Small-bowel capsule endoscopy: A ten-point contemporary review

Table 2 Head-to-head trials of small-bowel capsule endoscopy systems

Ref.	Country	Centre	Objective(s)	Study type	Design	CE type	Outcome(s)	Conclusion
Hart-mann et al[7]	Germany	Single centre	Head-to-head evaluation of technical performance and DY of two CE systems (PillCam®SB vs EndoCapsule®)	Prospective	►OGIB pts;	►PillCam®SB (Given®Imaging, Yoqneam, Israel);	►Pts enrolled: 40;	►Statistically non-significant trend for EndoCapsule® to detect more bleeding sources in pts with suspected small-bowel bleeding than PillCam®SB;
					►Pts randomized to undergo 2 CEs using different CE in random order		►CR: PillCam®SB 33/40 (82%); EndoCapsule® 40/40 (100%); P = NS;
							►Overall DY: PillCam®SB 26/50 (52%); EndoCapsule® 29/50 (58%); P = NS;
						►Pts randomized to undergo 2 CEs using different CE in random order	►DY (SB P2): PillCam®SB 22/50 (44%), EndoCapsule® 25/50 (50%), P = NS;
							►In all discordant SB P2findings (not detected by the PillCam®SB but detected by EndoCapsule®), PillCam®SB examinations were incomplete	►This is (likely) due to the longer recording time with EndoCapsule®
Cave et al[8]	United States	Multi-centre (4 centres)	Comparison of performance (DY in pts with OGIB): EndoCapsule®vs PillCam®SB	Prospective	►OGIB pts;	►EndoCapsule® (Olympus© America, Allentown, PA);	►Pts with OGIB (transfused or with haematocrit < 31%) (males) or < 28% (females): 63;	►Both devices are safe and have comparable DY within the previously reported range;
					►EndoCapsule® and PillCam®SB swallowed by each participant 40 min apart;		►Available data 51/63; 9 pts excluded for technical reasons + 3 pts for protocol violation;
							►24 videos read as normal, 14 as abnormal (from both CEs). Disagreement occurred in 13;	►Subjective difference in image quality favouring the EndoCapsule®;
					►Ingestion of CEs in randomized order;	►PillCam®SB (Given®Imaging, Yoqneam, Israel)	►No adverse events reported for either CE. Overall agreement: 38/51 (74.5%), κ = 0.48, P = 0.008;
					►Head-to-head comparison of CEs		Limitations: Although ingestion randomized, videos reading not blind (different shape of the image margin)	►Lack of electromechanical interference between 2 different CE
Kim et al[9]	South Korea	Single centre	Head-to-head evaluation of technical performance DY and of two capsule systems (PillCam®SB vs MiroCam®)	Prospective	►Pts referred to CE for various indications;	►MiroCam® (IntroMedic Co. Ltd., Seoul, South Korea);	►Pts enrolled: 24;	►MiroCam shows a longer operating time and a higher CR;
					►Each pt was randomly assigned to swallow 1 of 2 CEs, the second CE was swallowed once fluoroscopy indicated that first CE had reached the SB		►Mean operating time: MiroCam® 702 min; PillCam®SB 446 min, P < 0.001;
							►CR: MiroCam® 20/24 (83%); PillCam®SB 14/24 (59%), P = 0.031;	►Nevertheless, the 2 capsule systems showed comparable efficiency;
						►PillCam®SB (Given®Imaging, Yoqneam, Israel)	►DY: MiroCam® 11/24 (45.8%); PillCam®SB 10/24 (41.7%), P = 1.0;
							►DY (additive of both capsules): 12/24 (50%);	►Sequential capsule endoscopy with the MiroCam and PillCam SB produced slight (but NS) increase in DY
							►Concordance of findings among the two capsule systems 87.5%, κ = 0.74
Pioche et al[10]	France	Multi-centre	Head-to-head evaluation of the diagnostic concordance (κ value): PillCam®SB SB2 vs MiroCam®	Prospective	►OGIB pts;	►MiroCam®; (IntroMedic Co. Ltd., Seoul, South Korea);	►83 pts; drop-outs explained (10 technical issues), 73 pts/videos analysed;	►MiroCam® showed a slightly higher DY, difference not statistically significant;
					►Each pt ingested 2 CEs at a 1 h interval in a random order;		►31 concordant (-) ve cases (42.4%) and 30 concordant (+) ve cases (41.1%);
							►Satisfactory diagnostic concordance between the 2 systems (κ = 0.66);	►The 2 CE systems showed comparable efficiency for the diagnosis of OGIB
					►Videos read in a random order by 2 experienced (> 200 CEs) readers;	►PillCamSB2 (Given®Imaging, Yoqneam, Israel)	►DY similar among the 2 CE systems(PillCam®SB 2 vs MiroCam®: 46.6% vs 56.2%, respectively; P = 0.02);
					►Image-by-image review of cases of disagreement between the readers was performed by 3 expert readers		►SBTT longer with MiroCam®vs PillCam®SB (mean SBTT: 268 vs 234 min, < 0.05);
							►Reading time longer with MiroCam®vs PillCam®SB (mean reading time 40 vs 23 min, P < 0.05);
							►(+) ve diagnosis obtained in 46.6% vs 56.2% of pts with PillCam®SB2 vs MiroCam®, respectively;
							►PillCam®SB2 vs MiroCam®CEs identified 78.6% vs 95.2% of (+) ve cases, respectively, P = 0.02
Dolak et al[11]	Austria	Single centre	Head-to-head comparison (MiroCam®vs EndoCapsule®) of: CR of SB examinations, DY in SB disease	Prospective	►Pts referred to CE for various indications;	►MiroCam® (IntroMedic Co. Ltd., Seoul, South Korea);	►Pts enrolled: 50;	►The two capsule endoscopy systems were not statistically different with regards to CR and DY;
					►Each pt was randomly assigned to swallow either MiroCam® first, followed by the EndoCapsule® 2 h later, or vice versa;		►CR: MiroCam® 48/50 (96%) vs EndoCapsule® 45/50 (90%); P = 0.38;
							►DY in SB: MiroCam® 25/50 (50%) vs EndoCapsule® 24/50 (48%); P > 0.99;
						►EndoCapsule® (Olympus America, Allentown, PA)	►Concordance of findings among the two CE systems: 68%; κ = 0.50	►Moderate concordance, mainly caused by missed pathological findings (which affected both devices), needs consideration in clinical practice
					►All videos analysed by two investigators independently

DY: Diagnostic yield; CE: Capsule endoscopy; OGIB: Obscure gastrointestinal bleeding; pts: Patients; CR: Completion rate; NS: Not significant (statistically); SB: Small-bowel; P2: Refers to grading of angioectasias; SBTT: Small-bowel transit time.