Evidence-based medicine (EBM) has transformed healthcare. It replaced intuition and tradition with research, data and measurable outcomes. It helped standardise care, reduce unsafe variation and improve survival in many conditions. Few clinicians would argue against its importance.

But an uncomfortable question is beginning to surface: has our devotion to evidence-based medicine also created blind spots — and in some cases, harm?

Modern healthcare increasingly treats the “average” patient as the benchmark. Guidelines, targets and performance measures are built on population averages, and services are judged on how well they deliver those standards. Yet real people are not averages. They are complex, shaped by culture, poverty, trauma, relationships and circumstance. When systems optimise for the average, those who do not fit the model often struggle to receive care that truly meets their needs.

We have come to see evidence as neutral and objective, but evidence is shaped by the questions we choose to ask, the people we include in research and the interests that fund it. Clinical trials often exclude patients with multiple conditions, yet multimorbidity is now the norm. Minority ethnic communities, older adults and socially disadvantaged groups are routinely under-represented. The result is guidance built on populations that look very different from those seen in everyday practice.

Conflict of interest deepens the problem. Pharmaceutical companies fund many large trials and influence which questions are studied. Positive findings are far more likely to be published than negative ones. Guidelines then lean heavily on what is available in journals, not necessarily on what truly works in the messy reality of everyday healthcare.

Regulatory systems compound the issue. Organisations are measured against performance targets linked to funding and reputation. Predictably, behaviour shifts to protect performance metrics rather than improve outcomes. Workarounds emerge. Patients are moved through services to meet targets. Assessments are relabelled. Care is reorganised to satisfy indicators rather than human need. The numbers improve; the experience of care often does not.

Meanwhile, the complexity of real life continues to resist tidy categorisation. People take medicines differently from how trials assume. Chronic illness rarely exists alone. Poverty, housing instability and trauma influence outcomes as much as prescriptions do. Yet our evidence frameworks struggle to account for these realities.

And then there is us — clinicians and decision-makers — who are not immune to bias. Under pressure, we rely on mental shortcuts. We anchor on early impressions, follow senior opinions, avoid challenging consensus and stick with plans even when new information suggests change is needed. Stress, fatigue and cognitive overload shape decisions more than we like to admit.

None of this means evidence-based medicine is wrong. In many cases it has saved lives. Asthma care dramatically improved through evidence-driven prevention strategies. Countless treatments are safer and more effective because of rigorous trials.

But history also shows how evidence can mislead. Hormone replacement therapy was abandoned by many women following early interpretations of risk that later proved overstated. Opioid medicines were promoted as safe solutions for pain, contributing to a global addiction crisis with devastating consequences.

The problem is not evidence itself. The problem is believing evidence is complete, unbiased and universally applicable.

Healthcare now faces a choice. We can continue refining guidelines and performance measures, or we can widen our understanding of what counts as evidence and whose voices shape it. Lived experience, qualitative research and community perspectives must sit alongside trials and statistics. Research priorities should reflect public health need, not commercial opportunity. Negative findings should be published as routinely as positive ones.

Most importantly, we need to ask simpler, human questions: Could people access care? Was it delivered well? Did it make a meaningful difference to their lives?

Evidence-based medicine was never meant to replace judgement, compassion or context. It was meant to support them. If we forget that, we risk creating systems that perform well on paper while failing the people they exist to serve.

Perhaps the future of healthcare is not about abandoning evidence, but about rediscovering wisdom.