ABPN C-L Psychiatry common mistakes
ABPN C-L Psychiatry · Understand the traps and how to avoid them.
Administers and interprets a MoCA or Mini-Cog during an acute medical admission and documents 'cognitive impairment' or 'probable dementia' from the score, without first documenting arousal and attention (digit span, months backward, CAM) or obtaining collateral about baseline function.
A number is available and familiar, and the referral question ('is this dementia?') invites a score; acute illness, opioids, sleep deprivation, and hypoactive delirium depress performance, and the consultant wants to answer in a single visit.
What to do instead
Establish arousal and attention first (CAM, digit span), review medications, metabolic panel, and infection status, get collateral from family and nursing, state in the note that formal testing is deferred until delirium resolves, and re-test 4-6 weeks after discharge. If testing now, interpret the score as a floor rather than a baseline.
Treats the written consult request as the actual clinical question ('evaluate for depression') and answers it with a broad narrative note that ends in generic advice ('continue current medications, consider psychiatry follow-up'), without ever calling the referring clinician to ask what decision the consult is meant to inform.
Chart-based workflow and time pressure, plus a reluctance to 'bother' the primary team; trainees treat the note as the product rather than the decision being supported.
What to do instead
Before seeing the patient, call or message the requester: what is the question, what decision must be made, on what timeline, what has already been tried. State the reframed question in the first line of the note, answer it explicitly, and give prioritized, concrete recommendations (drug, dose, monitoring, who does what, when to call back).
Labels new tremor, clonus, hyperreflexia, and fever after an SSRI plus linezolid (or tramadol, methylene blue, ondansetron) as neuroleptic malignant syndrome, or attributes rigidity and autonomic instability in a mute patient to 'anxiety,' and then administers more antipsychotic.
The syndromes overlap and NMS is the pattern learned first in medical school; catatonia is under-taught; antipsychotics are the reflex response to agitation and rigidity on a medical ward.
What to do instead
Use time course and neuromuscular findings: clonus, hyperreflexia, and ocular clonus favor serotonin syndrome (stop serotonergic agents, supportive care, cyproheptadine if severe); lead-pipe rigidity, bradyreflexia, dopamine-antagonist exposure, and a days-long onset favor NMS (stop antipsychotic, dantrolene or bromocriptine, ICU support); use the Bush-Francis scale and a lorazepam challenge for catatonia, and avoid antipsychotics in malignant catatonia or NMS overlap.
Leaves a clozapine dose unchanged when the patient starts fluvoxamine or ciprofloxacin, stops smoking, or develops an infection with rising CRP; or adds a second QT-prolonging agent (haloperidol plus ondansetron plus methadone) without any ECG or electrolyte monitoring plan.
Attention is on the psychiatric symptom rather than the interaction; interaction checkers are not run or are dismissed; the new drug is ordered by the medical team and never reconciled with the psychotropic regimen.
What to do instead
Screen every new medical order against the psychotropic regimen using a CYP database (1A2, 2D6, 3A4, 2C19), know clozapine's narrow window and its sensitivity to smoking status and inflammation, reduce dose and check a level when adding an inhibitor, obtain ECG and electrolytes for QT combinations, and document the monitoring plan; involve pharmacy explicitly.
Consults on a quiet, withdrawn, poorly participating 78-year-old and diagnoses 'depression' or 'failure to thrive,' starting an antidepressant without a delirium workup.
Hypoactive delirium looks like depression; agitation is what usually triggers a consult, so the quiet patient is framed by the referring team as depressed; the CAM is not part of routine bedside assessment.
What to do instead
Test attention (digit span, CAM), review vitals, urinalysis, medications (anticholinergics, opioids, steroids, benzodiazepines), and obtain collateral on baseline and fluctuation; treat the multifactorial cause, avoid antipsychotics as a substitute for etiologic workup, and reassess cognition after resolution before making any mood or dementia diagnosis.
Concludes a patient lacks decision-making capacity because she refuses a recommended amputation or dialysis or because she carries a diagnosis of schizophrenia; conversely, documents capacity from the ability to state a choice alone.
An 'irrational' or risky choice feels like incapacity, and a serious psychiatric diagnosis is treated as global incompetence; the four abilities are not broken out separately, and the consult question is vague ('assess capacity').
What to do instead
Assess the four abilities for the specific decision: communicate a choice, understand relevant information, appreciate the situation and its consequences, and reason about options. Separate capacity from risk and from diagnosis, use teach-back or a decision aid, identify the exact decision and treat reversible contributors, and document each element plus who was involved. Capacity is decision-specific and may fluctuate.
Documents 'low risk' from a C-SSRS or PHQ-9 item 9 score and supports discharge, without a written formulation, collateral, means-restriction discussion, or a safety plan; never asks about firearms or medication stockpiles at home.
Scores feel objective and defensible; time is short; asking about suicide is uncomfortable; there is an assumption that medical inpatients are 'safe' because they are observed on the ward.
What to do instead
Combine structured screening with a documented formulation (risk factors, protective factors, acute warning signs, access to means), obtain collateral with permission, explicitly address firearm and medication access, write a safety plan with a follow-up appointment, and communicate the risk formulation verbally to the primary team and to the outpatient provider.
Recommends against or reduces opioids for a dying cancer patient because of a documented substance use disorder, or converts morphine to hydromorphone 1:1 and causes oversedation and respiratory depression; or labels escalating analgesic requests as drug-seeking.
Opioid phobia and the addiction frame dominate the consultant's thinking; equianalgesic tables are forgotten under time pressure; the concept of total pain and the distinction between dependence, tolerance, and addiction are not applied.
What to do instead
Distinguish physical dependence, tolerance, and addiction; treat pain and withdrawal as separate targets; use equianalgesic conversion with a 25-50% reduction for incomplete cross-tolerance and re-titrate; add adjuvants and co-manage with palliative care; use universal precautions (agreement, urine drug testing, prescription monitoring) rather than withholding analgesia.
Treats alcohol withdrawal with haloperidol or 'as-needed' antipsychotics alone, misses benzodiazepine withdrawal in a patient admitted for elective surgery who did not volunteer use, or gives a dextrose bolus before thiamine.
Antipsychotics are the default for agitation; substance history is stigmatized and not systematically asked; thiamine timing is treated as a trivial detail.
What to do instead
Screen every admission (AUDIT-C, brief benzodiazepine and opioid questions); use symptom-triggered benzodiazepine protocols (CIWA-Ar) for alcohol withdrawal and a structured benzodiazepine taper for benzodiazepine withdrawal; give thiamine before or with glucose; avoid antipsychotic monotherapy; involve addiction medicine early.
On learning of a pregnancy, advises abrupt discontinuation of lithium, valproate, or an antipsychotic, or tells the patient 'no psychiatric medication is safe in pregnancy'; alternatively continues valproate without any discussion or contraception planning.
Teratogenicity headlines and litigation fear drive advice; absolute versus relative risk is unfamiliar; there is no established collaboration with obstetrics, so the psychiatrist reasons alone.
What to do instead
Do an explicit risk-benefit analysis with the patient and the obstetric team: untreated illness carries substantial risk of relapse, suicide, and postpartum psychosis; valproate has the highest teratogenicity and should be avoided when possible; lithium's absolute Ebstein risk is small, and continuation with level monitoring and a fetal echocardiogram is often reasonable; never stop abruptly; plan postpartum prophylaxis and review lactation compatibility.