All steps of this systematic review will be performed according to the Cochrane Handbook (5.2.0). Preferred Reporting Items for Systematic Reviews and Meta-analyses Protocols (PRISMA-P) 2015 provide a guideline for the report of this protocol. [9]
Date sources
The published electronic literature will be searched in Embase, PubMed, and the Cochrane Library, Web of Science, China National Knowledge Infrastructure, Chinese Biomedical Literature Database, Wanfang Database (Wanfang) , Chinese Scientific and Journal Database (VIP). All the literature will be searched range from the establishment to February 2021.
The search strategy is developed according to published reviews.[12-17] The detail search strategy of MEDLINE (by PubMed) is listed in table 1, according to other different databases, the search strategy will be modified.
Table 1 Search strategy in PubMed
#1. “Constipation”[MeSH Terms]
|
#2. constipation[Title/Abstract]
|
#3. functional constipation[Title/Abstract]
|
#4. chronic constipation[Title/Abstract]
|
#5. dyschezia[Title/Abstract]
|
#6. colonic Inertia[Title/Abstract]
|
#7. idiopathic constipation[Title/Abstract]
|
#8. slow transit constipation[Title/Abstract]
|
#9. chronic functional constipation[Title/Abstract]
|
#10. Gastrointestinal disorder[Title/Abstract]
|
#11. Bowel disorder[Title/Abstract]
|
#12. Primary constipation[Title/Abstract]
|
#13. OR/#1-12
|
#14. “Massage”[MeSH Terms]
|
#15. “Acupressure”[MeSH Terms]
|
#16. “Manipulation, Chiropractic”[MeSH Terms]
|
#17. “Therapy, Soft Tissue”[MeSH Terms]
|
#18. “Musculoskeletal Manipulations”[MeSH Terms]
|
#19. Tuina[Title/Abstract]
|
#20. Massage Therap*[Title/Abstract]
|
#21. Zone Therap*[Title/Abstract]
|
#22. Rub[Title/Abstract]
|
#23. Massage*[Title/Abstract]
|
#24. Manipulation, Chiropractic[Title/Abstract]
|
#25. Therapy, Soft Tissue[Title/Abstract]
|
#26. Therapeutic Touch[Title/Abstract]
|
#27. Reflexotherapy[Title/Abstract]
|
#28. Rolfing[Title/Abstract]
|
#29. Reflexology[Title/Abstract]
|
#30. Manual* Therap*[Title/Abstract]
|
#31. Acupressure[Title/Abstract]
|
#32. massotherapy [Title/Abstract]
|
#33. OR/#14-32
|
#34. (#13 AND #33 )
|
Eligibility criteria
Study design
We will include RCTs which could avoid most bias.[10] Quasi-randomized RCTs, such as those allocating by registration order, which will be excluded. The language will limited to studies published in Chinese or English. There is no restriction on publication status.
Participants
The age of participants was defined as between 0 and 18 years old. All participants diagnosed with functional constipation according to the Roma Ⅲ or IV diagnostic criteria and clinical symptoms. There are no restrictions in ethnic distribution and gender.
Interventions
Intervention for treatment group was defined as using massage and interventions of massage combined with conventional treatment will be included.
Comparators
Intervention for control group was defined as received conventional treatment, which include ensure normal fiber intake, fluid intake and exercise, toilet training, PEG, lactulose,[4] and microecologics.[11]
Outcomes
The primary outcomes are 1) Overall response rate (according to stool frequency, time of defecation interval, stool consistency, difficult defecation, sense of incomplete evacuation, abdominal pain and bloating); 2) The mean number of bowel movements in the fourth week.
Secondary outcome measures include: 1) The symptom integrals (e.g. perceived ease of defecation, Bristol Stool Form Scale (BSFS), defecation time, abdominal pain or distension, stool consistency); 2) Adverse events; 3) Recurrence rate.
We defined the outcome assessment time point as the end of all treatment.
Exclusion criteria
Studies meet one of the following will be excluded: 1) Repeatedly published studies; 2) Studies of experiences, systematic reviews, conference papers, letters and animal experiments; 3) Massage was not only in the experimental group but also in the control group, comparison between massage and other massage-likeness therapies will be excluded; 4) Patients treated by massage or conventional treatment in addition to other treatments, such as Acupoint application, Oral traditional Chinese medicine, and Acupuncture; 5) Articles without full text or only with abstract or with data which are missed or cannot be used; 6) Participants were receiving pharmacological treatment in other study; 7) Participants were diagnosed with constipation due to structural or metabolic diseases; 8) Patients suffered from functional constipation with other diseases.
