Performance Reviews for Clinicians: What to Measure Beyond Productivity

A clinician review is fair when it rates the things you actually handed them in writing: the job description, the documentation policy, the productivity targets, the treatment-model expectations.

If an expectation was never written down, a review is the wrong place to introduce it — a surprise standard creates a disagreement you cannot document your way out of.

The domains worth rating:

• Documentation. Completeness and timeliness of notes, signatures and treatment-plan updates, measured against your written documentation policy — the record payers, auditors and any successor clinician may later rely on.
• Outcome measurement. Whether the clinician administers the standardized tools you have adopted, records scores and uses them in treatment planning — covered in detail below.
• Productivity. Delivered hours or caseload against the targets you set in writing, judged in context — also below.
• Work in supervision. Quality of case presentations, consultation-seeking, and how the clinician responds to feedback — the evidence comes from supervision notes and your own observations.
• Professional conduct. Schedule adherence, attendance, boundary and ethics-code compliance, and how they work with the rest of the team.
• Development. Progress on the goals set at the last review, so the cycle actually cycles.

Rating discipline is what makes the form fair.

Every rating needs at least one dated example you could show the clinician; if you cannot name one, the rating is not ready to give.

Rate observable work rather than personality — an "attitude" rating invites a disagreement you have no evidence for, while a rating backed by a returned-for-corrections documentation sample is hard to argue with.

The criteria should also match the role you actually hired.

If the job description is due for a rewrite before the next hire, start from the behavioral health hiring hub and rebuild the role before you rate it.

If your practice runs measurement-based care — standardized symptom tools administered during treatment, not just at intake — the data gives you an outcome section you did not have before.

It feeds a review in two ways: what the clinician does with the tools, and what the aggregated scores say.

The accreditation anchor is concrete.

Effective January 1, 2018, Joint Commission standard CTS.03.01.09 requires all Joint Commission-accredited behavioral health care organizations to assess outcomes of care using a standardized tool or instrument.

Under its elements of performance, the organization uses the tool to monitor each individual's progress, analyzes that data and uses it to inform the individual's care goals, and aggregates it to evaluate outcomes for the populations it serves.

Two scope limits keep that sentence honest.

It is an accreditation standard, not a law, and it binds the organization that holds Joint Commission accreditation — it does not tell you how to evaluate an individual clinician.

An accredited practice has to run measurement-based care; how outcome data appears in a performance review is your design decision.

Survey data shows why the "uses the data" part deserves its own line on your form.

In Joint Commission surveys from July 1, 2021 to March 31, 2022, the element requiring organizations to use measurement data to inform treatment goals drew 59 high- and moderate-risk findings — 4% of 1,570 full behavioral health surveys.

A common finding behind those numbers: organizations administering tools like the PHQ-9 and GAD-7 only at intake and discharge, so results never informed anyone's treatment goals, with inadequate clinician training among the contributing factors the Joint Commission lists.

Turn that into criteria a clinician controls: tools administered on schedule, scores documented, plans adjusted, results discussed in supervision.

Aggregate scores need more care: before you read an average as a measure of the clinician, ask what the case mix contributed — use aggregates to open a conversation rather than to close one.

And if you have not adopted measurement yet, say so in the review: until the data exists, documentation and supervision feedback carry the outcome section.

Productivity belongs in the review only when the target was written and communicated in advance — our guide to setting productivity targets covers that design.

In the review, judge each clinician against their own scheduled hours and case mix, not against whoever billed the most.

Then apply the disability-law boundary.

Under EEOC guidance interpreting the ADA, an employer is not required to lower production standards — qualitative or quantitative — that are applied uniformly to employees with and without disabilities.

What you may owe instead is an accommodation that helps a clinician with a disability meet the standard.

Scope check before you rely on that: the federal ADA covers employers with 15 or more employees for each working day in each of 20 or more calendar weeks in the current or preceding calendar year.

California's FEHA applies to employers regularly employing five or more persons, so a California practice below the federal threshold can still owe accommodation under state law.

The other states' thresholds were outside this guide's research — check yours before assuming none of it applies to you.

EEOC guidance also settles the medical-cause question in the employer's favor: an employer does not have to excuse poor job performance even if it was caused by a medical condition or medication side effects — which is why the EEOC urges employees to ask for accommodation before problems arise.

In review terms: rate the work as you find it, and if a health matter surfaces, the accommodation process runs alongside the review, not through it.

