Clinical outcomes of bowel preparation education strategies in colonoscopy: An evidence map of systematic reviews

Abstract

Background:

Bowel preparation before colonoscopy is a research focus and challenge in gastroenterology. Inadequate bowel preparation increases the risk of adenoma miss rate, healthcare costs, and morbidity and mortality from colorectal cancer. While health education shows promise in improving preparation quality, existing studies are limited by fragmented evidence, variable methodological quality, and insufficient strategy comparisons, constraining clinical efficacy. This study innovatively employs an evidence map to clarify the current distribution of evidence, core strategies, and practical gaps, providing a reference for optimizing clinical decision-making.

Methods:

Following the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews guidelines, we searched 8 significant databases (CNKI/Wan Fang/VIP/SinoMed/Web of Science/PubMed/CINAHL/Cochrane Library) built up to October 30, 2024, for inclusion of bowel preparation-related systematic evaluations/meta-analyses. Methodological quality was assessed using AMSTAR-2 and AMSTAR, and multidimensional data visualization was achieved by generating bubble plots via Pymeta.com.

Results:

We identified 20 systematic reviews and meta-analyses, clarified 6 core strategies: smartphone application, telephone, short message service, visual aids, newly designed brochure, and multi-method combination. Quality assessment revealed that only 10% of studies achieved moderate methodological quality; the remaining studies were of low or critically low quality. Overall, 6 strategies can effectively improve bowel preparation adequacy with smartphone applications and multi-method combination demonstrating superior efficacy. Among outcome measures, bowel preparation quality, adenoma/polyp detection rate ranked among the top three. Notably, patient-reported outcomes (compliance, satisfaction) improved more than clinician-reported outcomes, attributable to 4 explanatory factors.

Conclusion:

A multi-method combination has significant practical potential. This study proposes personalized combination strategies based on the characteristics of different populations, and their effectiveness needs to be further verified by more high-quality research in the future. Additionally, this study pioneers the visualization of the evidence landscape through an evidence map, establishing a direct mapping between educational strategies and outcome measures to provide a framework for optimizing clinical decision-making. These findings call for high-quality research to compare the efficacy and generalizability of strategy combinations, while exploring tailored approaches for diverse populations to advance clinical application of personalized, precision-based bowel preparation education programs.

1. Introduction

Colonoscopy is the gold standard for screening and diagnosing colorectal cancer. Bowel preparation is a crucial precolonoscopy step. Inadequate bowel preparation not only leads to incomplete colonoscopy and prolonged hospitalization, but also significantly pushes up healthcare costs (12% to 22%).^[1,2] Research data have shown that it results in adenoma underdiagnosis of up to 47.9% and is strongly associated with an elevated risk of postcolonoscopy colorectal cancer morbidity and mortality after colonoscopy.^[3–6] Guidelines recommend that adequate bowel preparation needs to be ≥ 90%, and updated guidelines from the European Society of Gastrointestinal Endoscopy in 2019 point to a bowel preparation failure rate of 18% to 35%^[7–9]; only 29% of endoscopy centers meet the minimum standard in a recent survey.^[10] Current bowel preparation standards remain suboptimal. Improving bowel preparation quality is a key concern for medical staff and is crucial for maximizing the benefits of colorectal cancer screening and avoiding the waste of medical resources.

Bowel preparation is complex and difficult for patients to remember. Additionally, the intake of large volumes of bowel cleansing agents can cause adverse reactions like diarrhea and abdominal pain, significantly impacting patient adherence. Non-adherence is an independent risk factor for inadequate bowel preparation, with 86.7% of patients with inadequate preparation failing to follow instructions strictly.^[11] Health education is an integral component of bowel preparation, encompassing structured interventions that enhance patient understanding of bowel cleansing protocols and promote proper execution to optimize endoscopic visualization and bowel health maintenance. Multiple studies have emphasized its critical impact on compliance and bowel preparation quality.^[12,13] Although medical professionals have made some progress in improving adherence and bowel preparation quality, there are still issues such as high heterogeneity of clinical evidence, lack of cross-strategy comparisons, and prominent repeated intervention studies. This highlights the need for systematic evidence.

Evidence map is a new type of evidence synthesis research method that can systematically collect, synthesize, and integrate existing evidence within a specific field. It aims to uncover research status, problems, development directions, and evidence gaps.^[14] While traditional systematic reviews (SRs)/meta-analyses (MAs) focus on statistical effect sizes and address specific clinical questions, evidence maps tend to summarize a research area and prioritize breadth of research over quantification of effect sizes.^[15] Bowel preparation in colonoscopy involves diverse educational strategies and multidimensional clinical outcomes, and its complexity makes it difficult to comprehensively assess the effectiveness of the strategies by traditional SRs/MA.^[16] The evidence map can simultaneously address the questions of “which health education strategies are effective” and “which outcome indicators are effective,” which is highly consistent with the objectives of this study. Therefore, we used an evidence map to integrate and analyze the existing evidence, providing a reference for optimizing clinical decision-making.

