Noam Angrist and colleagues set out a research agenda designed to make implementation itself an object of rigorous study. Their central claim is deceptively simple: the frontier of education research and practice is no longer just discovering what works, but understanding how to make it work repeatedly, equitably and at increasing scale. The authors write that there is growing evidence on what works, and that the task now is to translate that evidence into action and to study that translation systematically, so that effective programs can be delivered again and again to achieve what they describe as effective, equitable and enduring learning for all. In their framing, implementation is not an afterthought to research but a scientific discipline in its own right.

The scale of the problem gives the argument its urgency. Although most children around the globe now enroll in school, a large share of them complete years of schooling without mastering basic literacy and numeracy. Decades of randomized trials and rigorous evaluations have produced genuine breakthroughs: teaching at the right level, structured lesson plans, and cost-effective tutoring programs have all shown strong effects in carefully managed studies. Yet when governments adopt these interventions and roll them out across thousands of schools, effects often shrink dramatically or disappear altogether. The authors point to a diagnostic blind spot that makes this pattern hard to interpret: education research rarely measures how faithfully a program is implemented or how widely it is adopted, so it is often impossible to tell whether disappointing results reflect a flawed program or a flawed rollout.

This distinction between program failure and implementation failure is the conceptual heart of the new agenda. A curriculum reform that succeeds when delivered by a motivated NGO with intensive training and supervision may falter when a ministry of education must train tens of thousands of teachers, manage supply chains, monitor compliance and sustain funding across political cycles. Without data on fidelity, dose, reach and adaptation, evaluators cannot distinguish between a program that genuinely does not scale and a program that was scaled badly. The authors argue that this measurement gap has distorted the field’s accumulated evidence base, leading policymakers to abandon promising interventions prematurely or, conversely, to scale up programs whose real-world performance was never properly understood.

To address this, Angrist and colleagues call for stronger implementation science in global education, drawing explicitly on lessons from health care. Over the past two decades, implementation science has become an established field in global health, where researchers study the barriers and facilitators to delivering interventions such as vaccination campaigns, HIV treatment and maternal health programs. Health researchers have developed validated frameworks for measuring fidelity, adoption, penetration and sustainability, and they routinely build implementation questions into trial design from the outset. Education, the authors contend, has the scientific foundations to do the same: a mature experimental toolkit, growing administrative data systems and an expanding corps of researchers embedded within government systems. What has been missing is a shared framework that treats the journey from evidence to practice as a staged, studyable process.

That framework is the paper’s most concrete contribution. The authors propose a four-stage pathway for turning educational research into effective, equitable practice at scale: efficacy trials, effectiveness trials, policy plans and large-scale implementation. In the first stage, interventions are tested under tightly controlled conditions to establish whether they can work at all. In the second, they are evaluated in real-world settings across diverse populations, schools and implementers to establish whether they do work outside ideal circumstances. The third stage involves translating trial evidence into concrete policy plans that account for cost, political feasibility and delivery systems. The fourth stage is full-scale implementation, where the intervention meets the realities of national education systems, and where, crucially, researchers continue to measure, learn and adapt.

Each stage asks different questions, and the authors emphasize that the later stages have been chronically understudied. Efficacy and effectiveness trials are now relatively common in education research, but the transition from trial to policy and from policy to system-wide delivery remains poorly documented and poorly theorized. The framework insists that scalability is not a binary property of a program but an empirical question that must be tested: does the intervention work across diverse populations and real-world settings, and is it affordable and sustainable when delivered through ordinary government channels? Cost-effectiveness, the authors argue, deserves a central place in this calculus, because an intervention that produces modest learning gains at very low cost may deliver more total benefit at scale than a more powerful but prohibitively expensive alternative.

The equity dimension of the framework is equally deliberate. Programs that succeed on average can mask deep disparities, benefiting children in well-resourced urban schools while leaving behind those in remote, underfunded or crisis-affected settings. By embedding equity into every stage, from trial sampling to implementation monitoring, the agenda aims to ensure that scale-up narrows rather than widens learning gaps. The authors also highlight durability: education systems are subject to political turnover, budget shocks and shifting donor priorities, and interventions that depend on continuous external support rarely survive. Implementation science, in their view, should identify the design features and system conditions that allow gains to endure after the researchers and funders move on.

