Ongoing Professional Practice Evaluation is among the most frequently cited requirements in international hospital accreditation. It is also one of the least well understood — not because the requirement is obscure, but because meeting it properly demands a data discipline most medical administration departments were never resourced for.
What the requirement is actually asking
Stripped of framework language, OPPE asks a hospital to answer one question continuously, for every privileged practitioner: is this clinician's performance consistent with the privileges we have granted them?
Three words in that sentence do most of the work. Continuously — not at reappointment. Every — not the ones who have attracted attention. Consistent with the privileges granted — measured against what this practitioner is approved to do, not against a generic set of hospital indicators.
The four findings surveyors write up most often
The evaluation is not ongoing
The most common citation. A hospital produces evaluations dated within a fortnight of each other, six weeks before the survey, covering a two-year period. The intent is visible; the continuity is not. OPPE is meant to be a monitoring cycle — monthly, quarterly or annually depending on the metric and the specialty — that generates a record as it runs.
There is no comparison
A number without a reference point is not an evaluation. “Dr A had a 1.4% complication rate” means nothing until you know that the specialty peer median is 1.9% and the external benchmark is 2.1%. Many hospitals collect physician-level data and then present it in isolation, which makes the evaluation unfalsifiable in both directions — it can neither confirm competence nor identify concern.
The metrics do not match the privileges
Evaluating an interventional cardiologist on generic documentation-completeness indicators does not tell you whether they should retain their interventional privileges. Metrics have to connect to scope of practice, which means the privileging record and the evaluation record need to be the same system, or at least the same source of truth.
Nothing happens as a result
If every practitioner passes every cycle, surveyors reasonably conclude the process is not discriminating. A credible OPPE programme has defined triggers, a route into focused evaluation, and a record of at least some practitioners having been through it.
Choosing metrics that survive scrutiny
A workable OPPE metric set is usually built from four categories, weighted differently by specialty:
- Clinical outcome measures relevant to the privileges granted — complication rates, unplanned returns, mortality where volumes make it meaningful.
- Process and guideline adherence — adherence to agreed clinical pathways, prophylaxis timing, documented consent.
- Resource and utilisation patterns — length of stay, diagnostic ordering, blood product use, read against case mix.
- Professional behaviour — documentation timeliness and completeness, response to calls, participation in required activity, complaint and compliment records.
The constraint that catches most programmes is volume. In a specialty with four practitioners and low procedure counts, individual outcome rates are statistically meaningless, and presenting them as though they were is worse than not presenting them at all. In low-volume settings the weight has to shift toward process adherence, documentation and peer review, with outcome measures aggregated over longer windows.
A metric a physician does not accept as fair will not change behaviour, whatever the accreditation requirement says.
The physician's side of it
OPPE is frequently experienced by medical staff as surveillance conducted about them rather than a process conducted with them, and that perception is corrosive. It produces disputes over data quality at exactly the moment when the data most needs to be trusted.
Two design choices reduce that friction substantially. The first is transparency of definition: every metric should have a written specification — numerator, denominator, exclusions, data source — that a practitioner can read and challenge in advance. The second is continuous visibility: a physician who can see their own trend and peer position at any point in the cycle experiences the review as a conversation about a known picture rather than a verdict delivered from a spreadsheet they have never seen.
Where the data has to come from
This is the practical crux. OPPE demands physician-level data drawn from clinical systems, aggregated consistently, compared to peers and benchmarks, and refreshed on a cycle. Done manually, it consumes medical administration capacity that does not exist, which is exactly why the evaluations end up clustered before a survey.
The systems that make it sustainable share three characteristics: the indicator engine is shared with the organisation's wider KPI programme, so a measure is defined once; the practitioner record is shared with privileging, so scope and evaluation cannot drift apart; and the cycle is driven by the system, so a review that is due generates a task rather than relying on someone remembering.
A realistic starting point
Hospitals building an OPPE programme from scratch tend to do better by starting narrow and running properly than by starting comprehensive and running late. A defensible first cycle usually looks like: three to five metrics per specialty, agreed in writing with the clinical lead; a quarterly cadence; peer median as the primary comparator with an external benchmark where one credibly exists; and a documented trigger threshold that routes into focused evaluation.
That is a programme a surveyor can inspect and a physician can defend. Comprehensiveness can come in year two — continuity cannot be retrofitted.