WHY STOP? Quantifying Cognitive-Behavioural factors that influence the impact of PCR-POCT results on antibiotic cessation in ICU

Abstract

INTRODUCTION: Rapid Point of Care Tests for infection (POCT) do not consistently improve antibiotic stewardship (ASP) of suspected ICU infection. We measured 1) the effect of a negative PCR-POCT on antibiotic stop decisions, and 2) clinico-behavioural factors that prevent stopping.METHODS: Vignettes of antibiotic treated respiratory infection, with 4 distinct trajectories were presented to ICU clinicians: overall improvement, clinical improvement/biological worsening, clinical worsening/biological improvement, overall worsening. Initial and post PCR-POCT antibiotic decisions (stop or continue) /confidence levels were recorded. The PCR-POCT offer was voluntary but always presented and negative. Linear regression determined association of their final decision with influencing factors.RESULTS: Seventy clinicians responded. A negative PCR-POCT increased stop decisions in all scenarios (p<0.001) except improvement (already high); especially in discordant clin worse(49% pre-POCT vs 74% post-POCT). Inclination to stop was reduced by an ambiguous/worsening trajectory(p=0.015), initial confidence to continue(p<0.001), and involuntary receipt of POCT(p<0.001), not clinician experience or risk averseness. CONCLUSIONS: Negative PCR-POCT increases the inclination to stop antibiotics, particularly in ambiguous/worsening trajectories of ICU infection. Clinician intuition to continue and disinterest in POCT reduce its influence to stop. Highlighting and quantifying the predictive impact of behavioural-trajectorial factors can improve antibiotic stewardship and study design in ICU related infection.

Infection Point of Care Tests (POCT) in simulated vignettes can improve the education of Antibiotic Stewardship Programmes. An observational comparative study of Clinicians vs Medical Students

“…Participants were told that the POCT had high sensitivity and speci city, in line with commercial devices 20 . The result of the POCT was always negative (suggesting no active lung infection), with one exception: in the improvement vignette only, participants were subsequently told that the negative POCT was erroneous (a laboratory error) and retesting gave a positive result, as previously 25 .…”

“…We utilised the same four clinical vignettes as in Singh et al 25 (Appendix B), which depicted resolving/persisting or worsening VAP-infection after a course of antibiotics. The only amendment was the addition of normal ranges for biomarkers, deemed necessary during piloting (Appendix B).…”

“…The advertisements contained hyperlinks to the study website (hosted by XM Qualtrics, London, UK). We aimed to recruit the same number of participants (students) as our previous study (clinicians) 25 i.e. (N = 70).…”

“…learnt behaviour) 24 and we expect this to hold true for POCT-use and antibiotic stop decision making. We explored this as part of our previous study 25 but results were inconclusive. Speci cally: we found no evidence to suggest that years of experience exerts signi cant in uence upon POCT-use or antibiotic stop decision making.…”

“…Speci cally: we found no evidence to suggest that years of experience exerts signi cant in uence upon POCT-use or antibiotic stop decision making. However, the post hoc analysis was underpowered with unequal experience-related groups and inadequate distinction between degrees of experience 25 .…”