Effect of Postoperative Coffee Consumption on Postoperative Ileus after Abdominal Surgery: An Updated Systematic Review and Meta-Analysis

Background: Previous systematic reviews have not clarified the effect of postoperative coffee consumption on the incidence of postoperative ileus (POI) and the length of hospital stay (LOS). We aimed to assess its effect on these postoperative outcomes. Methods: Studies evaluating postoperative coffee consumption were searched using electronic databases until September 2021 to perform random-effect meta-analysis. The quality of evidence was assessed using the Cochrane risk-of-bias tool. Caffeinated and decaffeinated coffee were also compared. Results: Thirteen trials (1246 patients) and nine ongoing trials were included. Of the 13 trials, 6 were on colorectal surgery, 5 on caesarean section, and 2 on gynecological surgery. Coffee reduced the time to first defecation (mean difference (MD) −10.1 min; 95% confidence interval (CI) = −14.5 to −5.6), POI (risk ratio 0.42; 95% CI = 0.26 to 0.69); and LOS (MD −1.5; 95% CI = −2.7 to −0.3). This trend was similar in colorectal and gynecological surgeries. Coffee had no adverse effects. There was no difference in POI or LOS between caffeinated and decaffeinated coffee (p > 0.05). The certainty of evidence was low to moderate. Conclusion: This review showed that postoperative coffee consumption, regardless of caffeine content, likely reduces POI and LOS after colorectal and gynecological surgery.


Introduction
Postoperative ileus (POI), defined as the transient cessation of coordinated bowel motility, is a common cause of delayed return to normal bowel function after abdominal surgery (e.g., colorectal and gynecologic surgery), occurring in 10-15% of cases [1,2]. Delayed defecation associated with POI causes vomiting, bloating, and intolerance to food, and POI often leads to invasive interventions, such as nasogastric tube insertion [3]. POI increases postoperative length of hospital stay (LOS) and treatment-related costs [4,5]. POI and LOS are important postoperative outcomes because prolonged LOS and increased risk of morbidity due to POI have been shown to reduce patients' quality of life and increase hospital expenditures [4][5][6].
Coffee is the most widely consumed pharmacological substance worldwide [7]. Caffeine exerts anti-inflammatory effects on the gastrointestinal and cardiovascular systems, mediated by its antagonistic effects on A2A receptors on immune cells, such as T and B cells and macrophages [8,9]. Since the implementation of enhanced recovery protocols (ERPs), multimodal strategies have been used to improve the postoperative return of gastrointestinal function [10,11]. Recommendations regarding the use of postoperative coffee vary in various international ERPs [10,11]. Previous systematic reviews did not demonstrate that LOS and POI were statistically significantly reduced, because of the small number of trials [12][13][14][15]. In addition, it is unclear whether coffee or decaffeinated coffee is effective in treating POI [12].
Coffee, a popular and easily available beverage worldwide, could also be clinically significant if shown to prevent POI incidence in addition to shortening LOS. In terms of ERPs, colorectal and gynecological surgeries are treated similarly because of the manipulation of the bowel [10,11]. Therefore, the present updated systematic review and meta-analysis aimed to assess the effect of postoperative coffee consumption on POI after abdominal surgery, including colorectal surgery, cesarean section, and gynecological surgery.

Inclusion Criteria
Randomized controlled trials (RCTs) that assessed the effect of postoperative coffee consumption after abdominal surgery were included. No language, country, observation period, or publication year restrictions were applied. Review articles, case series, and case reports were excluded. The intervention of interest was postoperative 100-150 mL coffee consumption, three times per day, for 10-20 min. The control group consumed water, tea, or a placebo. The primary outcomes were time to first defecation (hours), LOS (days), and POI. The secondary outcomes were the time to first flatus (hours), the time to first bowel movement (hours), the time to tolerance of solid food (hours), and adverse events.

Search Method
The following electronic databases and trial registries were searched: MEDLINE (PubMed), Cochrane Central Register of Controlled Trials (Cochrane Library), EMBASE (Dialog) (Appendix B), the World Health Organization International Clinical Trials Platform Search Portal (ICTRP), and ClinicalTrials.gov (Appendix C). The reference lists were checked for studies, including international guidelines [10,11], as well as reference lists of eligible studies and articles citing eligible studies. The authors of the original studies were asked for unpublished or additional data if necessary.

