Appropriateness of gastrointestinal prophylaxis use during hospitalization in patients with acute myocardial infarction: Analysis from the China Acute Myocardial Infarction Registry

Abstract Background The current status of gastrointestinal prophylaxis (GIP) usage and its effects on hospitalized acute myocardial infarction (AMI) patients is not clear. We investigate the appropriateness of GIP usage and its relationship with clinical events in China. Hypothesis Appropriate use of GIP is not associated with increased adverse outcomes. Methods From January 2013 to September 2014, a total of 24 001 consecutive patients from 108 hospitals with AMI in China Acute Myocardial Infarction (CAMI) registry were analyzed. The appropriateness of GIP was evaluated using the current American College of Cardiology Foundation/American Heart Association (ACCF/AHA) and European Society of Cardiology (ESC) guidelines. The primary endpoint was in‐hospital gastrointestinal bleeding (GIB), while the secondary endpoints were in‐hospital and 2‐year follow‐up net adverse cardiovascular and cerebrovascular events (NACCE). Multivariate logistic regression analysis and Cox proportional hazard models were used to assess the effect of appropriate GIP. Results There were 16 413 (68.38%) AMI patients co‐medicated with GIP. Among 108 involved hospitals, only 35 (32.4%) hospitals prescribed more than 50% appropriate GIP. Totally, 59.7% (14 340) AMI patients received inappropriate GIP. Inappropriate GIP use was independently associated with use of GPIIb/IIIa receptor inhibitor and primary percutaneous coronary intervention (PCI). Moreover, appropriate GIP use was associated with decreased GIB risk (OR: 0.692, 95% CI: 0.507‐0.944, P = .0202) during hospitalization, while not with increased in‐hospital and 2‐year follow‐up NACCE. Conclusion The use of GIP is prevalent in patients with AMI in China but only 40% of hospitalized patients received appropriate GIP. Appropriate prophylactic therapy was associated with decreased GIB risk during hospitalization.

K E Y W O R D S acute myocardial infarction, appropriateness, co-medication, outcomes, proton pump inhibitors 1 | INTRODUCTION Antiplatelet and antithrombotic therapy are cornerstone in the treatment of acute myocardial infarction (AMI). 1 However, they have been associated with increased risk of major bleeding events, including gastrointestinal bleeding (GIB) and access site bleeding. GIB is related to an increased in-hospital mortality 2 ; the real-world rates of GIB vary from 1.3% to 3%. 3,4 The American College of Cardiology Foundation/ American Heart Association (ACCF/AHA) 5 and European Society of Cardiology (ESC) 6 guidelines recommended that gastrointestinal prophylaxis (GIP), especially proton pump inhibitors (PPIs), should be used in high-risk patients, such as those with a history of GIB, advanced age, concomitant use of warfarin, steroids, or Helicobacter pylori infection.
GIP use rate varies from 23% to 45% 7,8 as reported in large-scale, multicenter registry studies, with 30% to 60% GIP use were considered inappropriate. 9,10 In several cases, the use of PPIs may have a potential negative interference on antiplatelet effect of clopidogrel, because of the competitive inhibition of the CYP 2C19 isoenzyme, 11,12 with higher rates of adverse clinical events in some studies. 12 Our aim was to describe the current status of GIP use in AMI patients and assess the effect of appropriate GIP use on clinical outcomes in China.

| METHODS
The China Acute Myocardial Infarction (CAMI) registry is a prospective, nationwide, multicenter observational study of patients with AMI. The registry includes three levels of hospitals (provincial-, prefectural-, and county-level hospitals, representing typical Chinese governmental and administrative models) covering all provinces and municipalities across mainland China. The final inclusion criteria met the third Universal Definition for Myocardial Infarction (2012). 13 The CAMI registry was registered with ClinicalTrials.gov (NCT01874691) and was approved by the institutional review board of all participating hospitals. All patient data were protected at all times. Detailed descriptions about data management and quality control can be found in the methodological article about CAMI registry published previously. 13 The data analyzed in our research was from CAMI, which could provide a large-scale population.
We refer to the definition of appropriate GIP use according to the study of Morneau KM 14 and latest guideline. 6 Appropriate GIP in patients undergoing dual antiplatelet therapy (DAPT) was defined as the following: indicated for GIP and received PPIs or no indication and did not receive GIP. Inappropriate prophylaxis was defined as no indication represent yet received GIP or that GIP was indicated, but received incorrect prophylaxis (histamine-2 receptor antagonist [H2RA]) or no prophylaxis. 6 (Table S1) GIP indication was defined according to the guideline, 15 including advanced age (>75), concurrent use of anticoagulants, steroids, or nonsteroidal antiinflammatory drugs (NSAIDs), including aspirin and H. pylori infection. The primary endpoint was in-hospital GIB, which was defined as clinically evident GIB (gross hematemesis, heme positive coffee ground emesis, and heme positive melena). The secondary endpoints were GIB and net adverse cardiovascular and cerebrovascular events (NACCE) during a 2-year follow-up; NACCE was a composite of all-cause death, myocardial infarction, stroke, and BARC 3 or 5 bleeding. 16 Post-discharge study follow-up was conducted via centralized telephone interviews by trained personnel at 30 days, 6 months and 2 years. GIP use was identified at study baseline and at each study follow-up. Patients were excluded if they had changed their initial GIP status (appropriate or inappropriate) during the follow-up.
Patients with incorrect age and missing GIP data were excluded.   Table 1. Patients treated with GIP were inclined to be female, older, with a higher Killip class on admission and with a history of stroke, peptic ulcer, and GIB. On hospitalization, they were often treated with GPIIb/IIIa receptor inhibitor, heparin/low molecular weight heparin (LMWH), and DAPT. When compared with patients treated with PPIs, those treated with H2RA were likely diagnosed with non-ST elevated myocardial infarction, and proportion of patients who underwent thrombolysis was higher than patients treated PPIs.

