(Illustration by Maria Carluccio)
In the early 1990s, I was working in China opening a heart-surgery unit. We had four patients arrive in our 12-bed unit in the same week after being diagnosed elsewhere with “chronic flu.” When we evaluated them, we found that they did not have the flu or a lingering cold—they had heart failure. Fortunately, all four survived—but only because we caught what others had missed. Had their condition been diagnosed earlier, we would have had more treatment options, and the patients—who had paid for multiple “flu” visits—would have been better off clinically and financially.
Global health efforts have increasingly emphasized access to care while overlooking poor‑quality care, unwarranted variation in how doctors diagnose and treat the same condition—where one patient with chest pain receives an EKG and another is sent home with antacids—and the inefficiency that variation creates. The problem isn’t that we fail to recognize quality as important; it’s that we still don’t routinely measure the caliber of clinical practice in a valid, repeatable way.
Ask yourself: How good is your physician? It’s hard to know. Do you ever wonder if your doctor diagnosed you correctly? Research shows that roughly one time in six we doctors do not get the diagnosis right.
In a multicountry study published in Health Affairs, my research team at the University of California, San Francisco (UCSF), found that fewer than half of physicians in low‑ and middle‑income countries consistently followed evidence‑based guidelines when diagnosing and treating common conditions such as pneumonia, diarrhea, and preeclampsia—even when those guidelines were available. In the Philippines, for example, average provider adherence to clinical-quality standards was just 48 percent. More than half the time, patients there did not receive the correct care.
To be fair, doctors in these environments face constant shortages of time, training, and resources. Without clear, ongoing feedback, even the most well‑intentioned providers can fall back on outdated habits and inconsistent care.
From this and other studies, we know that significant variation in clinical practice directly leads to poorer patient outcomes. Although individual patients have unique needs, clinical decision-making should consistently reflect best practices, evidence‑based guidelines, and a culture of continuous learning. At Peabody Health Philanthropies (PHP), we believe that when we measure and improve the quality of care, we improve patient outcomes. Our extensive library of peer‑reviewed publications shows that improving quality often has a greater impact than expanding access alone. And now we have created a new platform to improve quality of care that has been tested, has proven effective, and is being applied around the world.
Scaling Quality
Since the 1990s, I have been developing and refining a way to improve care and measure and reduce waste. We at UCSF have built and validated an approach that is now used around the world. Our approach asks doctors to care for the same AI‑generated (simulated) patient—so large groups of providers face an identical case at the same time. We then provide itemized, individual, and confidential feedback with real‑time benchmarking against peers. This iterative, active‑learning process reduces unnecessary variation and teaches providers how to deliver better care, saving lives and lowering costs. The tool, now called the E-Patient Quality Improvement and Standardization (EQIS) platform, directly measures clinical decision‑making in a way that is reliable, repeatable, and scalable.
What began as in-person assessments in China has evolved into a validated, customizable online platform that helps medical professionals identify errors in their own practice by presenting simulated patients and immediate feedback on their clinical decisions. This teaching leverages adult active‑learning principles to reinforce evidence‑based decisions and improves patient outcomes in a sustainable, measurable way.
Originally developed and piloted in the United States in the early 2000s, the online EQIS platform showed early promise: Better measurement of clinical decision-making led to consistently higher quality in care and lower costs across large health systems. By the mid-2000s, these early successes prompted additional, rigorous international testing in settings such as the Philippines, Kyrgyzstan, and Eastern Europe, demonstrating that the tool is both effective and adaptable across cultures and resource levels.
Applying the tool in different contexts has taught us a lot. Our work with partners such as the World Bank, and in countries including the Philippines and across Eastern Europe, reveals that in low‑resource settings, variation in clinical practice is not just inefficient or expensive—it is deadly.
Through EQIS, we can measure and map this variation across entire health systems. In Kyrgyzstan, for example, EQIS scores varied by more than 30 percentage points between rural and urban providers. In other countries, performance improved by 24 percentage points over two rounds of EQIS participation—demonstrating how quickly providers can improve when given the right tools.
In our current work in Bangladesh, Malawi, and Vietnam, we are already seeing how regular measurement and targeted feedback are transforming clinical decision‑making and improving real‑world care, even under challenging conditions. In Vietnam, three rounds of EQIS have led to measurable increases in quality scores across 10+ sites. In Malawi, early data show a strong correlation between the number of cases completed and the speed of improvement in clinical performance.
Measuring quality this way is only the first step. Lasting change happens when providers feel supported, engaged, and safe as they improve their own care.
Our approach focuses on building trust and promoting learning rather than enforcing compliance. In most settings, participation in EQIS consistently exceeds 95 percent. Providers work through carefully constructed simulated cases that are detailed, locally relevant, and aligned with the patients they see daily.
This model works because it is supported by a philanthropy that understands systems change: Improvement takes time, engagement takes trust, and trust takes investment. By supporting providers, we empower them to become agents of change within their own health systems. Outside funders can accelerate progress by supporting in‑country initiatives and promoting change through their networks.
Since launching our global efforts in 2023, we have partnered with more than 15 clinical sites spanning rural clinics, regional hospitals, and medical-training institutions.
In Vietnam, we have completed four rounds of EQIS implementation with more than 250 providers. Clinical-quality scores have improved significantly, with some participants showing average gains of 20 to 30 percentage points from baseline. These improvements not only are statistically meaningful; they translate into more consistent care at the front lines. One provincial hospital, for example, saw pediatric pneumonia scores rise from 42 percent to 78 percent between the first and fourth rounds.
In Malawi, we observe a similarly strong correlation between case completion and improvement. Providers who engage with more than 10 simulated cases demonstrate faster and more sustained gains in clinical decision-making—underscoring the power of practice‑based feedback. Early adopters within the Malawi Ministry of Health have expressed interest in expanding EQIS into national continuing-education efforts.
Even in settings where providers are overworked and under‑resourced, participation remains high. This consistent engagement shows the real-world appeal and impact of EQIS.
Linking Quality to Outcomes and Cost
While EQIS demonstrates a clear connection between better measurement and improved health outcomes, long‑term success depends on more than clinical gains; it depends on a model that is scalable and sustainable.
We use a three‑stage sustainability model to ensure that EQIS can grow independently within local systems.
First, we establish feasibility by confirming that the tool functions effectively in a new setting. Can we collect high‑quality data? Can providers engage meaningfully with the platform? In each of our projects, the answer has been yes. Demonstrating feasibility builds trust and lays the foundation for long‑term local adoption.
Second, we support a locally viable business model. Once feasibility is established, we help partners build a plan to scale EQIS use through a structure that supports financial independence. PHP provides access to the platform through a free license; local partners manage implementation and operations.
Third, we pursue formal government approval for using EQIS in licensing and re‑licensing health professionals. Our goal is to integrate EQIS into national policy by working with ministries of health and licensing authorities.
When sustainability is prioritized from the outset, EQIS becomes a long‑term driver of better care embedded in national systems and guided by local leadership. In a global environment where donor funding is increasingly constrained, this model offers a reliable, country‑led pathway to lasting impact.
Right now, many donors and governments are focused on expanding health coverage. But access alone is not enough. Without improving the quality of care, expanded access risks delivering inconsistent, ineffective, or even harmful treatments.
The recent elimination of USAID funding makes this need even more urgent. We must rely on scalable, sustainable solutions that build local capacity and empower providers to deliver the best possible care. Organizations like PHP and tools like EQIS are essential to filling this gap. Quality cannot remain an afterthought in the global health conversation.
Read more stories by Dr. John W. Peabody, Steven Johnson & Othman Ouenes.