New sepsis treatment guidelines aim to reduce mortality from the infection.

An international group of experts in infectious diseases, critical care, emergency medicine, microbiology, pharmacology, and infection prevention presented a new interdisciplinary approach to improving sepsis treatment in hospitals. The position paper was published on August 25, 2026, in the journal Clinical Infectious Diseases and developed under the auspices of the Infectious Diseases Society of America with the participation of representatives from several major medical societies.

The authors emphasize that the problem of sepsis is not limited to how quickly a doctor makes a diagnosis and prescribes the first antibiotic. The work of the microbiology laboratory, the speed of drug delivery to the patient, the choice of antibiotic and infusion regimen, subsequent de-escalation of therapy, timely elimination of the source of infection, prevention of hospital-acquired infections, and the hospital's organizational structure all have a significant impact.

The new document therefore differs from standard recommendations for treating individual patients. Its primary objective is to identify specific changes hospitals can make to their processes to improve care for patients with sepsis. The expert group formulated recommendations in six areas: diagnostics, antibiotic therapy, epidemiological monitoring and quality indicators, complementary therapies, organizational infrastructure, and infection prevention.

It's particularly noteworthy that, alongside high-tech solutions like molecular diagnostics and electronic monitoring, the experts included extremely simple measures in the document. For example, a structured daily dental cleaning program for hospitalized patients is considered a potential way to reduce the incidence of hospital-acquired pneumonia, and therefore, the proportion of sepsis cases that develop in the hospital.

What is the document?	Characteristic
Publication type	Interdisciplinary multisociety position paper
Magazine	Clinical Infectious Diseases
Date of publication	August 25, 2026
Expert group	24 voting participants
Main directions	6
High Priority Tier 1 Recommendations	5
Moderate Priority Tier 2 Recommendations	15
Additionally	5 Promising Strategies Not Yet Recommended for Routine Implementation
The main goal	Improving hospital processes in sepsis

[1]

Why the old approach to sepsis was not enough

Sepsis remains a leading cause of death, disability, and healthcare costs. In recent years, significant efforts have been focused on early recognition and implementation of standard treatment strategies. In the United States, the SEP-1 quality indicator, which assesses the implementation of specific procedures in the first hours of sepsis treatment, has been a central element of this policy since 2015.

However, the authors of the new document note that the relationship between SEP-1 implementation and mortality reduction has been mixed. Large multicenter studies have not demonstrated a definitive reduction in mortality after the introduction of this system. This does not mean that early treatment is unimportant. Rather, the results indicate that a single set of actions taken early in the illness is insufficient to address the entire sepsis problem.

This is why experts suggest considering the entire patient journey: from obtaining blood samples and identifying the pathogen to selecting an antibiotic, monitoring its concentration, reassessing treatment, de-escalating therapy, and eliminating the source of infection. Furthermore, some sepsis cases arise during hospitalization due to hospital-acquired infections, meaning that preventing such infections should also be part of the sepsis management program.

A group of 24 specialists was formed to develop the recommendations. It included experts in adult and pediatric infectious diseases, intensive care, emergency medicine, hospital medicine, antimicrobial stewardship, microbiology, pharmacology, infection prevention, healthcare organization and quality of care improvement, as well as representatives of the patient community.

The recommendations are categorized by priority. Tier 1 indicates an intervention with the most compelling combination of evidence, likely clinical benefit, and feasibility for practical implementation. Tier 2 includes promising interventions for which additional questions remain regarding effect size, cost, feasibility, or potential adverse effects. A separate category is designated for technologies that appear promising but do not yet have a sufficient evidence base for widespread use.

Category	What does it mean?	Practical interpretation
Tier 1	High priority	The authors propose widespread implementation
Tier 2	Moderate priority	Sequential or selective implementation is possible
Emerging and promising	Promising technologies	There is not enough data yet for routine use.
Basis of assessment	Evidence + Potential Benefit + Feasibility	This is not only an assessment of the effectiveness of treatment

[2]

Rapid molecular blood diagnostics has become one of the top priority recommendations

One of the challenges of sepsis treatment is that physicians are forced to initiate antibiotic therapy even before the pathogen is accurately identified. Routine blood culture remains the most important diagnostic method, but even after the culture is positive, traditional identification of the microorganism and susceptibility testing often take another 48-72 hours. During this time, the patient may either receive an ineffective drug or, conversely, an unnecessarily broad combination of antibiotics.

Experts have assigned Tier 1 status to the use of multiplex nucleic acid amplification tests for already positive blood cultures. These panels can detect approximately 90% of common bloodstream infection pathogens within a few hours and simultaneously identify some key antibiotic resistance genes.

