Malnutrition, symptom burden and predictive validity of the Patient‐Generated Subjective Global Assessment in Central Australian haemodialysis patients: A cross sectional study

AIM: To (i) describe the prevalence of malnutrition among a cohort of central Australian, predominantly Indigenous, haemodialysis patients and (ii) determine the sensitivity and specificity of the Patient Generated Subjective Global Assessment total score for identification of malnutrition in these patients. METHODS: Cross‐sectional observational study of all patients attending haemodialysis units within the Central Australia Health Service. Patients were assessed using the Patient‐Generated Subjective Global Assessment. Chi‐Square tests were used to determine the association between nutritional status and location, age and gender. Receiver Operator Characteristic curves were used to ascertain the predictive validity for malnutrition of the total score. RESULTS: Indigenous patients comprised 98% of study haemodialysis patients (n = 249/253). One third were male, and 72% were aged between 30 and 59 years. Approximately 29% (74/253) were malnourished, and 93% (69/74) had a total score ≥ 4. The most frequently reported problems that kept malnourished patients from eating were early satiety (32%), no appetite (31%), diarrhoea (26%) and dental problems (24%). Money problems were reported by 32%, as were transport (20%) and depression (19%). The traditional tool cut off score of ≥9 had low sensitivity (50%) for detecting malnutrition. Instead, a score ≥ 3 is suggested due to a higher sensitivity (96%) and specificity (45%). CONCLUSION: Malnutrition was found to be common, and we suggest using a Patient‐Generated Subjective Global Assessment total score of ≥3 to improve the identification of malnourished individuals in this cohort of predominantly Indigenous haemodialysis patients. This will significantly increase referrals for dietetic intervention.