Mass General Brigham says it is saving more lives. Some doctors aren't so sure.
Mass General Brigham (MGB), a Boston-based academic health system, has claimed that mortality rates fell 16 per cent across its hospitals in 2025 compared with prior years. The system has prominently featured the figure in fundraising campaigns and corporate communications. A STAT News investigation has found that some physicians within the system question the underlying methodology.
MGB, which includes the principal teaching hospitals of Harvard Medical School and reports annual revenue exceeding $22 billion, is one of the largest academic health systems in the United States. The mortality-rate improvement claim is based on what is known as the 'Observed-to-Expected mortality ratio' (O/E).
Dr David Bates, MGB's Director of Data and Analytics, told STAT that 'our model is benchmarked against an expected mortality calibrated to our patients' age, illness severity and comorbidities, and the improvement is statistically significant.' Bates said the system had been transparent about the methodology in published academic papers.
Five physicians, however — including MGB intensivist Dr Susan Choi-Kain and former Director of Health Policy Dr Niteesh Choudhry — told STAT about limitations of the O/E framework. Choi-Kain said: 'The model does not capture 30-day post-discharge mortality, which renders a significant portion of real-world patient outcomes invisible.'
According to Choudhry: 'If expected-mortality modelling is not done carefully, a system can improve its numbers by directing its sickest patients to other institutions.' Choudhry said there had been changes in MGB transfer policies, but that these had not been clearly disclosed in publicly available reporting.
On statistical methodology, Harvard T H Chan School of Public Health Professor Dr Atul Gawande also weighed in. Gawande said: 'Methodological transparency is essential in hospital mortality reporting; the figures used in fundraising campaigns should be presented to the same standards as in academic publications.'
The US Department of Health and Human Services (HHS) standardised Medicare 30-day mortality rate, used in the federal Hospital Quality Report, places MGB near the national median in 2024 data. That points to more modest performance relative to MGB's internal data.
MGB President and CEO Dr Anne Klibanski, in a written response to STAT, said: 'Our internal metrics are consistent with methodologies considered standard in the academic literature.' Klibanski emphasised that the system's mortality reporting was 'a direct indicator of improving clinical care quality.'
The debate is part of a broader conversation among US health systems about how clinical outcome data should be used in hospital fundraising campaigns. The American Hospital Association (AHA) ethics committee in 2023 recommended that 'outcome data should not be used in fundraising communication without academic citation.'
Dr Choudhry and other critics are asking MGB to publish the methodology openly for all hospitals in the system and to commission external audit. MGB has announced it will release a methodology note publicly within the next six months. This article is general information; individual healthcare decisions and hospital choice should be discussed with a qualified healthcare professional.