Spiral enteroscopy versus single-balloon enteroscopy for the evaluation and treatment of small bowel disorders: a systematic review and meta-analysis

Abstract

Background/Aims

Device-assisted enteroscopy has advanced small bowel disorder management. We conducted this meta-analysis to compare the clinical and procedural outcomes between spiral enteroscopy and single-balloon enteroscopy.

Methods

A systematic search was performed on December 1, 2024, in the PubMed, Scopus, and Cochrane Library databases to identify studies that compared spiral enteroscopy and single-balloon enteroscopy. The outcomes included diagnostic and therapeutic yields, total procedure time, depth of maximum insertion, and adverse event rates.

Results

Five studies (including 496 patients) met the inclusion criteria. The diagnostic yield was similar between spiral enteroscopy and single-balloon enteroscopy (risk ratio [RR], 1.07; 95% confidence interval [CI], 0.96–1.20; p=0.24). The therapeutic yield also showed no significant difference (RR, 1.10; 95% CI, 0.45–2.69; p=0.83). The total procedure time was comparable (mean difference, –22.85 minutes; 95% CI, –46.83 to 1.12; p=0.06), although motorized spiral enteroscopy reduced the procedure time (p<0.001). Spiral enteroscopy achieved greater depth of maximum insertion (standardized mean difference, 1.33; 95% CI, 0.65–2.01; p<0.001). Adverse event rates were comparable (RR, 1.72; 95% CI, 0.80–3.70; p=0.16).

Conclusions

Spiral and single-balloon enteroscopies demonstrated similar diagnostic and therapeutic yields and safety. Spiral enteroscopy achieved a greater insertion depth, and motorized systems improved the efficiency in terms of procedure times.

Graphical abstract

INTRODUCTION

Small bowel disorders, such as bleeding, tumors, and Crohn’s disease, are associated with significant morbidity and can impact a patient’s quality of life.1 The small bowel’s unique anatomy, length, and relative inaccessibility to traditional endoscopes complicate diagnosis and treatment. Device-assisted enteroscopy (DAE) has improved small bowel evaluation by enabling direct visualization, biopsies, and therapeutic procedures, thereby enhancing diagnostic accuracy and treatment efficacy.2

DAE is available in the form of three platforms: double-balloon enteroscopy (DBE; Fujifilm), introduced by Yamamoto in 2001; single-balloon enteroscopy (SBE; Olympus Medical Systems Corporation), developed in 2007; and spiral enteroscopy (SE; Spirus Medical, LLC), introduced in 2008.3-5

DBE utilizes two balloons, one at the tip of the enteroscope and one on the overtube, to achieve deep intubation of the small bowel through a push-and-pull technique, allowing for the pleating of the intestine over the scope.6 In contrast, SBE employs a single balloon on the overtube, relying on a similar principle, but with less scope stability and maneuverability because of the absence of a second balloon. The balloon anchors the overtube, helping reduce bowel looping and allowing advancement of the enteroscope through sequential inflation and overtube shortening.7

Alternatively, SE utilizes a helical overtube that is manually rotated with a flexible enteroscope, allowing for rapid advancement and stable positioning during a thorough small bowel examination, both during insertion and withdrawal. The procedure can be performed either antegrade via the upper gastrointestinal tract or retrograde via the colon.8 In 2015, the novel motorized spiral enteroscopy (Olympus Medical Systems) was introduced to overcome the limitations of manual SE. It features a longer enteroscope (168 cm vs. 90 cm), shorter overtube (24 cm vs. 118 cm), and larger working channel (3.2 mm) with an integrated water jet system.9

Both SBE and SE are valuable tools for diagnosing and treating small bowel disorders. Previous studies compared SBE and SE; however, their findings lacked sufficient statistical power or consistency in terms of outcomes. Procedural aspects, such as diagnostic yield, therapeutic yield, procedure time, depth of maximal insertion (DMI), and the occurrence of adverse events, may vary between the two techniques, warranting further comparisons to guide clinical decision-making.

This meta-analysis compared the procedural aspects of SE and SBE. Specifically, we aimed to evaluate the differences in diagnostic yield, therapeutic yield, procedure time, DMI, and adverse events between the two modalities.

METHODS

Study design and protocol

This meta-analysis was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and the Cochrane Handbook (9th ed). The PRISMA checklist for this meta-analysis is available in the Supplementary Material 1. The protocol was registered in PROSPERO (CRD420251069513).

Search strategy

The database search was conducted on December 1, 2024, using PubMed, Cochrane Library, and Scopus. To further broaden the search, we manually screened the reference lists of relevant studies identified in PubMed. Keywords related to “spiral enteroscope” and “single-balloon enteroscope” were used in our search string. A detailed description of the search strategy and search term strings is provided in Supplementary Table 1.

