Table 3 Barriers and facilitators to guideline-concordant PPI use organized by the corresponding SEIPS element

Tools and technology

PPIs are low priority medications

Providers generally perceive PPIs as low priority medications. Even though side effects of PPIs are acknowledged, providers stated that PPIs are still considered low priority medications compared with, for example, antibiotics, where non-guideline-concordant use has far more reaching consequences.

Q1: The PPIs just do not sort of reach the threshold of this is important enough that I want to spend a lot of time tracking this down.

Poor awareness of ongoing intervention

Some providers were not well-informed about our ongoing PPI intervention, and some of them had not received PPI recommendations from the pharmacists.

Q2: I did have a pharmacist call me directly or speak with me directly, or Skype or message me directly. I think I can recall one or two occasions when there was like a brief pharmacist note put in the chart where they documented like, hey, this person is on a PPI. I reviewed the chart, and either we can de-escalate the dose or reduce the frequency. That is the most I can remember seeing is maybe one or maybe two chart interventions or documentation by the pharmacist.

Organization

The GERD assumption

Providers reported that for patients whose PPI is initiated in the outpatient setting, there is a general perception that these patients have a diagnosis of GERD and should be on a PPI. For such patients, providers were reluctant to assess the appropriateness of their PPIs during admission. This results in patients taking these medications unchecked for a long duration.

Q3: But I think more anecdotally, without having any numbers in front of me, I think the majority probably come into the hospital already on a PPI, usually for GERD symptoms.

Hierarchy and communication

Pharmacy providers stated that they encountered instances where they recommended PPI therapy de-implementation, but the recommendation ended only with medical trainees (residents) who were not willing to make any changes to the PPI without the authorization of their seniors. This happened often on surgical wards and resulted in delay or complete inaction about the PPI if the trainee provider was not able to get timely feedback from their senior.

Q4: Yeah. I think some of the barriers are, usually when I call a representative of the team to discuss it, I perhaps might be talking to the surgeon who is in their first year, and they might not necessarily feel really comfortable with stopping a PPI and wanting to kind of talk to, up the chain, talk to someone else and might not necessarily be aware of the evidence and things like that.

No EMR tool dedicated to PPIs

Providers noted the absence of an EMR tool dedicated to PPIs was a barrier to guideline-concordant PPI prescription.

Q5: Not that I come across as an inpatient provider. I guess I would not be surprised if there were something maybe that the PCPs use, but I do not, I personally don't get any reminders or alerts or anything.

Environment

Perception that chronic PPI use is an ambulatory care problem

Many providers perceive chronic PPI use an ambulatory care problem that should be handled by primary care providers (PCPs). Because of this, less effort is put towards evaluating appropriateness of PPIs for inpatients, particularly for those patients admitted while already taking a PPI.

Q6: A lot of the times, the impression that I get is the majority of them were already started outpatient, and they come into the hospital already on the PPIs.

Setting of PPI initiation

Providers reported that the PPIs initiated in outpatients were more of a problem than those initiated during inpatient admission. This is because providers felt they did not have sufficient details about the therapy. However, even for inpatient-initiated PPIs, there is no structured effort to ensure that they are stopped at discharge.

Q7: I think the bigger issue is those that are started on it inpatient, say for, whatever, stress ulcer prophylaxis, maybe those that are not necessarily on anticoagulants or on steroids, how do we make sure that they maybe get them stopped on discharge. I think that is a tougher issue.

Person

Unwilling to dispute another providers decision

Providers were reluctant to discontinue a PPI if it was started in the outpatient setting as they did not want to interfere with what the treating outpatient provider had started in the context of not knowing the full patient history.

Q8: I think as part of the med rec, when you see that someone is on a PPI as an outpatient, we are like, well, someone, their PCP or someone thought they needed to be on a PPI, and who am I to dispute that or argue against that?

Task

Time to review charts

As expected, many providers stated that time to review and find information needed to decide whether a PPI prescription is appropriate is a big challenge. Amidst many other tasks to attend to, providers find it impossible to spare time to fully evaluate a PPI prescription.

Q9: But it does take, it takes a lot of exploring. I know when there are pharmacy students, that's always helpful because the pharmacy student can look into some of those things a little bit more because they have more time.

Tools and technologies

Classic PPI therapy indications

Many providers reported that when the PPI was clearly indicated, they would prescribe it. This occurred in situations of classic PPI therapy indications such as high risk for gastrointestinal (GI) bleeding. However, providers do not necessarily follow up to verify the PPI gets discontinued when it is no longer indicated.

Q10: Or, and so it just sort of got overlooked, that maybe they should be on one, just for GI prophylaxis, and we start them because of that, because they are sort of identified as being high risk. And other times, like I said, they have an incident while they are with us that ends up requiring them, like they’ve had an active GI bleed, and then GI gets involved, and they end up being on a PPI as a result.

