American College of Physicians: Internal Medicine — Doctors for Adults ®


Hospital research moves from the basement to the bedside

The success of research at a community hospital depends on administrative buy-in and a multidisciplinary approach

From the October ACP Observer, copyright 2005 by the American College of Physicians.

By Paula S. Katz

Despite many efforts, solid evidence and well-accepted guidelines, too many patients at the University of Chicago were receiving unnecessary medications to prevent stress ulcers while they were in the hospital, as well as after discharge. Hospitalist Chad T. Whelan, ACP Member, saw in that impasse a research opportunity.

He devised a project to try to figure out what was causing the problem. As a result of his research, the institution now has an inexpensive educational intervention that focuses on the hospital's residents—and has reduced inappropriate prophylaxis rates.

Dr. Whelan's experience illustrates the fact that research is no longer something that takes place only in labs buried deep in hospital basements. A great deal of research these days occurs at the bedside—and is being spearheaded by physicians who view themselves first and foremost as clinicians, not researchers.

His project was based at an academic health center, Dr. Whelan pointed out at a recent Society for Hospital Medicine meeting. But more and more research, he said, is happening at community hospitals that see the benefit of supporting quality improvement research projects and interventions.

Getting buy-in

The key to getting hospital buy-in, said co-presenter Lakshmi Halasyamani, ACP Member, associate chair of the department of medicine at Saint Joseph Mercy Hospital in Ann Arbor, Mich., is to link your research to your institution's financial and performance measures. For community hospitals to back a physician's proposed research project, she said, it helps if the topic chosen addresses a process and population common to that institution.

According to Dr. Whelan, the success of his project was due in part to the fact that he didn't try to do it all on his own. Involving others—in his case, the residency program, other attending physicians and the pharmacy—was key.

For physicians working in community hospitals, he said, doing multidisciplinary research is even more important. Hospitalists in community hospitals tend to have less time to do research than those in academic settings, so they need to share the work to get it done. Dr. Halasyamani said that hospitalists may have to do research for free—at least at first—to convince administrators to support the next quality improvement project.

In addition to having people share the research load, engaging a wide array of professionals in a project may be crucial to its success. Nurses and pharmacists, for example, may already be collecting pertinent data.

Selecting the right topic is obviously of primary importance, but Dr. Halasyamani said the next step is just as critical: Deciding on a study plan that is feasible and of interest to hospital leaders.

Here are the types of studies she said are typically used for hospital-based research:

Randomized controlled trials. Randomized controlled trials can be conducted with randomization at the individual patient level, or cluster randomization can be performed with the hospital, service or physician group as the unit of randomization. Quality improvement research includes both types of randomization strategies, Dr. Halasyamani said, but cluster randomization avoids treatment group contamination and eases some administrative issues.

Before- and after-interrupted time series. This type of project often appeals to first-time researchers because it makes intuitive sense, doesn't require a lot of statistical help and can be done with or without a control group. The way these trials work is that a researcher looks at one point before and one point after a particular intervention to gauge the intervention's effect. Researchers might study whether the prevalence of venous thromboembolism prophylaxis changed, for example, by getting data before and after a mass e-mail reminder.

Cohort studies. These follow a defined group either prospectively for a given time period or retrospectively by looking back into medical records. Cohorts may be defined by their exposure status to an intervention.

Dr. Whelan first chose his project topic—stress ulcer prophylaxis—and then selected his study design. He looked at pre-and post-intervention data on stress ulcer prophylaxis because he had baseline information already captured on patient charts—and he could survey residents to find out how they made decisions about the stress ulcer prophylaxis they ordered.

Crunching numbers

The next step—analyzing the results—can be a stumbling block for some physicians. "Don't let the analysis and statistics be a deal breaker," Dr. Halasyamani said. "The goal isn't for you to become a statistician, but to know when to get statistical help." If you can't tap into in-house support, using an outside consulting service may make the most financial sense.

When Dr. Whelan analyzed his first set of data, he learned that most University of Chicago residents thought that steroids, sepsis and surgery were the chief risk factors for stress ulcer prophylaxis, even though none of the literature supported those opinions.

He also found that 80% of residents went to other residents for information, with only 40% going to attendings and 20% to the literature. And he learned that variation among teams was high and that residents didn't really know when to use ulcer prophylaxis. As a result, costs were unnecessarily high, patients suffered side effects and almost 25% of patients were discharged taking unnecessary medications.

In the past at the University of Chicago, standing orders, nursing reminders and pharmacy checks had all been used in previous quality improvement effotts. Given their limited success—and their high resource requirements—Dr. Whelan decided to implement a different intervention. He ended up using a survey tool to develop a one-hour, resident-led educational intervention to review literature and get group feedback, using the principles of practice-based learning and improvement. "Culture changed," he said, "by teaching the teachers."

His research on the effect of the intervention found a statistically significant and clinically important drop in rates of inappropriate prophylaxis both in the hospital and on discharge. "Everything we thought about was done on the cheap," he said. And how much has the intervention saved the hospital? About $50,000 a year, he said.

Paula Katz is a freelance writer based in Vernon Hills, Ill., who specializes in health care writing.


This is a printer-friendly version of this page

Print this page  |  Close the preview




Internist Archives Quick Links

Not an ACP Member?

Join today and discover the benefits waiting for you.

Not an ACP Member? Join today and discover the benefits waiting for you

ACP offers different categories of membership depending on your career stage and professional status. View options, pricing and benefits.

A New Way to Ace the Boards!

A New Way to Ace the Boards!

Ensure you're board-exam ready with ACP's Board Prep Ace - a multifaceted, self-study program that prepares you to pass the ABIM Certification Exam in internal medicine. Learn more.