How should I approach the initial assessment of a patient suspected to have bulimia nervosa?

Guideline-aligned answer with reasoning, red flags and references. Clinically reviewed by Dr Kola Tytler MBBS CertHE MBA MRCGP.

Posted: 16 August 2025 Guideline-Aligned (High Confidence) Clinically Reviewed
Dr Kola Tytler MBBS CertHE MBA MRCGPClinical Lead • iatroX

Initial assessment of a patient suspected to have bulimia nervosa should include:

  • Taking a detailed history focusing on eating behaviours, including binge eating episodes and compensatory behaviours such as vomiting, laxative or diuretic misuse, and excessive exercise.
  • Assessing physical health for signs of malnutrition, electrolyte imbalances, dental erosion, and other complications related to bulimia nervosa.
  • Evaluating mental health status, including screening for common comorbidities such as depression, anxiety, self-harm, obsessive-compulsive disorder, and substance misuse.
  • Considering any rapid weight changes, menstrual or endocrine disturbances, and gastrointestinal symptoms that may be associated with bulimia nervosa.
  • Reviewing social factors such as withdrawal from food-related situations and involvement in high-risk activities (e.g., professional sport, modelling).
  • Not relying solely on BMI or screening tools like SCOFF for diagnosis but using a comprehensive clinical assessment.
  • Assessing the need for emergency care if there is compromised physical health or suicide risk.
  • Providing psychoeducation about the disorder and discussing the limited effect of psychological treatments on body weight.
  • Referring immediately to an age-appropriate community eating disorder service for specialist assessment and treatment without delay.

During assessment, ensure coordination between primary and specialist services, and consider involving family or carers where appropriate, respecting patient confidentiality and Gillick competence in young people.

Educational content only. Always verify information and use clinical judgement.