Objectives To synthesise current evidence from systematic reviews (SRs) regarding the efficacy and safety of non-pharmacological interventions to prevent and treat pain in newborn infants.
Design Overview of SRs.
Data sources We searched PubMed, Embase, Cochrane Library, Web of Science, CINAHL, Chinese National Knowledge Infrastructure (CNKI), Chinese Biomedical Literature Database (CBM), Wanfang Database, Chinese Science and Technology Periodical Database (VIP) and Google Scholar to identify all relevant SRs published in the last 5 years.
Eligibility criteria for selecting studies We included SRs that evaluated the efficacy and safety of non-pharmacological interventions for neonatal pain.
Data extraction and synthesis Two reviewers independently extracted the data, assessed the methodological quality using a Measurement Tool to Assess Systematic Reviews (AMSTAR) 2 and graded the evidence quality with the Grading of Recommendations Assessment, Development and Evaluation (GRADE).
Results A total of 29 SRs were included in this overview, of which 28 focused on procedural pain and only 1 focused on postoperative pain. Based on AMSTAR 2, seven reviews were found to be of ‘high quality‘, eight of ‘moderate quality’, five of ‘low quality’ and nine of ‘critically low quality’. The GRADE results suggested that facilitated tucking, kangaroo care, sweet solutions, familiar odour or combined non-pharmacological interventions, such as a combination of sucrose and non-nutritive sucking, were effective and safe in reducing pain from medical procedures in neonates. However, sucrose alone was less effective than local anaesthesia or a combination of the two during circumcision.
Conclusions Facilitated tucking, small volumes of sweet solutions, kangaroo care and familiar odour were recommended. Scientific implementation strategies should be developed to promote the clinical use of these effective non-pharmacological interventions. Meanwhile, further rigorous trials and SRs are needed to identify the best non-pharmacological approaches for pain from common surgery and illnesses in neonates.
PROSPERO registration number CRD42021292583.
- Neonatal intensive & critical care
- PAIN MANAGEMENT
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information. All data relevant to this study are available in the manuscript and its supplementary materials. Any more information can be requested from the corresponding author.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Strengths and limitations of this study
This overview synthesised the latest evidence on non-pharmacological interventions for neonatal pain based on the findings of systematic reviews, which provided more extensive and direct evidence for clinical workers in neonatal intensive care units.
This overview was predesigned following the Cochrane Handbook, which reduced the risk of bias and increased the reliability of our overview.
Some information might have been missed since only English and Chinese studies were included in this overview.
As high-quality evidence emerges, the conclusions of this overview may be affected by newly published literature.
Infants admitted to the neonatal intensive care unit (NICU) experience a high prevalence of painful stimuli. On average, the number of daily acute painful events for hospitalised neonates can reach up to 26,1 and the cumulative time of persistent painful exposure is up to 57.61 hours.2 It is confirmed that both preterm and term infants can recognise, process and respond to painful stimuli.3 Neonatal pain exposure can induce a series of neurophysiological and behavioural changes, associated with adverse long-term effects, such as feeding difficulties, hyperalgesia, chronic metabolic diseases2 4 and even poorer cognitive scores, motor ability, and behavioural control ability in childhood.5
Neonatal caregivers should fulfil the obligation to provide newborns with analgesic treatment, given the well-established harmful impact of painful experiences in early life.6 Although both non-pharmacological and pharmacological methods can be used to alleviate pain and suffering in neonates, non-pharmacological approaches are recommended as the first-line treatment according to guidelines of neonatal pain management.7 8 Non-pharmacological analgesia is preferred not only because it is ethical but also because of its high benefit–risk ratio. Non-pharmacological therapies,9 comprising more than a dozen strategies such as non-nutritive sucking (NNS), sweet solutions, breast feeding, kangaroo care (KC) and music therapy, can reduce neonatal pain directly by blocking the transmission of nociception or activating descending inhibitory pathways, and indirectly by reducing the total amount of nociceptive stimuli to which infants are exposed.10 Furthermore, they also show greater advantages on account of their low risk and lack of side effects, ease of implementation, low cost, and nurse-friendliness.
