Intra-abdominal sepsis: new definitions and current clinical standards

Abstract

Purpose

The abdomen is the second most common source of sepsis and is associated with unacceptably high morbidity and mortality. Recently, the essential definitions of sepsis and septic shock were updated (Third International Consensus Definitions for Sepsis and Septic Shock, Sepsis-3) and modified. The purpose of this review is to provide an overview of the changes introduced by Sepsis-3 and the current state of the art regarding the treatment of abdominal sepsis.

Results

While Sepsis-1/2 focused on detecting systemic inflammation as a response to infection, Sepsis-3 defines sepsis as a life-threatening organ dysfunction caused by a dysregulated host response to infection. The Surviving Sepsis Campaign (SSC) guideline, which was updated in 2016, recommends rapid diagnosis and initiating standardized therapy. New diagnostic tools, the establishment of antibiotic stewardship programs, and a host of new-generation antibiotics are new landmark changes in the sepsis literature of the last few years. Although the “old” surgical source control consisting of debridement, removal of infected devices, drainage of purulent cavities, and decompression of the abdominal cavity is the gold standard of surgical care, the timing of gastrointestinal reconstruction and closure of the abdominal cavity (“damage control surgery”) are discussed intensively in the literature. The SSC guidelines provide evidence-based sepsis therapy. Nevertheless, treating critically ill intensive care patients requires individualized, continuous daily re-evaluation and flexible therapeutic strategies, which can be best discussed in the interdisciplinary rounds of experienced surgeons and intensive care medicals.

Introduction

In 2017, the World Health Organization (WHO) and the World Health Assembly adopted a resolution that emphasizes the importance of sepsis diagnosis, treatment, and prevention worldwide. As a complex disorder of global priority, sepsis has now moved further into the spotlight of medicine and medical research [1].

Disproportionately, most literature on the incidence, prevalence, and evidence of sepsis comes from developed countries, where up to 2.8 million deaths were attributable to sepsis in 2010 [2, 3]. Due to a suspected high number of unreported cases, the estimated incidence of sepsis is even higher [4, 5]. Recently published epidemiologic data suggest that sepsis causes one third to half of in-hospital mortality in the USA. While the available literature is mainly from high-income countries, the incidence of sepsis and sepsis-associated death is assumed to be even higher worldwide. Data on chest infections reveal that 90% of the associated worldwide mortality is from the developing countries [3, 4]. Estimates suggest that about 1400 patients die from septic diseases worldwide per day [6].

In 66% of all surgical patients with sepsis, an intra-abdominal infectious focus could be detected. In 85% of such patients, this localization of a septic source is even higher after elective surgical intervention and the development of sepsis in the postoperative clinical course [7, 8].

The new sepsis definitions

The primary sepsis definitions dating back to 1991 [9] defined sepsis as a systemic inflammatory response syndrome (SIRS) to an infection. The consecutive development of organ failure was termed “severe sepsis,” and SIRS complicated by cardiocirculatory failure was called “septic shock.” Due to its poor specificity, this Sepsis-1 definition was challenged ever since, resulting in its revision in 2001 [10]. However, although the International Sepsis Definitions Conference acknowledged its weaknesses, it confirmed the principal construct of the first definitions due to a lack of suitable alternatives. Nevertheless, it was recognized that SIRS criteria (Table 1) are unsatisfactory for describing the manifold appearance of sepsis, which is why the definitions were complemented by a list of possible symptoms, including the recognition of the fact that organ dysfunction can be the first detectable symptom. These extensions further reduced the specificity of the sepsis definitions, resulting in significant variance concerning estimated incidence and mortality [11]. Consequently, the sepsis definitions were revised again in 2014 and 2015, resulting in the publication of the Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3) in February 2016 [2]. Sepsis-3 no longer focus on the signs and symptoms of inflammation, which can reflect a reasonable healthy response to a systemic infection. Instead, Sepsis-3 emphasize that, in sepsis, the host response is not healthy, but dysregulated, resulting in organ dysfunction of sufficient severity to be life threatening. Concrete sepsis is defined as “life-threatening organ dysfunction caused by a dysregulated host response to infection” [2]. Organ dysfunction is measured by the Sequential Organ Failure Assessment (SOFA) score and is deemed “life-threatening” if the score is increased by ≥2 points [12] (Table 2). Thus, “abdominal sepsis” is now defined as an increase of the SOFA score of ≥ 2 points due to intra-abdominal infection [13]. If the patient requires the application of vasopressors to maintain a mean arterial pressure (MAP) of ≥65 mmHg (despite adequate volume resuscitation) and the serum lactate is ≥ 2 mmol/l, the clinical situation is defined as septic shock [13]. The term “severe sepsis” has been abolished and should no longer be used.