2. Methods

2.1. Design

Aligned with JBI guidelines, this evidence map is categorized as a scoping review. As a comprehensive analysis of existing literature, this study does not require ethical approval. It is registered in the Open Science Framework (Registration DOI: https://doi.org/10.17605/OSF.IO/2ZWK4) and adheres to the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews checklist.^[17]

2.2. Data sources and search strategy

A systematic search was performed in China National Knowledge Infrastructure (CNKI), China Biology Medicine (CBM), Wan Fang, VIP Database for Chinese Technical Periodicals (VIP), PubMed, Web of Science, Cochrane Library, Cumulative Index to Nursing and Allied Health Literature (CINAHL) from their inception to October 30, 2024. All search strategies were built using Exploded Medical Subject Headings and the appropriate corresponding text words. Table S1 (Supplemental Digital Content, https://links.lww.com/MD/Q166) outlines the search strategies.

2.3. Inclusion and exclusion criteria

The inclusion criteria were as follows: Type of study: SR, meta-analysis (MA) (the strength of evidence from SR/MA is generally superior to that of single original studies, which is the gold standard for efficacy evaluation^[15]); study population: patients who underwent colonoscopy requiring bowel preparation; intervention: any health education measures aimed at bowel preparation (including but not limited to brochures, videos, phone calls, etc); control measures and outcome metrics: no restrictions; language: limited to Chinese and English.

The exclusion criteria were as follows: duplicate publications; incomplete data; and conference literature.

2.4. Data selection

The literature screening process is divided into 3 steps. Firstly, the search results are imported into NoteExpress 26.0 software (developed by Beijing Aiqinhai Software Co., Ltd., Beijing, China) for checking, and the initial screening was based on the titles and abstracts of the included studies. Step 2: Read the full text of the screened literature to determine whether it should be included. The third step is to search the references of the included studies to see if there are any additional studies. The entire screening process is strictly completed according to the Preferred Reporting Items for Systematic reviews 2020 flowchart. Two reviewers independently complete literature screening, and when there are different viewpoints, they are discussed and resolved with a third reviewer.

2.5. Data extraction

Data extraction was performed independently by 3 researchers using a predesigned Excel sheet. Discrepancies were resolved through consensus or adjudication by the third reviewer. The first table recorded study characteristics: publication year, first author, country, objectives, search period, number of included studies, and sample size. The second table captured PICO (Population, Intervention, Comparison, Outcomes) elements and categorized study conclusions.

2.6. Quality assessment

SRs and meta-analyses (MAs) were assessed using AMSTAR-2 (16 items, 7 critical domains),^[18] Each item was rated as “Yes,” “Partial yes,” “No,” or “Not applicable,” with final quality classifications categorized as “high,” “moderate,” “low,” and “critically low”; For network MAs, we adopted the 11-item modified AMSTAR scale (11-item version) following Tang et al.‘s methodology, with each item scored as 1 point for “Yes” and 0 points for “No,” “Unclear,” or “Not applicable.^[19] Quality stratification was defined as high (9–11 points), moderate (5–8 points), and low (<5 points). Two researchers assessed quality independently, with a third resolving discrepancies.

It is worth noting that the evidence map describes the methodological quality of the included studies without criticizing low-quality evidence, aiming to clarify the quality of evidence and provide a reference for future study designs.^[16]

2.7. Data synthesis and statistical analysis

To evaluate the effectiveness of health education strategies, we categorized their conclusions into 5 types: effective, likely effective, uncertain, ineffective, and harmful, consistent with a prior study.^[20] Each category was defined and illustrated with examples (see Table S2, Supplemental Digital Content, https://links.lww.com/MD/Q166). Two researchers independently performed the categorization; discrepancies were resolved through discussion. When a study contained multiple health education strategies: subgroup-based conclusions were prioritized if available; otherwise, pooled effect sizes determined classification. Given that the included SRs/MAs often assessed multiple health education strategies and outcomes simultaneously, this study defined independent “education strategy-outcome measure” combinations as the unit of analysis (“unit of analysis,” i.e., the basic unit in a study that is analyzed independently). This terminology and concept adhere to guidelines from JBI and the Cochrane Handbook.^[21,22] Consequently, 1 SR/MA could correspond to multiple units of analysis.

We used Pymeta.com’s tool to generate an evidence map.^[23] The X-axis represents the conclusion classification/outcome indicators, while the Y-axis represents the results of the literature quality assessment/conclusion classification. Each bubble represents a unit of analysis. Bubble color denotes the strategy type. Bubble size corresponds to the number of primary studies included in the SRs/MAs. The number above the bubble indicates the ID of the included SRs/MAs. The number in parentheses indicates the total number of primary studies included in that SRs/MAs, consistent with the bubble size.

