ASH Guideline Champions Detail Ways to Improve Quality and Outcomes, Trim Hospitalizations

Quality improvement in health care is not easy. It takes buy-in from many stakeholders, and progress can be hard won. For the first time, the American Society of Hematology (ASH) singled out those who are making these efforts work.¹ During ASH’s 64th annual meeting and exposition in December 2022, a session chaired by Rachel Rosovsky, MD, MPH, director, Thrombosis Research, Division of Hematology at Massachusetts General Hospital in Boston, featured 3 Guideline Implementation Champions who led efforts to ensure ASH Clinical Guidelines were being met within or across institutions.

According to criteria from ASH, each project had to have a measurable impact and incorporate a unique change strategy. The speakers discussed collaboration across departments, any barriers to change, and how they overcame them. The 3 champions were able to improve health outcomes while reducing days or hours of hospitalization—achievements of interest to managed care.

E-consults for Heparin-Induced Thrombocytopenia
Jori E. May, MD, an assistant professor of hematology/oncology at the University of Alabama at Birmingham (UAB), outlined how a cross-disciplinary team developed a system for e-consults to address frequent deviations from guideline-supported care for patients with suspected and confirmed heparin-induced thrombocytopenia (HIT). May noted that HIT is associated with high resource utilization because of higher rates of bleeding and thrombosis.²

Guidelines published in 2018 are divided into those specific to the diagnosis phase and those that are situation specific, May said. Of primary importance, May said, was that when patients with suspected HIT have a low probability 4T score, the ASH guideline recommends against laboratory testing.

“[What] we saw at our institution—which unfortunately I think is pretty prevalent in most institutions—is frequent testing without consideration of a 4T score,” May said. Going further, an audit revealed frequent use of bivalirudin when 4T scores suggested it was not indicated, going against another guideline. Unnecessary bivalirudin use can result in bleeding and extended hospitalizations, May said. In addition, May said, “Warfarin was primarily used in our patients with HIT, with little use of direct oral anticoagulants [DOACs].”

To better comply with guidelines regarding HIT, an electronic consultation process was implemented to ensure a hematologist consulted in all cases involving suspected HIT and a positive enzyme immunoassay (EIA). In a 1200-bed hospital, there were 645 such tests a year, of which 101 were positive.
Eliminating 645 EIAs seemed overwhelming, but managing the 101 positive cases seemed doable, May said. “We identified advocates in pharmacy, our client division laboratory, and clinical services that were frequently testing for it. That included our cardiovascular surgery colleagues, intensive care unit, and hospitalist colleagues [who] were interested in engaging on this issue,” she said.

Defining desire outcomes based on ASH guidelines was key. “Where did we think we could make a difference?” May asked. The first goal was to decrease unnecessary bivalirudin use, and a second goal was to improve use of DOACs—especially among patients with confirmed HIT. Finally, the team sought to decrease length of stay among patients with HIT with an improved protocol, she said.

The new protocol allowed providers to order EIAs, as they had in the past. But to reduce bivalirudin use, a positive EIA automatically triggered an electronic consultation with the hematology team. This was a review of the medical record and a written recommendation in the chart. Hematologists did not see the patients, but sometimes they talked to the treating physician and made anticoagulation recommendations, May said, “particularly if we felt there was another cause of thrombocytopenia that might need attention.”

A positive EIA would prompt a serotonin release assay (SRA) to confirm suspected HIT, and the e-consult service would review the SRA to offer a final recommendation, both written and verbal—including additional evaluation for thrombosis or deep vein thrombosis as needed. The process was kept efficient: consult responses took an average of 76 minutes to complete.

After 12 months, the results were impressive: The hematology team ended up providing final recommendations for 89 patients. The number of days of bivalirudin use dropped from 148 days per month across the institution to 92.8 days per month (patient counts on bivalirudin also dropped). May noted that in many cases the patient lacked a low probability 4T score, and the hematology team offered assurances that nonheparin anticoagulation was not indicated.

The initiative made progress in following the ASH guideline to promote DOAC use as well. A year prior to the e-consult effort, 12 patients with HIT were discharged with warfarin (63%), compared with 7 on a DOAC (37%); after the change, 11 patients (85%) with HIT were discharged on a DOAC, compared with 2 (15%) on warfarin.

