Use of prokinetic agents in hospitalised adult patients: A scoping review

Gastrointestinal motility is important for adequate uptake of fluids and nutrition but is often impaired in hospitalised patients. Prokinetic agents enhance gastrointestinal motility and are prescribed for many hospitalised patients. In this scoping review, we aimed to systematically describe the body of evidence on the use of prokinetic agents in hospitalised patients. We hypothesised, that the body of evidence would be limited and derive from heterogeneous populations.


Editorial Comment
To assess the state of knowledge concerning prokinetic treatment in this context, this scoping review presents the breadth of reported clinical implementations, agents and outcomes in a systematic way. The authors demonstrate that more rigourous studies and stronger evidence are needed.

| INTRODUCTION
Adequate uptake of fluids and nutrition is essential, and malnutrition is associated with increased morbidity and mortality. 1,2 Delivery of nutrition depends on gastrointestinal (GI) motility, which is often impaired in hospitalised patients due to adverse drug effects, hyperglycaemia, immobility, impaired renal function, mechanical ventilation or as part of the underlying disease. 2,3 Prokinetic agents enhance GI motility and thereby improve propulsion of the intra-luminal content. 4 In patients outside the intensive care unit (ICU), prokinetic agents are used for a wide range of indications including gastroparesis, to prevent aspiration before anaesthesia, to promote GI function after surgery and before gastroscopy to improve visualisation. 5,6 In the ICU, prokinetic agents are primarily used to facilitate absorption of nutrients and drugs in enterally fed patients. 2,7 Different indications and outcomes have been assessed, but the evidence on the effects and safety of prokinetic agents has been questioned. 2 With this scoping review we aimed to systematically outline the body of evidence for the use of prokinetic agents in hospitalised patients, including those in ICU, to identify unanswered research questions. We hypothesised that the body of evidence would be limited and derive from heterogeneous populations.

| METHODS
We adhered to the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) statement 8 and the published protocol. 9 We posed the following research questions: • Which hospitalised adult patient populations receive prokinetic agents?
• Which prokinetic agents are used?
• What are the indications for the use of prokinetic agents?
• Which desirable and undesirable outcomes have been assessed?

| Eligibility criteria
We included all studies on prokinetic agents administered to adult hospitalised patients for any indication. The intervention and comparator of interest was any prokinetic agent compared with another, or no treatment or placebo. We included studies of any design except for literature reviews and case reports. We excluded studies on children, healthy subjects, animals as well as studies exploring functional dyspepsia, which we deemed beyond the scope of this review. Furthermore, we excluded studies addressing prokinetic agents for indications other than GI motility.

| Information source and search strategy
We systematically searched Medline, Embase, the Cochrane Library, Epistemonikos and clinicaltrials.gov. We tested and reviewed the search strategy before conducting the final search. Additionally, we manually screened the reference list of relevant studies and systematic reviews.
We identified several conference abstracts that were not published as full articles, and, where relevant, we contacted the authors to retrieve any unpublished articles/data. We updated the literature search on December 13, 2022. The search strategy for Medline is available in Data S1.

| Study selection
Reviewers used Covidence to organise and facilitate the study selection process. 10 Two reviewers (VC, EB) independently, and in duplicate, screened titles and abstracts of the identified studies and excluded irrelevant studies. Subsequently, we reviewed, in duplicate, the full text of potentially relevant studies. Any discrepancies were solved by discussion and consensuses, or arbitration with a third reviewer (MK, MHM). Articles in languages other than English, French, Italian and Spanish were translated using Google translate.

| Data extraction
Two authors (VC, EB) extracted data independently using a standardised, predesigned data extraction form. The extracted data comprised characteristics of the included studies (type, year, country, population, setting, interventions, comparator and reported outcome(s)).

| Outcome measures
Our outcomes of interest were all outcome measures provided in the included studies. We chose to exclude studies addressing GI tube placement as an outcome, since we found it beyond the purpose of this scoping review. We pragmatically classified outcome measures into patient-centred outcomes and non-patient-centred (surrogate) outcomes. 11

| Certainty of evidence
We used a modified Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach to assess the certainty of evidence. 12 The overall certainty of evidence was rated high, moderate, low or very low based on pragmatic judgement of the risk of bias based on study design, inconsistency, indirectness, imprecision and publication bias. We assessed risk of bias on a research question level, not on outcome level.

| Data synthesis
We summarised and presented data descriptively according to indications, and type of population (ICU/non-ICU) stratified by study design (clinical trial or observational study). Prokinetic agents were divided in groups according to drug class, and the percentage of studies assessing each prokinetic agent was reported.
Finally, outcomes were grouped based on our classification (i.e., patient-centred versus not).

| RESULTS
We included 102 studies with a total of 8830 patients ( Figure 1).

