EBN common mistakes
EBN · Understand the traps and how to avoid them.
When a scenario asks which study design best supports a causal claim about a diet-disease link, the candidate picks the design with the largest sample size or the most prestigious-sounding name (e.g., a large cohort or a meta-analysis of cohorts) instead of the randomized controlled trial.
Candidates conflate 'strength of evidence' with 'size of study' and forget the evidence hierarchy is organized by control of confounding and bias, not by N. Large observational studies are memorable and often cited in guidelines, so they feel authoritative.
What to do instead
Before answering, ask: 'Does this design assign the exposure?' Rank designs by their ability to control confounding: systematic review of RCTs > RCT > cohort > case-control > cross-sectional > case report > expert opinion. Practice by labeling 20 published abstracts with their design and the causal claim each can and cannot support.
The candidate reads a results paragraph reporting a p-value or a relative risk and interprets it as an absolute risk or as the probability that the intervention works, or declares a result 'important' purely because p < 0.05.
Statistical significance and clinical significance are taught together but tested separately; candidates memorize 'p < 0.05 means significant' without internalizing what a p-value is (probability of data given the null, not probability the null is true) or the difference between relative and absolute effects.
What to do instead
For every result, compute or request the absolute effect (e.g., NNT = 1/ARR) and the confidence interval. Train a reflex: state the baseline risk, the absolute change, and the CI width before judging importance. Review the definitions of p-value, CI, power, type I/II error, and relative vs absolute risk reduction until you can define each in one sentence.
When asked to grade a body of evidence or apply a guideline, the candidate cites the existence of a single positive trial or a single guideline recommendation as sufficient, ignoring consistency, risk of bias, directness, precision, and publication bias.
Candidates treat guidelines as monolithic and assume any recommendation is strong. They have not practiced applying a structured appraisal framework (e.g., GRADE domains) to a body of literature, only to a single paper.
What to do instead
Memorize the domains that downgrade evidence (risk of bias, inconsistency, indirectness, imprecision, publication bias) and those that upgrade it (large effect, dose-response, plausible confounding opposing the effect). Practice writing a two-sentence evidence summary that names the direction, magnitude, certainty, and population for three different topics.
The candidate confuses the roles of the four steps of nutrition assessment (e.g., labels a food-frequency questionnaire as a 'nutrition-focused physical finding' or treats a BMI calculation as a 'biochemical' indicator) and misclassifies data when writing a PES statement.
Assessment domains are memorized as a list rather than as a taxonomy with decision rules. Under time pressure, candidates grab whichever label sounds closest instead of asking what the data actually measure.
What to do instead
Build a one-page table with the domains (food/nutrition-related history, anthropometric, biochemical, clinical, client history) and sort 30 real data points into it. Then practice converting each data point into the correct diagnostic label, and check that every PES statement's 'E' comes from a domain consistent with the 'P'.
The candidate writes a PES statement whose problem is a medical diagnosis or a symptom (e.g., 'diabetes' or 'poor appetite') rather than an approved nutrition diagnosis term, uses vague etiology ('noncompliance', 'poor diet'), or omits the signs and symptoms entirely.
Candidates draft from the patient's chart narrative rather than from the standardized diagnostic terminology, and they collapse etiology and signs/symptoms because the distinction is not intuitive. The 'related to' and 'as evidenced by' clauses are often treated as decoration.
What to do instead
Memorize the structure: Problem (from the standardized intake/clinical/behavioral-environmental domains) related to Etiology (the cause you can intervene on) as evidenced by Signs/Symptoms (objective data). Practice rewriting 15 flawed statements, and apply the test: 'Can I intervene on the etiology? Do the signs/symptoms objectively document the problem? Is the problem a nutrition diagnosis, not a disease?'
When asked to set an intervention goal or energy/protein prescription, the candidate gives a single number with no basis, or applies an aggressive weight-loss or protein target to a patient whose clinical status (e.g., renal disease, critical illness, pregnancy, older adult with frailty) calls for a different approach.
Candidates memorize formulas (Mifflin-St Jeor, 25–30 kcal/kg, 1.2–1.5 g/kg protein) without the conditions that modify them, and they default to the 'textbook healthy adult' when the vignette contains modifiers they did not scan for.
What to do instead
Before calculating, list the modifiers in the vignette (age, disease state, renal/hepatic function, pregnancy, obesity class, activity, wound status). State the predictive equation, the activity/stress factors, and the rationale in one sentence. Practice 10 vignettes where the same patient is altered by one modifier, and justify each change.
The candidate treats monitoring as a repeat of the baseline assessment, choosing the same indicators regardless of the intervention, or fails to specify a measurable indicator, baseline, target, and timeframe in the monitoring plan.
Assessment and monitoring use overlapping data, so candidates reuse assessment content verbatim. They also skip the 'what would tell me this is working' question because the vignette already supplied a lot of data.
What to do instead
For each intervention, name the mechanism and the indicator that would move first (e.g., carbohydrate counting → postprandial glucose and HbA1c; protein restriction → BUN/creatinine and nitrogen balance). Write monitoring plans in a fixed format: indicator, baseline value, target, frequency, and the action if the target is not met.
The candidate selects a nutrition intervention for a chronic disease based on a single macronutrient rule (e.g., 'low-fat for heart disease', 'low-carb for diabetes') without matching it to the patient's readiness to change, cultural food patterns, health literacy, or the current guideline's actual emphasis.
Candidates study disease-by-disease nutrient rules in isolation and do not rehearse integrating behavior-change stage, social determinants, and patient preference into the plan. Guideline updates also shift emphasis (e.g., overall dietary pattern over single nutrients), and candidates rely on older memorized rules.
What to do instead
For each major chronic condition, write the current guideline's core recommendation in one sentence, then add two patient-specific adaptations. Practice using a behavior-change framework (stage of change, motivational interviewing) to justify why one intervention was chosen over an equally evidence-based alternative.
The candidate answers ethics and communication items with what is legally permissible or clinically efficient rather than what the code of ethics and client-centered practice require, e.g., disclosing information to a family member who calls, or continuing an intervention the patient has refused.
Candidates conflate 'not illegal' with 'ethical', and they have not rehearsed the specific principles (autonomy, beneficence, nonmaleficence, justice, veracity, confidentiality) against realistic pressure scenarios such as family requests, physician orders, or scope-of-practice boundaries.
What to do instead
Learn the code of ethics principles and, for each, write one scenario where it conflicts with another (autonomy vs beneficence; confidentiality vs duty to warn). Practice a scripted response for the common pressure points: verify identity and authorization before disclosing, document refusal and offer alternatives, and refer beyond scope rather than practicing outside it.
The candidate answers scope-of-practice and interprofessional items by deferring entirely to the physician's order or, conversely, by acting beyond the dietitian's role, rather than identifying the dietitian's specific responsibility within the team.
Candidates are unsure where nutrition care authority begins and ends, and the exam vignettes deliberately place a plausible-sounding order that conflicts with evidence or with the patient's goals. Deference feels safe; overreach feels proactive.
What to do instead
Map the dietitian's scope: nutrition assessment, diagnosis, intervention, counseling, and monitoring, plus the obligation to question orders that conflict with evidence or patient preference and to document the rationale. Practice responses that state the concern, the evidence, the proposed alternative, and the escalation path, all within role.