Systematic reviews (SRs), considered as high-quality evidence, have been increasingly developed to investigate the efficacy and safety of non-pharmacological interventions for neonatal pain. However, a large amount of information can make it difficult for clinicians in NICUs to make decisions across various analgesic interventions rapidly.11 It is also unlikely to assess the efficacy of all non-pharmacological strategies for various painful stimuli in only one systematic review owing to time and resource limitations. For example, a Cochrane SR has comprehensively evaluated the efficacy of multiple non-pharmacological interventions for acute procedural pain in neonates.12 However, it still does not include all types of non-pharmacological treatments, such as KC and music therapy, or all types of pain, such as postoperative pain and persistent pain. Furthermore, when conducting SRs, the reliability of their findings is susceptible to various risks of bias.13 The decision-makers would thus be misled if SRs were recommended for clinical practice without rigorous quality evaluation. Therefore, it is necessary to assess the quality of SRs and aggregate high-quality evidence to provide direct guidance for pain management practices in NICUs.
An overview is a comprehensive approach to summarise evidence from multiple SRs for a particular health condition in one document through a systematic literature search and strict quality assessment.14 Therefore, we conducted an overview to present comprehensive evidence on the efficacy and safety of non-pharmacological interventions for neonatal pain, which can aid evidence-based clinical decision-making and highlight current gaps in knowledge.
Methods and analysis
This overview was conducted and reported following the Cochrane Handbook for Systematic Reviews of Interventions15 and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement16 separately (for the PRISMA checklist, see online supplemental appendix 1). The study protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO, https://www.crd.york.ac.uk/prospero/). We developed this overview by following a predetermined protocol. Significantly, we reported pain assessment and non-pharmacological and pharmacological treatments separately. This study provides an overview of the efficacy and safety of non-pharmacological interventions for neonatal pain, while the other two reports are currently under review.
Inclusion and exclusion criteria
The inclusion criteria were (1) types of participants: preterm or term neonates who underwent one or more painful stimuli during their hospital stay in the NICU; (2) types of interventions for pain relief: non-pharmacological therapies, including but not limited to sucrose, glucose, breast feeding, NNS, KC, swaddling, music therapy and touch.9 Actually, whether sucrose and glucose are non-pharmacological or pharmacological analgesics is still controversial. We included sucrose and glucose as non-pharmacological treatments in this study according to the recommendations of the guidelines17; (3) types of outcomes: pain scores measured by a validated scale and incidence of adverse reactions were primary outcomes18; physiological, biochemical or behavioural indicators (eg, heart rate, respiratory rate, a saturation of peripheral oxygen in the blood, cortisol levels, cry duration and the proportion of time crying) were secondary outcomes; (4) types of reviews: SRs or meta-analyses in English or Chinese in which at least one randomised controlled trial (RCT) or non-randomised controlled studies were included accordingly. The exclusion criteria were as follows: SRs whose neonates’ data could not be extracted, duplicated publications, protocols of overviews, review comments, and conference abstracts.
An electronic literature search was conducted via PubMed, Embase, Cochrane Library, Web of Science, CINAHL, Chinese National Knowledge Infrastructure (CNKI), Chinese Biomedical Literature Database (CBM), Wanfang Database and Chinese Science and Technology Periodical Database (VIP). The search strategies were developed in collaboration with a librarian specialising in medical literature retrieval. The details of the search strategies with free-text words and subject headings are presented in online supplemental appendix 2. The reference lists of included studies and Google Scholar were screened for additional relevant SRs. The WHO Handbook for Guideline Development pointed out that guideline recommendations need to be based on the best available evidence.19 Moreover, evidence from high-quality SRs would be obsolete 3–5 years after publication.20 Thus, we limited the search period to the last 5 years, from November 2016 to November 2021.
Study selection and data collection
The reviewer (QS) conducted a comprehensive search according to a predeveloped standardised search strategy. After removing duplicate records with EndNote V.X9 (Beijing, China; Clarivate), two reviewers (QS and ZH) independently screened for candidates according to the prespecified selection criteria by reading the titles and abstracts. Full texts were retrieved for further screening. Finally, bibliographical references of the included studies were reviewed to identify possible SRs. Any disagreements after cross-checking were resolved by discussion or consultation with a third reviewer (HL) if consensus was not reached among the designated two reviewers.
