COLORECTAL SURGERY Does milk of magnesia impact length of hospital stay after major colorectal resection

Backgrounds: Milk of magnesia (MoM) has been reported to accelerate return of bowel function following surgery. However, there is insuf ﬁ cient evidence regarding the impact of MoM on postoperative recovery after colorectal surgery. We aimed to determine the impact of MoM on postoperative length of stay in patients undergoing colorectal surgery. Methods: All patients who underwent colorectal resection without an ileostomy between 2015 and 2018 were included. Patients were divided into two groups based on whether postoperative MoM (MoM) was administered or not, according to surgeons ’ prescribing preferences. Consecutive patients of surgeons who prescribe MoM were included in the MoM group, while consecutive patients of surgeons who prefer not to prescribe MoM served as the control group. Age, gender, preoperative comorbidities, surgical approach, length of stay, readmission, post-operative complications and mortality were evaluated and compared between the groups. Results: A total of 3292 patients were included; 523 (15.9%) patients were prescribed MoM. Patients in the MoM group were found to be signi ﬁ cantly older, with higher BMI and ASA class, and more often undergoing operations for cancer resection or colostomy creations, than the control group. Postoperative complications were comparable between the groups. On multivariable linear regression, MoM use was associated with a 14.1% reduction in length of stay (MoM group 4 (2; 8), control group 5 (3; 8 P = 0.006)). Conclusion: MoM as adjunct medication in the postoperative period following colorectal surgery is associated with reduced length of stay, without an increase in postoperative complications.


Introduction
Delayed return of gastrointestinal function increases length of hospital stay (LOS) and postoperative complications. To date, a number of agents such as chewing gum, prokinetics (metoclopramide, domperidone, cisapride, erythromycin), cholinergic agonists, mu receptor antagonists, and oral magnesium oxide have been studied to accelerate return of bowel function and gastrointestinal motility. 1,2 MoM is a magnesium hydroxide based, low-cost Mg supplement (wholesale cost of one dose MoM is 10 cent) without significant side effects. Although it has been reported that aggressive bowel stimulation with milk of magnesia (MoM) and biscolic suppositories are associated with early return of bowel movements, there is insufficient evidence regarding impact of MoM on postoperative bowel function, complications and LOS. 3,4 The aim of this study is to determine the impact of MoM on postoperative length of stay in patients undergoing colorectal resection.
We hypothesized that MoM as an adjunct medication after colorectal surgery is associated with a reduction in post-operative length of stay without increasing morbidity or readmissions.

Methods
All patients who underwent colorectal resection without an ileostomy for any indication, between 2015 and 2018 at the Cleveland Clinic, Department of Colorectal Surgery were included in this retrospective study. This study was approved by the institutional review board and informed consent was obtained from the patients.
Patients were divided into two groups based on whether they received MoM postoperatively or not. Consecutive patients of surgeons who prescribe MoM were included in the MoM group, while consecutive patients of surgeons who's preference was not MoM as part of the postoperative care served as the control group. In patients who received MoM, 15-30 ml of MoM were administered once daily starting on postoperative day #1 and continued until the patient had a bowel movement. All patients were discharged based on established discharge criteria, including tolerance of a solid diet and after the passage of flatus and/or bowel movements were documented. During the study period, a comprehensive Enhanced Recovery Protocol (ERP) was instituted at our institution, starting in October 2016. Hence, patients were divided into subgroups based on whether they were operated before the implementation of ERP or not. This allowed us to evaluate the effect of MoM independently and within the context of an ERP.
Exclusion criteria were an underlying preoperative intestinal motility disorder, undergoing any procedure without resection of colon or rectum and any procedure with ileostomy construction since pro-motility agents are usually avoided in patients with ileostomy. Age, gender, American Society of Anesthesiology (ASA) classification, preoperative comorbidities, procedure type, surgical approach, operative time, postoperative hospital length of stay (LOS), readmission, postoperative complications and mortality were evaluated and compared between the groups.

Statistical analysis
All relevant patient characteristics, operative and postoperative data were presented as mean (standard deviation), median (25th, 75th percentiles) or frequency (percent). Student's t-test or nonparametric Wilcoxon rank sum tests were used for continuous factors and chi-squared test or Fisher exact test was used to compare categorical variables between MoM and control groups based on normalcy of distribution and cell counts. Based on clinical expertise, extension of resection (colectomy vs. coloproctectomy), diagnosis (cancer vs. others) and patient characteristics including age, gender and ASA physical status class were adjusted in the model. The associations between taking MoM and the occurrence of ileus and readmission were assessed using multivariable generalized estimating equation (GEE) logistic regression models, adjusting for the same covariates mentioned above. Given the limited number of mortality cases, it was compared between MoM and control groups with an unadjusted GEE logistic regression model. All comparisons were made at a significance level of 0.05, and all analyses were performed with R version 3.6.1.

