Design. This qualitative study was conducted in collaboration with one of the five emergency departments at a Finnish university hospital and the adjoining laboratory center.

Methods. The study was conducted in four stages: 1) a semi-structured pre-interview in three focus groups; 2) education in small groups; 3) a test period when Point-of-Care tests and instruments were independently used by nurses; 4) a semi-structured post-interview in three focus groups. All seven participants were interviewed twice. The deductive-inductive content analysis was used. Perceptions were considered and coded according to the testing process framework. The Point-of-Care testing process includes the pre-preanalytical, preanalytical, analytical, post-analytical and post-post-analytical phases. Each separate phase includes different actors that can hamper patient safety.

Results. The deductive-inductive content analysis revealed that patient identification during the preanalytical phase was error-prone even after data entry was automated. Perceptions towards quality assurance in the analytical phases were either positive or negative. An instrument’s ease of use, independently performing the analysis of quality control samples, success in sample collection and proper examination of a patient’s clinical status were shown to increase nurses’ trust in the Point-of-Care test results.

Principal conclusions. It seems indisputable that Point-of-Care tests will soon be an integral part of nurses’ work descriptions. Information about the frequency of various error types is insufficient if problems exist around Point-of-Care testing following adequate education and support from laboratory staff.