However, one of the most important findings of the paper is that rapid testing alone is insufficient. In a meta-analysis of 88 studies covering 25,682 clinical cases, rapid diagnostic tests, coupled with active participation in antibiotic stewardship programs, were associated with lower mortality compared to traditional diagnostics: the odds ratio was 0.72. Furthermore, patients received optimal treatment more quickly and had shorter hospital stays.

Moreover, when rapid diagnostics were used without the active involvement of antibiotic specialists, no survival benefit was observed. This is an important organizational finding: getting test results quickly is not enough—someone must see them in a timely manner, interpret them correctly, and modify treatment. This is why the new recommendation combines technology and antimicrobial stewardship into a single process.

Experts also caution against the opposite extreme—the uncontrolled administration of blood cultures to every patient. Clinical decision support systems can help determine when initial or repeat cultures are truly necessary. Such diagnostic stewardship can reduce unnecessary testing, false positives, and subsequent unnecessary antibiotic use.

Diagnostic problem	What do the experts suggest?
Blood cultures take time.	Rapid molecular testing of positive culture
We need to identify the pathogen faster.	Multiplex nucleic acid amplification
It is necessary to identify individual resistance mechanisms	Identification of key resistance genes
A quick result may remain without action	Mandatory participation in antimicrobial stewardship
Too many unfounded crops	Electronic decision support
Prioritize rapid testing of positive cultures	Tier 1

[3]

In septic shock, antibiotics are recommended to be administered within an hour - but this rule does not automatically apply to all cases of sepsis.

One of the most important recommendations of Tier 1 is to reduce the time between the recognition of septic shock and the administration of the first antibiotic. Experts suggest that hospitals organize processes so that antibiotics begin to be administered less than one hour after the recognition of shock.

The rationale for such urgency is quite compelling, particularly for the most severely ill patients. In large observational studies, each additional hour of delay in antibiotic therapy for septic shock was associated with a 7-14% increase in the likelihood of death. This does not necessarily mean that the delay itself completely explains the increased mortality, but the consistency of results across studies makes prompt administration of an appropriate antibiotic for shock a high priority.

The authors also made an important distinction between sepsis with shock and sepsis without shock. For patients without shock, the evidence for a strict time threshold is significantly weaker: several studies have found no increase in mortality even when antibiotics were administered within six hours of recognition. Therefore, experts do not propose automatically extending the one-hour rule to all patients with suspected sepsis.

This distinction has practical implications. Applying the rule too aggressively to all patients can increase unnecessary antibiotic use in people who subsequently have no confirmed infection. The new concept attempts to simultaneously address two opposing objectives: treating patients with septic shock as quickly as possible and not turning any suspected infection into a reason to immediately prescribe the broadest possible antibiotic regimen.

Experts also suggest separately monitoring the time from drug administration to the actual start of intravenous infusion. The practical goal is less than 30 minutes. The authors emphasize that this is an organizational benchmark, not a proven biological threshold: it was chosen primarily to ensure that hospitals can realistically meet the more general goal of "antibiotic administration within an hour of septic shock recognition."

Situation	Proposed landmark	Priority
Septic shock: from recognition to the first antibiotic	<1 hour	Tier 1
From antibiotic prescription to infusion initiation in septic shock	<30 minutes	Tier 1
Sepsis without shock	A hard universal hourly threshold is not proposed.	Individual assessment
Why is it important to distinguish between these groups?	The benefits of urgency are most compelling in shock.	Reducing the risk of unnecessary therapy

[4]

Not only the choice of antibiotic is important, but also the order and method of its administration

Another high-priority recommendation concerns β-lactam antibiotics against Pseudomonas aeruginosa. For critically ill patients with sepsis, experts recommend using a prolonged infusion after the initial loading dose, if this regimen is appropriate for the specific antibiotic and patient. This is due to the pharmacology of β-lactams: their effectiveness depends largely on the time the drug concentration remains above the minimum inhibitory concentration of the microorganism.

One of the key pieces of evidence was BLING III, a large study of more than 7,200 intensive care unit patients receiving piperacillin/tazobactam or meropenem. In the main analysis, continuous infusion did not provide a statistically significant reduction in 90-day mortality: 24.9% versus 26.8% with standard intermittent administration. However, clinical recovery was more common, and an adjusted mortality analysis showed a benefit for continuous infusion.

Even more compelling was a pooled analysis of 17 randomized trials, including BLING III. The authors of this meta-analysis calculated a 99.1 percent probability that extended β-lactam infusion reduces 90-day mortality; the relative risk was 0.86. The estimated number of patients needed to be treated this way to prevent one death was approximately 26. It was this combined data that formed the basis for assigning a Tier 1 recommendation.