One trap to design out of the form entirely.

Penalizing an employee for work missed during leave taken as a reasonable accommodation — counting it against a productivity ranking is the EEOC's own example — is retaliation and a denial of accommodation.

Before you rate productivity or attendance, check whether an open accommodation explains part of the numbers.

Pre-licensure associates and certified paraprofessionals work under supervision that a board or certifying body defines, and that supervision is not your performance evaluation.

They are two tracks with two documents, and the distinction is easiest to keep when you can see it side by side:

Board law explains why.

California's LPCC statute, Business and Professions Code 4999.12(m), defines clinical supervision of a counselor associate to include monitoring and evaluating the supervisee's assessment, diagnosis and treatment decisions and providing regular feedback — and says plainly that consultation or peer discussion is not supervision and does not qualify as supervised experience.

That is California's definition for counselors; other states and professions write their own definitions, so check the board that issued your associate's title before you assume what supervision requires.

The ABA side already runs on this logic.

The BACB Ethics Code requires behavior analysts to collect and document ongoing performance-monitoring data on supervisees, give timely feedback, and — when performance problems arise — develop, communicate, implement and evaluate an improvement plan with clearly identified procedures.

That is an ethics duty of the supervising behavior analyst, not an employment-law requirement imposed on you.

The same code requires supervisors to continually evaluate their own supervisory practices, using feedback from others and client and supervisee outcomes, and to document those self-evaluations — feedback is meant to run in both directions.

For RBTs, the BACB lists developing performance expectations, observation with behavioral skills training and feedback, and ongoing evaluation of the effects of supervision among supervision activities.

Practically, this makes the formal review easier, not harder: for supervised roles the evidence accumulates continuously in supervision, and the review summarizes the record instead of introducing surprises.

The paperwork mechanics — who supervises, what the agreement says — are covered in our guides to clinical supervision agreements and RBT supervision for employers.

When a review documents a real gap, a performance improvement plan converts concern into a plan with terms.

It should name the expectation that was missed, with the evidence; the measurable standard that closes it; the support you will provide — training, schedule change, closer supervision; the check-in dates; the end date; and what happens if the standard is not met.

You set those terms when you write the plan — no template arrives from outside.

If the clinician is a behavior-analytic supervisee, part of the structure is already required of their supervisor: the BACB's improvement plan for performance problems must carry clearly identified procedures, and it must be communicated, implemented and evaluated.

Mirror that shape for everyone else and your plans get more consistent on their own.

Keep the plan about performance.

The ADA limits when you may ask a current employee about a disability — such inquiries have to be job-related and consistent with business necessity — so a PIP is not the place to probe for a diagnosis, and a clinician does not owe you one to stay in the plan.

If a health condition enters the conversation, route it to your accommodation process; as above, EEOC guidance does not require you to excuse poor performance because a condition caused it.

Document every check-in: date, what improved, what did not, both signatures.

A plan that succeeds gets a written close-out saying so.

One that fails becomes the first half of a separation record — which is the point where our guide to terminating a clinician takes over.

A review that cannot be reconstructed later did not, for practical purposes, happen.

The written record should contain the criteria used, the evidence with dates, the ratings, goals for the next period, the clinician's written response, and signatures with the date.

Give the clinician the document before the meeting and room to answer in writing afterward.

Record disagreements as their response rather than arguing them out of the file — a review with a documented dissent is more credible than one that pretends agreement.

Use the same form and the same criteria for everyone in the same role, and when you deviate from your own process, write down why at the time.

Keep medical material out of the review document entirely.

EEOC regulations require any medical information you obtain about an employee to be kept on separate forms in separate confidential medical files, and they limit what managers may be told: supervisors and managers may learn necessary work restrictions and accommodations, not the diagnosis.

A review form that quotes a diagnosis is a confidentiality mistake with a date on it.

The file's second job shows up later.

If the employment ends, the review trail — ratings, dated examples, the improvement plan, the check-in notes — is the history a separation decision stands on, and the quality of that history is decided now, not at separation.

Finally, unless your accreditor or program license sets an interval, pick a cadence you can sustain and hold to it.

A yearly written review with shorter check-ins in between is a workable rhythm for a practice; whatever you choose, the useful rule is that no rating in it should ever be a surprise — which, for supervisees, is exactly what continuous supervision feedback is for.