Recent discussions on the economics of innovation in rapid diagnostic testing for bacterial infections in the hospital setting have inspired me to once again take pen (keyboard) in hand. In most of these discussions, the key variable in the utility of any diagnostic test has been ignored. I’m talking about basic aspects of physician thinking and behavior,
The basic premise is that if accurate, bedside tests were available either in the hospital or in the physician’s office, we could avoid broad spectrum empiric antibacterial therapy and focus on the specific pathogen causing the infection. There are, though, a number of flaws that undermine this premise. The first is that antibacterials that are more specific for either pathogens or resistance mechanisms exist and can replace so-called broad-spectrum agents. In general, this is simply not true. Even amoxicillin, to which there is a high level of resistance among Gram-negatives, has a relatively broad spectrum. The other assumption is that reducing the use of “broad-spectrum” antibacterials and increasing the use of pathogen-specific antibacterials will lead to a slower rate of rise of resistance. While this may be a reasonable hypothesis, the quantitative effect remains to be determined and there will be other variables at play (eg – how long did it take for vancomycin resistance to emerge in the clinic?).
The other part of the discussion is economic. It seems that the technology to provide rapid tests capable of identifying pathogens in primary specimens such as blood, sputum and urine exist today. But the investment to bring them to market is much greater than the projected market since it is likely to be cheaper to use “broad-spectrum” empiric therapy than it would be to use the diagnostic tests to narrow therapy assuming such pathogen-specific therapy existed. The implication is that broad-spectrum empiric therapy is “bad.” We need to understand that a major reason for this lack of a market is based on well-justified physician hesitancy.
An additional issue relates to hospital and physician practice logistics. As I have noted in the past, if the diagnostic is not easily used (like a covid antigen test or a strep test for e.g.) and is designed to be carried out in a laboratory, there is the delay between obtaining the specimen, transportation to the lab, carrying out the test and reporting the result. In most cases, this delay would eliminate the potential advantage of the rapidity of the test.
But let’s talk about physician behavior. Speaking as a physician who has spent time in the ICU caring for critically ill patients, I think I am qualified to address this issue. I admit that this experience is now quite dated. But in recent years I spent time working in a local hospital on antimicrobial stewardship and based on that experience, I don’t think things have changed that much since my own ICU attending days.
Physicians see patients as individuals and the take their responsibility for each one extremely seriously. When a patient becomes acutely and severely ill in the hospital and where infection is a likely cause, I am not sure that even a reliable bedside test would be enough to prevent a physician from using additional or broad-spectrum antimicrobial therapy to be sure that all pathogens, likely and not so likely, will be attacked.
Our best approach, in my view, will be to discontinue antibiotic therapy as soon as we can based on good clinical evidence. Unfortunately, that evidence is still lacking for many infections. It goes without saying that in the absence of further evidence of bacterial infection in the cold light of day, antibiotics should be discontinued in any case.
This approach is validated by multiple clinical guidelines discussing appropriate empiric antibiotic therapy for various infections. I don’t see physicians changing their approach to empiric antimicrobial therapy of severely ill hospitalized patients regardless of the accuracy or rapidity of new diagnostic tests – and I’m not sure they should change.