Eligibility criteria and primary outcomes

Our inclusion criteria were any observational or clinical trial that compared single-balloon and SE (motorized or manual) in patients with confirmed or suspected small bowel disorders. Case reports, editorials, conference abstracts, and animal studies were excluded.

The primary outcomes of our study were diagnostic yield (defined as the number of positive findings among the patients tested), therapeutic yield (the ability to perform therapeutic interventions during enteroscopy), procedure duration, DMI, and incidence of adverse events.

Adverse events in the included studies were reported as composite outcomes, as defined by the American Society for Gastrointestinal Endoscopy lexicon for reporting endoscopic adverse events.10

Screening and data extraction

The screening process comprised two phases: title and abstract screening, followed by full-text screening. In both phases, two reviewers independently assessed each article using the Covidence platform (Covidence.org). The first authors resolved these discrepancies. Only studies that met the inclusion criteria were included in the analysis.

Data from the selected studies were extracted into a standardized Excel spreadsheet and organized into two sections: an overview of the study (including baseline characteristics) and a review of the outcomes. Continuous data are reported as means with standard deviations or medians with interquartile ranges. Two reviewers independently extracted and revised the data, and a third reviewer performed the final review.

The procedure time and DMI reported by Pal et al.11 were initially reported as medians with ranges. These values were converted to means and standard deviations using the method described by Wan et al.12

Risk of bias assessment

Because of the mixed designs of the included studies, we used two tools. For randomized interventional studies, we used the Cochrane Collaboration’s Risk of Bias 2 (RoB 2) tool.13 Each study was assessed across five domains: randomization process, intervention assignment, missing outcome data, outcome measurement, and selective reporting. Each domain was categorized as ‘low risk,’ ‘high risk,’ or ‘some concern.’

For non-randomized interventional studies, we used the ROBINS-I tool.13 This tool has seven domains: confounding, participant selection, intervention classification, deviations from intended interventions, missing data, outcome measurements, and reporting bias. Each domain was rated as low, moderate, high, or critical risk, and an overall risk of bias was assigned.

Data analyses

We conducted our analyses using Review Manager (RevMan Web; The Cochrane Collaboration). A random-effects model was used. The effect size was calculated using the risk ratio (RR) for binary outcomes and mean difference (MD) for continuous outcomes, both of which were reported with 95% confidence intervals (CIs). The standardized MD (SMD) was used for the DMI because of variations in the criteria used to define the DMI across the studies. These definitions are summarized in Supplementary Table 2.14-16 Statistical significance was set at p<0.05. A sensitivity analysis was performed to assess the influence of each study on the overall effect size for all outcomes.

Heterogeneity was evaluated using the I² statistic and chi-squared test; however, owing to the small number of included studies, the reliability of the heterogeneity statistics might be limited.13 We intended to assess publication bias if the number of included trials approached or exceeded ten. Unfortunately, this was not possible because of the limited number of studies included in our review.

RESULTS

Study selection results and background characteristics

The database search yielded 141 articles. Of these, 11 full-text papers were analyzed for inclusion, and only five papers, with a total of 496 included patients, were deemed eligible for full-text data extraction and analysis.11,14-17 The selection process is illustrated in Figure 1.

Fig. 1.

Overview of the selection process.

Of the five included studies, one was a retrospective cohort study,14 one was a prospective cohort study,11 and three were randomized clinical trials.15-17 Two studies used a manual SE,14,15 while the remaining three used a motorized SE.11,16,17 An overview of the included studies is presented in Table 1.11,14-17

Table 1.

Overview of the included studies

Study	Year	Type of study	Country	Total (n)	No. of procedures		Age (yr)*		Female (n, %)		BMI (kg/m2)*		Prior abdominal or pelvic surgery
					SE	SBE	SE	SBE	SE	SBE	SE	SBE	SE	SBE
Khashab et al.14	2010	Retrospective cohort	USA	92	53 (manual SE)	52	54±17	55±16	23 (45.0)	26 (63.4)	-	-	15 (29.0)	19 (46.0)
Moran et al.15	2018	RCT	USA	30	13 (manual SE)	17	53.1±18.3	55.8±20.3	3 (23.1)	14 (82.4)	26.6 (6.5)	25.1±6.7	2 (15.4)	6 (37.5)
Pal et al.16	2023	RCT	India	125	62 (NMSE)	63	40 (18–65)	45.5 (18–63)	21 (33.9)	24 (38)	22.45 (16.3–34.2)	21.9 (14.2–26.6)	-	-
Pal et al.11	2023	Prospective cohort	India	177	55 (NMSE)	146	40 (16–74)	35 (7–75)	13 (28.2)	53 (40.4)	-	-	-	-
Rughwani et al.17	2023	RCT	India	72	35 (NMSE)	37	52.2±16.7	47.3±18.3	9 (25.7)	15 (40.5)	22.48 (3.5)	21.24±4.8	-	-

BMI, body mass index; SE, spiral enteroscopy; SBE, single balloon enteroscopy; USA, United States of America; RCT, randomized clinical trial; NSME, novel motorized spiral enteroscopy.

