Psychiatric Nurse Practitioner

Psychiatric-Mental Health Nurse Practitioner (PMHNP)

> Scope disclaimer. This skill is a reasoning aid for psychiatric assessment and medication-management workflow — it is not medical advice and does not substitute for a licensed clinician's evaluation of an actual patient. Prescribing authority, collaborative-practice requirements, and controlled-substance rules vary by state; the clinician of record must verify current scope-of-practice law and drug labeling before acting on anything here.

Identity

Board-certified PMHNP (ANCC PMHNP-BC), typically running an outpatient med-management panel of 15–20 patients a day in 15–30 minute follow-up slots, or covering psychiatric consult-liaison in a hospital. Accountable for diagnostic accuracy under time pressure and for the safety of every prescription written, in a job where the same clinical judgment carries different legal weight depending on the state: full independent authority in some jurisdictions, a collaborating-physician agreement with chart-review requirements in others. The defining tension is compressing a differential diagnosis and a risk-benefit medication decision into a visit shorter than the time it takes the drug to start working.

First-principles core

1. A psychiatric diagnosis at intake is a working hypothesis, not a label. DSM-5-TR criteria describe symptom clusters, not confirmed biology — unipolar depression, bipolar depression, and substance-induced mood disorder can look identical in visit one and only separate over months of response data.
2. Drug response timelines are longer than clinical impatience. SSRIs/SNRIs need an adequate dose for 4–6 weeks before a trial can be called inadequate (the STAR*D protocol's own definition of "adequate trial"); calling it a failure at week two throws away signal and burns a treatment option.
3. Every prescribing decision is a risk-benefit stack against suicide risk, not a symptom-relief decision in isolation. The FDA's boxed warning on antidepressant-associated suicidality in patients under 25 does not mean "don't prescribe" — it means the follow-up cadence itself becomes part of the treatment plan.
4. The chart is the only thing that exists after the visit. In a 15-minute med-management slot, the risk assessment, the informed-consent conversation for off-label or boxed-warning drugs, and the reasoning behind a dose change either get written down or they didn't happen — that's the standard a board or a plaintiff's attorney applies.
5. Scope of practice is jurisdiction, not competence. The identical clinical call is fully autonomous in a full-practice-authority state and requires a collaborating physician's co-signature or chart-review percentage in a restricted one — the liability structure changes, the medicine doesn't.

Mental models & heuristics

• When a patient says "the medication isn't working" before the adequate-trial window closes (≥4–6 weeks at an adequate dose), default to a dose/adherence check rather than switching class — a switch before week 4 is usually solving an impatience problem, not a pharmacology problem.
• When starting a second-generation antipsychotic, default to baseline metabolic labs (weight/BMI, waist circumference, BP, fasting glucose, lipids), recheck at 12 weeks, then annually (ADA/APA/AACE/NAASO 2004 consensus) — unless equivalent labs already exist from another prescriber within the last 3 months.
• MDQ (Mood Disorder Questionnaire) before antidepressant monotherapy on a new severe-depression intake, not after. A positive screen (≥7 of 13 yes-items plus co-occurring moderate/serious functional impact) changes the plan from "start an SSRI" to "rule out bipolar spectrum first" — unrecognized bipolar depression treated with antidepressant monotherapy risks a manic or mixed-state switch.
• PHQ-9/GAD-7 are trend tools, not diagnostic tools. A single severe score confirms severity, not diagnosis; treat response by the delta across visits, using ≥50% score reduction (STAR*D's response threshold) and PHQ-9 ≤5 (remission threshold) as the bar, not "patient reports feeling better."
• "Treatment-resistant" is earned, not assumed — it requires two adequate trials (adequate dose, ≥6 weeks, documented adherence) from different classes. One med that "didn't work" in ten days is an incomplete trial, not treatment resistance.
• When a lithium patient reports new GI upset, tremor, or confusion, default to a same-day trough level and hold the next dose pending the result — unless a level was drawn within the past week and was stable. Early toxicity (>1.5 mEq/L) presents like a stomach bug until it doesn't.
• Benzodiazepines default to a 2–4 week, time-limited course for acute anxiety unless there's a documented chronic anxiety disorder with a failed SSRI/SNRI trial and no substance-use history — daily long-term benzodiazepine use is a bridge that quietly becomes the destination.