Data were independently extracted by two reviewers (QS and ZH) using a predefined spreadsheet in Microsoft Excel 2019, including authors, review title, year of publication, number of studies included, analgesic interventions, outcomes, quality evaluation method and conclusion. The extracted data were cross-checked by two reviewers (QS and ZH) to eliminate input errors. Differences were resolved through mutual discussion and consensus.
Quality assessment and strength of evidence
Two qualified reviewers (QS and ZH) trained in the Fudan University Center for Evidence-based Nursing (A JBI Centre of Excellence) independently evaluated the methodological quality of the included SRs using a Measurement Tool to Assess Systematic Reviews (AMSTAR) 2.13 Then, the results were cross-checked, and disagreements were resolved by group discussion or were arbitrated by a third reviewer (XL). The checklist consisted of 16 items, 6 of which were identified as critical domains (Item 4, 7, 9, 11, 13 and 15) based on the AMSTAR 2 guideline and group discussions.13 Each item was evaluated as ‘yes’, ‘partial yes’, ‘no’ and ‘no meta-analysis conducted’ according to compliance with the standard. The methodological quality of SRs was determined by weaknesses in the critical domains instead of generating an overall score. The overall quality was categorised as ‘high (no or one non-critical weakness)’, ‘moderate (more than one non-critical weakness)’, ‘low (one critical flaw with or without non-critical weaknesses)’ and ‘critically low (more than one critical flaw with or without non-critical weaknesses)’. Systematic reviews of moderate or high quality can provide an accurate and comprehensive summary of available studies.
The strength of evidence for all outcomes was assessed by the reviewer (QS) based on the Grading of Recommendations Assessment, Development and Evaluation (GRADE).21 Then, the evaluation results were reviewed for correctness by another qualified reviewer (XL) who was trained in the Lanzhou University GRADE Center (China). If the GRADE system was applied in a systematic review, it was also graded by the authors. The overall quality of evidence was rated as follows: ‘high’, ‘moderate’, ‘low’ and ‘very low’. Evidence based on RCTs began as high quality and otherwise began with low quality. The evidence quality was downgraded one level for serious or two levels for very serious limitations if there was: risk of bias, inconsistency across studies, indirectness of evidence, imprecision of estimates and publication bias.22 A web version of the GRADE profiler Guideline Development Tool (GRADEpro GDT, https://www.gradepro.org/) was used to create a summary of findings to report the quality of the evidence.
A descriptive analysis was performed to synthesise evidence on the efficacy and safety of non-pharmacological treatments for neonatal pain and to clarify the gaps between existing evidence and the clinical practice of pain management in neonates.
Patient and public involvement
No patient or the public was directly involved in the development of this overview of SRs.
This review retrieved 331 records in total. After removing 116 duplicates, the titles and abstracts of 215 papers were reviewed for eligibility. A total of 172 articles were excluded, leaving 43 full-text screening of which 29 articles met the inclusion criteria and were included in this overview.23–51 Online supplemental appendix 3 table 1 lists the reasons for the exclusion of 14 studies. Figure 1 shows a PRISMA diagram of the literature selection.
Characteristics of the included SRs
The characteristics of the included SRs are shown in online supplemental appendix 3 table 2. The number of trials in the SRs ranged from 3 to 168. Eighteen SRs included RCTs only,26–28 31–33 35–39 41 42 44 45 47 49 50 while the other reviews included non-RCTs, case studies and descriptive studies. Regarding the participants, 10 SRs were specific to preterm infants,26 27 31–33 37 39 40 42 43 1 SR was directed at term neonates30 and the other 18 SRs included both term and preterm neonates.23–25 28 29 34–36 38 41 44–51 The types of pain involved in the included SRs consisted of postoperative pain (3.4%), single procedural pain (89.7%) and repeated procedural pain (6.9%). The most frequently performed procedures were heel lance (26.9%), venipuncture (16.7%), injection (11.5%), eye examination (9.0%) and suction (9.0%). A meta-analysis was performed in 17 SRs (58.6%). The outcomes of SRs included validated pain scores; physiological, behavioural and hormonal indicators; and adverse events. Eleven pain scores were included as primary outcome measures for analgesic efficacy, and the most commonly used scales were Premature Infant Pain Profile (27.1%), Neonatal Infant Pain Scale (19.8%), Neonatal Facial Coding System (13.5%), Douleur Aiguë du Nouveau-né (9.4%), Bernese Pain Scale for Neonates (5.2%) and Neonatal Pain Agitation and Sedation Scale (5.2%). Furthermore, the incidence of adverse events, such as bradycardia, tachycardia, desaturation, apnoea, nausea, vomiting and hyperglycaemia, was the main indicator for evaluating the safety of non-pharmacological therapies. Non-pharmacological interventions that were demonstrated to be effective and safe for reducing procedural pain in neonates included sweet solutions, NNS, breast feeding, KC, positioning (facilitated tucking or swaddling), maternal voice, music therapy, aromatherapy, acupuncture or a combination of these. Whereas, for the single study of postoperative pain, non-pharmacological interventions alone were shown to be insufficiently analgesic, and pharmacological strategies in addition are required.
