|Home | About | Journals | Submit | Contact Us | Français|
The BD Phoenix AP instrument reduced the manual setup time for the Phoenix system by 50%. For batches of 14 organisms, the average manual manipulation time per isolate was 89.5 s for BD Phoenix by the use of the AP instrument and 101 s for Vitek 2 (P < 0.001).
Labor shortages and financial constraints are forcing clinical microbiology laboratories to make more efficient use of available resources. Chief among these resources is human talent in the form of knowledgeable and skilled technologists (2). Automated systems for identification and susceptibility testing allow technologists' efforts to be redirected to tasks that better utilize their skills. Regulatory oversight by the U.S. Food and Drug Administration has resulted in similar levels of accuracy among systems when testing most common bacterial isolates (3, 4). However, it has been reported that the BD Phoenix system (BD Diagnostics, Sparks, MD) requires more hands-on technologist time to set up isolates than the Vitek 2 system (bioMérieux, Durham, NC)—179 versus 91 s per isolate (1).
The BD Phoenix AP instrument was recently introduced to automate a major portion of the front-end processing involved in the setup of isolates for the Phoenix system. The Phoenix AP instrument automatically adjusts the turbidity of the bacterial suspension to a 0.5 McFarland standard, prepares a dilution for susceptibility testing, and adds indicator to the susceptibility testing inoculum. The purpose of this study was to determine the reduction in manual manipulation time for panel preparation with the Phoenix System by using the AP instrument and to compare this new Phoenix workflow to the Vitek 2 system.
(This study was presented in part at the 109th American Society for Microbiology General Meeting [abstract C-212], Philadelphia, PA, 19 May 2009.)
Bacterial isolates were processed according to manufacturers' instructions for inoculation of Phoenix panels and Vitek 2 cards. Isolates were tested in 16 batches of 14 organisms and four batches of 35 organisms. A variety of organisms was chosen to represent those typically tested in clinical labs in order to allow for differences in adjustment of turbidity among species and strains. Six of the 14 organism batches were comprised of ATCC strains, as follows: Enterococcus faecalis 29212, E. faecalis 51299, Staphylococcus aureus 29213, S. aureus 25923, S. aureus 43300, Staphylococcus epidermidis 12228, Staphylococcus lugdunensis 49576, Enterobacter cloacae 700323, Escherichia coli 25922, E. coli 35218, Klebsiella pneumoniae 13883, K. pneumoniae 700603, Proteus mirabilis 7002, and Pseudomonas aeruginosa 27853. The 280 organisms tested for the remaining batches were clinical isolates, as follows: S. aureus (n = 70), coagulase-negative staphylococci (n = 30), enterococci (n = 40), P. aeruginosa (n = 20), E. coli (n = 37), Klebsiella spp. (n = 31), Enterobacter spp. (n = 23), Proteus spp. (n = 10), Serratia spp. (n = 10), and Citrobacter spp. (n = 9).
The two operators who performed the testing had not been involved in routine setup of isolates for either instrument but completed numerous practice runs on both instruments until proficiency was demonstrated. Each operator set up eight batches of 14 organisms, while the second operator recorded the time required to complete each individual step in the workflow. Timed steps were captured using a Microsoft Excel spreadsheet running a “stopwatch” macro. The mean time requirement for each individual step, overall time, and overall hands-on time were calculated. The measurements for the four batches with 35 organisms were for total time and total hands-on time only, not times for individual steps. For comparison purposes, two batches of 14 organisms were set up on the Phoenix instrument using the manual method without the AP instrument. Statistical significance of differences in time measurements was determined by the two-sample independent t test.
Table Table11 illustrates the individual steps and workflow for each instrument when setting up a batch of 14 isolates. The Vitek 2 Smart Carrier Station can hold seven isolates when testing for both identification and susceptibility testing; therefore, two fully loaded Smart Carriers were used for each batch. The time necessary to print barcode labels for Vitek 2 purity plates was not included in this study. The Phoenix AP instrument uses racks that can hold identification (ID) and antimicrobial susceptibility testing (AST) broths required to set up five isolates. For batches of 14, three racks were used, with the third rack holding broths for only four isolates. The AP hands-on workflow is divided into two sections: steps taken in preparing the rack to be placed on the AP, and steps taken after the rack is removed from the AP. For this study, each rack was prepared and loaded onto the AP in succession before removing the first rack for post-AP processing.
The times that were required to complete individual steps for each system are shown in Table Table1.1. Among the steps that can be compared between the two instruments, a clear difference was seen in the times required for preparing the bacterial suspension (mean of 42.9 s on Vitek 2 and 17.5 s on Phoenix AP). This difference can be explained by the need for the technologist to manually adjust the turbidity of the suspension for the Vitek 2 to the equivalent of a 0.5 McFarland standard.
Compared to the standard manual method, the Phoenix AP instrument reduced the hands-on time needed to inoculate panels for the BD Phoenix instrument from 177.7 s to 89.5 s (Table (Table2).2). However, this difference did not reach statistical significance because only two manual trials were performed.
Setting up Phoenix panels by using the AP instrument required less hands-on time than did that using the Vitek 2 (Table (Table2).2). For batches of 14 isolates, the mean manual manipulation time required per isolate was 101.0 s (range of 88 to 113 s) on the Vitek 2 and 89.5 s (range of 82 to 101 s) on the Phoenix AP (P < 0.001).
Both instruments required some “wait” time, in which no further manual steps could be taken until the instrument completed processing. On the Vitek 2, this involved occasionally having to wait until the instrument door was accessible before placing the second Smart Carrier Station on the instrument. When testing batches of 14 isolates on the Phoenix AP, there was always a wait time while the AP finished processing of the second and third racks before inoculation of the panels could continue. The amount of wait time on the Phoenix AP was largely dependent on the proximity of the initial bacterial suspension to a 0.5 McFarland standard (more time was required for greater dilutions). The mean amount of wait time when using the Phoenix AP was 6.6 min for a batch of 14 isolates, compared to 48 s with the Vitek 2 (Table (Table2).2). For this reason, total time to process a batch of 14 isolates was approximately 3 min longer on the Phoenix AP than on the Vitek 2 (P = 0.002). However, with either instrument, wait time could be utilized by performing other laboratory tasks.
Setting up panels in batches of 35 isolates allowed for more efficient use of hands-on time, as wait time could usually be utilized for further manual processing of isolates. Better utilization of wait time for Phoenix AP resulted in similar mean total times for the two instruments (58 min 3 s [Vitek 2] versus 57 min 36 s [Phoenix AP]; P = 0.648). The mean hands-on time per isolate on the Phoenix AP (87.3 s) was shorter than that on the Vitek 2 (93.6 s) but did not reach statistical significance with only four batches of 35 isolates tested (P = 0.083).
In conclusion, the BD Phoenix AP instrument reduced the hands-on processing time required to prepare panels for the Phoenix system by 50%. In comparison to the Vitek 2, the BD Phoenix AP instrument required 11.5 s less manual manipulation time per isolate when setting up batches of 14 organisms (P < 0.001). Additional studies comparing BD Phoenix AP workflow to other automated AST systems are needed.
This investigation was supported by a grant from BD Diagnostics, Sparks, MD.
S.S.R. has received research funding from Abbott Laboratories, BD Diagnostics, Cerexa, and Schering-Plough. All other authors have no conflicts of interest.
Published ahead of print on 10 March 2010.