However, this does not apply to all antibiotics or all patients. The primary data is from critically ill adults. Drugs with a long half-life, such as ceftriaxone, usually provide the necessary exposure without a multi-hour infusion. A standard loading dose should be administered before an extended or continuous infusion to ensure rapid achievement of therapeutic concentrations.

Another interesting recommendation: if a patient is prescribed both a β-lactam and vancomycin as empirical treatment for sepsis, hospital procedures should prioritize the β-lactam by default. Experts acknowledge that the evidence base here is largely observational, but they consider the change in sequence relatively simple, biologically sound, and potentially beneficial. Exceptions are possible when the clinical situation requires prioritizing MRSA coverage.

Antibiotic therapy strategy	Recommendation
First antibiotic for septic shock	Start as soon as possible, aim for <1 hour
After appointment for shock	Start infusion approximately <30 minutes
Antipseudomonal β-lactam in a critically ill patient	Loading dose → extended/continuous infusion
β-lactam + vancomycin	By default, β-lactam is administered first.
Second dose of antibiotics	Avoid delays in patient transfers and handovers
Reported allergy to penicillin	Use allergy assessment algorithms rather than automatically avoiding all β-lactams

[5]

Experts suggest addressing both undertreatment and overtreatment simultaneously

One of the main challenges of sepsis is the need to select an antibiotic before the pathogen is definitively identified. If the drug is ineffective against the microorganism, the patient is left without adequate treatment in the critical first hours. According to expert data, in approximately one in five patients with bloodstream infection, initial empirical therapy is insufficient against the subsequently identified pathogen.

But the opposite problem is no less widespread. About two-thirds of patients with suspected community-acquired sepsis receive drugs against MRSA or Pseudomonas aeruginosa, even though these pathogens are detected in less than 10% of patients. Overly broad therapy is associated with the risk of kidney damage, Clostridioides difficile infection, the selection of antibiotic-resistant bacteria, and other adverse effects.

Therefore, the new document suggests that hospitals track both indicators simultaneously: how many patients are receiving insufficient treatment and how many are receiving inappropriately broad treatment. This changes the philosophy of quality assessment. A good sepsis treatment program shouldn't simply demonstrate that almost all patients received powerful antibiotics promptly; it should demonstrate that the chosen regimen truly matched the individual patient's risk.

The next step is de-escalation. If, after several days, microbiological and clinical data do not confirm the presence of MRSA or resistant gram-negative bacteria, therapy should be reconsidered and, when safe, tapered. Experts suggest turning the frequency of such de-escalation into a measurable indicator of hospital performance.

The authors also note the common labeling of "penicillin allergy." Approximately 10% of American patients report an allergy to penicillins or cephalosporins, but more than 90% of them do not actually have a true allergy that would prevent the use of β-lactams. Uncritical acceptance of such labeling can lead to the prescription of less effective, more toxic, or unnecessarily broad alternatives.

Problem	The scale specified in the document	Proposed solution
Inadequate empirical therapy for bloodstream infection	≈1 in 5 patients	Improving the choice of initial antibiotic
Very broad anti-MRSA/anti-pseudomonas therapy	≈2/3 of patients with suspected community-acquired sepsis	Assess the real risk of resistant pathogens
Really detectable resistant pathogens	<10%	Avoid unnecessary spectrum expansion
Lack of timely de-escalation	Varies significantly between hospitals	Control and feedback
Reported allergy to β-lactams	Often not true	Allergy assessment protocols

[6]

Antibiotics alone are not enough: the source of infection must be eliminated as soon as possible.

Antibiotics aren't always able to cure sepsis if a source of infection persists in the body, constantly releasing microorganisms and inflammatory signals. Examples include an abscess, an infected catheter, necrotic tissue, or an infected lesion in the abdominal cavity. In such situations, so-called source control—drainage, surgery, removal of the infected device, or another procedure—is necessary.

According to estimates in the document, approximately one-third of hospitalized patients with sepsis require some form of source control procedure. However, hospitals measure the time to such intervention far less frequently than the time to antibiotic administration. The expert group proposes changing this situation.

In one of the largest studies, which included 4,962 adult patients with community-acquired sepsis, eliminating the source within six hours of sepsis onset was associated with lower adjusted 90-day mortality. This association was most pronounced for gastrointestinal, intra-abdominal, and soft tissue infections. However, the authors of the new paper emphasize that the data remain observational and do not allow for a universal six-hour threshold to be established for all situations.