The methodological quality of included SRs
Details of the methodological quality of the included SRs are presented in online supplemental appendix 3 table 3. Seven reviews (24.2%) were found to be of ‘high quality’,32 33 39 44 45 49 50 eight reviews (27.6%) of ‘moderate quality’,25 26 29 34 35 46 48 51 five reviews (17.2%) of ‘low quality’23 27 37 38 47 and nine reviews (31.0%) of ‘critically low quality’.22 28 30 31 36 40–43 Only two SRs met all criteria of AMSTAR 2.45 49 Most of the remaining SRs did a good job on Items 1, 3–6, 8, 9, 11–14 and 16, especially did well in clearly describing the PICO (population, intervention, control group and outcome) questions, performing study selection and data extraction in duplicate, presenting the included studies in adequate detail, assessing the risk of bias (ROB) with a satisfactory technique and discussing the likely impact of ROB on the results, providing a satisfactory explanation and discussion of any heterogeneity observed in the SRs, and reporting potential sources of conflict of interest. However, most studies showed shortcomings in explicitly stating that the reviews were conducted following a well-developed protocol and reported any significant deviations from the protocol, providing a list of excluded studies with justification, reporting the funding sources for the studies included in the reviews, and investigating publication bias. A graphical representation of the methodological quality of the included SRs is presented in figure 2.
Quality of evidence in included SRs
The evidence quality of the primary outcomes extracted from included nine meta-analyses32 34 35 39 42 45 46 49 51 of moderate-to-high methodological quality is displayed with downgrading justification in online supplemental appendix 3 table 4. The results of the GRADE system showed that 6 outcomes (9.1%) were rated as high quality, 27 (40.9%) as moderate quality, 30 (45.5%) as low quality and the other 3 (4.5%) as very low quality. The evidence quality was downgraded mainly because of methodological limitations, significant heterogeneity and a small sample size failing to meet the optimal information size.
Efficacy of non-pharmacological interventions for neonatal pain
We summarised the effective non-pharmacological interventions for reducing neonatal pain from the included meta-analyses with relatively high quality and presented them in figure 3. These analgesic interventions were mainly used to reduce pain from heel lance, venipuncture, intramuscular (IM) injection, endotracheal suctioning, retinopathy of prematurity (ROP) examination, gastric tube insertion, bladder catheterisation, circumcision and echocardiography.
Regarding heel lance, there was low-to-high quality evidence that showed a significant reduction in pain scores for neonates in the facilitated tucking (FT),32 skin-to-skin care (SSC)/KC45 or sucrose (20%–33%)49 group compared with routine care or no-treatment control. Sucrose (24%) was superior to breast feeding or laser acupuncture during heel stick.49 The moderate-to-high quality evidence indicated that sucrose (24%–30%), sucrose (24%–33%) and sucrose (20%–25%) were separately effective in reducing pain during venipuncture, ROP examination and IM injection.49 Based on low-to-moderate quality of evidence, FT position could result in statistically significant decreases in pain during endotracheal suctioning,32 while sucrose (24%)49 showed benefits in improving pain scores during gastric tube insertion, bladder catheterisation and echocardiography; combined non-pharmacological interventions, such as sucrose combined with NNS, were more beneficial than either method alone. In addition, SSC, sweet solutions and familiar natural and artificial odours appeared to reduce pain from non-specific painful procedures.34 46 51 The low-quality evidence suggested that sucrose (24%) alone was less effective than an eutectic mixture of local anaesthetics cream or a combination of those two during circumcision.
