UHS DES Entrance Exam common mistakes
UHS DES Entrance Exam · Understand the traps and how to avoid them.
Answers anatomy items from memory of where a structure sat on a textbook diagram, so when the stem describes relations in words - for example, what lies immediately medial, posterior or inferior to a named structure at a named level - the candidate picks whichever option looked closest in the picture.
Anatomy is usually revised visually and passively, but the exam tests verbal three-dimensional relations. Picture memory records the layout of one illustration, not the anterior-to-posterior or medial-to-lateral ordering, so options that are merely near-neighbours feel correct.
What to do instead
Rebuild each region as a written relation ladder at named levels ("from medial to lateral: …", "from superficial to deep: …") and redraw cross-sections from memory on blank paper. Then self-quiz by describing one structure's relations aloud, without looking, before checking.
Recites normal values, curve shapes and equations accurately but stalls on perturbation stems ("what happens to X if Y is doubled / if this channel is blocked"), ending up guessing the direction of change rather than deriving it.
Physiology gets learned as a list of constants instead of as a set of controlled systems. Without practice in reasoning about the controlled variable, the sensor and the effector, there is no method for predicting the direction of a response.
What to do instead
For each system write four lines: controlled variable, sensor, effector, and the direction of correction. Then drill five perturbation questions per topic and justify each answer with the feedback loop before checking the key.
Names pathways and net ATP yields but cannot state the rate-limiting enzyme, its allosteric activators and inhibitors, the vitamin-derived cofactor at each step, or which step a deficiency or toxin blocks - and so misses items that ask exactly those things.
Pathways are memorised as sequences of names, which gives a false sense of mastery; regulatory and cofactor detail is treated as optional trivia rather than the part the exam actually probes.
What to do instead
Redraw each pathway with only three annotations per step: the regulated enzyme with its modifiers, the cofactor or vitamin, and the clinical block (deficiency, toxin, drug). Then test yourself by naming the block from the clinical picture, in reverse.
Selects the option containing the familiar buzzword from the stem rather than the option that satisfies every stated feature - choosing any granulomatous or any malignant option even when the site, stain, age or time course in the stem excludes it.
Pathology items are built from look-alike diagnoses, and rapid keyword matching feels efficient under time pressure. The candidate stops reading options once one triggers recognition.
What to do instead
Treat every option as a hypothesis and eliminate it with one stated feature ("this option requires … but the stem says …"). Build contrast tables for the classic look-alike pairs (granuloma types, renal tumours, thyroid lesions, ulcer types) with a single discriminator column.
Answers by recalling which drug "goes with" a disease, so items framed in reverse - which drug caused this adverse effect, which drug is contraindicated given this comorbidity, which drug is the antidote - are answered by elimination or guesswork.
Drugs are learned as disease-drug pairs rather than as a causal chain. Without a mechanism-to-effect-toxicity link, reverse-phrased and contraindication items have nothing to attach to.
What to do instead
Make one card per class: molecular target, clinical effect, characteristic toxicity with its mechanism, one absolute contraindication, and the antidote or reversal agent. Drill reverse items (toxicity → drug, comorbidity → drug to avoid) rather than only forward ones.
Learns organisms as isolated disease-name pairs, then cannot answer items that turn on Gram reaction, shape and arrangement, oxygen requirement, selective or enriched medium, toxin, or the nearest look-alike organism.
Microbiology is memorisation-heavy, so candidates memorise the shortest association available; the discriminating properties that generate the harder items are skipped because they feel like detail.
What to do instead
Keep one grid row per organism with fixed columns (Gram, morphology, oxygen, medium, key antigen or toxin, disease, first-line drug, main look-alike and the one feature that separates them). Cover a row and reproduce it, then build five-item quizzes from the look-alike column.
Commits to a diagnosis after the first two sentences of a long clinical vignette and answers without checking what the final line actually asks, which may be the next investigation, the mechanism, or the most appropriate initial therapy rather than the diagnosis.
Time pressure plus strong pattern recognition rewards early closure; distractors are deliberately written to be right for the diagnosis but wrong for the question asked, so the candidate feels confident while being wrong.
What to do instead
Read the question line first, then the stem, highlighting key positives, key negatives and the single decisive abnormal value. Write the question type (diagnosis / mechanism / next step / drug of choice) beside the item before looking at the options, and reject any option that answers a different question type.
Chooses definitive operative management for a patient described as still unstable or under-resuscitated, and mixes up "most likely diagnosis" with "best next step in management" when the stem asks only one of them.
Surgical revision focuses on procedures and operative detail, so the resuscitation-first sequence and the staged structure of surgical decision-making are less practised than the operations themselves.
What to do instead
Classify every stem into diagnosis, next step, or definitive management before reading options. Then apply the rule that the answer must be the intervention appropriate at this moment in the pathway, and check whether each option would come before or after stabilisation.
Anchors on the most dramatic finding in the scenario (the seizure, the low blood pressure, the abnormal lab) and picks treatment for it, while missing an immediately reversible cause or the actual life threat; in peri-arrest stems, orders ward-paced investigations and delays time-critical intervention.
Emergency scenarios are written with a distracting dramatic feature, and candidates transfer the slower diagnostic style learned on the wards. There is no fixed priority framework being applied under time pressure.
What to do instead
Run a fixed script on every emergency item: ABCDE, then the immediate threat, then reversible causes (the 4 Hs and 4 Ts, plus glucose, temperature, toxins, tension and tamponade), then disposition. Before selecting, ask "what kills this patient in the next five minutes?" and prefer the option that addresses that.
Confuses paired epidemiological and biostatistical concepts under exam pressure - incidence with prevalence, sensitivity with positive predictive value, relative risk with odds ratio, cohort with case-control - and computes values from a remembered formula without laying out the data first.
These terms are read many times but rarely used in calculation, so recognition substitutes for understanding. Without drawing the table, the row and column totals get swapped and the wrong denominator is used.
What to do instead
Draw and label the 2x2 table (exposed/unexposed, diseased/not, with totals) before touching any formula, write the formula symbolically, then substitute. For study-design items, first identify the sampling unit and whether sampling was exposure-first or outcome-first.