Purpose: This study investigates the validity, reliability, and detection limit of a visual examination method utilizing the Tyndall effect to enhance visible detection of particles. The suitability of the method for compatibility testing of intravenous fluids in a hospital pharmacy context is discussed.
Methods: A panel of 20 inspectors examined 20 samples, with and without particles, using two light sources (halogen lightbulb in a focused desk lamp and a red pocket laser pointer). The samples contained particles of different origin (precipitate, polystyrene standards), varying size, and concentrations. Light obscuration and turbidimetric measurements were used to obtain numeric references. The samples were divided into rejection probability zones, and the validity (sensitivity, specificity, and likelihood ratios) and reliability (inter-rater agreement coefficients Fleiss' kappa and Gwet's AC1) were estimated.
Results: The sensitivity of the laser pointer for detecting microprecipitates was quite high; however, it also showed a high false rejection rate. The specificity was slightly higher for the focused desk lamp than the laser pointer. The likelihood ratios were not within the recommended limits of a useful test, indicating that the method could not securely confirm the presence/absence of particles in the samples. The inter-rater agreement coefficients indicated fair to moderate agreement between the inspectors.
Conclusions: The validity and reliability were not satisfactory for either of the light sources. The visual detection limit seemed to be around 5 μm, although we propose that an exact detection limit is not that relevant for compatibility testing. Based on the current findings, the visual examination method cannot be recommended as the sole method for judging compatibility of parenteral nutrition and drugs, but rather in a program of several methods. In the hospital pharmacy, the method may be a resource, together with theoretical considerations, in situations where other methods are unavailable; however, use of in-line-filters is essential to protect the patient.
Lay abstract: Many patients under intensive care are in need of several intravenous drugs simultaneously. These drugs cannot be given in the same infusion line unless compatibility has been documented. Incompatibilities can result in, for example, precipitation of particles. Injected particles can harm the patient and should be avoided. Visual screening of blends of drugs for possible incompatibility, using a focused light source to enhance visual detection based on the Tyndall effect, could be a quick and easy methodology to identify incompatibility. In the following study the objective was to investigate how reliable visual inspection, with the utilization of the Tyndall effect, is at detecting particles and precipitations in blends of intravenous drugs and parenteral nutrition mixtures. Twenty inspectors each examined 20 different samples with two different light sources. Some of the samples were without particles (clean), and some contained different types and degrees of particle contamination. The inspectors' judgment of the samples was recorded and validity and reliability parameters were calculated to evaluate the method's suitability. The conclusion was that because of false positive and negative findings the visual inspection method alone is not enough to securely document compatibility/incompatibility, but it is more suitable as support together with additional methods.
Keywords: Incompatibility; Inter-rater Reliability; Parenterals; PocketLaser pointer; Sub-visual particles; TPN; Total parenteral nutrition.
© PDA, Inc. 2015.