The remaining figures were generated using Prism 10 (GraphPad, San Diego).

3. Results

3.1. Literature selection

A total of 1265 records were identified. After removing 444 duplicates, title/abstract screening excluded 773 irrelevant records. Full-text assessment of 48 potentially eligible articles resulted in the final inclusion of 20 SRs/MAs. The study selection process is detailed in Figure 1.

Figure 1.

PRISMA flow diagram. PRISMA = Preferred Reporting Items for Systematic reviews and Meta-Analyses. This figure is adapted from the official PRISMA 2020 flow diagram template released by the PRISMA Group in 2020. The original template is available at: https://www.prisma-statement.org/prisma-2020-flow-diagram. The use of this template complies with the Creative Commons Attribution 4.0 International License (CC BY 4.0), see: https://creativecommons.org/licenses/by/4.0/.

3.2. Study characteristics

A total of 20 SRs/MAs were included, comprising 9 MAs, 2 network MAs, and 9 SRs.^[24–43] All studies except Kurlander et al.‘s SR (no quantitative analysis due to heterogeneity) reported effect sizes.^[40] First authors were predominantly from China (n = 12), followed by the United States (n = 3), with Thailand, Spain, Italy, Canada, and Greece each contributing 1 study. Included original studies per review ranged from 4 to 47 (see Table 1).

Table 1.

Summary of SRs and MAs included in the evidence mapping.

ID		Country	Study design	Research aim	Search period	Type of study	Number of studies included	AMSTAR-2 score
1	Wonggom, 2023[24]	Thailand	SRM	Updated meta-analysis on bowel preparation innovative technology instructions.	Database inception – February 2022	RCT	47	Moderate
2	Losurdo, 2023[25]	Italy	SRM	Comparative efficacy: Visual aids/booklets vs standard instructions for bowel preparation adequacy.	Database inception – March 2023	RCT	6	Critically low
3	Li, 2023[26]	China	MA	To assess the impact of WeChat-based educational interventions on bowel preparation quality.	Database inception – May 2021	RCT	4	Critically low
4	He, 2023[27]	China	MA	To assess the impact of telephone instruction on bowel preparation quality.	Database inception – April 2022	RCT	9	Critically low
5	Peng, 2022[28]	China	SRM	Absence of meta-analysis on supplemental education for improving adequate bowel preparation rates in outpatients. Need for a systematic review and meta-analysis due to conflicting RCT results.	Database inception – August 2021	RCT	21	low
6	Li, 2022[29]	China	MA	To assess the impact of smartphone education on bowel preparation quality.	Database inception – August 2021	RCT	12	Critically low
7	Li, 2022[30]	China	MA	To assess the impact of short message service on bowel preparation quality.	Database inception – November 2021	RCT	7	Critically low
8	El Bizri, 2021[31]	Canada	SRM	To synthesize and critically evaluate the evidence on existing mHealth technologies to support patients in preparing for colonoscopy.	1996–2017 (updated November 28, 2019)	RCT	10	low
9	Chandan, 2021[32]	US	SRM	To evaluate smartphone app and digital audiovisual education on ADR and bowel preparation quality, assessing compliance and age-specific efficacy.	Database inception – May 2020	RCT	13	Critically low
10	Ye, 2020[33]	China	SRM	To assess the appropriateness of educational videos for colonoscopy patients.	Database inception – June 2019	RCT	8	Critically low
11	Tian, 2020[34]	China	MA	To evaluate the effectiveness of an enhanced patient education intervention in detecting polyps and adenomas in colorectal cancer compared to the standard patient education.	Database inception – June 2019 (Updated August 2019	RCT	10	Critically low
12	Guo, 2020[35]	China	MA	To evaluate the impact of enhanced education on improving the quality of bowel preparation and other outcomes	Database inception – April 2019	RCT	18	Critically low
13	Gkolfakis, 2019[36]	Greece	SRM	To understand various intervention measures for substantial improvement of intestinal preparation in hospitalized patients.	Database inception – July 2017 (Updated June 2018)	RCT, observational study	6	low
14	Desai, 2019[37]	US	SRM	To evaluate the impact of smartphone applications on bowel preparation compared to standard patient education.	Database inception – January 2018	Comparative	6	Critically low
15	Lu, 2018[38]	China	MA	To evaluate the impact of health education on the quality and compliance of patients’ bowel preparation, and provide evidence-based support and reference for clinical nursing practice.	Database inception – March 2017	RCT, CCT	15	Critically low
16	Guo, 2017[39]	China	MA	To evaluate the effects of diverse enhanced educational interventions on bowel preparation quality, ADR, PDR, and related clinical outcomes.	Database inception – August 2015	RCT	8	Critically low
17	Kurlander, 2016[40]	US	SR	To evaluate the impact of a patient education intervention on the quality of bowel preparation.	Database inception – August 2014	RCT, quasi-experimental	7	low
18	Chang, 2015[41]	China	MA	To assess patient education on bowel preparation quality, PDR, and repeat colonoscopy needs due to incomplete exams.	Database inception – November 2014	RCT	9	Critically low
19	Shi, 2020[42]	China	Network Meta-analysis	To provide optimal bowel preparation for outpatients.	May 2013 – May 2018	RCT	10	low
20	Tian, 2021[43]	Spain	Network Meta-analysis	To address the issues that traditional meta-analysis cannot solve.	Database inception December 2019 (Updated February 2020)	RCT	23	Moderate

