Introduction: High-quality colonoscopy relies on adequate bowel preparation (BP), but inadequate bowel preparation (IBP) occurs in 25% of colonoscopies. Current meta-analyses on IBP risk factors are limited by a relatively limited number of studies and inclusion of data reported as unadjusted risk factors. Since then, numerous recent studies exploring risk factors for IBP have been published with inconsistent results. This meta-analysis aimed to precisely assess predictors of IBP.

Methods: We searched PubMed, Embase, and Web of Science databases from inception until July 2023. Only studies that used a multivariate model to assess risk factors for IBP in adult participants were included. Studies reporting only unadjusted risk factors were excluded. Pooled adjusted odds ratios (OR) with 95% confidence intervals (CI) for risk factors reported in 3 or more studies were obtained within a random-effects model.

Results: 146 studies with 242,263 participants who underwent colonoscopy were included. The IBP rate was 18.3%. Boston BP scale was the most used scale to assess BP quality. 31 unique risk factors were analyzed. Sociodemographic predictors of IBP were Medicaid insurance 2.12 (1.46-3.09), low education 1.41 (1.19-3.92), tobacco use 1.29 (1.21-1.38), black race 1.27 (1.03-1.56), male gender 1.19 (1.12-1.26), and age 1.01 (1.01-1.02), especially age older than 65 1.28 (1.05-1.55), BMI 1.12 (1.08-1.17). Comorbidity-related predictors of IBP were cirrhosis 2.44 (1.62-3.67), gastrointestinal motility disorders 2.34 (1.41-3.99), poor functional status 2.31 (1.73-3.09), stroke/dementia 2.04 (1.46-2.86), constipation 1.89 (1.66-2.16), diabetes mellitus 1.79 (1.60-2.01), ASA class greater than 3 1.76 (1.13-2.76), prior abdominopelvic surgery 1.58 (1.31-1.92), and heart disease 1.36 (1.10-1.69). Medication-related IBP predictors were tricyclic antidepressant use 2.30 (1.30-4.09), opioid use 1.65 (1.41-1.93), and calcium channel blockers 1.54 (1.09-2.18). BP/procedure-related IBP predictors were increased BP-to-defecation interval (OR 2.17 [1.03-4.56), non-adherence to low-residue diet 2.14 (1.80-2.55), lack of split-dose BP 1.97 (1.48-2.63), prior IBP 1.99 (1.46-2.70), inpatient status 1.74 (1.37-2.20), increased BP-to-colonoscopy interval 1.69 (1.43-2.00). White race, hypertension, IBD, prior gynecologic surgery, alcohol use, and surveillance colonoscopy were not associated with increased risk of IBP. All risk factors are summarized in Table 1 and significant risk factors are highlighted in Figure 1.

Conclusions: Our meta-analysis focused on adjusted, multivariate risk factors to provide precise estimates of the most important risk factors for IBP. Our findings could help develop a validated prediction model to identify high-risk patients for IBP, potentially improve colonoscopy outcomes, reduce need for repeat colonoscopies and the associated healthcare costs.

John Guardiola

Indiana University School of Medicine

Appearances