Data Collection and Analysis
Two independent reviewers (J.W. and A.M.) performed screening, data extraction, and assessment of the risk of bias using the Risk of Bias 2 tool [17] and assessed the quality of evidence based on the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach [18]. Disagreements between the two reviewers were discussed, and if necessary, a third reviewer (K.K.) was consulted.
The relative risk ratios (RRs) and the 95% confidence intervals (CIs) were calculated for the binary variables, POI, and adverse events. The mean differences (MDs) and 95% CIs were calculated for the continuous variables, LOS (days), the time to first defecation (hours), the time to first flatus (hours), the time to first bowel movement (hours), and time to tolerance of solid food (hours). Intention-to-treat analysis was performed for dichotomous data as far as possible. For continuous data, missing data were not imputed based on the recommendations of the Cochrane handbook [19]. In cases where missing data were not known after contacting the original authors, the standard deviation was calculated using the method provided in the Cochrane handbook [19] or a previously validated method [20]. A random-effects meta-analysis was performed using Review Manager software (RevMan 5.4.2).

Assessment of Heterogeneity and Reporting Bias
Statistical heterogeneity was evaluated by visual inspection of the forest plots and calculating the I 2 statistic (I 2 = 0-40%, might not be important; 30-60%, moderate hetero-geneity; 50-90%, substantial heterogeneity; and 75-100%, considerable heterogeneity) [19]. When there was substantial heterogeneity (I 2 > 50%), we assessed the reason for the heterogeneity. Cochrane's chi 2 test (Q-test) was performed on the I 2 statistic, and a p-value less than 0.10 was defined as statistically significant. We searched the clinical trial registry system (ClinicalTrials.gov and ICTRP) to assess any reporting bias. Potential publication bias was evaluated through visual inspection of the funnel plots.

Overall Risk of Bias
Müller [21] Low Some concerns Some concerns Some concerns Some concerns Some concerns Dulskas [22] Some concerns Some concerns Some concerns Some concerns Some concerns Some concerns Piric [23] Some concerns Some concerns Some concerns Some concerns Some concerns Some concerns Göymen [24] Some concerns Low Low Some concerns Some concerns Some concerns Güngördük [25] Some concerns Low Low Some concerns High High Mohamed [26] Some concerns High High Some concerns Some concerns High Rabiepoor [27] Some concerns Low Low Some concerns Some concerns Some concerns Hasler-Gehrer [28] Low Some concerns Some concerns Some concerns Low Some concerns Hayashi [29] Low Low Low Some concerns Low Some concerns Bozkurt Koseoglu [30] Low Some concerns Some concerns Some concerns Low Some concerns Güngördük [31] Low Some concerns Some concerns Some concerns Low Some concerns Kanza Gül [32] Low Low Low Some concerns Some concerns Some concerns Parnasa [33] Low Some concerns Some concerns Some concerns Low Some concerns Table 3 summarizes the findings of the GRADE approach. The certainty of the evidence was low to moderate due to the high risk of bias and inconsistency.
Coffee reduced the time to first tolerance of solid food.
CI, confidence interval; MD, mean difference; RR, risk ratio. * The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). GRADE Working Group grades of evidence; High certainty: We are very confident that the true effect lies close to that of the estimated effect. Moderate certainty: We are moderately confident in the estimated effect. The true effect is likely to be close to the estimated effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the estimated effect is limited: The true effect may be substantially different from the estimated effect. Very low certainty: We have very little confidence in the estimated effect. The true effect is likely to be substantially different from the estimated effect. a Downgraded because of a high risk of bias. b Downgraded because of inconsistency due to substantial heterogeneity. c Downgraded because of imprecision due to the small sample size.

Time to First Flatus (Hours)
Coffee reduced the time to first flatus after abdominal surgery (MD −4.27 h; 95% CI = −8.28 to −0.26; I 2 = 96%) ( Figure A1). There was no statistically significant difference between colorectal surgery, cesarean section, or gynecological surgery in the subgroup test (p = 0.36).

Time to Tolerance of Solid Food (Hours)
Coffee reduced the time to tolerance of solid food after colorectal surgery, cesarean section, and gynecological surgery (MD −10.11 h; 95% CI = −14.26 to −5.95; I 2 = 95%) ( Figure A3).

Additional Analyses
In subgroup analyses of caffeinated vs. decaffeinated coffee ( Figures A5-A11), there were statistically significant differences between caffeinated and decaffeinated coffee for the time to first defecation (p = 0.02) and the time to tolerance of solid food (p = 0.04). However, when analyzed by surgery type, there were no statistically significant differences between caffeinated and decaffeinated coffee for the time to first defecation after colorectal surgery (p = 0.14) or cesarean section (p = 0.51) ( Figure A12) or for the time to tolerance of solid food after cesarean section (p = 0.35) ( Figure A13). The results of the sensitivity analysis, excluding studies using imputed statistics, were consistent with the original results except for the time to first flatus ( Figures A14-A16).
Regarding publication bias, the funnel plots were symmetric, suggesting a no-potentialno-publication bias ( Figure A17).