The optimal timing of intervention depends on the location of the infection, the severity of the condition, the pathogen, and the availability of surgery or interventional radiology. Therefore, the recommendation is not formulated as a requirement to "operate on everyone within six hours," but rather to measure delays and identify their causes.

It's especially important to consider hospitals where necessary procedures cannot be performed on-site. In such facilities, the quality of care should also be assessed by the speed of patient transfer to a center capable of performing surgery or intervention.

Possible source of infection	Source control example
Abscess	Drainage
Infected or necrotic tissue	Surgical treatment
Infected vascular catheter	Removing a device
Intra-abdominal infection	Surgical or interventional intervention
There is no necessary specialist in the hospital	Rapid patient transfer

[7]

Corticosteroids are not recommended for all sepsis, but for specific groups of patients.

A separate section is devoted to corticosteroids. Experts do not recommend their use in every patient with sepsis. Two Tier 2 recommendations apply to more clearly defined situations: severe community-acquired pneumonia and refractory septic shock.

In severe community-acquired pneumonia, the evidence has long remained conflicting. However, the CAPE-COD study showed that intravenous hydrocortisone at a dose of 200 mg per day, initiated within the first 24 hours of admission to the intensive care unit, reduced the likelihood of progression to mechanical ventilation and 28-day mortality. Several subsequent meta-analyses also supported a possible reduction in mortality.

However, this therapy is not suitable for all patients with pneumonia. The document primarily addresses severe illness with acute hypoxemic respiratory failure requiring intensive care. Suspected influenza pneumonia, aspiration chemical pneumonitis, or invasive fungal or mycobacterial infection may alter the benefit-risk balance. Therefore, the recommendation remains at Tier 2.

The second situation is persistent septic shock, in which blood pressure cannot be maintained without significant doses of vasopressors. The authors suggest that hospitals develop protocols for the timely administration of corticosteroids to such patients. As a practical example of refractory shock, they cite the need for norepinephrine or epinephrine at a dose of at least 0.25 mcg/kg/min for at least four hours; the standard regimen is hydrocortisone 200 mg daily.

But even here, experts remain cautious: different clinical trials have yielded inconsistent results regarding the magnitude of mortality reduction. Therefore, the main novelty of the document lies not in declaring steroids a universal treatment for sepsis, but in its proposal to ensure their timely use where there is already a reasonable likelihood of benefit.

Clinical situation	Corticosteroids
Any sepsis	Not routinely offered
Severe community-acquired pneumonia	A protocol for carefully selected patients is recommended.
Refractory septic shock	Protocolized use is recommended.
An example of a regimen for persistent septic shock	Hydrocortisone 200 mg/day
Priority level	Tier 2

[8]

Chlorhexidine and a regular toothbrush can help prevent sepsis.

An unusual feature of the new document is its strong focus on sepsis prevention after hospitalization. The authors note that hospital-acquired sepsis often occurs as a result of a nosocomial infection, and its mortality rate is approximately twice that of sepsis that begins outside the hospital. Furthermore, pre-existing sepsis can impair immune defenses and increase vulnerability to secondary infections.

In a large prospective cohort of intensive care unit patients, more than one in eight patients admitted with sepsis subsequently developed a hospital-acquired infection. The most common complications were central line-associated bloodstream infections, pneumonia, and intra-abdominal infections. These complications were strongly associated with increased mortality.

One proposed strategy was daily chlorhexidine skin preparation combined with universal or targeted nasal decolonization for MRSA in intensive care unit (ICU) patients and in some non-ICU patients with invasive devices. In the REDUCE-MRSA study, this strategy reduced the incidence of bloodstream infections caused by any pathogen by approximately 43% compared to standard care.

Even more surprising is the recommendation to organize daily toothbrushing for all hospitalized patients. It was based on a meta-analysis of 15 randomized trials: regular toothbrushing reduced the incidence of hospital-acquired pneumonia, especially in patients on mechanical ventilation, and was associated with a shorter duration of ventilation, shorter intensive care unit stays, and lower mortality in the intensive care unit.

Experts assigned this measure Tier 2 status because the most compelling evidence so far has been obtained in critically ill patients and those on mechanical ventilation, and there is no direct evidence yet of a reduction in sepsis-specific mortality. Nevertheless, this is a good example of how sepsis prevention can begin not with expensive equipment, but with the systematic implementation of a very simple care procedure. [9]

Preventive strategy	Evidence/Purpose
Prevention of catheter, urinary tract and ventilator-associated pneumonia infections	The main basis for the prevention of nosocomial sepsis
Chlorhexidine washes	Reducing bloodstream infections in high-risk groups
Nasal decolonization of MRSA	Adjunct to chlorhexidine in appropriate patients
Daily brushing of teeth	Reduction in hospital-acquired pneumonia
Priority of two complementary strategies	Tier 2

[10]

Artificial intelligence for sepsis prediction hasn't yet received the green light.

