MEASUREMENT OF TOTAL PROTEIN IS NOT SUITA BLE FOR DIAGNOSING HYPOALBUMINEMIA IN CRITICALLY ILL PATIENTS

In clinical practice, the administration of supplementary albumin ofte n depends on the measured plasma concentration of total protein (TPC), A TPC of less than 5 g/dl is generally accepted as an indication for albumin therapy, assuming an albumin concentration of less than 2.5 g/ dl. However, a physiological relation between TPC and albumin cannot b e expected in critically ill patients, and thus, measurement of TPC ma y be misleading as an indicator for the use of albumin, Therefore, we investigated the sensitivity and specificity of TPC testing for diagno sing hypoalbuminaemia requiring treatment. Methods. In this prospectiv e study, 210 consecutive patients were included. Protein electrophores is was performed three times a week; the second electrophoresis was se lected for evaluation. Applied statistical analysis revealed the numbe r of positive total protein tests indicating hypalbuminaemia requiring treatment (sensitivity) and the number of negative with tolerably red uced albumin concentrations (specificity). Results. Of the investigate d patients, 27.6% had normal TPCs between 6.2 and 8.0 g/dl, In 81.9% o f cases an albumin concentration below 3.5 g/dl was found, while 43 pa tients had a concentration below 2.5 g/dl. The sensitivity and specifi city of TPC measurement for the diagnosis of clinically relevant hypoa lbuminaemia (albumin concentration <2.5 g/dl) was calculated at differ ent cutoff points for total protein. With a TPC of 6.0 g/dl, the sensi tivity was 0.96 and the specificity 0.44. With a cutoff point of 5.0 g /dl, the sensitivity was reduced to 0.65 and specificity increased to 0.86, Finally, with a TPC of 4.0 g/dl sensitivity was 0.25 and specifi city almost 1. Conclusions. Depending on the cutoff point for TPC, a r elevant albumin requirement would frequently not be detected. In other cases. a need for albumin would be assumed from a reduced TPC even th ough the albumin concentration still exceeded 2.5 g/dl. Therefore, det ermination of TPC is not a suitable indicator of the need for albumin replacement. As a result, we suggest routine determination of albumin concentrations instead of TPC.