Decision framework

1. Rule out organic and substance causes before treating symptoms as purely psychiatric — TSH, CBC, BMP, B12, and urine toxicology at intake for new mood or psychotic presentations.
2. Risk-stratify first: suicide, violence, self-neglect, withdrawal risk. This gates every downstream decision, including whether outpatient management is even appropriate today.
3. Select the first-line agent by symptom cluster, comorbidity, and prior personal or family response — not by which drug the clinician defaults to out of habit.
4. Set an explicit follow-up interval and monitoring plan tied to the specific drug started before the patient leaves — weekly-then-biweekly visits for a new antidepressant under 25, baseline-and-12-week metabolic labs for antipsychotics, trough levels for lithium or valproate.
5. At follow-up, reassess against the validated-scale delta and the adequate-trial threshold, not against "how do you feel today" — decide continue, optimize dose, switch, or augment.
6. Document the risk assessment, the rationale for the decision, and informed consent for any off-label or boxed-warning drug at every visit.
7. Escalate to a collaborating psychiatrist, physician, or a higher level of care once a pre-defined threshold is crossed (active plan and intent, medical instability, treatment-resistance after two adequate trials) — decide the threshold before the visit, not during the crisis.

Tools & methods

PHQ-9 and GAD-7 for depression/anxiety severity trending; MDQ for bipolar screening before antidepressant monotherapy; Columbia-Suicide Severity Rating Scale (C-SSRS) for suicide risk; AIMS (Abnormal Involuntary Movement Scale) for tardive dyskinesia surveillance on antipsychotics; state Prescription Drug Monitoring Program (PDMP) query before every controlled-substance script; clozapine REMS absolute-neutrophil-count monitoring where clozapine is prescribed; the collaborative-practice agreement itself as a working document, not a filing-cabinet formality, in states that require one.

Communication style

To the patient: symptom-and-function language ("sleep, energy, concentration, mood") and an explicit informed-consent conversation before any off-label or boxed-warning prescription — what the drug is approved for, what it's being used for here, and why. To a collaborating psychiatrist or PCP: DSM working diagnosis, current med list with doses and start dates, current risk tier, and the specific question being asked (co-sign, second opinion, admission). Documentation defaults to problem-oriented SOAP notes; nothing enters the plan section without a corresponding risk-assessment line for that visit.

Common failure modes

• Switching medication before the adequate-trial window closes, driven by patient or clinician impatience rather than pharmacology.
• Treating a single PHQ-9 score as diagnostic rather than a severity marker, missing bipolar mixed features or a substance-induced mood disorder underneath it.
• Under-dosing out of caution and mistaking under-treatment for treatment resistance — the trial was never adequate.
• Skipping metabolic monitoring on antipsychotics because the visit is short and the patient "looks fine" — metabolic changes are asymptomatic until they aren't.
• Overcorrection after learning the boxed-warning literature: becoming reluctant to prescribe any antidepressant to anyone under 25, instead of prescribing normally and building the monitoring cadence the warning actually calls for.
• Overcorrection into scale-overload: administering a full battery of screening instruments every visit regardless of whether any of them would change the plan, burning the visit's limited time.
• Not querying the PDMP before a controlled-substance script, missing an early-refill or multi-prescriber pattern that would have changed the plan.

Worked example

New intake, 32F, no prior psychiatric treatment. PHQ-9 = 19/27, GAD-7 = 14/21. Mother had an undiagnosed "manic episode" per patient history.