IdenƟficaƟon
Screening Eligibility F I G U R E 1 Preferred reporting items for systematic reviews and meta-analyses flow chart.

| Indications
In the ICU setting, the most studied indications were feeding intolerance with 40 studies (68%), 13 67,69,70 and to promote recovery of bowel dysfunction postoperatively, two studies (3%). 85,111 In the non-ICU setting, five groups of different indications were studied, with the majority investigating the effect of prokinetic agents before endoscopy to improve visualisation, 16 studies The modified overall certainty of evidence was judged to be low due to risk of bias, inconsistency and imprecision.

| Type of prokinetic agents
In total, 13 different prokinetic agents were assessed, and all agents except itopride, 26,85 TAK-954 19 and camicinal 17,24 were studied in both ICU and non-ICU settings ( Table 2). The most studied prokinetic agent in both ICU and non-ICU settings was metoclopramide (49%) 14 The administration of prokinetic agents varied considerably with respect to dose and frequency. Routes of administration used, included the intravenous, enteral, oral and intramuscular route. In one study, the agent was administered rectally 27

| Outcome measures
In total, 58 different outcomes were assessed in the ICU setting and 87 in the non-ICU ( Figure 3A,B Serotonin receptor agonists
The most reported non-patient-centred adverse events in both ICU and non-ICU setting were laboratory test results, including liver enzymes and electrolyte disturbances (n = 9). 17 In total, 13 different prokinetic agents were studied; some experimental drugs and others, commonly used agents. The most widely studied prokinetic agent was metoclopramide, which was prescribed for a broad range of indications.
According to the 2018 European guideline the recommended first-line prokinetic agent to treat feeding intolerance in the ICU is erythromycin. 118 We found erythromycin to be the second most studied prokinetic agent.
It has previously been described that treatment with erythromycin for 2-4 days is significantly associated with better enteral feeding tolerance. 118 However, treatment with prokinetic agents in general for feeding intolerance in the ICU has, in a previous meta-analysis, not been found to prevent the development of pneumonia, nor reduce LOS ICU or mortality. 2 Three of the studied prokinetic agents (cisapride, mosapride and prucalopride) have, in greater parts of the world, been withdrawn from the market and require special permission to use due to serious side effects, especially cardiac arrhythmias 119 This has led to a growing interest in finding new, safer prokinetic agents. 4 As previously outlined, the most studied indication in the ICU was feeding intolerance, but the definition varied significantly. In some studies, it was defined as a range of GI symptoms. In others it was defined as a specific gastric residual volume, ranging from 100 to 500 mL, measured using several different methods, for example, manual aspiration, gastric emptying scintigraphy, manometry and 13 C-octanoic acid breath tests. The varying definitions and measuring methods makes it difficult to compare studies and may explain why the prevalence of feeding intolerance within the same population varies between 4.6% and 86.1%. 120 Additionally, the clinical relevance of gastric residual volume is questionable. In 2016 the American Society for Parenteral and Enteral Nutrition (ASPEN) guidelines recommended that gastric residual volume should no longer be used to monitor tolerance of enteral feeding, due to a lack of correlation with the incidence of pneumonia, regurgitation and aspiration. 121 In the studies conducted outside the ICU, different indications for the use of prokinetic agents were reported, with most studies assessing use of prokinetic agents before endoscopy to improve visualisation. The use of erythromycin improves visualisation and decreases the need for a second look, but the evidence is limited to few studies. [122][123][124] We found considerable variation in the assessed and reported

| Strengths and limitations
The strengths of our scoping review include a comprehensive and systematic literature search with no language restrictions, a prepublished protocol 9 and adherence to the PRISMA-ScR statement. 8 Our review also has some limitations. First, we cannot be sure that our search string identified all relevant studies, although we also manually searched reference lists for relevant studies wherever possible. Second, trials in Chinese and Polish were translated using internet-based translating tools, which may have led to some misinterpretations. Third, some studies were only available in abstract form, and it was not possible to acquire further data. Finally, we used a modified GRADE approach to assess the certainty of evidence on a research question level, as previously done. 125,126

| CONCLUSIONS
In this scoping review, we found that the studies addressing prokinetic agents in hospitalised adults had considerable variations in indications, drugs and outcomes assessed. The overall certainty of evidence was judged to be low.

ACKNOWLEDGMENTS
None. The authors received no financial support for the research, authorship, and/or publication of this article.

CONFLICT OF INTEREST STATEMENT
The authors declare no conflicts of interest.