In Denmark the dieticians can be organized centrally and called upon if needed, or locally in larger units, or very locally related to bed-units. The organization form is based on tradition not on measurements of quality. We wanted to compare the very local form with the traditional centralized.

Methods: Randomized, unblinded, parallel design. The intervention group (IT) was nutritionally handled by the dietician, and the control group (C) by the nurses as previously with the possibility for dietetic help from the center.

All patients were screened by NRS-2002 and had their estimated needs calculated (equations in IT, tables in C). The IT was in addition assessed for adapted weight-loss, counseled with a plan for nutrition and followed-up daily by the dietician if at nutritional risk. All patients were re-screened weekly and had their intake registered. Hand-grib-strength (HGS) was measured in both groups at admittance and discharge, and weight three times weekly.

The primary endpoint was re-admission rate to hospital within 2 months.

Inclusion Criteria:

• adult
• acutely admitted to hospital

Exclusion Criteria:

• dementia
• unable to comply with the protocol

Results handled without knowledge of randomization group