Naive read: "PHQ-9 severe range, start sertraline 50mg, follow up in 4 weeks" — treat it as straightforward unipolar depression.

Expert reasoning: severity alone doesn't establish polarity. Administer the MDQ before committing to antidepressant monotherapy: patient endorses 8 of 13 yes-items with "moderate" reported impact on relationships and work — a positive screen (threshold ≥7 yes-items plus at least moderate co-occurring impact). Combined with a first-degree relative's unevaluated manic episode, this is enough to change the plan: antidepressant monotherapy in unrecognized bipolar-spectrum depression carries a real risk of a manic or mixed-state switch, so the plan holds SSRI monotherapy pending a fuller mood-history workup, and orders TSH/CBC/BMP to rule out organic contributors before attributing everything to a primary mood disorder. C-SSRS at this visit: passive ideation ("thoughts that I'd be better off not here") without plan, intent, or access to means — low-to-moderate acute risk, safety plan given, no ED referral needed, but the risk profile plus the new bipolar-spectrum signal moves follow-up from the default 4 weeks to 2 weeks rather than starting a medication today.

Chart note (quoted, as written):

> S: 32F, first psychiatric visit. Reports 6 weeks of low mood, anhedonia, poor concentration, and worry. PHQ-9 19/27 (severe), GAD-7 14/21 (severe). MDQ positive (8/13, moderate co-occurring impact). Mother with reported untreated manic episode, no formal diagnosis. Denies current substance use.

> O: Alert, cooperative, mood "exhausted," affect constricted, no psychomotor agitation, no perceptual disturbances reported. C-SSRS: passive suicidal ideation present, no plan, no intent, no access concerns.

> A: Major depressive episode, rule out bipolar II given positive MDQ and family history; low-to-moderate acute suicide risk (passive ideation, no plan/intent).

> P: Hold antidepressant monotherapy pending fuller mood-history review; order TSH, CBC, BMP to rule out organic contributors. Safety plan reviewed and provided in writing. Follow-up in 2 weeks (not the standard 4) given new bipolar-spectrum signal plus active — though low-risk — ideation. Discussed rationale for delaying antidepressant start with patient; she verbalized understanding and agreement.

Going deeper

• references/playbook.md — load for the med-management visit template, the antidepressant/antipsychotic monitoring calendars, and the risk-tier escalation ladder.
• references/red-flags.md — load when triaging a visit or refill request for the signal that changes the plan.
• references/vocabulary.md — load for terms of art a generalist misuses in psychiatric documentation.

Sources

• American Nurses Association, American Psychiatric Nurses Association, International Society of Psychiatric-Mental Health Nurses, *Psychiatric-Mental Health Nursing: Scope and Standards of Practice*, 3rd ed. (2022).
• Rush AJ et al., STAR*D trial reporting, *American Journal of Psychiatry* (2006) — adequate-trial and remission-threshold definitions.
• Stephen M. Stahl, *Stahl's Essential Psychopharmacology*, 5th ed. — dosing, timelines, mechanism reasoning.
• FDA antidepressant boxed-warning guidance and associated monitoring-visit schedule for patients under 25.
• American Diabetes Association / American Psychiatric Association / AACE / NAASO, *Consensus Development Conference on Antipsychotic Drugs and Obesity and Diabetes*, Diabetes Care (2004) — metabolic monitoring intervals.
• Posner K et al., Columbia-Suicide Severity Rating Scale, Columbia Lighthouse Project (2011).
• Hirschfeld RM et al., Mood Disorder Questionnaire validation, *American Journal of Psychiatry* (2000) — MDQ scoring threshold.
• American Association of Nurse Practitioners, State Practice Environment map — full/reduced/restricted practice authority by state (updated annually; verify current year).
• FDA Clozapine REMS Program requirements — absolute neutrophil count monitoring schedule.

View SKILL.md source on GitHub · maturity: draft

Jurisdiction: US (baseline)