Discussion
This systematic review and meta-analysis demonstrated that postoperative coffee consumption likely reduces the time to first defecation, LOS, and POI after abdominal surgery. This trend is similar to the trends after colorectal and gynecological surgeries. Additionally, there was no difference in LOS and POI between caffeinated and decaffeinated coffee intake. This updated evidence is beneficial to both patients and surgeons regarding the practical endpoints of LOS and POI.
In previous systematic reviews [12][13][14][15], coffee accelerated the postoperative recovery of gastrointestinal function but did not reduce POI and LOS. The present review in 13 RCTs with 1246 patients extends the findings of previous reviews, showing a novel benefit of coffee for POI and LOS, in addition to standard ERPs. Preventing POI and shortening LOS can potentially affect the quality of life of patients and reduce their social costs by approximately 40-50% per patient [4][5][6]. In addition, preventing POI and shortening LOS has the potential to reduce hospital expenditures by US$750 million per year [4,5]. On average, the incidence of POI was 60% lower in the coffee group (POI: 6.9%) than in the non-coffee control group (16.5%). With postoperative coffee consumption, LOS was reduced by 1.5 days. Given that other consensus data show that ERPs reduce morbidity (RR 0.78) and LOS (−3.1 days) and opioid antagonists, which are frequently used to improve the postoperative course, reduce POI (32%) and LOS (−0.3 days) [34,35], the improved POI and LOS following coffee intake appear to be meaningful in the clinical setting.
The mechanism underlying the effect of coffee on POI is not fully understood. The factors may be caffeine and other substances in coffee, mainly phenolic antioxidants of chlorogenic acid [36]. Caffeine acts positively on inflammation, activating ryanodinesensitive Ca 2+ channels by releasing Ca 2+ from the sarcoplasmic reticulum and inhibiting cyclic guanosine monophosphate degradation, thereby promoting nitric oxide synthesis in the endothelium and enhancing caffeine-induced endothelium-dependent vasodilation [37][38][39]. Caffeine promotes postoperative recovery of gastrointestinal function through vasodilation [32,40]. Chlorogenic acid has beneficial effects on inflammation and pain [41]. Chlorogenic acid has an anti-inflammatory effect by potently inhibiting the production of tumor necrosis factor-α and interleukin-6 by peripheral blood mononuclear cells [42,43]. In addition, chlorogenic acid inhibits edema formation leading to pain and improves pain following inflammatory responses [42]. These effects may prevent POI and/or lead to shorter LOS.
There were no differences in the recovery of postoperative gastrointestinal function between caffeinated and decaffeinated coffee. These results suggest that caffeine and noncaffeine substances may have a positive effect on POI. In previous studies, both caffeinated and decaffeinated coffee similarly reduced the risk of various cancers and death from all causes [44,45]. The results of our study were in accordance with those of previous studies. However, caution should be exercised when interpreting the results due to the small number of studies involving decaffeinated coffee.
In the present review, there were no reports of adverse events related to coffee, although the caffeine group had a higher postoperative systolic blood pressure (mean 120 mmHg) than that of the control group (mean 100 mmHg) [32,46]. The amount of coffee used in this study was a common amount, and considering the safety of coffee, which is widely used, it is not a phenomenon that should be of great concern [47]. Whether hypertensive patients need to refrain from coffee consumption after surgery requires further study.
Our study showed that the certainty of the evidence was low to moderate because of the high risk of bias and inconsistency based on the GRADE approach. The overall risk of bias was high because the concealment of the allocation sequence was unclear, and the outcomes of interest, POI and LOS, were not included in the protocol. Further studies are needed to clarify allocation concealment and clarify outcomes, such as POI and LOS, in protocols. Additionally, the definitions of POI and LOS were unclear and may be affected by blinding and socioeconomic confounds. In the present review, many studies reported that POI was the indication for reinsertion of the nasogastric tube. POI and LOS should be clearly defined and recorded by blinded outcome assessors. When interpreting our results, heterogeneity in variables such as age, comorbidities, and surgical invasiveness in each population undergoing the procedure should be considered. In the case of cesarean section, the impact of coffee on LOS after cesarean section may be small because the hospital stay is short to begin with [48,49]. In the case of colorectal surgery, coffee had a relatively weak effect on POI, which may be due to other factors related to POI, such as postoperative exercise and nutrition [35,50]. This review has additional limitations. First, the dose-response relationship between coffee consumption and outcomes was not evaluated. In the studies included in this review, the amount of coffee consumed was 100-150 mL, three times per day over 10-20 min. Second, the characteristics of coffee consumers, such as the relationship between regular and non-regular coffee drinkers, have not been clearly reported. Third, our results may not be generalizable to all populations because the compounds in coffee may vary by region, bean type, roast, and brewing method. Furthermore, none of the studies included data collected from children or low-income countries.