The authors are candid about the cultural shift their agenda requires. Education economists, psychologists and evaluators have built impressive evidence on what works, but the incentives of academic research reward novel findings from clean trials rather than the slower, messier work of studying delivery. Governments, meanwhile, often lack the time, data and technical capacity to diagnose implementation failures before scaling decisions are made. Bridging this divide means embedding researchers within ministries, building measurement of implementation into routine administrative data, and designing evaluations that answer the questions implementers actually face. The payoff, the authors argue, is a virtuous cycle in which more science makes it into implementation and more implementation is made scientific.

Ultimately, the agenda is a bet that the next large gains in global learning will come not from new interventions but from better delivery of known ones. If the field can systematically study how programs are adopted, adapted, funded and sustained, the growing library of proven interventions could finally translate into learning gains for the hundreds of millions of children who currently sit in classrooms without acquiring basic skills. As the authors put it, implementation science can support the ultimate goal of all children learning. The framework they propose offers the field a roadmap for getting there, one carefully studied stage at a time.

Subject of Research: Implementation science framework for scaling effective education programs from research trials to government delivery

Article Title: A research framework for turning education evidence into practice

Article References: A research framework for turning education evidence into practice. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: implementation science, global education, education policy, learning outcomes, randomized trials, scale-up, education research, teaching at the right level, cost-effectiveness, equity, government school systems, evidence-based practice

The two dominant technologies at the heart of this debate are intravascular ultrasonography (IVUS) and optical coherence tomography (OCT). IVUS uses a miniaturized ultrasound transducer mounted on a catheter to generate cross-sectional images of the vessel wall, penetrating deeply enough to visualize the full plaque burden and the external elastic lamina. OCT, by contrast, employs near-infrared light to achieve resolution roughly ten times finer than ultrasound — on the order of 10 to 20 micrometers — allowing clinicians to identify thin-cap fibroatheromas, macrophage infiltration, microchannels, calcium thickness, and dissection planes with unprecedented detail. The trade-off is penetration depth: OCT light struggles to see beyond roughly 1 to 2 millimeters of tissue, whereas ultrasound reaches the adventitia. Together they offer complementary windows onto the pathology that angiography misses.

The cumulative randomized evidence now indicates that imaging-guided percutaneous coronary intervention (PCI) delivers measurable prognostic benefit over angiography guidance alone, but with important heterogeneity. The Korean IVUS-XPL and ULTIMATE trials showed reduced major adverse cardiac events with IVUS guidance in long lesions and in mixed populations. RENOVATE-COMPLEX-PCI and IVUS-ACS extended this signal to complex lesions and acute coronary syndromes. On the OCT side, OCCUPI demonstrated fewer device-oriented composite endpoints in complex lesions, and OCTOBER showed that OCT guidance dramatically improved outcomes in true bifurcation stenting, reducing target-vessel failure by roughly a third compared with angiography. Notably, ILUMIEN IV in North America and IVUS-CHIP in Europe failed to reach their primary endpoints, underscoring that trial design, protocol intensity, operator expertise and patient mix all shape whether imaging’s theoretical advantages translate into clinical gains.

One of the most consistent findings across the pooled datasets is that the benefit of imaging scales with lesion complexity. Simple, short, non-calcified lesions in large vessels rarely benefit measurably — the angiogram usually suffices. But in left main coronary interventions, true bifurcations, long diffuse disease, chronic total occlusions, and heavily calcified plaques, imaging changes management decisions in a substantial fraction of cases: stents are sized larger, placed longer, and dilated more aggressively. In the OCTOBER bifurcation population, for example, OCT uncovered extensive unintended stent deformation after side-branch techniques, deformities invisible on the angiogram that imaging-guided optimization could correct. Meta-analyses, including network meta-analyses by Stone and colleagues and by Giacoppo and colleagues, converge on the conclusion that both OCT and IVUS guidance reduce cardiac mortality, stent thrombosis and repeat revascularization relative to angiography alone, without meaningful differences between the two modalities in head-to-head comparisons such as OCTIVUS and OPINION.

The mechanistic explanation for the survival benefit is procedural. Imaging-guided PCI consistently results in larger post-dilatation balloons, larger and longer stents, greater minimal stent area, and more complete coverage of the diseased segment. Registry data going back to the HORIZONS-AMI substudies and the seminal work of Fujii have linked stent underexpansion and residual reference stenosis to acute and late stent thrombosis — the most feared complication of PCI, carrying high mortality. Imaging makes these failure modes visible before the patient leaves the catheterization laboratory. Achieving prespecified optimization criteria — typically a minimal stent area of 5.0 square millimeters by OCT or 5.5 square millimeters by IVUS in distal references, plus full lesion coverage — has been repeatedly associated with improved long-term outcomes, although the review’s authors stress that pursuing these numeric thresholds must never come at the cost of procedural safety, for instance by forcing oversized balloons into vessels at risk of rupture.