| Appropriateness of GIP use
Overall, 9086 (40.36%) patients received appropriate GIP use. The appropriate use of GIP was greater than 50% in 35 (32.4%) hospitals and only one hospital prescribed over 90% appropriate GIP to AMI patients (Supplementary Figure S1B).

| DISCUSSION
The main findings of the present study were as follows: (a) The GIP use rate during hospitalization in AMI patients was 68.38% and the F I G U R E 1 Flow chart for selection of study population T A B L E 1 Baseline clinical data in patients with and without GIP

| The impact of appropriate GIP use on clinical outcomes
Although several studies showed that GIP, especially PPIs, can reduce GIB risk, 20,21 pharmacokinetic studies demonstrated a decrease in platelet inhibition of DAPT. 22,23 Because both PPIs and clopidogrel are prodrugs that need to be metabolized by cytochrome P450 (CYP) in liver, co-medication would inhibit the conversion of clopidogrel to its active metabolite and altered its antiplatelet properties, which may increase the risk of ischemic events. 12,24 Therefore, routine use of GIP is not recommended by the current guideline 5 for patients with low risk of GIB, who have much less potential to benefit from prophylactic therapy. 15 Previous studies have focused on the effect of PPIs, but the results were inconsistent regarding their safety and effectiveness. 25 However, few studies have investigated the impact of appropriate prophylactic therapy on clinical outcomes. In our study, those treated with appropriate GIP could benefit from prophylactic therapy, with decreased GIB risk (OR: 0.692, 95% CI: 0.507-0.944, P = .0202). Furthermore, appropriate use was not associated with an increased risk for NACCE (OR: 0.942, 95% CI: 0.839-1.059, P = .3195). Some clinical studies showed that inappropriate GIP use independently associated with adverse outcomes. 9,18 In our study, the results indicated that appropriate GIP use did not increase the risk for ischemic events, such as MI, stroke, and other endpoints. This was because patients who were prescribed GIP without indication were enrolled inappropriate group, while they were classified as control group against patients without GIP in previous study. These patients without indication benefited few from GIP, while the side effect of co-medication seemed prominent which increased the patients' number of ischemic events in inappropriate group. In contrast, patients without indication who did not receive GIP were classified as appropriate group, which lower the ischemic events in this group. Therefore, in the final regression analysis, the OR was less than one for MI and stroke (without significant statistic difference). However, this emphasized the importance of compliance to the guideline.
Although, we did not find decreased GIB risk after the 2-year follow-up, the NACCE events were similar between the appropriate GIP use and inappropriate use groups. This indicated that appropriate GIP use only play a role in decreasing GIB risk during hospitalization.
It also suggested that concomitant use of GIP, when prescribed appropriately in patients receiving DAPT, was not associated with adverse clinical outcomes in long-term follow-up. These results were consistent with those of another study involving PCI patients. 17 T A B L E 2 In-hospital endpoints incidence and adjusted OR among DAPT patients  26 further studies should identify the best PPI for AMI patients in China. In addition, GIP was not prescribed to patients routinely after discharge in our study, therefore we failed to find the association between long-term outcomes and PPIs co-medication.

| CONCLUSION
The GIP use is prevalent in patients with AMI in China while only 40% hospitalized patients received appropriate GIP. Appropriate prophylactic therapy was associated with decreased GIB risk during hospitalization. Clinicians should pay more attention to latest guidelines and prescribe appropriate GIP to AMI patients.