Variation in phlebotomy techniques in emergency medicine and the incidence of haemolysed samples

Background: Phlebotomy is a potential cause of preanalytical errors. We have observed phlebotomy in routine practice in a busy Emergency Department, to see how current practice compares with optimal blood sampling. Methods: Phlebotomy episodes were audited and compared with standard procedures. A computer-based search of the number of haemolysed samples from Emergency Medicine and hospital inpatients was reviewed. Results: Four different ways of taking blood were observed: cannulation and a syringe (38%), cannula with evacuated tube and adaptor (42%), syringe and needle into vein (14%) and evacuated tube system used conventionally (6%). Where a syringe was used, two methods of transfer into the sample tube were observed; needle kept on with cap piercing (77%) and needle and evacuated cap both removed (23%). On 20 out of 50 phlebotomy episodes (40%), the potassium-EDTA tube was filled prior to the biochemistry serum gel tube. A search of the laboratory computer records for ward-based phlebotomy found 30 of 1034 samples were haemolysed (2.9%). In the 50 phlebotomy episodes in the Majors area of the Emergency Department, 24% produced a haemolysed sample (P < 0.0001). For samples taken from all areas of Emergency Medicine over a seven-day period, 52 of 485 were haemolysed (10.7%; P < 0.0001). Conclusions: This study has shown that phlebotomy techniques in the Emergency Department deviate from standard practice significantly. This may well be a reason for the much higher frequency of haemolysed samples and with the wrong order of collection the possibility of potassium-EDTA-contaminated samples.