Patient subgroups add further nuance. In acute coronary syndromes, imaging characterizes the culprit plaque, identifies plaque rupture and erosion, detects residual thrombus, and helps distinguish which lesions truly warrant stenting — particularly valuable when multiple lesions compete for attention. Patients with diabetes mellitus, whose plaques tend to be diffuse and negatively remodeled, and those with chronic kidney disease, who face elevated event rates after PCI, appear to derive particular prognostic benefit from imaging-guided optimization in complex anatomy, as shown in analyses from ULTIMATE and from the JAMA Network Open cohort of Kwon and colleagues. Conversely, in straightforward lesions, the added contrast injections, catheter exchanges, cost, and procedural time may not be justified. Current guidelines reflect this gradient: the 2024 European Society of Cardiology chronic coronary syndrome guidelines give imaging a class IIa recommendation for stent sizing and optimization, while the 2025 American ACC/AHA acute coronary syndrome guideline similarly endorses imaging use in selected settings.

The barriers to wider adoption are less scientific than economic and cultural. Surveys by the European Association of Percutaneous Cardiovascular Interventions and the Japanese association CVIT documented persistent underuse outside East Asia, where imaging guidance in complex PCI has become routine. A white paper by Escané and colleagues catalogued the reasons: reimbursement gaps, longer procedure times, unfamiliarity with image interpretation, and organizational inertia. Health-economic analyses complicate the picture in imaging’s favor — the RENOVATE-COMPLEX-PCI cost-effectiveness analysis suggested that, despite upfront device costs, avoided repeat revascularizations and reduced infarctions can offset expenditures, particularly in complex anatomy where the event-rate reduction is largest.

The review also charts where the field is heading. Hybrid catheters combining IVUS and near-infrared spectroscopy can map lipid-rich, potentially vulnerable plaques, and the PROSPECT II natural-history study showed that such non-culprit lipid-rich plaques predict future events, opening the door to preventive local therapy. Computational advances promise to shrink the learning curve: artificial-intelligence algorithms now perform automated lumen and external elastic membrane contouring, stent-apposition analysis, calcium scoring, and even OCT-derived fractional flow reserve, validated in trials such as FLASH and FUSION. Combining optical flow ratio with post-stent physiology assessment may eventually let one imaging pull-through yield both anatomical and functional verdicts on the result, reducing contrast load and procedure time simultaneously.

For clinicians, the practical message distilled by Biccirè and colleagues can be organized around four questions: why, when, how and which. Why: because imaging converts an educated guess into measurement, and the randomized evidence shows this conversion lowers cardiac death, stent thrombosis and repeat procedures in the right patients. When: preferentially in left main, bifurcation, long, calcified and chronic total occlusion lesions, in acute coronary syndromes with ambiguous culprits, and in patients with diabetes or chronic kidney disease undergoing complex PCI. How: with disciplined pre-implantation sizing, deliberate post-dilatation, and verification against validated optimization thresholds, without compromising safety. Which: either OCT or IVUS produces comparable outcomes in most settings, so the choice can follow availability, operator expertise and specific needs — OCT’s superior resolution for bifurcation morphology and plaque characterization, IVUS’s deeper penetration for sizing large vessels such as the left main. What began as a technology searching for proof now stands, for complex interventions, on one of the firmer evidentiary foundations in interventional cardiology.

Subject of Research: Use of intracoronary imaging to guide percutaneous coronary intervention in randomized clinical trials

Article Title: Intracoronary imaging in percutaneous coronary intervention: why, when, how and which from large, randomized trials

Article References: Biccirè, F. G., Gonzalo, N., Hahn, J.-Y., Jang, I.-K., & Räber, L. (2026). Intracoronary imaging in percutaneous coronary intervention: why, when, how and which from large, randomized trials. Nature Reviews Cardiology. https://doi.org/10.1038/s41569-026-01342-3

Image Credits: AI Generated

DOI: 10.1038/s41569-026-01342-3

Keywords: intracoronary imaging, percutaneous coronary intervention, optical coherence tomography, intravascular ultrasound, stent optimization, randomized trials, bifurcation lesions, acute coronary syndrome, stent thrombosis, left main PCI, coronary artery disease, interventional cardiology