The clinical characteristics of patients in the CPG and RPG subgroups are compared in Table 1. No significant differences were observed between these two groups for clinical features, including sex (P = 0.059), body mass index (BMI; CPG: 23.88 ± 3.32 kg/m2 vs. RPG: 23.66 ± 3.05 kg/m2, P = 0.315), American Society of Anesthesiologists Physical Status (ASA-PS; P = 0.346), operating time (minutes) (CPG: 166.31 ± 57.70 vs. RPG: 161.80 ± 39.93, P = 0.115), diabetes mellitus (DM, P = 0.302), non-tuberculosis (TB) pulmonary disease (P > 0.999), nephrology disease (P = 0.491), liver disease (P = 0.816), cerebral disease (P = 0.457), or TB (P = 0.770). However, significant differences between groups were observed for hypertension (HTN, P = 0.020) and cardiology disease (P = 0.046). According to pathologic reports, patients who underwent CPG had more advanced forms of gastric cancer than those who underwent RPG, as assessed by T-stage (P = 0.003), N-stage (P = 0.042), and total stage (P = 0.002). As a result, the CPG group had more extensive LND (P < 0.001), gastrectomy (P < 0.001), and omentectomy (P = 0.002) than the RPG group.

The RPG method significantly decreased CRPD3 values compared to the CPG method. (CPG: n = 1137, 90.13 mg/L, 95% CI 86.87–93.30 vs. RPG: n = 305, 75.49 mg/L, 95% CI 63.95–81.69, P < 0.001) (Fig. 3a). In subgroup analysis, the RPG method significantly decreased CRPD3 values in the STGD1+ group (CPG: n = 635, 84.49 mg/L, 95% CI 80.53–88.45 vs. RPG: n = 236, 70.01 mg/L, 95% CI 63.92–76.09, P < 0.001, Fig. 3b). No significant advantages of using RPG over CPG were observed for the STGD2 (CPG: n = 305, 92.51 mg/L, 95% CI 86.48–98.55 vs. RPG: n = 49, 91.03 mg/L, 95% CI 73.06–109.01, P = 0.861), TGD1+ (CPG: n = 118, 101.11 mg/L, 95% CI 90.66–111.56 vs. RPG: n = 11, 101.50 mg/L, 95% CI 60.72–142.28, P = 0.983), or TGD2 (CPG: n = 61, 117.35 mg/L, 95% CI 102.02–132.70 vs. RPG: n = 9, 102.79 mg/L, 95% CI 60.33–145.25 P = 0.495) groups.

Subgroup analysis identifying the advantage of reduced port gastrectomy. a CRPD3 reduction. b Subgroup analysis of CRPD reduction. c Bowel recovery. d The hospital stay. CPG conventional port gastrectomy, RPG reduced port gastrectomy, cSTGD1+ conventional port subtotal gastrectomy with D1+ dissection, rSTGD1+ reduced port subtotal gastrectomy with D1+ dissection, cSTGD2 conventional port subtotal gastrectomy with D2 dissection, rSTGD2 reduced port subtotal gastrectomy with D2 dissection, cTGD1+ conventional port total gastrectomy with D1+ dissection, rTGD1+ reduced port total gastrectomy with D1+ dissection, cTGD2 conventional port total gastrectomy with D2 dissection, rTGD2 reduced port total gastrectomy with D2 dissection

The RPG method significantly shortens bowel recovery in the STGD1+ (CPG: n = 651, 2.86 days, 95% CI 2.81–2.92vs. RPG: n = 236, 2.66 days, 95%CI, 2.57–2.74, P < 0.001, Fig. 3c), STGD2(CPG: n = 305, 3.03 days, 95% CI 2.95–3.12 vs. RPG: n = 48, 2.44 days, 95% CI 2.22–2.65, P < 0.001), and TGD1+ (CPG: n = 118, 3.05 days, 95% CI 2.92–3.18 vs. RPG: n = 11, 2.55 days, 95% CI 2.08–3.01, P = 0.026) groups. There was no significant benefit of RPG compared with CPG for the TGD2 group (CPG: n = 61, 2.92 days, 95% CI 2.74–3.10 vs. RPG: n = 9, 2.67 days, 95% CI 2.12–3.21, P = 0.313).

Similar to the above results, the RPG method significantly shortens the hospital stay in the STGD1+ (CPG: n = 653, 5.53 days, 95% CI 5.41–5.66 vs. RPG: n = 236, 4.39 days, 95% CI 4.21–4.56, P < 0.001, Fig. 3d), STGD2 (CPG: n = 305, 5.67 days, 95% CI 5.50–5.84 vs. RPG: n = 49, 4.80 days, 95% CI 4.33–5.26, P < 0.001), and TGD1+ (CPG: n = 118, 6.47 days, 95% CI 6.11–6.82 vs. RPG: n = 11, 4.82 days, 95% CI 3.88–5.76, P = 0.007) groups. No significant advantages of using RPG over CPG were observed for the TGD2 group (CPG: n = 56, 5.52 days, 95% CI 5.11–5.92 vs. RPG: n = 9, 5.33 days, 95% CI 4.00–6.67, P = 0.740).

On univariate analysis (Table 2), the following factors were significantly associated with increasing CRPD3: male sex, age ≥ 65 years, BMI ≥ 25 kg/m2, ASA-PS ≥ 2, CPG, D2 dissection, T-stage > 1, TG, total omentectomy, node metastasis, HTN, and DM. Multivariate analysis was performed to estimate each parameter’s contributions to changes in CRPD3. Significant factors associated with increasing CRPD3 were male sex (P < 0.001, reference value female sex), BMI ≥ 25 kg/m2 (P < 0.001, reference value BMI < 25 kg/m2), CPG (P = 0.0109, reference value RPG), D2 dissection (P = 0.0174, reference value D1+ dissection), TG (P < 0.001, reference value STG), total omentectomy (P = 0.042, reference value partial omentectomy), HTN (P < 0.001), and DM (P = 0.0138). Finally, linear regression was performed to estimate CRPD3 levels after minimally invasive gastrectomy to treat gastric cancer (Table 3). Every variable included in the linear regression model was significant: male sex (32.56 ± 2.79 mg/L, P < 0.001), BMI ≥ 25 kg/m2 (10.21 ± 2.94 mg/L, P < 0.001), RPG (− 8.97 ± 3.39 mg/L, P = 0.008), D2 dissection (9.26 ± 3.10 mg/L, P = 0.003), TG (14.79 ± 4.97 mg/L, P < 0.001), total omentectomy (9.53 ± 4.67 mg/L, P = 0.041), HTN (11.84 ± 2.98 mg/L, P < 0.001), and DM (9.66 ± 3.85 mg/L, P = 0.012). The pie chart shows the proportional contribution of each parameter on the CRPD3. (Electronic Supplementary Material, Fig. 1).