BSMMU Residency Phase-A common mistakes
BSMMU Residency Phase-A · Understand the traps and how to avoid them.
Spending the first 40-50 seconds on calculation-heavy biochemistry and physiology numericals (creatinine clearance, anion gap, oxygen delivery, base excess), then rushing the remaining one-liners with only a few seconds each.
Numericals feel 'high value' and candidates over-invest in them because they are familiar from textbook practice; under timed conditions the sunk-cost effect keeps them stuck rather than moving on.
What to do instead
Adopt a two-pass rule: on the first pass answer only items solvable in under 25 seconds, flag every numerical, and return to flagged items in the last third of the allotted time. Practise with a visible per-question timer so the 25-second cut-off becomes automatic.
Answering anatomy items from a single memorised fact (for example a nerve's root value) without checking the stem's qualifier words such as 'most commonly', 'except', or 'directly passes' before selecting.
Pre-clinical recall is stored as isolated associations rather than as relationships, so the first matching fact captures the answer and the qualifier is never parsed.
What to do instead
Underline or mentally box the qualifier in every stem before reading options, and rehearse anatomy as relations (structure to relations to clinical lesion) rather than as single facts. Do ten stems daily where the only difference is the qualifier.
Choosing the single most 'textbook-correct' diagnosis in internal medicine vignettes while ignoring the one abnormal value or time course in the stem that makes the classical picture atypical (for example chronicity, drug history, or a discordant investigation).
Pattern-recognition from undergraduate study rewards the prototypical presentation, and candidates stop reading once they recognise a syndrome.
What to do instead
Read the entire vignette before forming a diagnosis, then explicitly list the two or three findings that do not fit the leading diagnosis. If a discordant finding cannot be explained, re-rank the options instead of defaulting to the prototype.
Managing surgical items by naming the operation or investigation the candidate would personally choose, rather than the next step dictated by the resuscitation and prioritisation sequence (airway, then haemorrhage control, then imaging, then definitive surgery).
Surgical knowledge is learned as procedures and diagnoses, while examiners test sequencing and decision thresholds; candidates conflate 'definitive management' with 'next step'.
What to do instead
For every surgical stem, first classify the question as resuscitation, investigation, or definitive management, then answer within that category. Drill ATLS-style and acute-abdomen algorithms as ordered lists so the sequence, not the diagnosis, drives the answer.
In obstetrics and gynaecology items, applying a memorised threshold (blood pressure, glucose, haemoglobin, Bishop score) without checking the gestational age, parity, or whether the stem describes the mother or the fetus.
Thresholds are memorised as numbers detached from the context that defines them, and obstetric stems deliberately vary gestation and parity to test exactly that discrimination.
What to do instead
Before using any cut-off, write the three context variables (gestation, parity, maternal versus fetal) in the margin. Build a one-page table of thresholds with their stated context and review it weekly rather than memorising bare numbers.
Answering paediatric items with adult physiology or adult drug dosing, especially for fluid maintenance, resuscitation volumes, and normal ranges that change with age.
Adult medicine is studied longer and more intensively, so adult patterns are more accessible under time pressure; paediatric values are also age-banded and harder to recall as a set.
What to do instead
Maintain an age-banded reference sheet (neonate, infant, child, adolescent) for vital signs, fluid rates, and key drug doses, and consciously state the patient's age band before answering any paediatric calculation or management item.
Treating community medicine items as general knowledge and answering from intuition, without applying the specific study-design, bias, or screening-criterion framework the stem is testing.
The domain feels descriptive and non-clinical, so candidates do not prepare structured frameworks (validity, confounding, sensitivity versus predictive value, levels of prevention) and answer by plausibility.
What to do instead
Prepare a fixed decision framework for each recurring topic: study design selection, measures of association, screening test properties, and outbreak investigation steps. Apply the framework explicitly to every item rather than judging options by how reasonable they sound.
In forensic medicine and toxicology, choosing the answer that reflects the clinical severity of the poison or injury rather than the medicolegal construct the question asks about (manner of death, survivability, or the specific statutory criterion).
Clinical training dominates, so the clinically dramatic option feels correct; medicolegal categories are learned late and thinly, and the distinction between cause, mechanism, and manner of death is easily blurred.
What to do instead
Classify each forensic stem before answering: is it asking for cause, mechanism, manner, or a legal criterion? Revise toxicology by antidote and by toxidrome in parallel columns, and practise distinguishing these four constructs on short stems until the classification step is automatic.
Answering ethics and professionalism items by personal moral preference or by what is customary locally, instead of applying the four-principle framework, consent capacity criteria, and confidentiality exceptions in a stated order.
Ethics is perceived as opinion-based and receives little structured revision; candidates also conflate legal duty with ethical obligation and local practice with accepted standards.
What to do instead
Use a fixed sequence for every ethics item: identify the conflict, apply the relevant principle (autonomy, beneficence, non-maleficence, justice), check capacity and consent elements, then check confidentiality exceptions. Revise the capacity criteria and consent elements as a memorised checklist.
Selecting a recent-advances option because the technology or drug sounds new and impressive, without verifying that the stem's stated indication, guideline status, or level of evidence actually matches it.
Recent-advances material is encountered through headlines and promotional summaries rather than guidelines, so familiarity is mistaken for evidence, and candidates cannot distinguish established practice from experimental or superseded practice.
What to do instead
For each recent advance, record three fields: the exact indication, the guideline or regulatory status, and whether it is established, emerging, or experimental. Answer only from options that match all three fields, and treat any option whose status is unverified as a distractor.