*Presented as mean±standard deviation or median (range)

Risk of bias assessment

All randomized clinical trials included in this analysis had some concerns, primarily because the investigators were aware of the intervention, which could potentially have impacted the outcome assessment.15-17 In contrast, the non-randomized cohort studies exhibited a moderate risk of bias, particularly due to confounding factors and participant selection.11,14 The detailed risk of bias assessment is provided in Supplementary Tables 3 and 4.11,14-17

Outcomes results

The diagnostic yield was reported in five studies involving 217 patients in the SE group and 316 patients in the SBE group. The pooled RR was 1.07 (95% CI, 0.96–1.20; p=0.24), suggesting comparable diagnostic performance between SE and SBE (Fig. 2). Heterogeneity was moderate (I²=35%). A subgroup analysis comparing manual SE to motorized SE did not show a significant effect modification (p=0.06, Fig. 3).

Fig. 2.

Diagnostic yields. M-H, Mantel-Haenszel; CI, confidence interval.

Fig. 3.

Diagnostic yield subgroup analysis (manual vs. motorized spiral enteroscopy). M-H, Mantel-Haenszel; CI, confidence interval; SE, spiral enteroscopy; SBE, single balloon enteroscopy.

The therapeutic yield was reported in only three studies, comprising 100 patients in the SE group and 107 in the SBE group. The RR was 1.10 (95% CI, 0.45–2.69; p=0.83), indicating a similar therapeutic yield between the two modalities (Fig. 4). Substantial heterogeneity was observed (I²=70%). However, owing to the limited number of studies, the heterogeneity estimates were limited in scope. Subgroup analysis comparing manual SE to motorized SE was not possible because of the limited number of studies.

Fig. 4.

Therapeutic yields. M-H, Mantel-Haenszel; CI, confidence interval.

The total procedure time was reported in four studies involving 163 patients in the SE group and 169 patients in the SBE group. The pooled MD was –22.85 minutes (95% CI, –46.83 to 1.12; p=0.06), favoring SE, but it did not reach statistical significance (Fig. 5). A considerable level of heterogeneity was observed across the studies (I²=95%), although the small number of studies limited the accuracy of the heterogeneity estimation. A subgroup analysis comparing manual SE to motorized SE showed a significant modifying effect (p<0.001), with a significantly shorter procedure time in the motorized SE group (Fig. 6).

Fig. 5.

Total procedure time. SD, standard deviation; IV, intravenous; CI, confidence interval.

Fig. 6.

Total procedure time subgroup analysis (manual vs. motorized spiral enteroscopy). SD, standard deviation; IV, intravenous; CI, confidence interval; SE, spiral enteroscopy; SBE, single balloon enteroscopy.

Regarding the DMI, data from three studies comparing SE and SBE (n=128 vs. 132) showed an SMD of 1.33 (95% CI, 0.65–2.01; p<0.001), indicating a significantly greater insertion depth in the SE group than in the SBE group (Fig. 7). Substantial heterogeneity was present (I²=81%); however, the small number of studies limited the certainty of the heterogeneity estimates.

Fig. 7.

Depth of maximal insertion. SD, standard deviation; IV, intravenous; CI, confidence interval.

Finally, adverse events were reported in only four studies (total n: SE, 163; SBE, 169), with no statistically meaningful difference observed between the two groups (RR, 1.72; 95% CI, 0.80–3.70; p=0.16, I²=28%) (Supplementary Fig. 1).

Sensitivity analysis

Sensitivity analysis was performed by removing one study at a time to assess its impact on the overall effect size for all outcomes. However, no significant changes were observed.

DISCUSSION

This meta-analysis found no significant differences between SE and SBE in terms of diagnostic yield, therapeutic yield, total procedure time, or adverse events. However, in the subgroup analysis, motorized SE was associated with a shorter procedure time than manual SE. Additionally, SE demonstrated a greater DMI than SBE.

