Does Peptide Nucleic Acid In Situ Hybridization (PNA FISH) technology improve antimicrobial use?

Background: The early and appropriate use of antimicrobial agents for patients with suspected infection is critically important to optimize clinical outcomes. Some labs have used PNA FISH technology to provide additional information beyond the blood culture (BC) gram stain (GS) to help guide early therapeutic decisions. While some studies have shown that this test can improve appropriate drug selection, it is not clear if a similar benefit can be found in other settings. The Kaiser Regional Reference Lab (RRL) began using the PNA FISH in 2005.

Objective: We performed a retrospective evaluation to examine how the PNA FISH result affected physician prescribing behavior.

Methods: The RRL processes BCs from 11 medical centers scattered throughout Southern California. All + BCs between 25 Jan – 14 Feb 2010 with gram positive in clusters, gram positive cocci in chains and yeast were evaluated by one author. During the period between 23 Feb – 1 Mar 2010, only BC with gram positive cocci in clusters were evaluated independently by two authors; in addition, the length of time from the +BC to the time of the PNA FISH result was recorded. Charts were reviewed to determine the presumptive diagnosis, antibiotic (abx) therapy when the positive BC GS was called, and how the clinician subsequently used the PNA FISH result. Responses were categorized into one of five scores:  0 =  therapy could have been altered based on either the GS or PNA FISH but was not (inappropriate therapy), 1 = therapy was not altered by PNA FISH because clinical picture was clear,  2 = pt was discharged or transferred when the BC turned +,  3 = cases where abx decisions should have been altered by PNA FISH but wasn’t,  4 = PNA FISH result was used to help guide treatment.  Scores of 0 - 3 were considered situations where the PNA FISH was either not used or where results had no effect on antimicrobial decisions.

Results: 262 charts were reviewed - 180 in the first period and 82 in the second. More than 80% of charts received scores of 0 - 3 indicating PNA FISH was not useful in abx decision making (0 = 6.5%, 1 = 55.3%, 2 = 8%, 3 = 16.4%). The PNA FISH result was used in the management of the remaining 13.7% of cases. During the second period, complete scoring agreement between reviewers occurred in 98% of cases. The average time to release PNA results from the time of positive blood cultures was 7.3 hours (range 3-13 hours).

Conclusions: This study demonstrates that in the vast majority of cases, providing the PNA FISH result did not appear to assist antimicrobial decision making. In most instances, the clinical presentation dictated the abx choice and was not altered by the PNA FISH (score 1). Education and more timely PNA FISH results could potentially improve the clinical utility of the test (scores 0, 3). The introduction and use of this technology alone may not improve antimicrobial use.