Conclusions
The findings of this updated systematic review and meta-analysis indicate that postoperative coffee consumption, with or without caffeine consumption, may reduce POI and LOS after colorectal surgery, cesarean section, and gynecological surgery. The findings suggest that patients and surgeons should preferably use postoperative coffee to reduce POI. More RCTs are needed to verify the effect of postoperative coffee consumption because the evidence for its consumption is limited by variations in surgeries.

Conflicts of Interest:
The authors declare no conflict of interest in association with the present study.

Section and Topic Item Checklist Item Location Where Item Is Reported
Selection process 8 Specify the methods used to decide whether a study met the inclusion criteria of the review, including how many reviewers screened each record and each report retrieved; whether they worked independently; and if applicable, details of automation tools used in the process.

2, 3
Data collection process 9 Specify the methods used to collect data from reports, including how many reviewers collected data from each report; whether they worked independently; any processes for obtaining or confirming data from study investigators; and if applicable, details of automation tools used in the process.

2, 3
Data items 10a List and define all outcomes for which data were sought. Specify whether all results that were compatible with each outcome domain in each study were sought (e.g., for all measures, time points, analyses), and if not, the methods used to decide which results to collect.

2, 3 10b
List and define all other variables for which data were sought (e.g., participant and intervention characteristics, funding sources). Describe any assumptions made about any missing or unclear information.

2, 3
Study risk-of-bias assessment 11 Specify the methods used to assess the risk of bias in the included studies, including details of the tool(s) used; how many reviewers assessed each study and whether they worked independently; and if applicable, details of automation tools used in the process.

Effect measures 12
Specify for each outcome the effect measure(s) (e.g., risk ratio, mean difference) used in the synthesis or presentation of results.

2, 3
Synthesis methods 13a Describe the processes used to decide which studies were eligible for each synthesis (e.g., tabulating the study intervention characteristics and comparing against the planned groups for each synthesis (item #5)).

2, 3 13b
Describe any methods required to prepare the data for presentation; synthesis, such as handling of missing summary statistics; or conversions.

2, 3 13c
Describe any methods used to tabulate or visually display results of individual studies and syntheses. 2, 3 13d Describe any methods used to synthesize results and provide a rationale for the choice(s). If meta-analysis was performed, describe the model(s), method(s) to identify the presence and extent of statistical heterogeneity, and software package(s) used.

2, 3 13e
Describe any methods used to explore possible causes of heterogeneity among study results (e.g., subgroup analysis, meta-regression).

2, 3 13f
Describe any sensitivity analyses conducted to assess the robustness of the synthesized results.

2, 3
Reporting bias assessment 14 Describe any methods used to assess the risk of bias due to missing results in a synthesis (arising from reporting biases).

Overall Risk of Bias
Müller [20] Low Some concerns Some concerns Some concerns Some concerns Some concerns Dulskas [21] Some concerns Some concerns Some concerns Some concerns Some concerns Some concerns Piric [22] Some concerns Some concerns Some concerns Some concerns Some concerns Some concerns Güngördük [24] Some concerns Low Low Some concerns High High Mohamed [25] Some concerns High High Some concerns Some concerns High Rabiepoor [26] Some concerns Low Low Some concerns Some concerns Some concerns Hasler-Gehrer [27] Low Some concerns Some concerns Some concerns High High Hayashi [28] Low Low Low Some concerns Low High Güngördük [30] Low Some concerns Some concerns Some concerns High High Parnasa [32] Low Some concerns Some concerns Some concerns Low Some concerns

Overall Risk of Bias
Müller [20] Low Some concerns Some concerns Some concerns Some concerns Some concerns Dulskas [21] Some concerns Some concerns Some concerns Some concerns Some concerns Some concerns Piric [22] Some concerns Some concerns Some concerns Some concerns Some concerns Some concerns Güngördük [24] Some concerns Low Low Some concerns High High Mohamed [25] Some concerns High High Some concerns Some concerns High Hasler-Gehrer [27] Low Some concerns Some concerns Some concerns High High Bozkurt Koseoglu [29] Low Some concerns Some concerns Some concerns High High Güngördük [30] Low Some concerns Some concerns Some concerns High High Parnasa [32] Low Some concerns Some concerns Some concerns High High