Cholesterol in the ICU: a cheap and reliable marker for illness severity?

Introduction An intensive care unit frequently deals with acute-phase responses of patients to several noxae, whether infectious or not. Clinicians gather circumstantial evidence about improvement or deterioration of their patients every day. There is, however, no single parameter which informs the physician whether the initiated treatment is right or complications are developing. Evidence for infectious complications is important because the potentially evolving sepsis is associated with morbidity and mortality [1]. Fever has been seen as an early marker for infectious complications, although its positive predictive value for having an infectious complication has been questioned [2]. Also the combination of fever with other systemic inflammatory response syndrome criteria (tachycardia, tachypnoea and leucocytosis) is not very specific for infection and can occur in noninfectious conditions [3]. As an addition to clinical signs, daily inflammatory laboratory parameters as C-reactive protein (CRP) and procalcitonin can help to identify infectious complications. One can argue that if there is a definite improvement in the patient, routine measurements are not cost-effective. However, when the patient is not responding or responding slowly to treatment, infectious complications have to be considered. Measurements of CRP, procalcitonin and leucocytes have been linked to infectious complications. It was shown that a CRP ≥ 50 mg/L in combination with systemic inflammatory response syndrome (SIRS) in the first week of admission was a good predictor of infectious complications [4]. Another paper stated that patients with an increase of 8.7 mg/L in consecutive daily CRP measurements had an 88% risk of infection [5]. The much more expensive procalcitonin is also an excellent parameter for early detection of infectious complications. Several overview articles explain the value of procalcitonin in the identification of infectious clinical problems [6].