Table 1 Diagnostic criteria of sepsis (from International Guidelines for the Management of Severe Sepsis and Septic Shock 2012)

Table 2 The SOFA score reflects organ dysfunction, which now defines the term “sepsis” according to the new Sepsis-3 definition [13] (in contrast to the complex SOFA score, the qSOFA score is an everyday tool for clinicians to categorize patients rapidly, e.g., in the emergency room)

The key consequences from the Sepsis-3 definition [13] are:

• 1. The formal diagnosis of sepsis relies on the detection of organ dysfunction based on the SOFA score.

• 2. The continuum of infection, sepsis, severe sepsis, and septic shock has been abolished in favor of the reduction to infection, sepsis, and septic shock only.

• 3. The concept of SIRS can still be used to describe a systemic response to a sterile hit (pancreatitis, trauma, etc.) or an infection, and its appearance should trigger a screening for infectious foci.

• 4. Sepsis is more than inflammation; it is a complex, life-threatening organ dysfunction resulting from dysregulated host response.

As the SOFA score is not always comprehensively available outside of intensive care units (e.g., in the emergency room or on surgical wards), the authors of Sepsis-3 suggested the quick (q) SOFA as a screening tool for sepsis. Resulting from a retrospective analysis of large databanks, the qSOFA consists of three easy-to-evaluate criteria:

• 1. Alteration in mental state (Glasgow Coma Scale < 15).

• 2. Respiratory rate ≥ 22 breaths/min.

• 3. Systolic blood pressure ≤ 100 mmHg.

The literature reveals that patients who meet these criteria have prolonged hospital stay and increased risk of death [4, 11]. If the patient fulfills two criteria, admission to intensive care is obligatory. Since its introduction in clinical medicine, several trials have underlined the specificity of the new stratification scores SOFA (intensive care) and qSOFA (emergency department, normal wards, outpatient department, emergency medical service) for predicting the mortality of the septic patient. Compared to qSOFA and the former Sepsis-2 criteria (“the SIRS criteria”), the SOFA score has the highest predictive values for intensive care unit (ICU) mortality, which Raith et al. recently analyzed in an impressive collective of 180,000 patients [14]. Despite the advantages of SOFA and qSOFA for risk stratification of patients with organ dysfunction (and sepsis!), the “old” SIRS criteria remain an important tool for surgeons and intensivists for everyday rounds. Compared to SOFA and qSOFA scores, the SIRS criteria have the highest sensitivity for early detection of inflammation and infection, which is of major surgical importance both in the emergency setting and in the postoperative phase [14].

The SIRS criteria (Sepsis-2) help detect inflammation/inflammatory/infectious complications; qSOFA (emergency department) and SOFA (ICU) scores identify the septic patient, defining the need for intensive care.

As a consequence of the Sepsis-3 definition, the number of patients with sepsis has decreased in retrospective analyses, as organ failure does not result from sepsis, but defines it.

Donnelly et al. applied the new Sepsis-3 criteria to a cohort of > 30,000 patients from the REGARDS (REasons for Geographic and Racial Differences in Stroke) Study [15]. While 1526 patients fulfilled the SIRS criteria (in-hospital mortality, 9%), only 1080 were septic according to the SOFA score (13%), and only 378 patients had a positive qSOFA score (23%!) [15]. Last year, Shankar-Hari analyzed the influence of the new sepsis criteria on the epidemiology of sepsis in detail. While Sepsis-3 identifies a similar population of severe septic patients according to Sepsis-2, the identification of the population of patients with septic shock is more specific when the new criteria are applied [16]. While the qSOFA score has the highest predictive value for the critically ill septic patient, we still have to treat patients with infection (SIRS criteria) and the risk of sepsis development according to the sepsis guidelines.