Safety of non-pharmacological interventions for neonatal pain
Eleven of the included 29 SRs mentioned the adverse events of non-pharmacological interventions in treating neonatal pain.25 33 34 36 39 42 44 45 47 49 50 Most studies on non-pharmacological therapies (eg, facilitated tucking, KC, breast feeding, aromatherapy, acupuncture, sweet solutions or a combination of those) for neonatal pain reported no or minimal adverse effects. Meanwhile, the number of minor adverse events was similar across the groups, suggesting no contribution to adverse events of non-pharmacological interventions.
This overview of SRs aimed to provide comprehensive evidence of non-pharmacological treatments for neonatal pain. Through a systematic search, it was found that the current SRs for non-pharmacological analgesia focused on procedural pain (28 SRs), and rarely on postoperative pain (only 1 SR). Based on AMSTAR 2, 13 out of the 29 SRs were rated as moderate or high quality. The results of the GRADE suggested that facilitated tucking, KC, sweet solutions, familiar odour or combined non-pharmacological interventions, such as a combination of sucrose and NNS, were effective and safe in reducing pain from medical procedures in neonates. Although there is no evidence that non-pharmacological interventions are recommended alone for postoperative pain, we do know that sucrose in combination with local anaesthesia appears to be effective in reducing pain for newborn circumcision. There is sufficient evidence of altered brain development and increased pain sensitivity following repeated exposure to painful procedures during the newborn period.52 Leaving neonates suffering from painful stimuli without intervention is inappropriate, and further placebo-controlled trials are considered unethical.46 Hence, the following non-pharmacological methods were recommended for clinical use and control treatment in clinical trials based on the results of this study.
FT is described as the method where neonates are placed in a flexed position in either lateral, supine or prone positions with parents’ or professionals’ hands on the baby’s hands and feet to control the entire body and provide support.53 It was mentioned in a systematic review that facilitated tucking by parents (FTP) was the best position for procedural pain relief in preterm infants.25 The authors recommended that FTP in side-lying be administered from 15 min before the beginning of the painful procedures to 15 min after the painful stimuli.25 One advantage of FT is that parents or health professionals can easily learn and perform it because of its minimal technical challenges.54 Another advantage is that it is effective for extremely or very premature infants without requiring any other abilities, such as mature sucking ability.55 Moreover, it can be used when newborns are unable to be transferred from the incubator or bed.26
SSC, also known as KC, refers to the way the mother holds a diaper-clad infant upright on her breast at approximately 60°, providing maximum skin-to-skin contact between the baby and mother.45 Studies have shown that KC can not only reduce pain and stabilise the physiological and behavioural responses of the neonates, but it also strengthens the mother–infant bonding.56 57 Furthermore, no significant difference in pain scores was found between the mother and other providers (father and another woman).45 If mothers could not participate in KC, fathers or other women can serve as alternatives. As for the number of minutes in KC, no analyses were conducted to identify its effect on neonatal pain response. The duration of KC before painful procedures varied from 2 min to 3 hours. The most commonly used scheme was 15 or 30 min of KC before and throughout the painful stimuli.45 There was one limitation, however, to be noted for KC. Contextual challenges such as heavy workload and parental anxiety have restrained its utilisation in NICUs.58
One of the most commonly studied interventions is administering sucrose via syringe, dropper, pacifier or any other way for pain relief in neonates.59 The dose and concentration of sucrose, administration time and the method of delivery varied among studies. The optimal timing and volume for sucrose intervention for pain relief in preterm and term neonates have not yet been determined. A Cochrane SR that included 74 trials with 7049 infants recommended that 24% sucrose solution could be used approximately 2 min before the painful stimulus.49 The minimally effective dose of 24% sucrose during a single painful procedure in neonates was found to be 0.1 mL in a multicentre randomised controlled study.60
Olfactive stimulation interventions using a familiar odour refer to exposing infants to either natural odour or artificial odour with habituation during painful procedures.34 The natural odour consists of maternal odour, breast milk odour of a newborn’s mother or other mothers and amniotic fluid odour.61 On the other hand, the artificial odour comprises formula milk, lavender odour and vanilla odour.62 63 The habitation of artificial odour refers to the preliminary exposure to an odour for a while before the painful procedure to improve the effectiveness of the artificial odour.34 And the habitation time spans between 8 and 18 hours.34 Compared with other non-pharmacological treatments, the advantage of olfactive stimulation is that it can be easily implemented anytime during hospitalisation with little preparation or cost and without requiring parental presence. Meanwhile, it can be easily applied with other effective methods, thus demonstrating an enhanced effect.64 The results of SRs included in this study recommended that a pad with maternal breast milk or an artificial odour with a previous period of habituation could be used during painful procedures.34 However, concerning the intervention, the best artificial odours, the method of releasing the odour, the time required for odour habituation and the distance from the odour position to the neonate’s nose are inconclusive.