ADR = adenoma detection rate, CCT = controlled clinical trial, MA = meta-analysis, PDR = polyp detection rate, RCT = randomized controlled trial, SR = systematic review, SRM = systematic review with meta-analysis.

3.3. PICO questions

We provided detailed information on PICO, as shown in Table S3 (Supplemental Digital Content, https://links.lww.com/MD/Q166).

3.4. Study participants

Among the 20 included studies, 17 focused on adult patients aged ≥ 18 years, while 3 studies did not restrict age.^[25,33,41] Three studies enrolled only outpatients,^[28,42,43] 1 included only inpatients,^[36] and the remaining 16 did not specify whether participants were outpatients or inpatients.

3.5. Intervention characteristics

The intervention measures in the 20 SRs/MAs can be divided into 6 strategies: Smartphone applications (SA): 9 studies involve this method,^{[24,26,28,29,31,32,37,42,43]} mainly including the use of social media software, newly developed bowel preparation software, camera applications; Telephone(T): including telephone reminders, telephone reeducation, which appeared in a total of 6 studies^{[24,27,28,40,42,43]}; Short message service (SMS): 5 studies remind patients to complete bowel preparation or send health education content through SMS^{[24,28,30,42,43]}; Newly designed brochure(NDB): 5 studies designed educational manuals different from the control group^{[25,28,36,40,43]}; Visual aids (VA): refers to technical tools that facilitate information transmission, knowledge understanding, or behavior change through visual media such as videos, dynamic simulations, virtual environments. 7 studies applied video and virtual reality (VR) technologies^{[24,28,32,33,40,42,43]}; Multi-Methods Combination (MMC): it refers to simultaneously using ≥ 2 complementary approaches (e.g., SMS plus tele-counseling, customized booklets with phone follow-ups).^{[24,28,31,32,34,35,38,39,41]} This definition addresses issues like unclear methods in source studies or trial designs with multiple modalities that can’t be categorized singularly.

Control group protocols were largely consistent: 13/20 SRs/MAs implemented health education via oral explanations and written guidance, while 8 studies lacked detailed intervention descriptions.^{[25,28,32–34,36,37,41]}

3.6. Outcome characteristics

According to Clinical Outcome Assessment (COA) criteria, the 15 indicators in this evidence map were categorized into patient-reported outcomes (PROs) and Clinician-Reported Outcomes (ClinROs) (as shown in Table 2).^[44] Only Tian (2020) used polyp/adenoma detection rates (PDR/ADR) as primary outcomes^[34]; 19 studies focused on bowel preparation quality as the main outcome measure. We extracted the adequate bowel preparation rate from 12 SRs/MAs.^{[24–30,32,35–37,39]} It was found that all intervention groups were superior to the control group, but the qualified rate was <90%, which did not meet the minimum standard required by the guidelines.^[9] The intervention group had the highest qualified rate of 89.8%, as shown in Figure 2.

Table 2.

Outcome measures included in the studies.

Outcome measure		Corresponding literature numbers	Frequency
Clinician-Reported Outcomes (ClinROs)	1. Bowel preparation quality	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 13, 14, 15, 16, 17, 18, 19, 20	19
	2. Incomplete colonoscopy rate	3, 13, 18	3
	3. CIT	3, 4, 5, 6, 8, 12, 16	7
	4. Withdrawal time	2, 3, 4, 5, 6, 8, 9, 10, 12, 16	10
	5. ADR	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 14	13
	6. PDR	4, 5, 6, 7, 10, 11, 12, 16, 18, 19, 20	11
	7. CIR	2, 3, 4, 6, 7, 16	6
	8. CDR	11	1
Patient-Reported Outcomes (PROs)	9. Willingness to repeat bowel preparation	3, 4, 5, 6, 8, 9, 13, 16, 20	9
	10. Abdominal discomfort	3, 4, 6, 12, 13, 16, 20	7
	11. Nausea/vomiting	3, 4, 6, 9, 12, 13, 16, 20	8
	12. Adherence	1, 3, 8, 9, 12, 13, 14, 15, 16, 19, 20	11
	13. Satisfaction	1, 13, 14, 20	4
	14. Sleep quality	3, 20	2
Others	15. No-Show Rate	5, 7, 8	3

ADR = adenoma detection rate, CDR = cancer detection rate, CIR = cecal intubation rate, CIT = cecal intubation time, PDR = polyp detection rate.

