By Bob Kronemyer

Antibiotic stewardship in the ICU should be a critical part of a patient’s care, but it is not easy to practice in this setting because the patients are so vulnerable, the need is immediate, and usually all the information needed to choose an antibiotic is not available.

“As intensivists, we also have to deal with the consequences of multidrug-resistant pathogens,” said Richard Wunderink, MD, a professor of medicine in the Division of Pulmonary and Critical Care at Northwestern University Feinberg School of Medicine in Chicago. “However, we are partly to blame for these pathogens because of our use of antibiotics and the way that we use them, especially in situations where we do not practice good antibiotic stewardship.”


One of the major barriers to successful antibiotic stewardship is fear of missing the potential pathogen, which drives much of excessive antibiotic use, according to Dr. Wunderink. Also important is the lack of information, particularly early in the course of care. However, rapid diagnostic tests are helping to provide that information sooner.


“Cultures usually take about 72 hours, during which time you are working without a clear diagnosis, and therefore a tendency to continue to cover broadly until the results come back,” said Dr. Wunderink, who was the senior author of an article on antibiotic stewardship in the ICU (Chest 2019 Jan 25. [Epub ahead of print]).

“In the ICU, there is a feeling that there is a lack of leeway or time to let things unfold compared to a more stable patient,” said Sarah Doernberg, MD, an associate professor of medicine at the University of California, San Francisco School of Medicine, and the medical director of antimicrobial stewardship at UCSF. “Hence, there is more risk aversion in the ICU settings toward missing an organism or missing an infection. This results in broader antibiotic use, which typically is appropriate. But it can also increase adverse effects for the patient and increase antibiotic resistance in general.”

If the patient stabilizes with a current set of antibiotics, even if it is probably excessively broad and without a diagnosis, clinicians “are afraid to rock the boat because the patient improved on the current antibiotics,” Dr. Wunderink said. “There is a reluctance to change course.”


Time for Reassessment

Computerized decision support such as automated prompts to reassess antibiotics, patterns of resistance for particular pathogens and results from previous treatments hold promise for enhancing stewardship programs. A case in point are antibiotic time-outs, which can be triggered by the electronic health record, remind a clinician that it is time re-evaluate the need for a particular antibiotic. “After a designated number of hours of treatment with a particular antibiotic, it makes sense to re-evaluate, once the dust settles in a critically ill patient,” Dr. Wunderink said.


When a patient is being admitted to the ICU, “we are trying to cover all the bases at that point,” he said. “Once the patient has been stabilized, we reassess antibiotic usage.”


Other crucial times to reassess are when results of cultures and diagnostic tests return. “Return of diagnostic tests is also an opportunity to reassess antibiotic usage,” said Dr. Wunderink, also the medical director of the medical ICU at Northwestern Memorial Hospital in Chicago. “This can help us to either exactly define the infection and the cause or rule out certain organisms, thus not continue to cover empirically for those infections.”

Also, by sharing outcomes, “one learns there is a natural variation among physicians,” Dr. Wunderink said. For instance, discovering that other clinicians stopped antibiotics at a certain time or did not use broad-spectrum antibiotics for a particular scenario in which patients fared well is valuable in helping clinicians understand how they are managing a patient compared with their peers.

ICU-Specific Protocols

Bonnie Falcione, PharmD, BCPS-AQ ID, a clinical pharmacist in antibiotic management and critical care at UPMC Presbyterian Hospital in Pittsburgh, said clinicians at her institution do a good job practicing antibiotic stewardship in the ICU. “They are very much aware and actively work as teams,” she said.


Dr. Falcione attributes the high success rate partly to the fact that the hospital has had an antimicrobial stewardship program since 2002. “We have ICU-specific guidelines and protocols that address appropriate use of antimicrobials.”


Dr. Falcione is excited about rapid diagnostic and laboratory testing to reduce inappropriate antibiotics. “However, the coordination of these technologies to bring them to the bedside, particularly in the ICU, is going to make the difference going forward,” she said.