For surgeons, the SIRS criteria remain the most important tool for detecting inflammation, infection, and complications! The new definition of sepsis must NOT lead to any delay in diagnostic or therapeutic approaches!

Pathogenesis and risk factors

Figure 1 summarizes the general and independent risk factors for infections and sepsis. Additional to these “general” risk factors for sepsis, the surgical patient is permanently threatened by surgical complications caused by impaired healing of anastomoses or sutures for abdominal closure. Several trials have analyzed patient-related risk factors that lead to impaired healing, resulting in increased anastomotic leakage, surgical-site infections, and intra-abdominal sepsis. These factors, in part, overlap with the general risk factors, but are of major importance for abdominal surgery. Besides intraoperative complications and episodes of intraoperative hypotension, patient-related factors such as male gender, age, smoking, and diabetes mellitus correlate with increased anastomotic leakage rate. The same holds true for medication (corticosteroids, chemotherapeutics, immunosuppressants) and radiation (Fig. 1 and 2).

Fig. 1

Simplified summary of risk factors for sepsis development [4]. In contrast to other medical specialties, the amount of risk factors is increased in visceral surgery due to the fact that impaired postoperative healing with failure of complex surgical reconstructions leads to anastomotic leakage and dramatically increased sepsis rates

Fig. 2

Schematic workflow for the interdisciplinary intensive care treatment of surgical patients with intra-abdominal sepsis. Therapy is based on the three columns: surgery, antimicrobial therapy, and intensive care treatment

Diagnosis

Early identification of the septic patient

The clinical presentation depends on the site of infection. While general symptoms such as fever (or hypothermia), tachycardia, and tachypnea reflect the SIRS criteria, additional signs such as altered mental status, oliguria, change in the skin with elongated capillary refill time, elevated liver enzymes, pathologic coagulation, etc., should be recognized on everyday rounds, but especially should lead to further diagnostics (Table 1). Early diagnosis of any postoperative complication is life-saving, but can be masked by “normal” postoperative symptoms such as abdominal pain or gastrointestinal paralysis with nausea. Compared to community-acquired secondary peritonitis, abdominal pain, tenderness, and fever occur less often in postoperative or ongoing peritonitis [17]. Improving the time to diagnosis requires not only an experienced surgeon but also the establishment of interacting interdisciplinary rounds on intensive and intermediate care additionally consisting of anesthesiologists, pain therapists, antibiotic stewards, etc.

As early markers of infectious complications after surgery, the specificity of several molecular markers has been tested. More than 30 studies have evaluated the predictive value of C-reactive protein (CRP) for surgical complications in the postoperative phase. Typically, CRP peaks between postoperative day (POD) 2 and POD 3 (about 12–24 h after interleukin (IL)-6 peaks) and declines to baseline level on POD 5. Persistent elevation of CRP can indicate septic complications in the postoperative phase [18]. For colorectal resection (about 100 mg/l on POD 5) and pancreatic surgery (140 mg/l on POD 4), cut-off values have been suggested, which must lead to further diagnostics. The role of procalcitonin (PCT) has been controversially discussed in the literature. In contrast to single cut-off values, the PCT clearance kinetics appear to be a better indicator for diagnosing septic complications or for predicting clinical outcome: persistently increased plasma PCT levels are associated with infection or septic surgical complications. While systemic infections are in line with up to 5000-fold increase within 4 h, the located sources of infection can be PCT negative. To date, it remains nebulous whether PCT can distinguish between (“sterile”) SIRS and sepsis. The CAPTAIN trial [19] revealed that none of the circulating biomarkers (including PCT) discriminated better between sepsis and SIRS than CRP alone. In contrast, PCT is a helpful tool for monitoring a patient with intra-abdominal infection [20]. Reith et al. published a trial on 246 patients with abdominal sepsis after surgery. A PCT reduction from POD 1 to POD 4 was a good predictor of clinical improvement.

As a marker of when to end antimicrobial therapy, PCT guidance can reduce treatment duration and even reduces mortality [21]. In the trial by de Jong et al., a PCT reduction of 80% of the initial value or serum PCT < 0.5 μg/dl were the cut-off values for ending antibiotic treatment [21].

