Table 2. Subanalysis: rectal NSAIDs vs. placebo in reducing the number of pancreatitis in subgroup of patients with a risk factor.

	Subanalysis	Studies [ref.]	Patients	Risk Ratio (M-H-Fixed-95% CI)	I2
Age
	Young patients	2 [12], [18]	527	0.50 (0.30–0.83)	0%
	Old patientsa	2 [12], [18]	570	0.57 (0.32, 1.05)	0%
Sex
	Females	2 [12], [18]	513	0.54 (0.34–0.85)	0%
	Males	2 [12], [18]	327	0.54 (0.23, 1.24)	8%
Sphincterotomy
	Yes	4 [12], [15], [16], [18]	786	0.53 (0.35–0.83)	0%
	No	4 [12], [15], [16], [18]	572	0.41 (0.23–0.72)	0%
Suspected or confirmed SOH
	Yes	2 [12], [15]	662	0.50 (0.31–0.83)	10%
	No	2 [12], [15]	274	0.36 (0.17–0.77)	0%
Pancreatic duct injection
	Yes	3 [13], [15], [18]	454	0.29 (0.13–0.63)	0%
	No	1 [18]	348	0.62 (0.21–1.80)	-
Prophylactic pancreatic stentb
	Yes	1 [12]	496	0.61 (0.38–0.98)	-
	No	5 [11], [12], [16]–[18]	960	0.45 (0.28–0.70)	0%

SOH: sphincter of Oddi hypertension; ERCP: endoscopic retrograde cholangiopancreatography; NSAIDs: non-steroidal anti-inflammatory drugs.

^aThe two different studies used different cut-off values to separate young and old patients, one using 45 years and the other 60 years.

^bProphylactic measure, not risk factor.