Our study specifically compared SE and SBE. A previous meta-analysis that compared balloon enteroscopy with SE included eight studies, seven using DBE, and one using SBE.18 Both groups showed similar diagnostic and therapeutic success rates and comparable DMIs. The only significant difference was a shorter procedure time in the SE group. These findings align with our results, which also showed a shorter procedure time, although this was limited to motorized SE. The only notable difference from our results was the DMI, which was significantly higher in the SE group than in the SBE group in our analysis.

Our results showed that DMI was greater with SE. May et al.19 compared DBE to SE and found that, while SE had a shorter procedure time, DMI was superior in the DBE group (310 cm vs. 250 cm; p=0.004). However, their study was limited by its observational design, small sample size of 18 patients, and potentially variable operator experience with DBE. Conversely, an observational study by Khashab et al.14 that was included in our analysis, found a greater DMI and shorter procedure times with SE, although the diagnostic yields were similar.

A larger, non-randomized, multicenter study by Rahmi et al.,20 involving 241 patients, compared DBE and SE and found no significant differences regarding diagnostic yield, procedure time, or DMI. To minimize operator bias, the procedures were performed by less-experienced endoscopists (those with fewer than ten cases per arm). The authors concluded that SE is a safer option than DBE for small bowel disorders, offering similar diagnostic and therapeutic outcomes. These results further support the use of SE in the management of small bowel disorders.

Motorized SE is a relatively recent advancement that differs from manual SE in that it utilizes a motorized mechanism to drive the spiral overtube.21 Unlike manual SE, which requires manual rotation of the overtube, motorized SE automates advancement and withdrawal through a motorized spiral segment that is controlled by a foot pedal.

The key advantage of a motorized SE is its user-friendly interface, which makes it easier to learn and less physically demanding than manual SE. These factors may explain the shorter procedure time observed for motorized SE in our study.

Moreover, the learning curve for motorized SE appears to be shorter, allowing for faster skill acquisition, even among less-experienced endoscopists. This ease of use, combined with reduced operator fatigue, may facilitate the wider adoption of motorized SE, particularly in centers where advanced endoscopic expertise is limited. In such settings, motorized SE could provide an effective and scalable solution for deep small bowel evaluation and intervention.

Although no direct comparison was made between manual and motorized SE, Pal et al.16 compared motorized SE to SBE and found that motorized SE achieved a higher technical success rate (96.8% vs. 80.7%, p=0.08) in reaching deeper small bowel regions (distal jejunum/proximal ileum). Motorized SE was also associated with increased diagnostic and therapeutic yields, including greater DMIs and improved depth-to-time ratios.

In summary, SE demonstrated diagnostic and therapeutic outcomes similar to those of SBE, with comparable rates of adverse events. Motorized SE was associated with shorter procedure times. Finally, SE was associated with greater DMIs than SBE.

The small number of studies and limited sample sizes are the key limitations of this analysis. The lack of randomized controlled trials influenced the mix of study designs used in our review. This may introduce variability owing to differences in protocols and patient selection criteria. Selection bias is particularly likely in cohort studies in which treatment decisions were not randomized and may reflect underlying patient conditions or physician preferences. The heterogeneity assessment was not sufficiently powered due to the small number of studies and sample sizes; therefore, we cannot have full confidence in its estimates.13 DMI was reported in only three studies, with variations in their definitions; hence, we used SMD for the evaluation. The subgroup analysis comparing motorized and manual SE was also underpowered because of the small number of studies and participants. Additionally, an assessment of publication bias was not possible because approximately ten studies are typically required for a reliable evaluation of publication bias.

Owing to these limitations, it is difficult to make recommendations for clinical practice. Randomized controlled trials with large sample sizes can provide valuable insights. Future studies should involve larger multicenter randomized controlled trials to enable more definitive comparisons. Additionally, secondary outcomes, such as patient comfort, learning curves, and cost-effectiveness, should be explored to provide a comprehensive understanding of both techniques. These efforts will support better clinical decision-making and enhance the use of SE and SBE in managing small bowel disorders.

In conclusion, this meta-analysis found no significant differences between SE and SBE in terms of diagnostic and therapeutic yields, total procedure time, or adverse events. However, in the subgroup analysis, motorized SE was associated with shorter procedure times.

SE was associated with greater DMIs than SBE. Further high-quality comparative studies with larger sample sizes are needed to confirm these findings.

Supplementary Material

Supplementary Material 1. Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2020 checklist.

Supplementary Table 1. Database search results.

Supplementary Table 2. DMI definitions in the included studies.

Supplementary Table 3. Cochrane risk-of-bias tool, ROB 2: assessment the risk of bias in randomized control trials.

Supplementary Table 4. ROBINS-I: assessment the risk of bias of cohort studies.

Supplementary Fig. 1. Adverse events.

Supplementary materials related to this article can be found online at https://doi.org/10.5946/ce.2025.184.