Most impressively, the average length of stay for the HIT population was 39 days; after the e-consult shift, it fell to 17 days. “And the range was on the shorter side, which we’re quite excited about,” May said.

More changes are afoot. UAB is working with a new assay to replace the SRA and provide a faster, more accurate result. May offered advice for taking on quality initiatives. “One key thing for us was starting small and manageable. In an ideal world, perhaps we would be able to review every EIA order and make sure that only those that were productive were actually performed,” she said. “But we were a small team, and we didn’t think that that was feasible for us.
“At the same time, even though you’re starting small, I encourage you to think big,” she continued. “If you are able to make improvements, thinking forward to how you might expand those improvements is also important.”

Reducing Excess Aspirin Among Patients Anticoagulated for VTE
Jordan K. Schaefer, MD, a clinical professor of hematology at Rogel Cancer Center at University of Michigan Health in Ann Arbor, presented a project that involved 6 clinical sites. The project was launched between October 2017 and June 2018 and has sustained meaningful change over 5 years, involving more than 4000 patients.³

The effort to reduce excess aspirin use among patients being treated for anticoagulation for venous thromboembolic disease (VTE) was a project of the Michigan Anticoagulation Quality Improvement (MAQI) Initiative, which combined the efforts of 5 urban sites and 1 rural site. It began with a simple idea: Look for patients who have no obvious reason to be on aspirin alongside warfarin or a DOAC—no recent myocardial infarction, no heart valve replacement—but are nonetheless taking it despite recent guideline updates that recommend against taking aspirin this way, because of the risk of bleeding.

The first step, Schaefer said, was to evaluate the results of adding aspirin to anticoagulant therapy. “Among our warfarin-treated patients,” he said, “we found the number needed to harm was about 36. So for every 36 patients treated with combination therapy, we saw it led to an increased major bleeding event compared [with] patients on warfarin monotherapy.”

And adding aspirin didn’t seem to provide any benefit. The same pattern was found among patients taking combination therapy with DOACs, Schaefer said.
He noted that the 2020 guideline specifically recommends suspending aspirin therapy—not merely not starting it—if patients with deep vein thrombosis or pulmonary embolism had been taking it for cardiovascular risk modification. “For about a third of our patients with guideline discordant therapy, our goal is to try to develop an intervention to bring these patients’ treatment more in line with ASH guidelines,” Schaefer said.

Staff at the 6 centers went to work. One unique element of the MAQI effort is that it allowed centers to customize approaches to identifying patients based on size and resources—some used pharmacists to help, some used electronic medical records. “We evaluated outcomes of the intervention based on the percent of patients with inappropriate aspirin use per month,” Schaefer said, and they also tracked site specific barriers to identification.

The fact that aspirin is an OTC medication was acknowledged as an issue to tracking it. So was the fact that it’s not always clear which specialist—if any—ordered the aspirin in the first place. Finally, so many patients see aspirin as not very harmful that this could be a barrier to taking it out of regimens, Schaefer noted.

“We kind of started with the low-hanging fruit,” Schaefer said. Patients at anticoagulation clinics with no history of vascular disease, no history of heart valve replacement, or no heart transplant were identified, and some were simply asked to complete forms on aspirin use. Shared decision-making between patients and providers was encouraged, as was coordination among multiple providers, he said.

Between the start of the initiative and the time of data cutoff in October 2022, there had been quite a shift. The mean level of aspirin use in the 24 months prior to the intervention was 19.1%, and just prior to the intervention it was 22.1%. Four years later, it was 12.1%, and the most recent data showed the rate had fallen to 8%. (The cohort had 4017 patients: 43.7% were men, average age was 58 years, 49.1% were obese, and 21.5% had cancer.)

Prior to the intervention, the centers had a mean of 2.4 bleeding events per month, compared with 1.7 events per month post intervention. There was also a statistically significant reduction in major bleeding, with 0.34% of patients experiencing major bleeding events preintervention, compared with 0.25% of patients post intervention. Emergency department visits and admissions for bleeding also fell post intervention, Schaefer said.

“The magnitude of the change is small, but considering the thousands of patients, this is significant,” he said. “Also, with the associated with health care costs and complications for patients experiencing bleed, we see some significant impact,” he said.

