You Can’t Predict Everything: What the Research Says About Doing This Work Well

If you’re a mental health provider, there’s a good chance you carry a quiet, specific fear that most clinicians rarely say out loud: What if I miss something? What if the thing I didn’t catch is the thing that mattered?

That fear deserves to be taken seriously and so does the research on where it comes from, what it distorts, and what it doesn’t actually say about your competence.

Hindsight makes the unpredictable look obvious

One of the most consistent findings in the clinical risk literature is hindsight bias: once an outcome is known, people, including trained psychiatrists, systematically overestimate how predictable it was beforehand. A study published in the Journal of the American Academy of Psychiatry and the Law found that clinicians reviewing case files after a bad outcome rated the risk as more foreseeable than clinicians reviewing the identical file without knowing what happened. That same body of research has found that simply telling an evaluator a patient went on to be violent makes them judge the treating clinician’s original decisions as less reasonable, even when nothing about those decisions actually changes.

This isn’t a character flaw in the people doing the judging. It’s how human cognition works under the weight of a known outcome. But it means that the internal replay so many clinicians do after a hard case, such as “The signs were right there, how did I not see it,” is running on a distorted instrument. The signs are only “right there” now, because now you know where to look.

It’s also worth saying plainly: the research on violence risk assessment consistently shows that prediction, even by experienced clinicians using validated tools, has real and well-documented limits. Studies have found that clinicians’ confidence in a risk assessment doesn’t reliably track its accuracy. That’s not a reason to stop assessing risk carefully. It’s a reason to stop expecting yourself to be a perfect instrument.

The legal standard was never “perfect”

Many clinicians hold themselves to an unspoken standard that no court or licensing board actually applies to them. Malpractice law is explicit on this point: it holds providers to a standard of reasonable care, not perfect care. That distinction was sharpened further in 2024, when the American Law Institute updated its Restatement of the Law governing medical malpractice to center negligence determinations on reasonable care rather than simply the customary practices of one’s peers. The standard a psychiatric or mental health provider is held to is what a reasonably prudent practitioner in similar circumstances would have done with the information available at the time, not the outcome that later became known, and not the theoretical best-case treatment with unlimited resources and hindsight.

Practically, that standard is built on things that are within your control: a documented and clinically sound risk assessment, appropriate consultation or referral when a case exceeds your setting’s capacity, adherence to your organization’s protocols, and decisions that a respected portion of your professional community would recognize as reasonable. It is not built on an expectation of never being wrong about a human being’s future behavior. No field of medicine holds its practitioners to that.

The toll this takes on you is real, and it’s common

There’s a name for what a lot of clinicians experience after a patient’s outcome goes badly, even one they had no part in causing: second victim phenomenon. It describes the guilt, intrusive replay, sleep disruption, and eroded confidence that can follow an adverse event. Estimates suggest roughly half of healthcare workers experience some version of this over the course of a career, and surveys of clinicians have found a troubling gap: a meaningful share report distressing work-related events in any given month, most say they’ve watched a colleague struggle with the same thing, but only a fifth feel they have adequate support available.

If you’ve ever left a hard case rattled, replaying old files, or quietly wondering if you’re cut out for this work, that reaction is not a sign you’re in the wrong field. It’s a documented, common response to doing emotionally weighty work in a system that doesn’t always build in enough support for the people doing it.

What actually builds confidence here

Lean on consultation and documentation as your actual safety net, not as paperwork. A well-reasoned, documented risk assessment is both good clinical practice and the concrete record of reasonable care if it’s ever questioned.

Notice when you’re doing a hindsight review on yourself, and name it. If you catch yourself reconstructing a past case with information you didn’t have at the time, that’s the bias at work, not evidence of a missed obvious sign.

Talk to peers. The research on second victim phenomenon is consistent that peer support is the intervention that helps most, and the one most clinicians say they don’t have enough of. A five-minute conversation with a colleague after a hard week counts.

Hold the standard you’re actually accountable to: Reasonable, informed, good-faith care, rather than the standard hindsight can make you feel like you should have met.

None of this is a case for complacency. Good clinicians keep sharpening their judgment precisely because the stakes are real. But sharpening judgment and carrying an impossible burden are two different things, and only one of them is asked of you. You do this work inside real limits: limits on what any human being can foresee, limits on the information available in the room, limits on what one provider can hold alone. Practicing well inside those limits, with sound judgment and the support of your peers, is not a lesser version of the job. It’s the job.

References

Hindsight Bias Among Psychiatrists, Journal of the American Academy of Psychiatry and the Law — jaapl.org/content/35/1/67

Examining Systemic and Interpersonal Bias in Violence Risk Assessments of Patients in Acute Psychiatric Care, Psychiatric Services, 2024 — psychiatryonline.org/doi/full/10.1176/appi.ps.20240108

Accuracy and Confidence in Clinical Assessment of Psychiatric Inpatients’ Risk of Violence, ScienceDirect — sciencedirect.com/science/article/abs/pii/S0160252798000326

Medical Malpractice: Coming Changes and Their Impact on Psychiatry, Psychiatric Services, 2025 — pubmed.ncbi.nlm.nih.gov/40432374

Second Victim Phenomenon: Impact on Healthcare Professionals, Organizational Responsibility and Support Strategies, 2025 — pmc.ncbi.nlm.nih.gov/articles/PMC12283087

Second Victim Phenomenon’s Impact in Male and Female Healthcare Workers: A Scoping Review, International Journal for Quality in Health Care, 2025 — academic.oup.com/intqhc/article/37/2/mzaf034

Peer Support Targeting the Second Victim Phenomenon: Implementation and Outcomes, PMC, 2025 — pmc.ncbi.nlm.nih.gov/articles/PMC11821368

Back to Blog Home

Latest Articles

• You Can’t Predict Everything: What the Research Says About Doing This Work Well
• Does Your EMDR Training Meet the Standard? Here’s What EMDRIA Approval Means
• Why Thousands of Clinicians Are Exploring Psychedelic Therapy for Stuck Clients
• When Seasonal Depression Doesn’t Lift With Spring
• Are Continuing Education Expenses Tax Deductible for Mental Health Professionals?

Featured Course

NAMI Ask the Expert: Good News for People With Mental Illness As the Great American Smokeout Nears OD031

Free Course

Ethics And Boundary Issues E097J