Results
A total of 3292 patients were included; 523 (15.9%) patients received MoM. Patients in the MoM group were found to be significantly older, with higher BMI and ASA class (Grades 3 and 4), and more often undergoing operations for cancer resection or colostomy creations, than the control group. Postoperative complications were comparable (Table 1).
Results from the multivariable linear mixed effect model evaluating the association between MoM administration and LOS are shown in Table 2. After adjusting for age, sex, ASA class, extension of resection (colectomy vs. coloproctectomy), surgery date (before vs. after ERP implementation), and diagnosis (cancer vs. others): MoM administrations was associated with a 14.1% reduction in median LOS compared with the control group (MoM group 4 (2; 8), control group 5 (3; 8 P = 0.01)). Summarized in Tables 3 and 4 are the results of GEE logistic regression models to test the effect of taking MoM on the occurrence of ileus and readmission. After adjusting for age, sex, ASA class, extension of resection, before vs. after ERP group, and diagnosis category, there was no significant association between MoM and occurrence of ileus (OR = 1.06, P = 0.73) or readmission (OR = 1.15, P = 0.42).

Discussion
This large, single-institution retrospective study demonstrated that MoM as adjunct medication in the postoperative period following colorectal surgery was associated with reduced LOS, without an increase in postoperative complications. After adjustment for confounders, MoM use was associated with a 14.1% reduction in LOS without an increase in occurrence of ileus, readmission or mortality.
Our ERP pathway was systematically and completely implemented in our department by 2016. Our subgroup analyses shown that before     the implementation of ERP, MoM is significantly associated with reduced LOS (median: 5 vs. 6 days, P = 0.04). However, after the implementation of ERP, though LOS was reduced in the MoM group (median: 3 vs. 4 days, P = 0.34), the difference was not statistically significant. Although length of stay was reduced in the MoM group, postoperative complications were not increased. Ileus and readmission rates were similar between the groups both before and after the era of ERP. We think that MOM perhaps only works superiorly in patients not on an ERP or that the subgroup analysis was not powered to show a significant difference.
Impact of magnesium-based laxatives and ERP on LOS have been studied previously. In a study assessing the transit rate of the gastrointestinal tract by a scintigraphic method, it has been shown that gastrointestinal functions were returned earlier in patients undergoing colonic resections with early oral nutrition and laxatives. [5][6][7][8] More recent data show that ERP is safe and effective tool to optimize postoperative recovery. 9 It has been well established that ERP improve postoperative outcomes and reduce LOS without increasing postoperative complications and readmissions. 10 However, there is lack of evidence supporting the routine use of magnesium-based laxatives as an adjunct to ERP despite being used frequently after colorectal surgery by many surgeons. 11 On the other hand, some surgeons may be apprehensive using laxatives in the early postoperative period especially after construction of a new anastomosis which could potentially lead to risk of leak and sepsis. This study demonstrated that there is no significant difference between the groups in terms of anastomotic leak (P = 0.43). Further studies are required to prospectively institute safety and efficacy of MoM use especially in regard to anastomotic leaks after minimally invasive colorectal resections with anastomosis.
In a prospective trial, Fanning et al., 3 stated aggressive bowel stimulation with MoM and biscolic suppositories was associated with early return of gastrointestinal function and early discharge following radical hysterectomy without an increase in postoperative complications. Hendry et al. 12 conducted a randomized clinical trial in patients undergoing hepatic resection to evaluate the effectiveness of magnesium hydroxide and/or oral nutritional supplements within an ERAS protocol in promoting early return of gastrointestinal function. They emphasized that laxative agents can provide early return of colonic function in patients undergoing hepatic resection. However, these findings are limited by small sample size and needed to be confirmed in patients with an intestinal anastomosis as stated above.
Andersen et al. 13 conducted a double-blinded, placebo-controlled randomized study to assess the impact of magnesium oxide on the gastrointestinal function of patients who underwent colonic resection with a primary anastomosis. Contrary to our findings, there were no significant differences between the groups in terms of LOS. However, this study was limited by small sample size and only open colectomies, which may have limited the benefits of magnesium oxide.
Our study has several limitations. The major limitation of the study is that surgeons who prescribed MoM routinely may be more predisposed to discharge patients earlier and thus surgeon factor may have an impact on the length of stay. However, standardized discharge criteria, with return of bowel function as an established criterion, should in theory prevent treatment bias to influence LOS. Second, although, our study has shown that MoM shortens LOS only in patients not on an ERP, subgroup analysis may not be powered enough to show a significant difference. Last, heterogeneity between the groups such as increased stoma formation or emergent patients may affect results even though multivariate analysis has been performed. Further studies with more homogeneous groups may yield better results.
In conclusion, MoM as an adjunct medication in the postoperative period following colorectal surgery was associated with a reduced LOS, without an increase in postoperative complications. Given its low cost and excellent safety profile, we recommend use of MoM to be considered a part of ERP after colorectal surgery.