A Network Meta-analysis Of The Treatment Of Chronic Non-specific Low Back Pain By External Therapy Of Traditional Chinese Medicine Based On GRADE Rating

Objective:To rank external treatments in traditional chinese medicine and provide new idea for curing CNLBP,this study conduct a Network Meta-analysis(NMA)of 8 external treatment in traditional chinese medicine(tui-na,acupuncture,moxibusion,gua-sha,fireneedle,electroacupunture,warming acupuncture)with GRADE approach,by assessing the outcome of VAS,ODI,RMDQ and efficiency.Hoping this study could provide new evidence in evidence-based medicine.Methods:The literature was searched to August 2020 of electronic databases including CNKI,WanFang,VIP,PubMed,Embase and The Cochrane Library by computer.Two independent reviewers extracted data from studies after selected RCTs with the criteria of inclusion and exclusion.Determined the quality of including studies and risk of bias by Jadad AR and The Cochrane Collaboration’s tool for assessing risk of bias in RCTs.Using Statal4.2 to report network map,league table,rankogram and comparison-adjusted funnel plot.The NMA quality was assessed using GARDE approach.Results:(1)10520 records was identified from databases searching,of 55 RCTs(N=5023)met eligibility criteria and were included in the NMA.A total of 12 non-sugical interventions were studied in these RCTs:tui-na,acupuncture,moxibusion,gua-sha,cupping,fire-needle,electroacupunture,warming acupuncture,oral medicine,physiotherapy,conventional treatment,sham acupuncture.(2)According to Jadad AR,37 trials were high quality and 18 trials were low quality studies.34(61.8%)trials mentioned using random table form randomization,21(61.8%)trials conducted allocation concealment,only 1(1.8%)trial reported using blind method,12(21.8%)trials conducted blindness of assessor,54(98.2%)trials reported data completely and all trials doesn’t mention other bias.(3)NMA outcome:①VAS:There were 59 statistically significant comparisons(P<0.05)and the outcome of fire needle was better than other interventions.Bellow were the effect and 95%CI of fire needle VS other interventions:acupuncture[SMD=-1.49,95%CI(-2.70,-0.27)],physiotherapy[SMD=-1.65,95%CI(-2.98,-0.32)],oral medicine[SMD=-1.79,95%CI(-2.92,0.66)],sham acupuncture[SMD=-1.98,95%CI(-3.48,-0.48)],electroacupunture[SMD=2.41,95%CI(-3.90,-0.93)],conventional treatment[SMD=-2.64,95%CI(-3.93,-1.34)].②Efficiency:There were 15 statistically significant comparisons(P<0.05)and fire needle is better than other interventions except gua-sha,bellow were the effect and 95%CI of fire needle VS other interventions:acupuncture[RR=0.2 4,95%CI(0.11,0.51)],medicine[RR=0.19,95%CI(0.10,0.37)],conventional treatment[RR=0.15,95%CI(0.03,0.84)],physiotherapy[RR=0.1 0,95%CI(0.04,0.31)].③ODI:There were 18 statistically significant comparisons(P<0.05)and gua-sha could be the best treatment than other interventions.Bellow were the effect(9 5%CI)of gua-sha VS other interventions:conventional treatment[SMD=-2.34,95%CI(-4.12,-0.56)],oral medicine[SMD=-2.63,95%CI(-4.80,-0.47)].④RMDQ:There were 21 statistically significant comparisons(P<0.05)and fire needle could be the best treatment than other interventions.Bellow were the effect(95%CI)of fire needle VS other interventions:acupuncture[SMD=-3.88,95%CI(4.67,-3.09)],electroacupunture[SMD=-3.86,95%CI(-4.84,-2.88)],tui-na[SMD=-4.00,95%CI(-4.91,-3.10)],oral medicine[SMD=-4.45,95%CI(-5.51,-3.39)],physiotherapy[SMD=-4.69,95%CI(-5.57,-3.82)],conventional treatment[SMD--4.72,95%CI(-5.57,3.86)].(4)The GRADE quality of VAS,efficiency,ODI,RMDQ were very low.All indirect estimates of four outcomes were evaluated and below were the down grading factors:(1)all were indirect estimates.(2)the sample size were too small.(3)the rank of interventions were unstable.(4)the prediction intervals weren’t match the confidence intervals.Conclusion:(1)External treatment in traditional chinese medicine could be better than conventional treatment and oral medicine on the part of reliving pain,efficiency and improving lumbar dysfunction.(2)Fire needle could be the best intervention for CNLBP on pain reducing while gua-sha could be the best treatment on efficiency.(3)The GRADE quality of 4 outcomes were very low,meanwhile the most quality of included trails were also low and the number of samples were small.To get more evidence on the results,it still needs more high quality and large samples of RCTs.