Organization

Pharmacy residents and students able to support the intervention

The initial chart reviews to determine PPI appropriateness were carried out by a pharmacy resident or intern. Recommendations were then communicated to the inpatient pharmacist, who reached out to the patient’s treating team. Although this created some delays, it helped save the inpatient pharmacist’s time, which promoted the intervention.

Q11: But it does take, it takes a lot of exploring. I know when there are pharmacy students, which is always helpful because the pharmacy student can look into some of those things a little bit more because they have more time.

Ready availability of pharmacists

Providers noted that the ready availability of pharmacists at the facility and a close working relationship with them facilitated guideline-concordant PPI prescriptions. Providers would easily consult pharmacists if they needed help with medication reconciliation.

Q12: We work very closely with pharmacists, I would say that, I mean, if it weren't for pharmacists, we consult them for just about everything, so…

Person

Patient’s willingness to make changes to their PPI medications

Through patient education about PPI therapy, we noted that patients were willing to have their PPI therapy changed if necessary. We encountered only two patients who insisted that their PPI therapy could only be changed by the outpatient provider who had initiated it.

Q13: If they don't have a compelling indication, it likely that at one point in time they were started on it for something like GERD or indigestion, it's usually just a conversation with the patient about trialing, reducing the dose or trialing sort of like a taper to step them down and off of it and using it as needed.

Providers’ acknowledgment of risk of PPI adverse events

Providers reported that they recognize that PPIs have adverse events and are willing to make the necessary interventions to ensure that guideline-concordant PPI prescriptions happened. However, the motivation is low, as PPIs are perceived to be low-priority medications.

Q14: I think we recognize that PPIs are not without their risk. I think people don't sort of look at them as a completely benign medication.

Task

Acceptance of PPI recommendations by providers

Many of the providers were willing to make a PPI recommendation suggested by pharmacists. This facilitated the flow of the intervention.

Q15: I think out of all the options we have; I mean, I do think when pharmacists reach out to a medical team, and they say like I think we need to change this, I think most of the time, we agree with them. I think most of the time, they are right.

Tools and technology

Forced functions in the EMR

Some providers stated that a forced function in the EMR can be an effective strategy. This would ensure that the provider thinks about the PPI before initiating it or continuing it and prevents the possibility of simply clicking through without making any changes to the PPI therapy.

Q16: I think people will just click through it without thinking. So, a forced function is needed

Pharmacy-driven intervention

Many providers agreed that any effective PPI intervention should be pharmacy-led, where pharmacists perform the PPI review and provide recommendations on the course of action to providers. This could be done through pharmacy notes to providers and, more importantly, through verbal communication between pharmacists and providers face-to-face, by phone, or through another platform, such as Skype.

Q17: I think having the pharmacists heavily involved and reviewing it and coming up with a recommendation and then potentially reaching out to the medical team at what seems to be a convenient time, like potentially later in the morning after we are rounding, I think seems good. I think even just a note, I mean just putting a note in the chart and adding the residents as additional signers.

Organization

Intervene at both admission and discharge

Providers stated that an ideal PPI intervention should focus on medications the patient is taking at admission and those the patient is taking at discharge. This provides an opportunity to assess PPI prescriptions initiated in the outpatient setting and those initiated during admission.

Q18: I think a thorough medication review on discharge would probably be very beneficial, because I think there is a lot of protocol that happens in the hospital, and people get placed on PPIs because of protocols, especially in like ICU situations or very acute, very ill patients where it's just part of the protocols to help heal people, essentially, and they don't necessarily need them long term. And so, I think identifying, you would probably find you could stop a big percentage of PPI use from continuing in perpetuity if you med rec them on discharge as to whether or not they truly met criteria to use the medication moving forward. So, a discharge medication reconciliation is probably even more important than the admission one for appropriateness of use.

Specific intervention/recommendations

Providers mentioned that they are more likely to respond to and intervene in PPI therapy if there is a specific intervention in place. This should state, for example, how long the patient has been on a PPI and any side effects experienced, and it should clearly suggest what the provider needs to do about the PPI.

Q19: Well, something that, it's hard sometimes to try to track down why patients were on PPIs. So, in the future, going forward, if there is somehow a way for us to know exactly why someone was on the medicine, so it's clearly documented or clearly documented how long they should be on it, I think that would help.

Involve resident physicians

Providers recommended that involvement of resident physicians in PPI interventions is likely to increase the likelihood of the intervention happening, as residents enter most medication orders.

Q20: I think like if we wanted to sort of maximize the likelihood of it happening and the educational aspect of it, I think reaching out to the residents, either the intern, like the PGY1, or the more senior resident, PGY2 or PGY3, I think would be the most effective.