Documented dietitian consultation in hospitalized underweight adults varies more by admitting diagnosis than by nutritional severity: a retrospective single-center study

Background Disease-related malnutrition is common in hospitalized adults and is associated with worse outcomes, but nutritional care begins only when a dietitian sees the patient. Without a protocolized admission screen, referral rests on clinician discretion. We described documented dietitian consultation among underweight adults at a Saudi tertiary center with no admission nutrition screening protocol. Methods We screened 126,267 admissions between 2017 and 2022 and analyzed 381 non-pregnant, non-smoking adults aged 18–55 years without cancer, diabetes, hypertension, or critical illness, with a recorded stay of at least 5 days and admission body mass index below 18.5 kg/m 2 ; only the first eligible admission per patient was retained. The outcome was a completed dietitian consultation documented in the record. We described its prevalence, timing, and distribution across ICD-10 chapters, and modeled correlates with Firth penalized logistic regression. Length of stay was examined descriptively. Reporting followed STROBE. Results A consultation was documented in 141 of 381 patients (37.0%, 95% confidence interval [CI] 32.3–42.0), and within 24 h in 54 (14.2%, 95% CI 11.0–18.0; 38.3% of those consulted). Consultation ranged from 6.5% (blood and immune-system disorders) to 60% (congenital conditions; omnibus p = 0.0005). In the multivariable model, consultation was associated with body mass index (adjusted odds ratio 1.36 per 1 kg/m 2 lower, 95% CI 1.20–1.56) and admitting diagnosis (global penalized likelihood-ratio p < 0.001). Relative to respiratory admissions, blood and immune-system diagnoses were least likely to be consulted (adjusted odds ratio 0.17, 95% CI 0.04–0.56; adjusted absolute difference 23.7 percentage points, 95% CI 10.7–36.2). There was no statistical evidence of a length-of-stay difference by consultation timing (median difference 0.8 days shorter with early consultation, 95% CI 4.0 shorter–2.1 longer; p = 0.63). Conclusion Under discretionary referral, consultation varied roughly ninefold across admitting diagnoses and was associated more strongly with diagnostic category than with nutritional severity. Because no protocol defined which patients should have been seen, these data cannot establish whether 37% represents under-provision. Standardized admission screening with a defined referral trigger could reduce this dependence and requires prospective evaluation.