Roel Bogie

Chapter 3

SMI rate in LST-NG/LST-G Author and year

OR [95% CI]

Kudo (2015) Zhao (2014) Kim (2013) Nakae (2012)

1.75 [0.73 , 4.23] 1.39 [0.56 , 3,47] 4.70 [1.44 , 15.41] 0.81 [0.26 , 2.49] 1.55 [0.41 , 5.84] 1.17 [0.13 , 10.33] 1.98 [0.57 , 6.85] 1.01 [0.16 , 6.56] 1.34 [0.88 , 2.03] 6.30 [1.71 , 23.16] 1.93 [0.29 , 12.61] 1.07 [0.06 , 20.69] 0.58 [0.01 , 33.50] 2.69 [1.57 , 4.60] 2.35 [1.29 , 4.27] 12.89 [0.65 , 254.19] 1.89 [1.48 , 2.42]

Kim (2012)

Rotondano (2011)

Kim (2011) Kaku (2011) Oka (2009) Huang (2009) Chiu (2009) Urban (2008) Tantau (2008) Kudo (2008) Uraoka (2006) Teixeira (1996)

Total:

I 2 = 4.8%

12 8 OR SMI in LST-NG vs LST-G 0 4

16

Figure 3.9: Forest plots showing pooled odds ratio (OR) of a LST containing submucosal invasion in the non-granular vs granular subtype.

A total of 24 studies examined the prevalence of LSTs by granular vs non-granular endoscopic subtype. Pooled prevalence of granular LSTs was 61% (95% CI: 54 – 68, I 2 : 96.8%, Figure 3.7 ). Eight studies examined the distribution of the four endoscopic LST subtypes. 1, 19, 26, 30, 32, 50, 54, 57 The pooled prevalences of homogenous granular, nodular mixed granular, flat elevated non-granular and pseudo-depressed non-granular LST subtypes were 35.4% (95% CI 27.2 – 43.6, I 2: 96.0%), 26.1% (95% CI 18.5 – 33.8, I 2: 97.0%), 33.0% (95% CI: 22.8 – 43.2, I 2: 97.5%) and 5.5% (95% CI: 3.2 – 7.8, I 2 : 91.8%), respectively.

60