Figure 2.

Adequate bowel preparation rate.

3.7. Quality assessment

According to the AMSTAR-2, quality assessment was conducted on 18 SRs/MAs. One study provided a list of excluded literature, and 1 study did not conduct a MA. Overall, none of the included studies were of “high” quality, only one was of “moderate” quality, four were of “low” quality, and 13 were of “critically low” quality. The evaluation results and item conformity rates are presented in Figure 3. Two network MAs were assessed using AMSTAR, revealing one as low quality and the other as moderate quality. As shown in Table S4 (Supplemental Digital Content, https://links.lww.com/MD/Q166).

Figure 3.

(A) Compliance with AMSTAR-2 items. (B) Methodological quality of the included SRs/MAs (AMSTAR-2). MAs = meta-analyses, SRs = systematic reviews.

3.8. Evidence map

No studies reported “harmful” effects. To present the bubbles included in the evidence more clearly, only “Effective,” “Likely effective,” “Ineffective,” and “Uncertain” were retained.

3.8.1. Education strategies – bowel preparation quality

Except for Tian (2020), the remaining 19 SRs/MAs (representing 36 units of analysis) evaluated bowel preparation quality (as shown in Figure 4).^[34] Four units of analysis concluded that strategies (SA, VA, SMS) were “ineffective”; 3 indicated “Likely effective” results for strategies (SA, T, NDB); and 29 units of analysis (80.6%) concluded that 6 strategies were “effective.” Notably, all units of analysis involving MMC showed significant improvement in bowel preparation quality. In terms of methodological quality, among these 19 SRs/MAs, 5 studies were assessed as low quality, two as moderate quality, and the remaining as very low quality.

Figure 4.

Education strategies – bowel preparation quality.

3.8.2. Education strategies – ClinROs

Among 56 units of analysis on ClinROs (in Figure 5), the effectiveness conclusions were polarized: 46.4% (26 units) found the strategies “effective,” 44.6% (25 units) “ineffective,” with only a few being “uncertain” (7.1%, 4 units) or “likely effective” (1.8%, 1 unit). Specifically, for the 3 indicators ADR, cecal intubation time (CIT), and withdrawal time, 53.8% (7/13), 57.1% (3/7), and 70% (7/10) of units of analysis found that education strategies were “effective” or “likely effective.” SA showed outstanding performance in ADR and CIT, while MMC was more prominent in withdrawal time.

Figure 5.

Education strategies – ClinROs. ClinROs = clinician-reported outcomes.

For PDR, although only 37.5% (6/16) of the units reported effective results, all 3 units using the T strategy found it effective (despite moderate to critically low-quality evidence), suggesting potential value. The incomplete colonoscopy rate (33.3%, 1/3 effective) and cecal intubation rate (33.3%, 2/6 effective) had weak evidence due to the small-sample sizes and critically low-quality literature. For the cancer detection rate, only one unit of analysis (MMC-based) was included, which concluded “ineffective” and should be interpreted cautiously.

Overall, from a strategy perspective, SA and MMC covered the broadest range of indicators. SA showed consistent efficacy – effective for ADR and CIT, ineffective for cecal intubation rate and PDR, and minimal controversy overall. MMC showed some variability (e.g., in CIT, four different conclusions), but demonstrated good improvements in ADR and withdrawal time. T showed strong potential in PDR. SMS, NDB, and VA generally showed “ineffective” results and should be cautiously applied.

3.8.3. Education strategies – -PROs

Figure 6 shows that the overall effectiveness rate of education strategies on PROs was 63.2% (36/57), with satisfaction (85.7% (6/7)) and adherence (88.9% (16/18)) showing the greatest improvements. All strategies except NDB were effective in improving adherence, supported by strong data consistency and confidence.

Figure 6.

Education strategies – PROs. PROs = patient-reported outcomes.

Seventy percent (7/10) of the units of analysis supported the effectiveness of strategies in improving willingness to repeat bowel preparation, with SA and MMC showing significant effects. However, the two units using T (moderate and very low quality) were ineffective.

About 55.6% (5/9) of units found that SA, MMC, and SMS relieved nausea/vomiting; Results were mixed for abdominal discomfort: 57.1% (4/7) of small-sample units found SA, T, and MMC ineffective, while two larger units (moderate and very low quality) confirmed the effectiveness of SA and MMC. Due to small sample sizes, conclusions on sleep quality (66.7%, 2/3, SA/T strategies) and no-show rate (33.4%, 1/3, SMS strategy) should be interpreted cautiously.