Despite the rapid development of artificial intelligence, experts deliberately did not include automated sepsis prediction systems in Tier 1 or Tier 2. Such algorithms continuously analyze vital signs, laboratory data, and other electronic medical record information in an attempt to detect incipient sepsis before a physician can.

Some observational and quasi-experimental studies do show that such systems can improve intermediate outcomes—for example, by speeding up antibiotic administration or increasing compliance with treatment protocols. However, compelling randomized data demonstrating improved patient clinical outcomes is still lacking.

Additional problems include false alarms, medical staff fatigue from the large number of alerts, the lack of transparency in the operation of some models, and a decrease in accuracy after transferring an algorithm from the hospital where it was developed to another institution. This is especially dangerous if the system begins to over-report sepsis, thereby encouraging the unnecessary use of broad-spectrum antibiotics.

Similar caution applies to new immune response assays, molecular detection of the pathogen directly from blood without prior culture, and broad molecular panels for pneumonia diagnostics. These can speed up the acquisition of information, but faster diagnosis does not necessarily equate to improved survival.

Therefore, experts suggest that these technologies be used primarily in pilot programs and studies with clear action algorithms and outcome monitoring. This approach demonstrates an important principle of the new document: technological innovation alone is not sufficient grounds for widespread adoption; clinical benefit must be demonstrated.

Promising technology	Status in the document
Automated sepsis prediction and AI	Not yet for widespread routine implementation
Immune response tests	Further research into clinical benefit is needed.
Search for the pathogen directly in the blood without culture	Speeds up diagnosis, but benefit to outcomes has not been proven
Multiplex panels for the lower respiratory tract	Possible in certain high-risk groups
Procalcitonin to determine the duration of antibiotics	Promising, but evidence remains mixed

[11]

What does the new document change?

The experts' key conclusion is that modern sepsis treatment must go far beyond the well-known principle of "give the antibiotic quickly." Urgency remains critical in septic shock, but treatment success also depends on the correct drug selection, how quickly it physically reaches the patient, the route of administration, and whether treatment is reconsidered following new findings.

The balance between under- and over-antibiotic therapy is becoming equally important. The new approach effectively views both scenarios as systemic failures: not only is it dangerous to prescribe an active antibiotic too late, but also to unnecessarily continue the broadest treatment after the data already allow it to be narrowed.

Another significant shift is shifting some of the focus from the actions of individual physicians to the hospital's overall organization. The electronic record-keeping system, microbiology team, pharmacists, infectious disease specialists, nurses, intensive care specialists, and infection prevention teams must work as a unified system. The authors explicitly recommend creating an interdisciplinary sepsis program management structure that includes specialists in infectious diseases, antimicrobial stewardship, and clinical microbiology.

However, the document does not provide evidence that implementing all recommendations simultaneously will automatically reduce mortality in every hospital. Some proposals are based on randomized trials and meta-analyses, while others rely primarily on observational data, clinical plausibility, and expert consensus. Therefore, the authors categorized the recommendations by priority level and specifically identified technologies for which there is insufficient evidence.

As a result, the position paper proposes viewing sepsis not as a single emergency event in the first hours of illness, but as a continuous process of infection management: quickly identify a patient at risk, promptly administer the correct antibiotic, identify the pathogen, optimize dosing, eliminate the source of infection, promptly narrow treatment, and, if possible, prevent further infection. This systematic approach, rather than a single new technology, is the key message of the paper.

Key High Priority Recommendations

Tier 1	The essence
Rapid molecular diagnostics of positive blood cultures	Use in conjunction with antimicrobial stewardship
Antibiotic for septic shock	The goal is less than 1 hour after recognition
From prescription to start of infusion	Practical guideline: less than 30 minutes
Antipseudomonal β-lactams	Extended infusion after loading dose in critically ill patients
β-lactam + vancomycin	By default, administer the β-lactam first unless there is a clinical reason to change the order

[12]

News source

Rhee C, Masur H, Klompas M, Bell T, Busch LM, Chiotos K, et al. IDSA/ACEP/ASM/PIDS/SCCM/SHEA/SHM/SIDP Multisociety Position Paper: Hospital Strategies to Improve Sepsis Outcomes. Clinical Infectious Diseases. Published online August 25, 2026. DOI: 10.1093/cid/ciag438

This is a position paper on the organization of hospital care, not a randomized trial of a new treatment method, and not a replacement for individual clinical guidelines for the management of a specific patient.