The HIPAA Security Rule requires every covered entity and business associate to conduct an accurate and thorough assessment of the potential risks and vulnerabilities to the confidentiality, integrity, and availability of electronic protected health information. That assessment is the risk analysis. It is the foundation the rest of the Security Rule stands on: safeguards, policies, and training all flow from what the analysis finds.
It is also a requirement the Department of Health and Human Services keeps citing in enforcement. The Office for Civil Rights has run a dedicated Risk Analysis Initiative, and on June 18, 2026, it announced a $450,000 settlement with the Spencer Gifts employee health plan, its 14th enforcement action under that initiative, for failing to conduct an accurate and thorough risk analysis (Nixon Peabody). The pattern in these settlements is consistent: the organization either had no analysis, had an old one, or had one that didn’t cover everywhere electronic protected health information lives.
This article walks through what a defensible risk analysis contains. It is general information, not legal advice.
What “accurate and thorough” means in practice
A defensible risk analysis:
- Covers the full environment. Every system, device, and location where electronic protected health information is created, received, maintained, or transmitted — including personal devices, cloud services, and vendors. The most common failure is an analysis that covers the EHR and misses everything else.
- Identifies threats and vulnerabilities specifically. Not “malware is a risk” but which threats apply to which systems, and which vulnerabilities exist on each. Generic language is the tell of an analysis that was copied, not conducted.
- Assesses likelihood and impact. For each threat-vulnerability pair, how likely it is and what happens if it occurs. The ratings need a stated method, applied consistently.
- Determines risk levels. The analysis produces a risk rating per item, so the follow-up list has a defensible order.
- Documents everything. The analysis itself, the method, the evidence reviewed, and the people involved. If it isn’t written down, it didn’t happen — that is how an investigator will read it.
What a risk analysis is not
- Not a checklist. A yes/no questionnaire is not an analysis. It produces no risk ratings and no method.
- Not a tool output alone. Security Risk Assessment tools and scanners produce useful inputs. An analysis requires judgment applied to those inputs: what the findings mean for this organization, in this environment.
- Not a one-time event. The Rule requires the analysis to be reviewed and updated as the environment changes — new systems, new vendors, new locations, new threats. An analysis from three years ago describes a practice that no longer exists.
The follow-up list matters as much as the analysis
The analysis produces a list of risks. What happens next is the risk management plan: which risks get fixed, in what order, by whom, by when — and which risks are accepted, by whom, in writing. An analysis with no follow-up list is a document that proves you knew and did nothing. The follow-up list, kept current, is what turns the analysis into a program.
