounds formed by wet wounds under closed dressings contribute to the infiltration of polymorphonuclear leukocytes. A systematic review by Burhan et al. [41] based on 3599 chronic trauma patients found that NPWT decreased the rate of surgical site infection, controlled the inflammation and edema, created a good local environment for immune function cells to play phagocytosis and sterilization, so as to prevent wound infection. Zhang et al. [42] found that in the treatment of diabetic foot ulcers, compared to non NPWT, NPWT can also reduce the wound healing time in the treatment of diabetes foot ulcers, and has similar safety. These results were similar to our findings, where NPWT also reduced the detection rate of positive bacteria in patients with pediatric burns. In addition, NPWT can also reduce the edema of the wound, which is also one of the reasons for delayed wound healing. In general, NPWT treatment had little effect on the activities of daily living. It is worth mentioning, because children’s active, but also may bring the negative pressure drainage drainage tube bending or pulled off cause blockage or leakage, and and adhesion of discomfort and increase the patient’s activities [43], therefore, will focus on communication with the families of children with postoperative nursing, get children to cooperate, at the same time grasp the anatomical and physiological characteristics of children [44], regulation is applicable to the optimum pressure value of children, the sealing condition of negative pressure material was closely observed to ensure the best treatment effect [45]. Although the application of NPWT is widely effective, it may produce some potential adverse complications. If the NPWT device is left on the wound for more than 72 h [46, 47], granulation tissue has the potential to ingrowth and cause the wound bed to rupture, bleed, and produce pain at the time of dressing removal. In addition, any retained foam fragments may cause adverse reactions [48]. It is worth noting that although Frear et al. [27] reported that the NPWT group had more adverse reactions, they mainly exist in the long-term scar management process after two healing, and did not report specific reasons. Only the number of people was simply recorded, and we did not contact the authors to obtain more relevant details, which also led to bias in the results to a certain extent. However, overall, we believe that the treatment of NPWT in pediatric burns is relatively safe, with a higher probability of complications/adverse reactions. Similar to other treatment methods, and no serious complications have been reported. Our study also found that NPWT significantly reduced the incidence of complications and the total cost of treatment in pediatric burn patients. Lin et al. [49] found that NPWT significantly increased the overall grade rate and decreased the infection rate in the first week after burn skin grafting. However, they also found no significant difference in the overall incidence of complications compared to standard treatment. An RCT [50] found that NPWT did not benefit in wound healing or cost after primary knee arthroplasty. But in another clinical trial [51] found that Closed incision negative pressure wound therapy brought about lower invasive wound complications and significantly more non invasive complications to the surgical site after total joint replacement surgery. On the contrary, a clinical study [52] has used individualized comfortable nursing on patients with orthopedic traumas and significantly reduced the round healing time after receiving NPWT, the postoperative pain level, the incidence of complications, and improve patient satisfaction, this suggests that one of the key focuses of applying NPWT in pediatric burn wards in the future is postoperative care [53]. The cost of negative-pressure devices varies from country to country. Two retrospective studies [54, 55] found that VAC is one of the crowd cost-effective burn or chronic wound support therapy in America, but it seems to be cheaper in China [56, 57]. We suggest that although the initial cost of NPWT is higher than that of standard dressings or standard treatments [58], wound management options that include NPWT may help mitigate the costs associated with delayed healing, the reasons for this may include portable NPWT devices, which was associated with long-term cost savings and potential reductions in the cost of care [59, 60]. Although NPWT has the potential to reduce medical costs, but the key to reduce the pediatric burn the hospitalization expenses of is still should prevent children burns, wounds after response timely and reasonably, reduce the length of hospital stay, and thus fundamentally prevent children burn accidents, reduce the medical burden [61]. NPWT can be directly used to repair superficial skin wounds or scar repair, which can reduce tension, fix the graft, and improve the wound bed [62]. Our study found that NPWT significantly reduced scar scores after wound healing compared with conventional treatment in pediatric burn patients. In a meta-analysis [63] of determining the effects of invasive negative pressure wound therapy (iNPWT) on scars, there was also moderate evidence that iNPWT reduced scar width and improved patient and observer satisfaction with scar, but the study did not focus on g pediatric patients. Al-ishaq et al. and Timmermans et al. [64, 65] found no advantage of NPT in reducing the rate of skin necrosis in patients undergoing breast reconstruction and tumorplastic surgery, but may have a potential role in reducing scarring, However, Falkner et al. [66] found that NPT improved the healing and prognosis of anterior lateral thigh (ALT) flap donor site. However, in pediatric patients with complex wounds, it has been reported that the scars of patients after applying NPWT show better results on the visual scar scale than those applying traditional wet to moist dressing [67]. Similarly, Incisional negative pressure wound therapy was also found to improved the quality of cosmetic suture wounds and reduced the degree of postoperative scar hyperplasia [68]. The number of randomized controlled trials on the application of NPWT in pediatric burns to observe its effect on wound scar is too small and the quality is too low, further verification of this conclusion is needed in the future. Overall, this study is the first systematic review and meta-analysis to focus on the safety and efficacy of NPWT in pediatric burns. Our findings provide a summary of experience and inspiration for the future treatment of pediatric burns with NPWT. The total sample size included in this study is relatively large, which better represents the pediatric burn population, however, there are some shortcomings in this study. Each study encompasses diverse transplant sites/sources and various stages of burn treatment (including simple negative pressure treatment, skin grafting combined with negative pressure treatment, burn wound treatment, or post-burn scar treatment), which may impact the study outcomes. Furthermore, uncontrollable variables such as the surgeon’s skill level in performing these surgeries can also influence the