'Passionate Champion' Key to Inpatient Glycemic Control

Miriam E Tucker

April 08, 2015

NATIONAL HARBOR, MD — A multidisciplinary team, administrative support, real-time data, and a dedicated "champion" are among the features that characterize hospitals with the best glycemic-control practices, a new study shows.

"The ability to gain the win-win of decreasing hyperglycemia and hypoglycemia isn't as simple as just having moderate targets or implementing order sets. It's a complex process," Cheryl W O'Malley, MD, program director, internal medicine, at Banner Good Samaritan Medical Center, Phoenix, Arizona, said March 31 here at the Society of Hospital Medicine (SHM) 2015 Annual Meeting.

O'Malley and colleagues surveyed a total of 94 US hospitals that were participating in one of two SHM glycemic-control quality-improvement initiatives and had uploaded at least 12 months of data for all inpatients who had point-of-care glucose testing.

The researchers divided the hospitals into quartiles based on the number of patient blood glucose values within target range (70-180 mg/dL) and then surveyed the nine best-performing hospitals to determine the characteristics associated with their success.

"One of the real learning points was that the hospitals had active glycemic-management teams that would go out on a daily basis and respond to individuals having lows or highs and do one-on-one education with nursing [staff] and physicians," Dr O'Malley told Medscape Medical News.

The background of the team members — which included physicians, nurses, and pharmacists — didn't matter as much as did the approach. "All were considered experts and aware of best practices. They would create a very nonconfrontational conversation about what they would suggest for that patient to try to help improve their trend….When [providers] had a better understanding and were able to respond in case-by-case instruction, that seems to be associated with improvements, as it was one of the characteristics at all [top] sites," she said.

The paper won the "top-innovation" oral presentation at the conference, session moderator Ethan Cumbler, MD, from the University of Colorado Anschutz Medical Campus, Aurora, told Medscape Medical News.

"The methodology that they used…seeks to learn from excellence where it exists in a large group of organizations that are struggling with a similar problem with similar resources," he said. He noted that this research approach is sometimes called "positive deviance."

"What we loved about it was that rather than focusing on those hospitals having the most difficulty and asking how to we bring them up to average, that particular project was starting with the hospitals that were doing spectacularly well not only at lowering high blood glucose but also avoiding low blood glucose…and then asking what are they doing that every other hospital could learn from, so that they could not only come up to average but could actually rise to that level of exceptional performance…That's what made this project so exciting for the judges."

"Passionate Champions" Among the Keys to Success

Of the 94 hospitals, 74 were participating in SHM's Glycemic Control Mentored Implementation (GCMI) program, which provides one-on-one mentoring, a listserve, webinars, tool kits, and an online data registry with feedback. The other 20 came from another SHM quality initiative, Electronic Quality Improvement Programs (eQUIPS), which involves many of the same features as GCMI except without the mentoring aspect.

All 94 hospitals had uploaded at least a year's worth of data by February 27, 2014. Nearly half (45%) were community nonteaching hospitals, another 38% were community teaching hospitals, and a small proportion (7%) were academic medical centers. The majority had either 201 to 499 beds (51%) or less than 200 beds (31%).

The data included all point-of-care blood glucose values for medical and surgical units, including both noncritical (all 94 hospitals) and critical care (from 84 of the 94 hospitals). Pediatrics, obstetrics, emergency, dialysis, and perioperative units were excluded. In all, the data comprised 1,129,300 patient days in 727 non-ICU units and 286,573 patient days in 303 ICUs.

Blood glucose data were calculated based on the average percent of patient stays that were within the targets, as well as the values of 180 mg/dL or greater, less than 70 mg/dL, or less than 40 mg/dL during patient stays. The researchers conducted the surveys by email and clarified responses through structured interviews.

Of the top nine best performing hospitals, six were community nonteaching, two were community teaching, and one was an academic medical center. None had more than 499 beds.

All of the top performers had a "passionate champion" who organized, led, and "provided vision" for the glycemic-control efforts. Hospitalist physicians filled this role at three of the hospitals, registered nurses or nurse practitioner/certified diabetes educators at another three, and at one hospital each the champions were a pharmacist, an ICU nurse, and an endocrinologist.

All nine top performers also had multidisciplinary committees, comprising nursing, pharmacy, physicians, and information technology, as well as boots-on-the-ground glycemic-management teams that would do the daily one-on-one interventions with providers to target glycemic outliers. At only one site did team members write orders or take on more active management. Mostly, their role was "more just advising and education," Dr O'Malley noted.

Pharmacists were involved in all the top nine hospital committees, and an endocrinologist participated on nearly half of the committees. In addition, approximately half of the committees included local nursing champions who would act as liaisons with other stakeholders on clinical units.

All sites reported having at least initial hospital administration support in the way of funding for uploading data and/or for other items such as physician salaries, nursing education time, and IT support.

All of the top programs had established paper- or electronic-medical-record–linked protocols and insulin order sets with clear dosing guidance, and seven of the nine top performers required use of those order sets. Most had developed a process for real-time tracking of data with dashboards reflecting SHM data or internally developed metrics.

Changing Practice

The most common barrier, according to survey respondents, was getting physicians to change their prescribing practices. Some of the approaches that didn't work were alerts about "best practice" and grand rounds or staff meetings. These things "increased awareness but didn't change practice," Dr O'Malley noted.

What did work, she said, were the sharing of data with benchmarks, nonthreatening one-on-one discussions between the "champion" and the individual provider, and "constructive and nonconfrontational case-based feedback in real time."

In speaking with Medscape Medical News, Dr O'Malley acknowledged that these data may not represent hospitals in general because all of the hospitals in the original study sample were already uploading data within one of the SHM programs.

"All data were from hospitals that were uploading data….Many hospitals don't do that and don't even have good measurements of how they're doing. At least the other [SHM-program-participating] hospitals all had data to monitor how they're doing in comparison with others, so when they realize they're not a top performer, it gives them some benchmarks of places to be able to make improvements. So having that data ends up being important."

But Dr Cumbler sees the study results as generalizable. "As I looked at the things that they were able to identify, none of them were especially resource-intensive. I would say that many of the techniques that they identified in those best-performing hospitals would be widely applicable….What I heard was less about the individual protocol and more about the concepts of champions and culture."

The study was funded by an unrestricted grant from Glytec. Dr O'Malley has no other relevant financial relationships, and Dr Cumbler has none either.

Society of Hospital Medicine 2015 Annual Meeting. March 31, 2015; National Harbor, Maryland. Abstract 285.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.