BRIDGE trial suggests quality improvement interventions may increase use of evidence-based therapies

The BRIDGE Cardiovascular Prevention Cluster Randomized Trial found that a multifaceted quality improvement educational intervention significantly increased the use of evidence-based therapies for cardiovascular disease in a low- and middle-income setting.

Otavio Berwanger, MD, PhD, director of the Research Institute HCor at Heart Hospital in Sao Paulo, Brazil, presented the results of BRIDGE CV, a multifaceted intervention to narrow the evidence-based gap in the treatment of high cardiovascular risk patients.

Although large-scale, randomized trials have established the efficacy of antiplatelet therapy, statins, ACE inhibitors and ARBs for the management of high-risk cardiovascular disease, adopting these modalities in practice is suboptimal, particularly in low- and middle-income settings, which account for up to 80 percent of the global CVD burden. Quality improvement interventions have rarely been evaluated in these settings.

BRIDGE CV enrolled 1,619 stable patients with high-risk, established CV disease, including prior coronary artery disease, prior stroke or symptomatic peripheral artery disease from 40 clusters of outpatient clinics from tertiary hospitals or primary care units in Brazil. Clusters were randomly assigned to a multifaceted QI intervention (18 clusters, 726 patients) or to routine care (22 clusters, 893 patients). Allocation to the interventions was concealed and adjudication of the endpoints was blinded.

Patients in both groups had screening and randomization visits as well as additional visits at one, six and 12 months. The primary endpoint of the study was adherence to antiplatelet therapy, statins and ACE inhibitors or ARBs at 12 months in an all-or-none approach.

In addition to the primary endpoint, secondary endpoints at 12 months included:

  • Individual components of the primary endpoint
  • Percentage of eligible patients with LDL below 70 mg/dL and below 50 mg/dL
  • Adherence to high-dose statins by patients with no contraindications
  • Adherence to beta-blockers in patients after myocardial infarction
  • Combined endpoint of CV mortality, non-fatal MI, or non-fatal stroke, and control of risk factors such as hypertension, diabetes or smoking cessation

The QI intervention was based on behavioral marketing. The first intervention included three steps.

  1. In a pre-physician visit, a case manager evaluated each patient using a one-page form with four colored sections to capture anti-hyperlipidemic, anti-diabetic, anti-platelet and anti-hypertensive medications. Case managers were trained in treatment guidelines. Care gaps, such as lack of recommended medical treatment, were noted with stickers
  2. Established reminders and physician prompting before the physician visit.
  3. During the physician visit, a similar four-colored, one-page decision support tool summarized treatment guidelines to facilitate the intervention. During the discussion, Berwanger said guidelines are “great, but they are complicated and long documents,” hence the use of a one-page summary.

During the second part of the intervention, audit and feedback reports were generated monthly for individual measures, such as lipid levels and medication use, as well as for the all-or-none adherence to antiplatelet therapy, ACE inhibitors/ARBs and
statins. Information on lifestyle modification was provided for patients.

The two intervention groups were generally well matched for baseline characteristics, with a median age of about 65 years; 63 percent were male, and about half had a previous MI. Cluster baseline characteristics were also well balanced.

The intervention significantly improved the primary endpoint (complete adherence to statins, anti-platelet therapy and ACEi or ARB at 12 months, with 73.5 percent complete adherence in the intervention group versus 58.7 percent in the control group (P=0.01).

The intervention also significantly improved the secondary endpoints of statin adherence, with 93.6 percent adherence in the intervention group versus 81.7 percent in the control group (P<0.01). It also improved adherence to anti-platelet therapy, with 94 percent in the intervention group versus 86.3 percent in the control group (P<0.01). Smoking cessation also significantly improved after the intervention.

There were no significant differences between treatment groups for adherence to ACEi or ARB, use of high-dose statins, use of beta-blockers post-MI, LDL less than 70 mg/dL, blood pressure below 140/90 millimeters of mercury, systolic blood pressure less than 120 mm Hg, or HbA1c ≤7.0 percent in patients with diabetes. There was no significant difference between groups in major adverse cardiovascular events.

