To the best of our knowledge, this is the first study to explore the prevalence and characteristics of inappropriate prophylactic use of PPIs during the perioperative period and its associated factors. In our study, the inappropriateness rate of perioperative prophylactic use of PPIs was 27.75%, which was lower than in previous studies (Chen et al. 2022; Zhang et al. 2021; Bez et al. 2013). Different departments, study population size, different evaluation criteria, and study duration may lead to different inappropriateness rates (Liu et al. 2021). For example, previous studies had focused on patients admitted to specific surgical departments; however, our research was focused on patients in the perioperative period, which was also our biggest highlight. Meanwhile, the underprescription of PPIs was not considered in our study because our study population was limited to patients who had already used PPIs to prevent perioperative SRMD. In addition, we had taken some measures to improve the rationality of perioperative prophylactic use of PPIs prior to the study.

In our study, the most common problem was drug use without indication, which was also a major problem faced by other studies (Ali et al. 2019; Chen et al. 2022; Bez et al. 2013). Owing to concern about the risk of perioperative gastrointestinal bleeding, clinicians often prescribe PPIs to patients. In fact, the risk of perioperative gastrointestinal bleeding is only 4%, so there is no need to prescribe PPIs to prevent SRMD in low-risk perioperative patients (Li et al. 2022). Superior efficacy for acid suppression and the availability of generic formulations have led prescribers to favor the use of PPIs. This has also led to the overuse and abuse of PPIs (Savarino et al. 2018). A total of 25–70% of patients lack indications for PPIs use, resulting in unnecessary expenditure of close to £2 billion each year (Forgacs and Loganayagam 2008). In addition, overuse and abuse of PPIs are associated with a variety of adverse events in patients, such as acute kidney injury, Clostridioides difficile infection, pneumonia, and bone fractures (Savarino et al. 2017). Therefore, PPIs should only be used in patients with risk factors for SRMD during the perioperative period based on domestic and foreign guidelines and expert recommendations (American Society of Health-System Pharmacists 1999; Hospital Pharmacy Committee of Chinese Pharmaceutical Association 2020; Writing Group of Expert Consensus on the Preventive Application of Proton Pump Inhibitors 2018; Bai et al. 2018; National Health Commission of the People’s Republic of China 2020). Criteria for appropriate use include one serious risk factor or two potential risk factors, as shown in Table 1.

In this study, inappropriate usage and dosage were mainly reflected in the frequency of medication. The current literature showed that there was a contradiction in the frequency of administration, which focused on once or twice a day (American Society of Health-System Pharmacists 1999; Hospital Pharmacy Committee of Chinese Pharmaceutical Association 2020; Writing Group of Expert Consensus on the Preventive Application of Proton Pump Inhibitors 2018; Bai et al. 2018; National Health Commission of the People’s Republic of China 2020). PPIs maintain a longer-lasting acid inhibition, which is due to the fact that the proton pump cannot be recovered once it is deactivated, and its acid secretion can only be restored after the formation of a new proton pump. From the perspective of pharmacodynamics, the acid inhibition effect of PPIs can be maintained for 16–18 h (Savarino et al. 2018). Based on the above considerations, PPIs were administered once a day as an evaluation criterion in this study.

There is no clear standard for the duration of PPIs, but it is suggested that if patients can tolerate adequate enteral nutrition or have taken food, the clinical symptoms begin to improve as indications for drug withdrawal. Our study showed that 12 (9.2%) patients had an inappropriate duration of medication, which was mainly reflected in patients who continued to use PPIs even after discharge. Although the proportion of this problem is small, we should pay attention to it. Long-term use of PPIs may increase the risk of adverse reactions, especially in the absence of guidance from doctors or pharmacists after discharge. In our study, 78.4% of patients received intravenous PPIs, but only 3.1% of patients had a problem with the inappropriate route of administration, which was significantly lower than in previous studies (Li et al. 2022). This was attributed to the fact that patients admitted to otorhinolaryngology-head and neck surgery, neurosurgery, thoracic surgery, and other departments could not take medication orally or had gastrointestinal dysfunction after surgery.

Our study showed that oral dosage form of PPIs, discharge medication of PPIs, and junior doctors were associated with an increased risk of inappropriate prophylactic use of PPIs during the perioperative period, which was another highlight of this study. To our surprise, up to 87 of the 102 patients who received oral dosage form of PPIs had a problem with inappropriate prophylactic use of PPIs. In response to this study’s findings, a number of changes will be implemented in our hospital to address inappropriate use of PPIs for SRMD prophylaxis. Clinical pharmacists will strengthen training for doctors on the oral dosage form of PPIs. In particular, we recommend the prescriber to evaluate the need for ongoing PPI therapy at the time of the switch from injectable to oral and discontinue the PPI if it is no longer needed. Of the 87 patients who received inappropriate oral PPI orders, 48 were due to inappropriate discharge medication. It should be given more attention that discharge prescriptions make up 36.6% of the inappropriate orders. This result again confirms that inappropriate PPIs continuation upon discharge is a common issue despite several guidelines and ongoing global attention in recent years to highlight the risks versus benefit. The inappropriate PPIs continuation upon discharge exposes patients to excess risk of adverse events. Clinical pharmacists will rigorously review the surgical discharge medication of PPIs. In teaching hospital settings, it is often the primary responsibility of the junior doctor to enter medication orders after reviewing the case with the attending doctor. We will continue to explore the reasons behind the association between junior doctors and inappropriate prophylactic use of PPIs. In view of the current result, in addition to the training and education of doctors, especially junior doctors, adding a reminder function to the doctor’s order system may be a good approach (Clarke et al. 2021; Fan et al. 2023). When surgeons prescribe PPIs, the system will automatically pop up a reminder interface, including the indication of PPIs, usage and dosage, and the course of treatment, which will help surgeons use PPIs rationally.

The concomitant use of antithrombotics was associated with a decreased risk of inappropriate prophylactic use of PPIs during the perioperative period. Eid also found a similar conclusion that the combination of anticoagulants had a protective effect on the rational use of PPIs (Eid et al. 2010). However, another study showed that the concomitant use of aspirin or anticoagulants promoted inappropriate stress ulcer prophylaxis (Issa et al. 2012). Other studies had found that glucocorticoids were associated with the irrational use of PPIs, but our study did not find a similar conclusion (Li et al. 2022; Schepisi et al. 2016). The inappropriate prophylactic use of PPIs was reduced as the length of postoperative hospital stay (longer than 15 days) lengthened. This may be due to the increased risk factors for prophylactic use of PPIs in patients with prolonged hospital stay after surgery. Previous studies had explored the relationship between the length of hospital stay and the irrational use of PPIs, which was different from our study (Li et al. 2022; Mayet 2007).

Our study has the following limitations. First, this was a retrospective, single-center study. However, we believe our result is representative based on our hospital scale. Second, the study population size was small. The outbreak of the novel coronavirus reduced the number of inpatients during the study period. Third, there was a large percentage of otorhinolaryngology-head and neck surgery patients, which might not be representative of surgical patients in general. Finally, our study did not assess the underprescription of PPIs and track patient comorbidities and outcomes such as bleeding or other adverse events.