Avoidant restrictive food intake disorder OSCE — boundary, safety and evidence-based care

Station objectives

1. Hold the boundary between ARFID and anorexia nervosa by screening explicitly for weight and shape overvaluation. [2] [3]
2. Assess medical and refeeding safety in the low-weight or falling-weight child. [6]
3. Classify the dominant ARFID driver pattern and choose pattern-matched evidence-based therapy. [3] [4]
4. Stop coercion and co-build a non-coercive, family-supported plan. [3] [5]

Candidate brief

You are the paediatric doctor in a general paediatric clinic. You have 10 minutes for Station A (13-year-old, sudden food refusal after choking, falling weight) and 12 minutes for Station B (8-year-old with autism, six accepted brand-specific foods, faltering growth, force-feeding). Examiners score safety, boundary-holding, evidence-based framing, and partnership language with exhausted caregivers. [3] [5]

Station A — Sudden food refusal after a choking scare, falling weight

Setup: A 13-year-old boy has refused all solids for six weeks after choking on meat at a family dinner. He now takes only smoothies and milk and has lost 5 kg. He is anxious at the table and panics if solid food is offered. He is clear he is not trying to lose weight and is upset about the weight lost. No purging, no body-image concern, no compulsive exercise. His BMI has dropped but remains above the 0.4th centile. [3] [1]

Expected actions:

• See the young person alone; state conditional confidentiality with its lawful limits. [3]
• Screen explicitly for anorexia-nervosa boundary: drive for thinness, fear of weight gain, body-image disturbance, purging, compulsive exercise — and document its absence. [2]
• Classify ARFID fear or aversive subtype: sudden onset, dated trigger, anticipatory anxiety and panic at the table. [3]
• Assess medical safety: hydration, ability to maintain fluids, hypoglycaemia risk, tempo of weight loss; screen for refeeding risk given the weight loss and consider an electrocardiogram if bradycardic. [6]
• Name the evidence-based therapy: CBT-AR with systematic desensitisation and graded, non-coercive exposure matched to the phobic mechanism — never forcing solids to prove safety. [4]
• Protect nutrition with dietetics; give a clear safety-net and the red flags that mandate urgent return. [3]

Station B — Autism, brand-specific selectivity, faltering growth, force-feeding

Setup: An 8-year-old boy with autism eats only six foods, all beige, from specific brands; he will not eat if the packaging changes. His weight has crossed two centiles down over a year. His mother has started force-feeding and mealtimes last ninety minutes and end in tears. He has no interest in losing weight and no body-image concern. [3] [7]

Expected actions:

• Validate the mother without blame before changing anything — she has been trying to protect his growth. [3]
• Explain that force-feeding amplifies the avoidance loop and worsens outcome; stopping coercion is treatment, not giving up. [3]
• Give the first prescription tonight: stop force-feeding, restore a calm meal structure of three meals and two to three snacks with no pressure. [3]
• Classify ARFID sensory-based subtype in autism; hold the anorexia-nervosa boundary by confirming no weight or shape overvaluation. [2]
• Order targeted investigations guided by the narrow diet and faltering growth — full blood count and iron studies, vitamin D, micronutrients by pattern including vitamin C for scurvy risk in autistic ARFID; plot serial growth. [7]
• Co-build the plan: dietetics to close gaps within accepted foods, CBT-AR or an FBT adaptation with sensory-based exposure changing one variable at a time, neurodiversity-affirming goals, treat comorbid anxiety, set shared functional goals. [4] [5]

Marking anchors

Clear pass: holds the ARFID versus anorexia nervosa boundary by screening explicitly for overvaluation; assesses medical and refeeding safety; classifies the dominant driver pattern; names CBT-AR or an FBT adaptation with the correct mechanism-matched exposure; stops coercion and validates the family without blame; gives a concrete safety-net and functional goals. [2] [4] [6] Borderline: validates well but offers vague therapy, defers the safety-net, or avoids the coercion conversation. Fail: prescribes force-feeding or pressure tactics; misses refeeding or medical-safety risk; fails to screen for anorexia nervosa; sets a perfect-plate rather than functional goal; dismisses the autism sensory reality; uses stigmatising language. [3] [6]

Debrief pearls

• The single screening question that catches crossover: does it bother you what you weigh or what your body looks like? [2]
• Stopping coercion is the most powerful first intervention in the coercion-amplified loop. [3]
• CBT-AR is pattern-matched — name the mechanism and the exposure type. [4]
• A safety-net with red flags and a next weight-check date is the contract that lets an ambulatory plan proceed. [6]