Overall, educational strategies were effective in improving PROs. SA and MMC were the most comprehensive and effective. T and VA were effective for adherence and satisfaction, but had a limited impact overall. SMS showed potential for improving nausea/vomiting, satisfaction, and adherence, though further validation is required. NDB showed poor effectiveness and had the weakest supporting evidence.

4. Discussion and conclusion

4.1. Discussion

4.1.1. Research findings

A rigorous selection process identified 9 SRs, 9 MAs, and 2 network MAs. Methodological quality assessment via AMSTAR-2 and AMSTAR showed 65% of studies were critically low-quality, with 3 key flaws: (1) lack of excluded literature lists (19 studies); (2) unregistered protocols (13 studies); (3) insufficient funding disclosure (12 studies). This aligns with previous studies, where 83.3% to 93.3% of SRs were critically low-quality, particularly for exclusion criteria transparency (item 7), even in journals with impact factors > 90 (only 11.2% adequately reported exclusions).^[45–47] Enhancing methodological rigor, especially adherence to item 7, should be prioritized.

Although this study demonstrated an overall bowel preparation rate <90%, this does not indicate that all 6 educational strategies underperformed. Heterogeneity arising from aggregating strategies with divergent implementation frequencies and types may obscure truly effective approaches. It is suggested that future SRs/MAs classify and evaluate each strategy’s compliance rate and refine the analysis of implementation details to more accurately compare differences in effectiveness and provide a basis for optimizing strategies.

All 6 strategies effectively improved bowel preparation quality, yet demonstrated lower efficacy for ClinROs than PROs (46.4% vs 63.2%). The analysis suggests that education strategies can effectively improve patients’ subjective experiences and reported outcomes by enhancing their cognition and compliance, thereby strengthening self-management. The limited improvement of ClinROs can be explained by the following reasons: low proportion of patients with inadequate preparation (Boston Bowel Preparation Scale (BBPS) scores ≥ 8 vs 6–7 showed comparable ADR/PDR/CIR). Only poor quality (single segment score < 2) significantly reduces lesion detection^[48]; most studies focus on improving bowel preparation quality, and the sample size is insufficient to test differences in other ClinROs; ClinROs are affected by uncontrollable variables such as the standardization of doctor operations and equipment performance; Although no significant advantages have been found in the short term, educational strategies may reduce the risk of colorectal cancer by increasing long-term screening participation rates.

The evidence map categorizes health education strategies into 6 types: SA, T, SMS, NDB, VA, and MMC. SA and MMC are the most comprehensive and effective educational strategies, while other strategies have unique advantages in different indicators. This article systematically analyzes the core characteristics of the 6 types of strategies and proposes a combination scheme based on population characteristics. SA establishes a knowledge-to-behavior management loop through integrated text/video instructions and smart medication reminders. It demonstrated favorable results in clinical outcomes, especially PROs. Most studies use existing programs (such as WeChat applets, official accounts), embed medication timers, and send users the knowledge and task reminders related to bowel preparation in the form of text or video 1 to 4 days before the colonoscopy and on the day of examination.^[49,50] This design conforms to the cognitive load theory, replacing the traditional memory-dependent model with an active prompt mechanism to reduce patients’ cognitive burden and behavioral omission risk.^[51] One study further customized an AI-enabled app that analyzes fecal images to personalize bowel cleanser dosage, exemplifying precision bowel preparation advancements.^[52] In summary, SA integrates behavior change techniques (e.g., goal setting, self-monitoring, feedback), promoting correct behaviors through enhanced psychological capability (knowledge transfer) and physical opportunity (medication reminders).^[53] While Park et al reported no association between bowel preparation quality and ADR, our study and others confirm SA significantly improves both outcomes.^[54–56] This discrepancy may stem from differing bowel preparation scales (Aronchick Scale vs the BBPS/Ottawa Bowel Preparation Scale used in the studies^[54]). Moreover, this strategy has been shown to improve PROs. It is worth noting that SA has technical applicability boundaries: benefits are limited for elderly patients and those with low digital literacy. Effectiveness also critically depends on application design – requiring functional trade-offs aligned with intervention goals, complexity limitation, minimized data input, and streamlined user processes.