Data collection and analysis
Selection of literature
Two authors will independently identify studies according to the inclusion and exclusion criteria. First, duplicate researches will be eliminated by using NoteExpress software. Second, screening the title and abstract, if necessary, reading the full article to examine if it should be included. NoteExpress software will also be used to manage the included studies. If there is disagreement during the screening process, discuss with the third author to make a decision. The selection process is performed as shown in figure 1.
Data extraction and management
Before data extraction, all the reviewers will discuss and develop a standardized data extraction form. Data from the included studies will be extracted by two authors independently. In multi-arm RCTs, we will extract data from two arms. We will select the group as the treatment group which contains the treatment of massage, and choose another group as the control group which the treatment was without massage.
The following information will be collected: 1) general information (title, publication year, authors, country, language, funding and journal source); 2) details of study (aim, design, method of randomization and allocation, inclusion and exclusion criteria); 3) patient characteristics (age, sex, sample size, diagnosis criteria, response criteria, the syndrome differentiation of TCM); 4) interventions and controls (type, duration and follow-up time) ; 5) outcome information which including primary outcome, second outcome and adverse effects. All information will be summarized in the predetermined structured data form.
When extraction finished, data will be checked with each other by the 2 reviewers. Any discrepancies should be resolved by negotiation between the 2 reviewers with the help of a third reviewer.
Assessment of risk of bias in included studies
Two authors will independently evaluate the risk of bias by using the Cochrane Collaboration bias risk assessment tool. The two authors will assess the risk of bias of sequence generation, allocation concealment, blinding of participants personnel and outcome assessment, selective outcome reporting, incomplete outcome data and other bias. The reviewers will categorize the evaluation into three levels of bias: low, high and unclear risk of bias.
Measures of treatment effect
For continuous outcome data, we calculated the mean difference (MD) and corresponding 95% confidence interval (CI). For the secondary dichotomous outcomes, we calculated the risk ratio (RR) and corresponding 95% CI.
Dealing with missing data
When the included article lacks some important information, we will try to contact the correspondence author through phone, e-mail or other contacts to get original data for analysis. If we can't get the information through the ways above, we will to evaluate the potential influence of missing data by the following strategies: 1) Worst-case situation analysis: all participants with missing data regarded as failures; 2) Extreme worst-case/best-case situation analysis: participants with missing outcome data in the intervention group regarded as failures and in the control group as success and vice versa.
Assessment of heterogeneity
We will evaluate heterogeneity using the standard χ2 test (α=0.1) and I2 test. The fixed effects model will be used if the p≥0.1 and if I2≤50%. If the p<0.1 or the I2>50%, Random-effects models will be used. We will perform a subgroup analysis to explore the possible causes. If the heterogeneity is more than 75%, a meta-analysis will not be performed. We will provide a narrative, qualitative summary.
Sensitivity analysis
We will perform the sensitivity analysis to appraise the reliability of the pooled results. First, we will evaluate the impact of including studies with a high risk of bias on the results of this review. We will combine all the included studies, and find out if the results are still consistent after excluding the studies with high risk of bias. Second, to clarify whether different models affect the results of data synthesis, we will combine the outcomes using fixed and random effects models, and check whether the results remain the same.
Assessment of reporting bias
We will use funnel plots to demonstrate reporting biases. If 10 or more studies are included in a meta-analysis, we will use Egger’s method to assess funnel plot asymmetry.
Data synthesis
The meta-analysis of intervention and outcome measures methods will be conducted by RevMan (Version 5.3.5) software. In order to estimate intervention effect, we will combine more than one trial which examine the same intervention and outcomes with comparable methods in similar populations. We will use the fixed-effect model to combine the data, when the statistical heterogeneity is low. While if the statistical heterogeneity is moderate, we will use the random-effect model. However, if the heterogeneity level much significant, descriptive analysis or subgroup analysis will be performed. We will combine RR of each study and calculate 95% CI for dichotomous data using fixed effects model, if no heterogeneity is detected. If significant heterogeneity is found, we will combine the data using random effects model and prudently explain the results. Moreover, we will provide a p value for a comparison of massage and conventional treatment. For continuous data, we will combine the MD of each study and compute 95% CI if the same outcome measurement is used; if not, we will combine the SMD instead.
Subgroup analysis and investigation of heterogeneity
When sufficient data are available, subgroup analysis will be performed to investigate heterogeneity. We will conduct subgroup analyses based on age, different types of massage, different types of therapeutic effect criterion, different types of control group (basic treatment, medication, microecologics). We will use the χ2 test to analyze the intervention effect, if the p<0.05, indicating statistically significant differences between subgroups.
Grading of evidence quality
We will use the Grading of Recommendations Assessment, Development and Evaluation to access the quality in cumulative evidence.[18] Risk of bias, imprecision, indirectness, heterogeneity and publication bias will be assessed, and the results will fall into four levels: high, moderate, low and very low.