CONTAMINATION OF INJECTION SITES BY LANDMARK PALPATION AFTER SKIN ANTISEPSIS

Abstract

When performing an intraarticular injection in the clinic setting the skin must first be cleaned with an antiseptic. This is typically done by spraying the skin with an alcohol based solution and allowing it to dry. Bony landmarks are then palpated to identify the correct insertion point for the needle. In the busy clinic setting this is sometimes done wearing sterile gloves, non-sterile gloves or no gloves at all. Potentially organisms from the palpating hand could contaminate the injection field and be introduced into the joint leading to a septic arthritis. We therefore looked at different scenarios often seen during intraarticular injections in clinic to see which was the least likely to contaminate the injection site.

In order to investigate the safest method of palpating bony landmarks whilst preventing infection we sprayed the entire volar surfaces, of both forearms, of fifty volunteers with alcohol. After the alcohol had dried, the subjects then palpated their own forearms in three separate areas with a naked digit, an unsterile-gloved digit and a digit itself sprayed with alcohol. Microbiology samples were taken using contact agar plates in each of the three areas as well as a control area, which had been sprayed but not touched. The number of bacterial colonies on the plates after incubation was then counted.

During transit to the incubator, three of the contact plate lids became dislodged. It could not be determined if further contamination had occurred and so all the samples from those volunteers were discarded. This left 188 contact plates (47 volunteers x 4 samples). The average number of colonies were, 14.5 for a naked digit, 3.5 for an unsterile glove, 2.0 for an alcohol sprayed digit and 1.7 for the control. Kolmogrov-Smirnov and Shapiro-Wilk tests were performed to assess the data for normality. The data was found to be highly skewed and therefore a Wilcoxon signed ranks test was performed comparing the three arms of the study with the control. There was a highly significant difference in the number of colonies between the naked digit and the control (p=0.0001) and to a lesser degree between the gloved digit and the control (p=0.030). No significant difference was found between the alcohol sprayed digit versus the control group (p=0.805).

In order to prevent contamination of an injection site after skin preparation the area should never be touched with a naked digit. We would also not recommend unsterile gloves often found in clinics. However, spraying your own fingers with antiseptic before palpating the injection site causes as much contamination as not touching it at all. This would seem to be a cheap and effective method as it avoids the cost of sterile gloves in clinic. We intend to extend the study further by adding an unsterile glove, which has been sprayed with alcohol. This may be the best solution of all.