In addition, this review identified the following critical clinical questions that need to be explored further: (1) What are the effective and safe non-pharmacological interventions for common postoperative pain, and prolonged pain in neonates? Neonatal pain is often classified as acute pain, prolonged pain and particular types of pain such as postoperative pain and mechanical ventilation pain.18 Although acute pain from minor medical procedures such as heel lance is most common during neonatal hospitalisation,65 other types of pain from major procedures, surgery, medical illness and even painful stimulation like postoperative mechanical ventilation also deserve attention. Yet the available evidence is scant on the latter.66 (2)How to adjust non-pharmacological strategies based on pain scores? The guideline or expert consensus for neonatal pain management recommends a stepped analgesic approach based on pain assessment results with prevention as the premise and non-drug treatment as the mainstay.7 8 However, current evidence focuses on preventive interventions for specific painful procedures. The use of pain scores to guide the clinical selection of appropriate non-pharmacological interventions is still lacking. More research is urgently needed to identify the best non-pharmacological methods for various painful stimuli and to establish a stepped strategy. Finally, scientific implementation strategies are required to effectively translate research evidence on non-pharmacological interventions into practice.
The results of this overview could serve as guidance for clinicians and researchers in the selection of suitable non-pharmacological interventions for neonatal pain. However, there are several limitations to be noted in this study. First, there might be some missing information because only studies in English and Chinese were included in this overview. Second, because this overview was primarily based on published studies, its conclusions may be influenced as new evidence emerges.
This was the first overview to comprehensively assess published SRs to investigate the efficacy and safety of non-pharmacological interventions for neonatal pain. It was concluded that small volumes of sweet solutions, facilitated tucking, KC, familiar odour or combined non-pharmacological interventions, such as a combination of sucrose and NNS were superior in reducing pain from medical procedures in hospitalised neonates. However, sucrose alone was less effective than local anaesthesia or a combination of the two during circumcision. Next, we need to explore scientific implementation strategies to promote the clinical application of these effective non-pharmacological interventions. Moreover, further rigorous trials and SRs are needed to identify the best non-pharmacological approaches for various types of pain, especially pain from common surgery and medical illnesses in neonates. Meanwhile, a stepped non-pharmacological intervention strategy based on pain scores should be explored and established accordingly.
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information. All data relevant to this study are available in the manuscript and its supplementary materials. Any more information can be requested from the corresponding author.
We would like to thank Professor Rong Ou and Professor Xiaorong Hou from the College of Medical Informatics, Chongqing Medical University (Chongqing, China) for peer-reviewing the retrieval strategies; Shan He, Yuting Yao, Songlin Ke, Yating Zhou and Lin Wang for article selection. Many thanks to Editage for helping to improve the quality of the English throughout our manuscript.
Contributors QS was the major researcher and involved in the design, implementation, analysis and reporting of this work. XZ served as the study guarantor supervising all study activities. QS, ZH, HL and XL selected articles, extracted data, performed the assessment of bias and graded the evidence. QS drafted the manuscript. All authors critically reviewed and revised the manuscript and approved the final manuscript.
Funding This study was supported by grants from the National Natural Science Foundation of China (No. 72074038) and the Chongqing Science and Technology Commission (No. cstc2019jscx-msxmX0157; No. cstc2022ycjh-bgzxm0275).
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.