What Compassion Does to a Patient You Will Never Meet

Most clinicians enter healthcare believing that compassion matters. Most healthcare administrators are skeptical of how much it matters. The skepticism is reasonable. We work in an evidence-based field. Soft constructs that resist measurement have a long history of being weaponized to justify projects that are not actually working, and patient experience scores have been used to drive everything from menu changes to architecture decisions without much consistent connection to clinical outcomes. When the chief financial officer asks whether compassion training will move metrics that the organization actually tracks, it is the right thing to do is not an answer that survives a budget meeting.

The good news, for everyone in that meeting, is that the answer is now better than that.

What the Numbers Actually Show

Stephen Trzeciak and Anthony Mazzarelli, an intensivist and a physician executive at Cooper University Health Care, spent several years synthesizing what the medical literature has actually documented about compassion's effects on patient outcomes. The result was Compassionomics, a 2019 book that pulled together findings across hundreds of studies. The aggregate signal is striking enough that the book has become assigned reading in some health system leadership programs.

Three lines of evidence are particularly hard to argue with.

The first is diabetes care. Mohammadreza Hojat and colleagues at Jefferson Medical College published a study in Academic Medicine in 2011 examining the relationship between physician empathy scores on the Jefferson Scale of Empathy and diabetes outcomes in their patients. Patients of physicians with high empathy scores had significantly better hemoglobin A1c control and significantly lower rates of acute metabolic complications than patients of physicians with low scores. The effect held after controlling for the typical confounders.

The Hojat finding is suggestive on its own. What makes it compelling is that Stefano Del Canale and colleagues replicated it at population scale the following year (Del Canale et al., 2012). They examined 20,961 diabetic patients in Parma, Italy, classified by the empathy scores of their assigned physicians. Patients of higher-empathy physicians had meaningfully lower rates of hospitalization for acute metabolic complications. The same effect, twenty thousand patients, different country, different healthcare system.

The second line of evidence is immune function and recovery from common illness. David Rakel and colleagues at the University of Wisconsin conducted a series of studies in which patients with common cold symptoms were randomized to clinicians varying in patient-rated empathy. Patients who rated their clinician's empathy highly had shorter cold duration, reduced cold severity, and measurable changes in immune function including interleukin-8 and neutrophil counts (Rakel et al., 2009, 2011). The patient's biological response to a viral infection was measurably different depending on how the clinician they encountered showed up.

The third line of evidence is adherence to treatment. Kelly Haskard Zolnierek and Robin DiMatteo, in a meta-analysis published in Medical Care in 2009, found that physician communication quality was strongly associated with patient adherence to treatment recommendations. The effect size translated into a substantial change in the odds of adherence between high-communication and low-communication physicians. Adherence in turn is one of the strongest predictors of clinical outcome across virtually every chronic disease management protocol we use. The chain from clinician compassion through patient experience through adherence to outcome is documented at every link.

Why This Is Hard for Clinicians to Believe

Most clinicians, hearing this, have a strange reaction. They believe it intellectually. They have trouble believing it operationally. The disconnect comes from how we were trained.

Medical and rehabilitation training emphasizes the technical content of care. Diagnostic accuracy. Procedural competence. Pharmacological knowledge. Documentation precision. The relational dimension of care has historically been treated as adjunctive, as the soft skin around the hard core of clinical work. We were taught to be kind. We were not taught that kindness changes A1c.

The Trzeciak and Mazzarelli synthesis is a structural correction to this training. It places the relational dimension of care inside the clinical core rather than around it. A physician encounter is not technical content delivered through a relational wrapper. The relational dimension is part of the technical content. It changes the dose of medication the patient actually takes, the cold duration the patient actually experiences, the metabolic complications the patient actually presents with months later.

For a chief financial officer in a value-based payment environment, this is operationally consequential. Patient adherence affects readmission rates. Adherence affects total cost of care. Adherence affects HCAHPS scores, which affect reimbursement under Hospital Value-Based Purchasing. Compassionate care, in the technical sense the literature uses, is upstream of every metric the organization is currently being financially penalized for missing.

The Patient You Will Never Meet

A diabetic patient three years from now will have an A1c that depends partly on whether their primary care physician was trained in compassion. A patient with a viral infection next month will recover faster or slower depending partly on whether their family physician's nervous system co-regulated with theirs during the encounter. A patient with a chronic condition will adhere or not adhere to a regimen depending partly on whether they felt seen.

You will not meet most of these patients. You will not see the A1c trajectory in five years. You will not learn whether the patient took the medication. The downstream outcomes of compassion are mostly invisible to the clinician who produced them, which is one reason the field has been slow to recognize how clinically consequential they are.

But the outcomes are there. They have been documented. The literature is substantial enough that ignoring it is no longer a defensible position for institutions making decisions about where to invest in clinical infrastructure.

Compassion training is clinical infrastructure. It is in the same category as the simulation lab and the continuing medical education budget. It changes outcomes the organization is responsible for and is increasingly being financially measured against. The evidence supports the investment. The patients who will benefit are the ones you have not met yet.

Care differently, not less.