Kuwait Board Internal Medicine common mistakes
Kuwait Board Internal Medicine · Understand the traps and how to avoid them.
When a vignette gives a serum sodium, glucose, and sometimes triglycerides or total protein, the candidate reports the measured sodium as the diagnosis instead of correcting for the osmotic effect, or corrects in the wrong direction (e.g., subtracts 1.6 mEq/L per 100 mg/dL of glucose above 100 rather than adding it, or applies the correction even when glucose is only mildly elevated).
Pseudohyponatremia from hyperglycemia is taught as a formula to memorize, but candidates rarely rehearse which direction the true sodium moves, and the older factor of 1.6 versus the modern 2.4 mEq/L per 100 mg/dL adds confusion. Under time pressure the measured value looks like the safe answer.
What to do instead
Work through the translocational hyponatremia logic from first principles: glucose is an effective osmole that pulls water out of cells, diluting extracellular sodium, so the measured sodium is falsely low and the corrected value is higher. Practice with three worked examples (glucose 250, 600, and 1200 mg/dL) and state aloud whether true sodium is higher or lower before calculating. Reserve pseudohyponatremia for marked hyperproteinemia or hyperlipidemia with normal measured osmolality.
On an ECG or rhythm strip described in text, the candidate commits to a diagnosis of ventricular tachycardia versus SVAR with aberrancy based on rate alone, without checking for AV dissociation, capture/fusion beats, concordance, or the Brugada or Vereckei criteria that the stem deliberately plants.
Wide-complex tachycardia is the highest-stakes ECG decision in internal medicine, and candidates default to 'VT is more common, so call it VT' or the reverse without applying a systematic algorithm. The stem usually gives one or two discriminating features, and the candidate reads past them.
What to do instead
Adopt a fixed checklist to run every time: (1) Is there AV dissociation (P waves marching through independently)? (2) Any capture or fusion beats? (3) Precordial concordance or extreme axis? (4) Was there a prior bundle branch block with identical morphology? Treat any positive feature as VT until proven otherwise and treat unstable wide-complex tachycardia as VT electrically. Rehearse the checklist on ten written descriptions until it is automatic.
When a patient presents with dyspnea and the stem provides a BNP or NT-proBNP, the candidate uses a single cutoff (e.g., 100 pg/mL) regardless of age, renal function, or obesity, or excludes heart failure because the value is only mildly elevated in a patient with obesity and AF.
BNP cutoffs are taught as a single number, but the interpretive thresholds shift with age and are lowered by obesity and raised by renal impairment and atrial fibrillation. Candidates also forget that BNP is a rule-out aid in ambiguous dyspnea, not a stand-alone diagnostic test.
What to do instead
Memorize the age-stratified NT-proBNP rule-out thresholds (e.g., 450, 900, and 1800 pg/mL for ages under 50, 50–75, and over 75) and remember that obesity lowers values while eGFR below 60 and AF raise them. In practice questions, always ask whether the result changes the pretest probability before committing.
In an arterial blood gas question, the candidate identifies the primary disorder but stops there, failing to calculate the expected compensatory response (Winter's formula for metabolic acidosis, the expected pCO2 for metabolic alkalosis, or the A-a gradient and the expected change in bicarbonate for acute versus chronic respiratory disorders), and so misses a superimposed second disorder.
Compensation formulas are memorized in isolation and rarely applied under time pressure, and candidates assume that any abnormal value is explained by the primary process. Mixed disorders are exactly what the exam tests.
What to do instead
For every blood gas, write down the primary disorder, then explicitly compute the expected compensation: Winter's formula (expected pCO2 = 1.5 × HCO3 + 8 ± 2) for metabolic acidosis, expected pCO2 rise of 0.7 per 10 mmHg acute and 0.4 per 10 chronic fall in PaO2 for hypoxia, and the 1-for-10 and 4-for-10 rules for metabolic alkalosis. If the measured value falls outside the expected range, name the second disorder.
When a stem describes a patient with suspected pulmonary embolism and gives a Wells score or a PERC rule result, the candidate jumps to CT pulmonary angiography without first applying the D-dimer in a low or intermediate pretest probability patient, or orders CTPA in a patient with a documented contrast allergy or renal failure without considering V/Q scanning.
Candidates conflate 'rule out PE' with 'order CTPA' and skip the pretest-probability step that determines whether a D-dimer is even useful. They also forget that the choice of imaging depends on contraindications.