T-based strategy demonstrates unique advantages in bowel preparation education through bidirectional communication and psychosocial support mechanisms. Authoritative language used by healthcare workers can enhance patient trust; open-ended questioning identifies individualized barriers (e.g., “Taking laxatives will cause excessive vomiting”), which can lead to adjustments in coaching strategies; empathic responses activate the patient’s secure attachment system (e.g., “I understand your concern...”), reducing avoidance behaviors; and clarifying misunderstandings (e.g., “Diarrhoea after taking medication is normal”) can reduce anxiety triggered by uncertainty.^[57] The results of this study showed that the T strategy avoided patients’ misunderstanding when watching videos or reading manuals on their own, effectively improving the quality of bowel preparation. At the same time, it was found that T has potential application for PDR, which is consistent with previous research findings.^[58,59] Current practice is mostly limited to a single instruction before colonoscopy and lacks continuous intervention at key stages (e.g., final dose reminder), which does not guarantee a response rate and makes the intervention less effective. However, its low technological threshold, instant feedback capability, and cultural adaptability to verbal communication make this strategy advantageous among the elderly (≥ 60 years old) and those with a low education level (≤ junior high school).^[60,61] Effective implementation likely depends on two prerequisites: a standardized communication framework (timing of telephone guidance, content design, training of guidance personnel’s communication skills, etc) and resource allocation planning (calculating nurse-patient ratios, avoiding overwork).

SMS is implemented through online platforms that do not require dependence on smartphones or networks, and is low-cost, high-coverage, and can promote health equity.^[62] Compared with other strategies, SMS delivers core points in standardized text, avoiding the bias of manual communication, and can send short, high-frequency reminders (e.g., “Please take the first box of laxatives at 8:00 pm tonight”) at key times, directly triggering patient behaviors and ensuring continuity of care. In addition, the cost-effectiveness of SMS is better than that of telephone and smartphone-based strategies, and it is easier for patients of all economic levels to use.^[63,64] The limitations of SMS are that it does not address barriers to understanding (e.g., misinterpretation of “clear liquids”) or correct errors, and that populations with low health literacy and older adults are less likely to be receptive to this strategy.^[65] Therefore, SMS can be used as a basic strategy in resource-limited settings.

VA optimizes information acquisition and comprehension effectiveness through the intuitiveness of visual perception. This category incorporates two types of tools, educational videos and VR technologies. Based on Mayer multimedia learning cognitive theory dual-channel model, visual-auditory bimodal information can overcome the capacity limitation of a single channel and achieve maximum knowledge integration.^[66] Educational videos promote deep encoding of the core knowledge of bowel preparation for patients by synchronously presenting visual operation processes (such as the preparation of bowel cleansing agents) and voice explanations; VR technology effectively stimulates emotional resonance and self-efficacy through 3-dimensional anatomical modeling (such as colonoscopy perspective simulation) and interactive situational training (such as virtual medication time management), forming sustained motivation for behavior change. This mechanism can well explain the results of this study: VA enables patients to achieve higher compliance and satisfaction, thereby improving the quality of bowel preparation, which is consistent with the findings of Au et al and Chen et al.^[67,68] VA has demonstrated significant universal advantages and sustainable application value in the field of health education, breaking through the limitations of traditional methods in terms of personalization and accessibility. This strategy requires a high level of content development: it is necessary to carefully design information presentation modes (explanatory/practical/narrative), optimize multimedia element combination strategies (timing ratio of static graphics and 3D animations), and establish a dynamic balance mechanism between evidence-based medical content and visual communication effectiveness. Through a rigorous content structure and theory-driven development process, the effectiveness of health education in improving clinical outcomes can be maximized.

NDB is an enhancement of traditional written materials, utilizing a modular information architecture with clear hierarchical levels to reduce the visual fatigue often caused by dense paragraph formatting. One notable feature is the inclusion of a “checkable process table,” which breaks the complex bowel preparation process into clear, actionable items. This allows patients to track their progress and provides positive reinforcement, reducing the likelihood of missing critical steps. This evidence map indicates that the NDB improves the bowel preparation quality, but shows limited effectiveness across other outcome measures. This finding differs from that of Ergen et al but aligns with the result of Guardiola et al.^[69,70] This discrepancy may be attributed to the limited and low-quality evidence included in the current study, which restricts the credibility of the findings. Therefore, no definitive conclusions can be drawn at this stage, and further high-quality research is necessary. As booklets inherently risk being misplaced, inconvenient to carry, or overly lengthy – potentially discouraging full patient engagement – the NDB is best deployed as a supplementary resource alongside other educational approaches.