results. The compliance of pediatric patients with different burn conditions towards NPWT treatment is an additional crucial limitation that necessitates consideration since our cohort primarily consists of pediatric patients, making compliance evaluation more challenging. Several factors such as the severity of the burns, comorbidities in the patients, technical differences in the application of NPWT, and the individual patients’ response to treatment could have contributed to the divergence in outcomes [69]. Although we performed a subgroup analysis and found that burn severity was not a source of high heterogeneity in our results, a more extensive and in-depth analysis could not be performed to find the true source of heterogeneity due to the uneven data levels of the included studies. In addition, the source of heterogeneity may also be related to many factors such as the compliance of the patients, the professional level of the nursing staff, and the timing of treatment. In the case of large skin lesions, children’s vital signs are often unstable, and too long operation time should be avoided. However, there were not enough data to evaluate the operation time of NPWT treatment and non-NPWT treatment respectively. Therefore, future research should increase exploration in this direction, as it may affect future NPWT management and help customize interventions that are tailored to individual needs [70]. Despite the obvious advantages of the NPWT technique, the importance of surgical debridement to remove necrotic tissue cannot be replaced. In addition, this study included in the test report is not complete in the process of the treatment of negative pressure value, this will also effect burn children wound healing process. These limitations notwithstanding, our study fills a significant gap in the literature and sets important groundwork for future research. Our findings underscore the potential of NPWT in managing paediatric burns, while also indicating areas where further scrutiny and robust research are needed.To draw stronger conclusions, further high-quality RCTs must be incorporated. Conclusions NPWT demonstrates a favorable role in the management of paediatric burns, offering a novel approach. This systematic review and meta-analysis provide support for the potential application of NPWT in reducing healing time, frequency of dressing changes, positive bacterial detection rate, incidence of adverse reactions, scar scale scores, and treatment costs while also enhancing the healing rate in paediatric burn patients. Supplementary Information Supplementary Material 1: Supplementary Figure 1. Forest plot of subgroup analysis of the overall weighted effect size of mild burns and moderate or severe burns on the healing time in pediatric burns. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. "1" refers to the patients conforming to mild burn, "2" refers to the patients conforming to moderate or severe burn. Supplementary Material 1: Supplementary Figure 1. Forest plot of subgroup analysis of the overall weighted effect size of mild burns and moderate or severe burns on the healing time in pediatric burns. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. "1" refers to the patients conforming to mild burn, "2" refers to the patients conforming to moderate or severe burn. Supplementary Material 2: Supplementary Figure 2. Forest plot of subgroup analysis of the overall weighted effect size of mild burns and moderate or severe burns on the healing rate in pediatric burns. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate."1" refers to the patients conforming to mild burn, "2" refers to the patients conforming to moderate or severe burn. Supplementary Material 2: Supplementary Figure 2. Forest plot of subgroup analysis of the overall weighted effect size of mild burns and moderate or severe burns on the healing rate in pediatric burns. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate."1" refers to the patients conforming to mild burn, "2" refers to the patients conforming to moderate or severe burn. Supplementary Material 3: Supplementary Figure 3. Forest plot of subgroup analysis of the overall weighted effect size of mild burns and moderate or severe burns on the detection rate of positive bacteria in pediatric burns. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. "1" refers to the patients conforming to mild burn, "2" refers to the patients conforming to moderate or severe burn. Supplementary Material 3: Supplementary Figure 3. Forest plot of subgroup analysis of the overall weighted effect size of mild burns and moderate or severe burns on the detection rate of positive bacteria in pediatric burns. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. "1" refers to the patients conforming to mild burn, "2" refers to the patients conforming to moderate or severe burn. Supplementary Material 4: Supplementary Figure 4. Forest plot of subgroup analysis of the overall weighted effect size of mild burns and moderate or severe burns on the incidence rate of adverse reactions in pediatric burns. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. "1" refers to the patients conforming to mild burn, "2" refers to the patients conforming to moderate or severe burn. Supplementary Material 4: Supplementary Figure 4. Forest plot of subgroup analysis of the overall weighted effect size of mild burns and moderate or severe burns on the incidence rate of adverse reactions in pediatric burns. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. "1" refers to the patients conforming to mild burn, "2" refers to the patients conforming to moderate or severe burn. Supplementary Material 5: Supplementary Figure 5. Forest plot of subgroup analysis of the overall weighted effect size of mild burns and moderate or severe burns on the scar scale scores in pediatric burns. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. "1" refers to the patients conforming to mild burn, "2" refers to the patients conforming to moderate or severe burn. Supplementary Material 5: Supplementary Figure 5. Forest plot of subgroup analysis of the overall weighted effect size of mild burns and moderate or severe burns on the scar scale scores in pediatric burns. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. "1" refers to the patients conforming to mild burn, "2" refers to the patients conforming to moderate or severe burn. Supplementary Material 6: Supplementary Figure 6. Forest plot of subgroup analysis of the overall weighted effect size of mild burns and moderate or severe burns on the treatment costs in pediatric burns. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. "1" refers to the patients conforming to mild burn, "2" refers to the patients conforming to moderate or severe burn. Supplementary Material 6: Supplementary Figure 6. Forest plot of subgroup analysis of the overall weighted effect size of mild burns and moderate or severe burns on the treatment costs in pediatric burns. The diamond symbol at the bottom of the forest plot represents the overall weighted estimate. "1" refers to the patients conforming to mild burn, "2" refers to the patients conforming to moderate or severe burn.