What to do instead
Build a decision tree: PERC-negative low-risk patients need no testing; low or intermediate Wells with a negative age-adjusted D-dimer excludes PE; high probability goes straight to imaging. When CTPA is contraindicated by contrast allergy or severe renal impairment, choose V/Q scan or lower-extremity ultrasound. Practice the tree until the branch points are automatic.
In a hepatology vignette, the candidate interprets an isolated elevated AST or ALT as alcoholic liver disease or as 'mild hepatitis' without calculating the AST:ALT ratio, checking the R factor (ALT upper limit of normal multiples) to classify hepatocellular versus cholestatic injury, or considering the pattern (e.g., a ratio above 2 with elevated GGT and macrocytosis suggesting alcohol).
Liver enzyme interpretation is taught as 'AST and ALT are liver enzymes,' but the pattern, ratio, and magnitude carry the diagnosis. Candidates anchor on the single highest value and ignore the accompanying alkaline phosphatase, bilirubin, and albumin.
What to do instead
For every liver panel, compute the R factor (ALT/ULN divided by ALP/ULN): above 5 is hepatocellular, below 2 is cholestatic, and 2–5 is mixed. Then check the AST:ALT ratio (above 2 suggests alcohol, below 1 suggests viral or NAFLD), and always review synthetic function (albumin, INR, bilirubin) to judge acuity. Write these three steps on your scratch sheet before reading the answer choices.
In an acute kidney injury question, the candidate labels the cause prerenal because the FENa is below 1%, without checking whether the patient received diuretics (which raise FENa in prerenal states) or whether the urine sodium is confounded by CKD, and without using the fractional excretion of urea or the urine sediment to distinguish prerenal from ATN.
FENa is taught as the definitive prerenal-versus-ATN test, but its performance collapses in patients on diuretics, with CKD, or with contrast-induced nephropathy. Candidates apply the cutoff mechanically.
What to do instead
When FENa is uninterpretable, switch to FEUrea (below 35% suggests prerenal) and the urine sediment (bland or hyaline casts in prerenal, muddy brown casts in ATN). Also check the response to a fluid challenge and the BUN:creatinine ratio. State the confounder explicitly in your reasoning before choosing an answer.
When a stem describes a patient with hyperkalemia and ECG changes, the candidate selects 'calcium gluconate' only after seeing peaked T waves, or chooses insulin-dextrose without recognizing that calcium is the first-line therapy for any ECG change, and forgets to address potassium removal (binder, dialysis) or to check for the underlying cause (e.g., ACE inhibitor, adrenal insufficiency, type 4 RTA).
Candidates memorize the hyperkalemia treatment list as an unordered set and pick the most familiar drug rather than sequencing by urgency. They also skip the underlying-cause step because the acute management feels sufficient.
What to do instead
Memorize the sequence: (1) stabilize the myocardium with IV calcium for any ECG change; (2) shift potassium intracellularly with insulin-dextrose and albuterol; (3) remove potassium with a binder or dialysis; (4) treat the cause. For every hyperkalemia question, ask what the ECG shows, what the cause is, and whether the answer addresses all four steps.
In a diabetes question involving DKA or HHS, the candidate calculates an anion gap but fails to correct it for hypoalbuminemia, or treats the closure of the anion gap as the endpoint of therapy without checking that the patient has transitioned to subcutaneous insulin and that the underlying precipitant (infection, nonadherence, SGLT2 inhibitor) has been addressed.
The anion gap is taught as the marker of ketosis resolution, but hypoalbuminemia lowers the measured gap and can mask residual ketones, and candidates forget that DKA resolution requires more than a normal gap.
What to do instead
Correct the anion gap for albumin (add 2.5 to the gap for every 1 g/dL the albumin is below 4). Define resolution as a gap below 12 with corrected values, pH above 7.3, bicarbonate above 18, and the patient able to eat, then overlap subcutaneous and IV insulin for 1–2 hours before stopping the drip. Always name the precipitant.
In an infectious disease vignette, the candidate selects an antibiotic based on the organism name alone without matching it to the site, the patient's allergy history, local resistance patterns, or the need for source control, and picks a broad-spectrum agent when a narrow one is indicated (or vice versa).
Antibiotic questions reward a structured approach (organism, site, host, resistance, source control), but candidates answer from a memorized list and ignore the stem's details about allergies, prior cultures, or a drainable collection.
What to do instead
Before reading the choices, write four lines: organism and likely resistance, site and penetration, host factors (allergy, renal/hepatic function, pregnancy), and source control. Then eliminate any option that fails one line. Practice with a set of ten stems covering bacteremia, pneumonia, UTI, and intra-abdominal infection until the four-line habit is automatic.