MMC achieves cross-modal efficacy through a synergistic effect model. This strategy has significant advantages in both ClinROs and PROs; however, evidence consistency in CIT is low, which may be directly related to the heterogeneity of different strategy combinations. Based on the research findings, we can speculate that MMC has important clinical value – supported by the 2025 US Multi-Society Task Force consensus and randomized controlled trial evidence (intervention: video + phone + routine education vs control: routine education; bowel preparation rate 92.3% vs 82.2%, P < .05).^[71,72] MMC should follow the principle of precise adaptation, tailoring communication methods based on population characteristics, while avoiding the risk of cognitive overload. However, previous studies have lacked in-depth exploration of strategy combination patterns and population adaptation strategies. Given that differences in the effectiveness of education strategies are essentially related to the complex mechanisms of patient behavior change, this study innovatively adopts the Capability-Opportunity-Motivation-Behavior (COM-B) model, integrating the characteristics of strategies across 3 dimensions: ability (knowledge/skill acquisition), opportunity (environmental support), and motivation (intrinsic drive).^[73] Based on population characteristics, the following strategy combinations are proposed: For populations with low digital literacy/elderly individuals, core obstacles are concentrated in limited abilities (difficulty decoding graphics and text, short-term memory decline) and limited opportunities (lack of smart devices). It is recommended to use structured face-to-face education combined with graphic and text manuals as the basic intervention, or face-to-face education combined with videos, overlaid with phone reminders to enhance compliance at the key node. Notably, providing simultaneous health education to elderly patients’ families can help build a social support network, significantly improving the bowel preparation rate.^[58] For the population with high digital literacy/younger individuals, the main challenges lie in limited opportunities (interrupted execution due to work interference) and insufficient motivation maintenance (procrastination tendency). In resource-rich scenarios, priority should be given to using AI-driven mobile health applications or VR technology to simulate the entire bowel preparation. In resource-constrained scenarios, a simplified combination of visual videos and SMS reminders is preferred.

4.1.2. Evidence gaps and future research directions

Evidence mapping revealed the following gaps and future research directions for research on health education regarding bowel preparation before colonoscopy: (1) Deficient methodological rigor: The limited number of high-quality SRs/MAs underscores the need for enhanced study designs. Future studies should ensure scientific transparency through prospective protocol registration and comprehensive documentation of excluded literature to minimize bias. (2) Nonstandardized intervention reporting: Incomplete methodological disclosures (e.g., timing/frequency specifications) compromise the validity of conclusions and the feasibility of replication. Future protocols should require standardized documentation to ensure implementation fidelity and translational reproducibility. (3) Lack of psychological state assessment: Existing assessments do not consider psychological states such as anxiety and depression. Although a painless colonoscopy alleviates physical discomfort, its invasive nature may still trigger stress reactions, such as procedural anxiety. (4) Limited exercise adherence monitoring: SRs/MAs predominantly assess dietary and bowel cleanser compliance while neglecting exercise adherence. Standardized exercise adherence metrics should be incorporated into future research protocols. (5)Lack of comparative effectiveness data between strategies: Future studies should conduct high-quality comparative analyses of strategy combinations’ effectiveness while also exploring evidence-based tailored strategies for diverse populations. (6) Unaddressed cost-effectiveness analyses: Economic evaluations have not fully compared the implementation costs and medical service accessibility of bowel preparation strategies. Cost-benefit comparative evaluations should be conducted to support clinical evidence-based applications.

4.1.3. Strengths and limitations

The strengths of this study include the comprehensive search and rigorous methodological evaluation, as well as the use of visualization to present the current state of research. Additionally, we provide a detailed analysis of the characteristics of existing health education strategies and offer innovative suggestions for combining strategies in different scenarios. This study has some limitations: (1) Due to limitations in the information available in the included literature, this evidence map could not categorize and discuss intervention strategies based on the frequency of health education received; (2) Given the vast number of studies in this field, only SRs/MAs that provide more comprehensive evidence and reliable results were included in this study; (3) The current literature on SRs/MAs related to the topic of this study is of low quality, which impacts the credibility of the conclusions; (4) Our findings are based solely on publicly available publications up to November 2024, and the results should be updated as new studies emerge; (5) Limiting the language to English and Chinese may introduce potential language bias.

4.2. Conclusion

This study successfully established a direct mapping relationship of “education strategy - outcome indicators” through evidence maps. The results demonstrated that education strategies have a better improvement effect on PROs than ClinROs, and SA and MMC have enormous clinical potential. Based on the Capability Opportunity Motivation Behavior model, this study proposes tailored strategy combinations according to the characteristics of different populations. These findings strongly advocate for conducting high-quality research to compare the effectiveness and generalizability of strategic portfolios, deepen exploration of population-tailored approaches, and balance clinical efficacy with resource utilization. This will facilitate the identification of optimal feasible intervention combinations to maximize patient benefits.

Author contributions

Conceptualization: Ningning Li, Wen Li.

Data curation: Mengdan Ye.

Formal analysis: Qihan Wu, Bingru Li.

Funding acquisition: Wen Li.

Investigation: Qihan Wu, Bingru Li.

Methodology: Ningning Li.

Project administration: Wen Li.

Software: Ningning Li.

Supervision: Wen Li.

Validation: Mengdan Ye, Yingying Wang.

Visualization: Mengdan Ye.

Writing – original draft: Ningning Li.

Writing – review & editing: Wen Li.

Supplementary Material

medi-104-e44644-s001.docx ^{(40.6KB, docx)}