AOCNP common mistakes
AOCNP · Understand the traps and how to avoid them.
On any item naming a leukemia, lymphoma, plasma cell dyscrasia, or gynecologic cancer, defaulting to solid-tumor TNM logic - e.g., calling Ann Arbor stage IV lymphoma 'metastatic to a distant organ,' or treating Rai stage 0 CLL as advanced disease.
TNM is learned first and becomes the automatic mental template; the alternative systems are memorized as isolated fact lists without the trigger that tells you which system is in play.
What to do instead
Before choosing an answer, name the staging system out loud from the histology (Ann Arbor = lymphoma, Rai/Binet = CLL, FIGO = gynecologic, ISS/R-ISS = myeloma, TNM = solid tumor), then recall the single variable that defines that system's stages (B symptoms and nodal regions, absolute lymphocyte count, beta-2 microglobulin, tumor size/nodes/metastasis).
Treating a single-arm phase II response rate as evidence of survival benefit, or naming overall survival as the endpoint that must be met for accelerated approval, or saying a phase I trial establishes efficacy.
Phase names are memorized as a sequence without attaching each phase to its question (safety/dose vs. activity vs. comparative benefit) and its typical design (single-arm vs. randomized).
What to do instead
Map each phase to one question before answering: phase I = dose, safety, MTD/DLT; phase II = does it shrink tumors in this histology (response rate, single-arm allowed); phase III = does it improve a hard outcome versus a comparator (randomized; OS is the gold standard, PFS/ORR may support accelerated approval with confirmatory follow-up).
Applying average-risk screening intervals to a patient whose stem contains a risk qualifier - e.g., a 10-year colonoscopy interval after a prior advanced adenoma or in Lynch syndrome, or starting mammography at 40 in a known BRCA carrier with no MRI.
The average-risk schedule is heavily overlearned and the risk qualifier is usually buried in a subordinate clause of the history rather than in the question itself.
What to do instead
Scan the stem for three triggers before selecting a test or interval - personal cancer/precursor history, family history (and age at diagnosis), and germline mutation - assign the patient to a risk tier first, then pull the interval from a memorized high-risk list kept separate from the average-risk list.
Choosing CA-125, CEA, CA 19-9, or LDH as the answer to 'what test should be used to screen for' or 'what establishes the diagnosis,' and accepting a PET/CT as diagnostic of a new primary without tissue.
These markers have strong associative links to specific cancers in memory, so they feel correct regardless of whether the question is asking about screening, diagnosis, monitoring, or surveillance.
What to do instead
Classify the purpose of the test first - screening in an asymptomatic population, diagnosis, staging, or surveillance of known disease - and only then match the test; tissue is required to establish a new primary unless the stem describes a validated exception (e.g., a marker-plus-imaging pathway in a specific clinical scenario).
In febrile neutropenia, choosing to repeat cultures, wait for results, or give a non-antipseudomonal regimen, or adding vancomycin reflexively to every case.
General infectious-disease habits (culture before treating, tailor to results) and the desire to be 'targeted' conflict with the oncology rule that time-to-antibiotic drives mortality; vancomycin overuse comes from confusing line infections with all neutropenic fevers.
What to do instead
Memorize one threshold-plus-timer pair: a single oral temperature of 38.3 C or a sustained 38.0 C over one hour with an ANC under 500 means an antipseudomonal beta-lactam within 60 minutes, cultures drawn on the way to the antibiotic, not before it. Add vancomycin only for catheter-site infection, skin/soft tissue infection, pneumonia, or hemodynamic instability.
Continuing a checkpoint inhibitor while 'watching' grade 2 or 3 immune-mediated colitis, hepatitis, or hypophysitis, or reserving steroids for grade 4 toxicity.
Cytotoxic chemotherapy habits carry over - hold only for grade 3 or 4 marrow toxicity - but immune-related adverse events have much lower intervention thresholds and are treated with immunosuppression rather than dose delay alone.
What to do instead
Replace the chemo reflex with a grade-based irAE table: grade 2 gastrointestinal or hepatic toxicity means hold the drug and start corticosteroids; grade 3 means hold plus high-dose steroids; grade 4 (excluding some endocrinopathies) means permanent discontinuation. Endocrine irAEs are managed with hormone replacement rather than drug discontinuation in most cases.
For a patient reporting 7-10/10 cancer pain, selecting a weak opioid or an NSAID, or refusing to escalate the dose because the patient 'is already on opioids,' and converting between opioids without adjusting for incomplete cross-tolerance.
Opioid-phobia transfers from general practice, and equianalgesic conversion is often memorized as a single ratio rather than a table with a safety adjustment step.
What to do instead
Separate the two concepts in your own words before answering - tolerance and physical dependence are expected pharmacology, not addiction - and treat severe pain with a strong opioid titrated to effect. For any switch, use the standard equianalgesic row for the two drugs involved and then reduce the calculated new dose by 25-50% for incomplete cross-tolerance.
Answering with the definitive diagnostic or definitive-treatment option when the question asks for the immediate next action - e.g., ordering MRI before giving dexamethasone in a patient with known malignancy, new back pain, and leg weakness.
The most sophisticated-sounding option feels like the best answer, and the frame word ('next,' 'immediate,' 'priority') is a single word in the final sentence that is easy to skim past.
What to do instead
Re-read the last sentence and underline its frame - next/immediate/priority versus best diagnostic test versus definitive management. Where a reversible emergency is on the differential (cord compression, SVC syndrome, tumor lysis, hypercalcemia), the stabilizing step precedes the confirmatory imaging or oncology treatment.
Prescribing intensive surveillance for an asymptomatic survivor beyond what the guideline recommends - routine PET in treated lymphoma, indefinite tumor-marker and CT testing after resected colon cancer, or serial scans with no symptom or marker trigger.
Surveillance is assumed to follow the same logic as treatment response assessment, and 'more monitoring is safer' is a strong intuitive pull that is not supported for most curatively treated cancers.
What to do instead
Anchor each answer to the named cancer's surveillance regimen and its stop date (for example, colon cancer: CEA every 3-6 months and CT annually, both for about 5 years), and apply the general principle that surveillance is visit- and symptom-driven with modality-specific intervals; escalating imaging in an asymptomatic patient with normal markers is usually the wrong choice.
Advising a patient of reproductive potential to finish chemotherapy first and address fertility afterward, assuming an older or very young patient does not need a referral, or telling a patient it is safe to conceive while on endocrine therapy.
Fertility is filed mentally as a post-treatment quality-of-life issue rather than a time-critical pretreatment decision, and coordination failures are easy to make when the question is framed around starting therapy quickly.
What to do instead
Trigger on any patient with reproductive potential facing gonadotoxic therapy: refer to reproductive endocrinology and document the discussion before cycle one, recognizing that oocyte or embryo cryopreservation takes roughly two weeks and ovarian suppression is an adjunct rather than a substitute. Also state that pregnancy is avoided during endocrine therapy and for the drug-specific washout interval afterward.