Modern immunological research has identified a panel of markers such as IL-6, IL-1α, tumor necrosis factor alpha (TNFα), high mobility group box 1 (HMGB)-1, matrix metalloproteinase 9 (MMP-9), vascular endothelial growth factor (VEGF), intercellular adhesion molecule 1 (ICAM-1), myeloperoxidase (MPO), methylglyoxal, and caspase-3 as sensitive indicators of sepsis development [22, 23]. As patients who have undergone major surgery are in a phase of hyperinflammation (SIRS), it remains unclear if these markers can help to detect complications in the postoperative phase. As an example, the cut-off values of IL-6 in the current literature vary widely between 12 pg/ml and 2760 pg/ml [24].

Using ultrasound, surgeons have a bedside tool to obtain a rapid overview of potential peritoneal pathological conditions. Ultrasound-guided diagnostic drainage of suspicious fluid collections, combined with therapeutic tube insertion on demand, help to diagnose conditions such as intra-abdominal abscesses, hematoma, and pancreatic fistula. As a modern diagnostic approach, the measurement of intraperitoneal cytokines might be another promising tool for determining and monitoring the inflammatory reaction in patients [25]. Several studies have shown up to 1000-fold higher local concentration of cytokines compared to plasma levels [24]. Once suspected, computed tomography (CT) scan is the diagnostic gold standard for both secondary and ongoing peritonitis, with diagnostic sensitivity of 97.2% [17, 23]. For secondary peritonitis, the peritoneal CT attenuation values can even predict hospital survival. In hospital non-survivors, the values are significantly lower than that in survivors [26]. Alternatively, positron emission tomography (PET)-CT scan could play an important role in further diagnosis of intra-abdominal sepsis, but is mainly restricted to septic foci of unknown origin [27] or spreading multifocal bacteremia (e.g., Staphylococcus aureus) [28].

Therapy

Early goal–directed therapy and volume resuscitation

The therapeutic principle of early goal–directed therapy (EGT), which was first introduced by Rivers in 2001, postulated a protocol-based approach for treating patients with septic shock [29]. Targets for therapeutic resuscitation were central venous pressure (CVP, 8–12 mmHg), MAP (>65 mmHg), urinary output (>0.5 ml/kg body weight), and central venous oxygen saturation (SaO₂, >70 mmHg). In contrast to the single-center study by Rivers et al., three large multicenter randomized controlled trials (ProCESS [30], ARISE [31], ProMISe [32]) showed no benefit for a protocol-based sepsis therapy in the early resuscitation phase. Compared to the standard therapy, EGT showed no 90-day survival benefit, no improved 1-year survival, but longer duration of ICU stay and increased vasopressors.

In cases of hemodynamic instability (systolic blood pressure < 90 mmHg, MAP <70 mmHg, or systolic blood pressure decrease of >40 mmHg) or serum lactate >4 mmol/l, the SSC guidelines recommend the rapid application of crystalloids (30 ml/kg) within the first 3 h after hospital admission (3-h bundle, Table 3). Many trials on fluid resuscitation compared crystalloids versus colloids and could not show any benefit for the (expensive) colloid solutions [33]. Instead, colloids may be nephrotoxic (except albumin). Whether balanced crystalloid solutions (Ringer lactate) or “simple” saline solution should be used is still being discussed. Compared with saline, a buffered crystalloid solution could not reduce the risk of acute kidney injury in critically ill patients [34]. As reported previously [35], a chloride-restrictive balanced solution is in line with decreased rate of dialysis and renal insufficiency. This is being analyzed by two ongoing trials (BaSICS [36], PLUS trial protocol [37]).

Table 3 The newest version of the SSC guidelines recommend certain diagnostic and therapeutic measures within a 3- and 6-h timespan after hospital admission (these easy and rapid steps of sepsis therapy are essential for surgeons’ education)

However, how is volume application monitored? Over the years, a protocol-based, highly standardized resuscitation strategy has been postulated. In contrast, the new guidelines only recommend patient-oriented, individualized volume substitution according to the patient’s fluid responsiveness, which can be examined by the passive leg raising test, for example. The predictors of inadequate fluid responsiveness are [38] heart insufficiency, hypothermia, deteriorated gas exchange, increased serum lactate (>4 mmol/l), immunodeficiency, and coagulopathy. In 2017, Marik et al. wrote about volume overload in the early phase of sepsis (>5000 ml) presenting the risk of increased mortality [39]. To date, how fluid can be substituted in sepsis remains a matter of debate. While the guidelines postulate substitution as long as the patient shows circulatory response/increased cardiac output (“fluid challenge”), Takala suggests that maintaining tissue perfusion should be the target parameter of modern, individualized fluid resuscitation [40].

