Preanalytical quality in clinical chemistry laboratory.

Background: Haemolysis is usually caused by inadequate specimen collection or preanalytical handling and is suggested to be a suitable indicator of preanalytical quality. We investigated the prevalence of detectable haemolysis in all routine venous blood samples in OPDs and IPDs to identify differences in preanalytical quality.

Methods: Haemolysis index (HI) values were obtained from a Vitros 5,1 in the routine clinical chemistry laboratory for samples collected in the outpatient department (OPD) collection centres, a hospital, and inpatient departments (IPD). Haemolysis was defined as a HI > or = 15 (detection limit).

Results: Samples from the OPD with the highest prevalence of haemolysis were 6.1 times (95% confidence interval (CI) 4.0 - 9.2) more often haemolysed compared to the center with the lowest prevalence. Of the samples collected in primary health care, 10.4% were haemolysed compared to 31.1% in the IPDs (p = 0.001). A notable difference in haemolysed samples was found between the IPDs section staffed by emergency medicine physicians and the section staffed by primary health care physicians (34.8% vs. 11.3%, p = 0.001).

Conclusions: The significant variation in haemolysis indices among the investigated units is likely to reflect varying preanalytical conditions. The HI is a valuable tool for estimation and follow-up of preanalytical quality in the health care laboratory.