In a subgroup analysis, there was better adherence to statins, anti-platelet therapy and ACEi or ARB in teaching units than in non-teaching units (P=0.02). Berwanger speculated that the personnel in teaching units might be more motivated to use the intervention tools.

He concluded that the tools tested in the trial could be used to develop QI programs to maximize the use of evidence-based interventions to manage the care of patients with established CVD, particularly in limited resource settings.

Trial to improve adherence to evidence-based stroke, TIA treatments yields mixed results

A multifaceted quality intervention didn’t significantly increase the composite adherence score for use of evidence-based therapies for acute ischemic stroke and transient ischemic attack patients, according to results presented Saturday in a Late-Breaking Clinical Science session.

Maria Julia Machline Carrion, MD, MHS, PhD, coordinator of the Bridge-Stroke Project at Heart Hospital in Sao Paulo, Brazil, presented results of BRIDGE stroke, a cluster randomized quality improvement trial conducted in Argentina, Brazil and Peru.

However, when using a more conservative all-or-none approach of complete adherence, the QI intervention improved adherence to evidence-based therapies and significantly increased the use of thrombolysis and smoking cessation education.

The uptake of interventions with established efficacy for managing patients with AIS and TIA has been suboptimal, particularly in low- and middle-income countries, and there are few robust QI trials in these settings, Carrion said.

The trial enrolled patients in 36 clusters that were randomly assigned to the multifaceted QI intervention (19 clusters with 817 patients) or to a control group following routine practice (17 clusters with 807 patients).

Clusters were hospitals with 24/7 emergency departments, central nervous system imaging capability and the ability to administer tissue plasminogen activator (Rt-PA). It included 1,624 consecutive patients with AIS or TIA who were admitted within 24 hours of the onset of symptoms.

The primary endpoint was a composite adherence score to 10 in-hospital quality measures, including early antithrombotics, Rt-PA within therapeutic window, deep vein thrombosis prophylaxis, door-to-needle time of less than 60 minutes, screening for dysphagia, assessment for rehabilitation, antithrombotics at discharge, anticoagulants for atrial fibrillation or flutter, statins for LDL greater than 100 mg/dL or not documented and smoking cessation education.

The secondary endpoints included complete adherence to 10 in-hospital quality measures, Rt-PA in patients admitted within 24 hours, antihypertensives, DTNT less than 45 minutes and 90-day clinical events (mortality, disability and stroke recurrence).

The QI intervention included the use of a colored wristband for patient identification and a printed poster of reminders that the patients should receive special care. A therapeutic plan/algorithm for care management included recommendations of evidence-based therapies for these patients.

Trained nurse case managers ensured all components of the QI intervention were used. Educational materials containing evidence-based recommendations were also provided. Periodic feedback reports were generated on adherence to quality measures.

Patient baseline characteristics were well matched between intervention and control groups. The majority of patients had AIS as a diagnosis. At baseline, intervention clusters were somewhat more likely to have a stroke unit, although all clusters in both groups had a stroke protocol available.

There was no significant difference in composite adherence score between the intervention group (85.3%) and the control group (77.8%; intraclass correlation coefficient, 0.32). Complete adherence to in-hospital quality measures was 49.2 percent in the intervention group and 25.2 percent in the control group (population average odds ratio = 2.59; 95% CI, 1.05-6.1; ICC = 0.25).

Of the in-hospital quality measures assessed, only Rt-PA use within the therapeutic window and smoking cessation education were statistically significantly better in the intervention group than in the control group (55.0% versus 39.9%, P=0.01; and 72.1% versus 48.5%, P=0.04, respectively).

BRIDGE stroke may be a “first step in quality improvement in Latin America,” said Carrion, who suggested tailoring interventions to each location. “The other component that is very important from my end is the audit and feedback system.”