Rapid diagnostics continue to evolve, according to Dr. Falcione. Accelerating the ability to grow the bacteria, which historically has been the rate-limiting factor, “will enable us to truly move forward by minimizing the time to identification of the bacteria, along with the phenotype,” she said.


Dr. Falcione, also an associate professor of pharmacy at the University of Pittsburgh Schools of Pharmacy and Medicine, is encouraged that institutions and ICUs continue to use their ICU-specific data to guide empiric therapy selection. “We know that the time to appropriate antibiotic use in critically ill patients, particularly those with septic shock, makes a difference,” she said. “So, the more we can identify appropriate empiric regimens specific to patient populations and subpopulations, the better we will be at identifying those regimens and starting them on time.”

In addition to selecting an appropriate drug, other key factors are proper dosing, method of administration and frequency. “ICU patients are particularly vulnerable to augmented renal clearance,” Dr. Falcione said. “Based on their IV fluid requirements, ICU patients may also end up with larger volumes of distribution.” Therefore, the same dose that is given to one patient compared with another with a similar body habitus may not result in the same drug concentration where the infection occurs.

Prolonging the infusions of drugs for optimal exposure of the antibiotic is important, too.

De-escalation is particularly challenging because these patients are so ill. “I think having a plan identified up front helps with adhering to limiting duration to the most appropriate timeline, followed by ongoing monitoring to ensure that the plan continues to be suitable,” Dr. Falcione said. “The availability of new technology to inform de-escalation is also changing practice. The technology continues to evolve with improved speed and scope. This may allow for quicker de-escalation.”

The literature indicates that the involvement of infectious disease (ID) consultations helps to improve markers associated with good patient outcomes, according to Dr. Falcione. ID consults are particularly important with complex patients, such as those with a history of multidrug resistance. The ID consult can help determine the duration of therapy for complicated patients. These consults can also assist in establishing and coordinating management and monitoring for longer term plans once the patient leaves the ICU, for ideal continuity of care.

Although there are some biomarkers that provide some information on the likelihood of a bacterial infection, “they are not perfect,” Dr. Doernberg said. “Until we have better markers of infection, we probably will not move the needle on empiric therapy all that far.”

In contrast, being able to identify susceptibilities directly from patient samples (blood, urine or sputum) in rapid time has already started to take hold with some molecular techniques. “You may be able to draw the blood sample and get an answer without needing to grow anything,” Dr. Doernberg said. “These various molecular methods answer not only what kind of bacteria, but an answer on which antibiotics are susceptible as well, all within hours versus days.”

An ID consult can help clinicians optimize rapid diagnostic results. “Various studies of rapid diagnostics have shown that without having a pharmacist or other ID specialist, or both, helping clinicians integrate the information quickly, the fact that it is recorded in the medical record six hours later does not necessarily change antibiotic behavior in six hours,” she said.

Dr. Doernberg believes that de-escalation goes hand in hand with some of the rapid diagnostics and stewardship approaches in reviewing the patient’s course of antibiotics. “We know that patients are often treated longer than needed. In the ICU setting, de-escalation probably makes more sense as a focus rather than empiric therapy because you really do need to be broad up front with these patients.”

Like Dr. Falcione, Dr. Doernberg believes in the value of an ID consultation, by focusing on whether there is an infection, and if so, identifying the infection, and starting patients on the correct antibiotics.

Dr. Doernberg also acknowledged the conundrum of weaning patients off antibiotics in the ICU. For example, if a patient is not doing well and not responding to antibiotics, clinicians are hesitant to stop the antibiotics because of the fear that the patient’s condition will worsen. Conversely, if the patient is improving, clinicians are hesitant to stop them because they attribute the improvement to the antibiotics. “So, it is a no-win situation,” she said.

The likely solution is to enlist an ID consultant, according to Dr. Doernberg, who can assess the situation, determine the best course of therapy and focus antibiotics based on the current information in the ICU setting.

“We also need physicians to become more comfortable with the uncertainly in the ICU and figure out strategies for how stewardship can be executed in this really high-stake setting and leading to good patient outcomes,” Dr. Doernberg said.