Driving quality improvement with nudges: True interventions in cardiology

Key Points

• Quality improvement research brings evidence‐based interventions to clinical practice.

• This single‐center study demonstrated that simple clinical guidelines, computer prompts, and education were able to reduce the duration of triple therapy after percutaneous coronary intervention.

• Behavioral interventions are an important tool to drive clinical improvement but must be tailored to respect physician autonomy and professionalism.

Quality improvement (QI) research has been defined as “the design, development and evaluation of complex interventions aimed at the re‐design of health care systems to produce improved outcomes.” ¹ Although better patient outcomes should be the ultimate goal of all clinical research (investigator grants, influence, and prestige notwithstanding), QI studies too often receive short shrift in the scientific literature. Compared with the gold standard of randomized controlled trials, QI studies are uncontrolled, real‐world interventions that are subject to multiple confounders and sources of bias. However, without implementation studies, data from randomized trials might never change practice for the better.

The persistent use of triple therapy (TT), that is, an anticoagulant with dual antiplatelet therapy (DAPT) for patients with percutaneous coronary intervention (PCI) and atrial fibrillation or venous thromboembolism, provides a clear opportunity for QI. Most PCI antiplatelet trials have excluded the 10%–20% of patients with atrial fibrillation, and the clinical presumption in the absence of evidence was that DAPT was necessary after PCI even in the presence of anticoagulation. Recently, multiple large randomized controlled trials including WOEST, AUGUSTUS, PIONEER‐AF‐PCI, and RE‐DUAL PCI have convincingly demonstrated that TT with warfarin has no ischemic benefits but causes more bleeding events compared with dual therapy with a direct oral anticoagulant (DOAC) and a single antiplatelet agent, usually clopidogrel. As a result, a 2020 American College of Cardiology expert consensus statement recommended that TT be limited only to patients at the highest ischemic and lowest bleeding risk, be used for 30 days or less, and preferentially utilize DOACs rather than warfarin. ²

To best implement these new recommendations, the Beth Israel Deaconess Medical Center (BIDMC) created a hospital‐wide QI intervention including (1) institutional guidelines approved by interventional cardiology, general cardiology, and pharmacy, (2) new prescribing prompts in their computer order entry system, and (3) the addition of a 1‐h case‐based discussion on antiplatelet agents after PCI to the curriculum for housestaff.

With these very simple interventions, Earle et al. ³ were able to demonstrate a change in prescribing behavior. Of the 15.3% of PCI patients included because of an indication for anticoagulation, 219 (51%) were preguideline change and 212 (49%) were postintervention. The mean time on TT decreased from 58.7 to 37.8 days (p = 0.02). Patients discharged on TT < 30 days increased from 24% to 37% (p = 0.019), while patients discharged on dual therapy increased from 26% to 33% (p = 0.02). Documentation of the planned duration of TT improved. The use of warfarin decreased from 50.2% to 29.2% with a concomitant increase in apixaban prescriptions. Most patients (>90%) were on clopidogrel. Among the patients with clinical follow‐up, there was a decrease in ischemic events postintervention (17% vs. 8.4%, p = 0.03), while bleeding and rehospitalization were not significantly different.

The authors and colleagues are commended for demonstrating that relatively simple interventions can drive at least some change in practice. The limitations of the study are multiple and expected, including its single‐center design, small and homogeneous population, and limited and incomplete patient follow‐up. The small size of the study means we need not make any conclusions about the clinical effectiveness of the intervention; the large‐scale randomized data are conclusive that TT increases bleeding even if not evident in this data set.

What makes for effective change in clinical medicine? How do we as a profession overcome longstanding practice patterns and assumptions? A growing literature on behavioral sciences has described “nudges” as subtle changes to the design of the environment or the framing of information that can influence our behavior (Figure 1). ⁴ Such gentle interventions respect clinical autonomy and specific patient situations and thus tend to be more readily accepted. The BIDMC intervention would be classified as a gentle nudge encompassing primarily education and framing, and as a result, decreased TT use only modestly. TT was still used in over 50% of patients, even if the proportion of TT < 30 days was increased. From the evidence against TT, it would be difficult to conclude that the interventions used by BIDMC were sufficient to achieve the evidence‐based goal of making TT rarely used and short‐term. A more forceful and effective nudge would have been to make the default a therapeutic substitution dual of therapy with a DOAC and P2Y12 inhibitor, and disallow TT prescriptions without a pharmacy consultation or multiple warnings.

Figure 1.

Nudges in health care, from Harrison et al., ⁴  open access [Color figure can be viewed at wileyonlinelibrary.com]

Nudges work for pharmacy interventions when providers have to interact with a computer order system, but what about within the cath lab? Doorey et al. ⁵ recently reported that education moderately improved compliance with closed‐loop communication, but that threats of accountability to nursing managers were ultimately necessary to optimize adherence. For professional physicians, similarly aggressive management risks contributing toward a loss of autonomy and burnout, and even “moral injury” when such interventions are aimed at financial goals rather than professional ethics (i.e., prior authorization). Nevertheless, if the goal is truly patient safety and welfare and the evidence is clear, behavioral interventions of all types should be welcomed as a necessity for QI.

CONFLICTS OF INTEREST

Dr. A. H. Seto has received research grants from Philips, Acist, and Pfizer, and is a speaker for Terumo, General Electric, and Janssen, and is a consultant for Frond Medical.