Table 3.

Characteristics of operated patient groups, variables studied, and limitations of each reviewed article

Authors	Surgeries, n	EEN pts, n	No-EEN pts, n	Variables similar between study groups	Variables significantly different between groups	Study limitations	Newcastle- Ottawa score
Li et al. [57]	780	219 - group 4*	332 - group 1 †29 - group 2 ‡ 128 - group 3 §	Pt gender and age at diagnosis and surgery, BMI, previous CD surgeries, disease duration, location and behavior, perforation/abscess, preoperative CSTs, immunomodulatory and anti-TNF treatment (compared with groups 2 ‡ and 3 §), surgical indication, approach and main procedure, operative time, and blood loss; groups were not matched	Group 4* had > drug-free interval than groups 2‡ and 3 §, at surgery	Retrospective study; no sample size calculation; pts with digestive intolerance may be sicker than EEN pts	Poor quality
Heerasing et al. [58]	114	38	76 - straight to surgery	Pt ethnicity, gender, age, smoking status and comorbidities, previous surgeries, CD duration, Montreal classification, abscesses, weight, preoperative serum albumin, CSTs, immunomodulatory and anti-TNF treatment, and surgical indication and approach; groups were matched for age at diagnosis, disease duration and behavior, and type of surgery	EEN pts had < BMI and > baseline disease activity and CRP, and more often ATB at surgery	Retrospective data collection; small sample size; missing data on CD course and severity; incomplete matching of groups	Fair quality
Beaupel et al. [59]	56	35 high-risk II	21 low-risk - straight to surgery	Pt gender, age, smoking status, BMI, CD duration, perianal lesions and previous intestinal resection, serum albumin, immunomodulatory and biologic therapy, type of surgery and approach, anastomotic technique, and intraoperative findings; groups were not matched	High-risk EEN pts had > rate of obstructive symptoms and/or weight loss >10% and/ or penetrating CD and/or CST therapy and > CD intractability	Retrospective study; small sample size; no control groups with different nutritional support; biased prescription	Poor quality
Zhu et al. [60]	74	46	28 - intolerant or noncompliant with EEN	Pt gender, HGB and WBC, previous surgeries, CD location and behavior, perianal lesions, CSTs, immunomodulatory and anti-TNF treatment at time of abscess diagnosis, abscess location, number of abscesses >3 cm, surgical approach, operative time, and intraoperative blood loss; groups were not matched	EEN pts had > serum albumin and < ESR and CRP at surgery	Retrospective study; small sample size; withdrawal of EEN before surgery in some pts and initiation of immunomodulatory or biological therapy after abscess resolution also before surgery, in others	Poor quality

EEN, exclusive enteral nutrition; pt(s), patient(s); BMI, body mass index; CD, Crohn's disease; CST, corticosteroid; >, higher; <, lower; CRP, C-reactive protein; ATB, antibiotic; HGB, hemoglobin; WBC, white blood cells; ESR, erythrocyte sedimentation rate.

^*Group 4 - preoperative EEN and immunosuppressant-free interval (CSTs, immunomodulators, and biologics) less than 8 weeks.

^†Group 1 - without preoperative immunosuppressant therapy within 8 weeks of surgery.

^‡Group 2 - immunosuppressant therapy at surgery.

^§Group 3 - preoperative immunosuppressant-free interval less than 8 weeks.

^IIHigh-risk patients: obstructive symptoms and/or penetrating CD and/or CST therapy and/or weight loss >10%. Criteria for application of the Newcastle-Ottawa Quality Assessment Scale to cohort studies are detailed in the supplementary material (Table 6).