Reducing Time to Delivering Analgesic Dose in Pediatric SCD
Andres Vasconez Samaniego, MD, now a fellow with Johns Hopkins University and the National Cancer Institute in Baltimore, Maryland, outlined how he and his former colleagues at Jackson Memorial Hospital and University of Miami in Florida, where Vasconez Samaniego was a resident, implemented a plan to dramatically reduce the time to give pain medication to pediatric patients with sickle cell disease (SCD) when they presented with vaso-occlusive episodes.⁴

Despite guidelines from both ASH and the National Heart, Lung, and Blood Institute that call for rapid assessment and administration of the first opioid-based pain medication within 60 minutes of arrival at the emergency department, Vasconez Samaniego and his fellow residents found that these young patients were waiting 90 minutes on average, well in excess of guidelines.

Besides the need to trim 30 minutes off the time until the first analgesic was delivered, the residents needed a system to ensure frequent reassessments—every 30 to 60 minutes—to stabilize these patients until they were ready to go home. Despite the fact that they are specifically exempt from CDC guidelines on opioids, many patients with SCD deal with medical providers who do not fully understand their need to manage acute and chronic pain, Vasconez Samaniego said.

In surveys, Vasconez Samaniego said, patients voiced that “there was lack of trust of their pain severity, and there were fears of drug-seeking behaviors” from the staff.

The team solved these challenges in a multipronged way. From the guidelines, they noted that pain medication should be delivered in a nonintravenous method, so they advocated for nurses to administer an initial dose of fentanyl via intranasal delivery to offer rapid relief. Second, they developed an SCD vaso-occlusive clinical pathway, which was uploaded on a mobile app and made available to all residents and staff. Residents were also given an orientation in its use.

At all times, the team aimed to trim the time for logistics. There was a need to balance better patient care with increased workload for nurses who would have to assess patients more frequently. Following the initial dose, a prescriber’s order sheet and an SCD power plan embedded in the electronic health record would let the physician take ownership from there. Patients would have to be checked for excessive sedation, and naloxone might be needed if too much opioid medication was used.

Between December 2019 and December 2020, the intervention drove tremendous change: the mean time until the first analgesic dose fell from 90 minutes to 45 minutes. Between September 2020 and February 2021, there were 64 patient encounters in which the patient with SCD was clinically qualified to receive intranasal fentanyl; of these 49 received it and 5 refused it; 10 did not receive it for another reason.

The effort shaved almost a full hour off time in the hospital for preintervention. Previously, the average length of stay was 5 hours, 21 minutes; after the introduction of intranasal fentanyl, the average length of stay was 4 hours, 29 minutes. The sickle cell occlusive pain episodes clinical has been applied in 90% of encounters.

There were some barriers. The nursing staff was concerned about opioid administration due to fears of respiratory depression, and patient refusal of medications was fueled by misconceptions about the drugs, Vasconez Samaniego said.

He noted the importance of this intervention as a health equity issue, since 88% of the patients were Black and 12% were Hispanic. Given the early success, this model is being repeated at other hospitals in the Miami area with pediatric emergency department, Vasconez Samaniego said.

“I have to give a shout-out to the nursing team who took ownership and appropriation of this project,” he said. Residents may move on, but nurses stay. Despite some initial fears, “they are taking care of these patients,” he said. 

References
1. American Society of Hematology Clinical Practice Guidelines. Accessed December 29, 2022. https://www.hematology.org/education/clinicians/guidelines-and-quality-care/clinical-practice-guidelines
2. May JE, Irelan PC, Boedeker K, et al. Systems-based hematology: highlighting successes and next steps. Blood Adv. 2020;4(18):4574-4583. doi:10.1182/bloodadvances.2020002947
3. Schaefer JK, Errickson J, Gu X, et al. Assessment of an intervention to reduce aspirin prescribing for patients receiving warfarin for anticoagulation. JAMA Netw Open. 2022;5(9):e2231973. doi:10.1001/jamanetworkopen.2022.31973
4. Samaniego WAV, Matheus C, Aguilar-Velez C, et al. A resident-led quality improvement initiative to optimize care for children and young adults with sickle cell disease vaso-occlusive pain with intranasal fentanyl in the emergency department. Blood. 2021;138(suppl 1):1892. doi:10.1182/blood-2021-152849