Home Patient Safety Breaking a bad habit

Breaking a bad habit

April 2013
Getty Images; mathisworks

Published in the April 2013 issue of Today’s Hospitalist

WHY DO HOSPITALISTS prescribe proton pump inhibitors (PPIs) to so many of their patients? Mark B. Reid, MD, a hospitalist at Denver Health Medical Center, suspects that much of the time, it’s just a matter of habit.

In a study published last year in the Journal of Hospital Medicine, Dr. Reid found evidence that may prove his hunch. “We like to think we do evidence-based studies and adjust our practice in response,” he explains. “But much of what we do today is the same thing we did yesterday. It’s really a habit.”

And in the case of inpatient PPIs, he adds, the habit is probably a bad one. Prescribing the drugs to hospitalized patients who don’t need them has been conclusively linked to increased incidence of C. diff and pneumonia. Taken long-term, the drugs have also been shown to have a range of detrimental side effects including irregular heartbeat and convulsions.

Nonetheless, acid reflux-suppressing drugs have become the country’s third most commonly prescribed class of medication. Dr. Reid and his colleagues found that 41% of the patients admitted to the medical service at Denver Health in 2008-09 received a PPI. But more than half of those patients had no valid indication anywhere in their medical record to justify that order.

The research team also reviewed the records from the University HealthSystem Consortium (UHC) on more than six million patients nationwide to make sure they weren’t just seeing a local phenomenon. But those national data revealed an even bigger problem: While researchers found no valid indication for 61% of Denver Health’s patients receiving a PPI while hospitalized, that percentage jumped to 73% of the UHC patients.

In an interview with Today’s Hospitalist, Dr. Reid points out that if hospitalists would take a minute to ask themselves why every patient needs a PPI, they might start to break their bad habit.

Doctors don’t set out to prescribe medicine that patients don’t need. Why has that happened with PPIs?
My theory was that we are using these medicines to prophylax against gastrointestinal bleeding. And in doing this study, we got to check that theory out. In about 56% of cases, somewhere in the chart the doctor wrote “prophylaxis” or “GI bleed prophylaxis” as the indication for or problem associated with the PPI order. Doctors are frequently using these medicines to prevent stress ulcers in hospitalized patients.

What’s wrong with that?
First, H2 receptor antagonists are the only FDA-approved medicines for stress ulcer prophylaxis. Of course, drugs sometimes get used outside their specific FDA indications, but this is a case where a huge number of prescriptions are written outside of FDA approval.

And as for who actually needs stress ulcer prophylaxis, it is a subset of ICU patients “and only while in the ICU. So the fact that you were in the ICU at the beginning of your hospital stay or last month does not mean you should stay on these medicines forever. For the last 20 or 30 years, stress ulcer prophylaxis has been on the ICU problem list. But that element of the assessment and plan has crept into the normal floor admission.

For most medicine doctors, you put a lot of thought into the first one or three or five problems a patient has, and then part of our work is routine: things everybody needs, housekeeping. These drugs have become part of that housekeeping element at the end of every note or every admission.

Curiously, the incidence of stress ulcers among hospitalized patients has actually gone down. Experts wonder if it is because we are treating them with prophylactic H2 blockers and other medicines. Or is it because patient care is better in terms of nutrition? Basically, we’re putting all this effort into preventing a disease that really doesn’t happen all that commonly.

Why hasn’t there been more pushback from, say, cost control or patients questioning their prescriptions?
Part of the reason these medicines have become so widely used is because most people don’t get a rash or get sick from them. If they have reflux symptoms, those get better. But if you don’t have reflux and you take a PPI, you probably feel next to nothing “and you still don’t have reflux.

It’s easy to give someone a pill like this where they won’t complain about it or even notice it. Also, these medicines have gone generic in the last several years, so their cost has fallen.

Has the inappropriate use of PPIs in your hospital grown since your study?
No, I think the trend has been toward more judicious use. A year ago, we looked at how it has been changing, and we saw the curve trending down. Education seems to be working, and physicians seem to be moving toward a more evidence-based practice. That says something about doing studies like this and getting the information out there to influence physician behavior.

Besides educating themselves, what can hospitalists do to reduce inappropriate PPI use? We did an intervention associated with this study. At Denver Health, we have a computer order entry system with an admissions order set for medicine inpatients. We realized that somewhere along the line, ordering a PPI had become a checkbox option at the time of admission. We had basically integrated the practice habit into our computer order entry.

So the first thing we did was take it off the checklist. You can still order anything you want, but we aren’t going to present it as a routine checkbox option.

It’s hard to say if the intervention worked because the curve was already trending down. An analogy would be if the murder rate was falling and you banned a type of weapon and the murder rate continued to fall. You couldn’t say the weapon ban did anything.

What else can hospitalists do?
The key is breaking the habit of using PPIs for every patient for stress ulcer prophylaxis. You need to turn this issue back into something you give a minute’s thought to. The first question is: Does the patient need prophylaxis against stress ulcer bleeding? The second question is, what is the right medicine for this purpose? And then the third question is, what is the endpoint or stop date?

Let’s say I see patients in the ICU who are on a ventilator, and they have an indication for stress ulcer prophylaxis but are allergic to an H2 blocker. This is an appropriate use of a PPI for stress ulcer prophylaxis.

But as I write the order for the PPI, I can also write in the chart that once the patient is extubated or at the time of discharge, stop the PPI. When we use these medicines for stress ulcer prophylaxis, we should think the use of these drugs all the way through to the “stop date,” like we do for antibiotics. Putting a little intellectual energy into these decisions would keep us from making them simply out of habit.

Deborah Gesensway is a freelance writer who covers U.S. health care from Toronto.