The P2 harm probability enumeration defines a six-level scale (0-5) measuring the conditional probability that a hazardous situation results in actual harm.
P2 answers the question: “Given that the hazardous situation has occurred, how likely is the patient to actually be harmed?” This separation from P1 enables more precise risk estimation, as many hazardous situations do not always result in harm.
While both P1 and P2 use 0-5 scales, they measure probability differently:
Factor
Measures
Scale Type
Range
P1 (Hazard Probability)
How likely the hazardous situation occurs
Frequency-based
≤1/100,000 to ≤1/1
P2 (Harm Probability)
How likely harm results given the hazardous situation
Percentage-based
≤5% to ≥96%
This deliberate difference reflects the ISO 14971 distinction between situation occurrence (best expressed as frequency over device lifetime) and harm occurrence (best expressed as conditional percentage per exposure).
Example: A use step where the hazardous situation is Likely (P1=4) but harm is Rare (P2=1) yields a product of 4. Using the HARA bucketing thresholds, a product of 4 maps to composite probability P=2 (Low). This means the risk matrix evaluates at a lower probability level than the hazard occurrence alone would suggest, reflecting that the hazardous situation rarely causes actual harm.
Risk controls can target P2 reduction specifically. For example:
Protective barriers may reduce P2 by preventing the hazardous situation from reaching the patient
Detection mechanisms may reduce P2 by enabling intervention before harm occurs
Design changes may reduce P2 by making the hazardous situation less likely to cause harm even when it occurs
When evaluating residual risk, assess the postHarmProbability field independently from postHazardProbability to capture controls that affect harm occurrence specifically.
