Intralesional steroid is beneficial in benign refractory esophageal strictures: A meta-analysis

Table 1 Main characteristics of the studies included

Study	Study design	Country	Parameter	Patients		Etiology of BRES	Follow-up (mo)	Complication
				Cases	Control			Cases	Control
Kochhar et al[13] 1999	Crossover	India	PDI	14	14	Mixed	23	1	0
Kochhar et al[14] 2002	Crossover	India	PDI	71	71	Mixed	59	0	0
Ahn et al[16] 2015	Crossover	New Zealand	PDI	25	25	Mixed	90	0	0
Nijhawan et al[16] 2016	Crossover	India	PDI	11	11	Corrosive	18	0	0
Dunne et al[17] 1999	RCT	United States	TNRD, DS	20	22	Mixed	60	0	0
Altintas et al[18] 2004	RCT	Turkey	TNRD	11	10	Mixed	48	1	1
Orive-Calzada et al[20] 2012	Cohort	Spain	TNRD	14	9	Mixed	45	0	1
Hirdes et al[19] 2013	RCT	Netherland	TNRD, DS	31	29	Anastomotic	33	5	1
Pereira-Lima et al[21] 2015	RCT	Brazil	TNRD, DS	9	10	Mixed	13	0	0
Camargo et al[22] 2003	RCT	Brazil	DS	7	7	Mixed	12	0	0
Rupp et al[23] 1995	RCT	United States	DS	22	21	Mixed	11	0	0

PDI: Periodic dilation index; NRD: Total number of repeat dilations; DS: Dysphagia score; RCT: Randomized controlled trial; BRES: Benign refractory esophageal stricture.

Table 2 Results of the Newcastle-Ottawa quality assessment scale for cross-over and cohort studies

	Item 1	Item 2	Item 3	Item 4	Item 5	Item 6	Item 7	Item 8
Ahn et al[21] 2015	-	N/A	+	+	-	?	+	?	Modified NOS
Kochhar et al[21] 2015	-	N/A	+	+	-	?	+	+	Modified NOS
Kochhar et al[21] 2015	+	N/A	+	+	-	?	+	+	Modified NOS
Nijhawan et al[21] 2015	-	N/A	+	+	-	-	+	+	Modified NOS
Orive-Calza et al[21] 2015	-	-	+	?	+	+	+	?	NOS

Item 1: Were the cases randomly selected subjects with BRES without significant exclusion criteria? Item 2: Were the controls randomly selected subjects with BRES without significant exclusion criteria? Item 3: Was there an endoscopic or radiological diagnosis of BRES? Item 4: Was the diagnosis of non-refractory BES excluded? Item 5: Were the cases and controls comparable? Item 6: Were the subjects and investigators blinded to the intralesional steroid treatment? Item 7: Was follow-up long enough (≥ 6 mo) for outcomes to occur? Item 8: Was there complete follow up of all subjects enrolled? For the above detailed items an answer of yes represented low risk, no represented high risk, while lack of description represented unknown risk of bias (- = high risk of bias; ? = unknown or moderate risk of bias; + = low risk of bias). BRES: Benign refractory esophageal stricture; BES: Benign esophageal stricture.

Table 3 Results of the quality assessment of randomized controlled trials by the JADAD scoring system

	Item 1	Item 2	Item 3	Item 4	Item 5	Overall	Quality
Dunne et al[17] 1999	1	-1	0	0	0	0	Low; 0
Altintas et al[18] 2004	1	-1	0	0	0	0	Low; 0
Hirdes et al[19] 2013	1	1	1	1	1	5	High; 5
Pereira-Lima et al[21] 2015	1	1	1	1	1	5	High ,5
Camargo et al[21] 2003	1	-1	1	-1	1	1	Low; 1
Rupp et al[21] 1995	1	-1	0	0	0	0	Low; 0

Item 1: Was the study described as randomized? (Yes = 1 point, No = 0 point); Item 2: Was the randomization scheme described and appropriate? (Yes = 1 point, No = -1 point); Item 3: Was the study described as double-blind? (Yes = 1 point, No = 0 point); Item 4: Was the method of double blinding appropriate? (Yes = 1 point, No = -1 point, if the answer of Item 3 was No, Item 4 is not calculable); Item 5: Was there a description of dropouts and withdrawals? (Yes = 1 point, No = 0 point). Low range of quality: 3 >, high range of quality: 2 <.

Table 4 Summary of findings

Outcomes	Intervention values	Control values	Number of patients	Quality of evidence (GRADE)	Comments
PDI	0.335/mo	1.355/mo	121	Very low	Only studies with cross-over design were analyzed
	MD: -1.12
	95%CI: -1.99 to -0.25
	P = 0.012
TNRD	n/a	n/a	165	Very low	Different length of follow up results in high risk of bias
	MD: -1.17
	95%CI: -0.24 to 0.05
	P = 0.057
DS	n/a	n/a	178	Very low	Different scoring scales were used and different lengths of follow up result in high risk of bias
	SMD: 0.35
	95%CI: -0.38 to 1.08
	P = 0.351

PDI: Periodic dilation index; TNRD: Total number of repeat dilations; DS: Dysphagia score; MD: Mean difference; SMD: Standardized mean difference.