Capillary blood draws in the NICU: the use of the Innovac quick-draw whole blood collection system versus traditional capillary blood draws.

PURPOSE: (1) To determine the rate of damaged and discarded capillary blood draws in the NICU; (2) to compare the rate of damaged and discarded samples between traditional capillary blood draws and the Innovac Quick-Draw device; (3) to determine whether in-service training for nurses on capillary blood draws decreased the rate of damaged and discarded blood samples. DESIGN: During Phase I of the study, the rate of capillary blood draws by the traditional method was determined. At the completion of Phase I, the manufacturer provided in-service training to senior nurses in the NICU with the use of the Innovac Quick-Draw device. Additional in-service training was also provided for the traditional capillary blood draw technique. Within a month of in-service training, an openly randomized study (Phase II) was carried out comparing traditional versus Innovac device capillary blood draws. SAMPLE: All infants admitted to the NICU between June 2008 and June 2009 were eligible to be in the study. There were no exclusion criteria based on weight, gestational age, or gender because the sampling method was the only variable being assessed. Phase I lasted two months, whereas Phase II lasted approximately four months. MAIN OUTCOME VARIABLE: Occurrence of damaged capillary samples with the Innovac device versus the traditional method. RESULTS: In Phase I, the rate of damaged and discarded samples was 10 percent (28/278). In Phase II, the rate of damaged and discarded samples for traditional and Innovac device was 7.2 percent and 10 percent, respectively. Comparisons between traditional and Innovac for different type of samples were as follows: complete blood count, 11.0 percent (12/104) vs. 13.4 percent (14/104); serum electrolytes, 6.4 percent (6/94) vs. 9.5 percent (9/95); C-reactive protein, 5.7 percent (4/70) vs. 8.0 percent (5/62); and liver panel, 5.3 percent (7/131) vs. 8.3 percent (9/108). There were no statistically significant differences of damaged and discarded samples for the overall or individual sample type comparisons.