Source control

Source control in intra-abdominal sepsis is based on four important elements: debridement, removal of infected devices, drainage of purulent cavities, and decompression of the abdominal cavity. Inadequate initial source control increases the 28-day mortality rate from 26.7% to 42.9% [4, 41, 42]. While the importance of rapid surgical source control is clear, evidence for the effectiveness of so-called damage-control surgery is lacking until today. According to the modern concept of damage-control surgery, which was first established for heavily injured victims in combat/military surgery, the first operation should be performed as short as possible, followed by secondary reconstruction of the gastrointestinal continuity or the abdominal wall. Ceresoli et al. [43] suggested that further studies are required to define the indications, timing, and techniques of damage-control surgery for patients with non-traumatic abdominal sepsis.

What is the rationale of damage-control surgery in the non-traumatic patient? In both trauma and intra-abdominal septic shock/severe intra-abdominal sepsis, the patient is threatened by a pathophysiological triad of coagulopathy, inflammation, and cardiovascular instability (“lethal triad”). As published recently by Lyons et al., hospital mortality increased progressively from 25.4% to 56.1% in patients without and with severe sepsis–associated coagulopathy, respectively [44]. Furthermore, cardiovascular instability with concomitant high levels of circulating catecholamines is associated with poor outcomes and severe side effects such as myocardial injury and peripheral ischemia [45,46,47]. From trauma surgery, the principle of an abbreviated initial surgical approach for controlling abdominal blood loss and contamination could lead to accelerated resuscitation of physiology within this critical early phase after damage. A retrospective case series showed that this approach resulted in improved survival of shocked patients.

Table 4 provides an overview of the five steps of an escalating approach in damage-control surgery. Only evidence level III and IV data exist to transfer the traumatic damage control concept to abdominal sepsis [48].

Table 4 The impact of the damage control concept is increasingly being evaluated for intra-abdominal source control

While evidence is low due to retrospective studies only, acceptance in several reviews and editorials has promulgated the concept widely. Regardless of the limitations of the studies, there is a general trend toward the adoption of damage control strategies for abdominal sepsis comparable to the experiences in trauma surgery in the early 1990s. In contrast to damage control for trauma or intra-abdominal hemorrhage, the concept has to be modified for intra-abdominal sepsis, especially in the presurgical phase (Table 4). There is growing evidence that patients with secondary peritonitis benefit from a damage control resuscitation phase prior to surgical intervention, which is strongly recommended by the SSC Guidelines 2016 (see 3-h bundle) [49] (Table 5).

Table 5 Comparison of damage control surgery concepts for trauma versus intra-abdominal sepsis

In the case of ongoing, persisting peritonitis after initial surgery, three different surgical strategies have been established:

• 1. Relaparotomy on demand.

• 2. Planned relaparotomy within 36–48 h.

• 3. Open abdomen technique.

In contrast to relaparotomy on demand, which is performed in cases of clinical deterioration of the critically ill patient, the approach of a planned relaparotomy is based on the a priori decision for re-do surgery independent of its necessity. In a landmark study, Ruler et al. reported no difference between “on demand” (n = 116) and “planned” (n = 116) laparotomy in terms of patient mortality (on demand, 29%; planned, 36%), but the on-demand group had significantly lower intervention rates and hospital costs [57]. Nevertheless, a scheduled relaparotomy might still be indicated in cases of mesenteric ischemia requiring planned reassessment of the intestinal viability. The alternative approach in these cases would be an on-demand decision. The latter requires an experienced surgeon and an interdisciplinary everyday relook of the patient’s clinical course. Neither the initial source of intra-abdominal infection nor the findings during the primary surgical source control could predict the demand for reintervention in 219 cases [58]. Any recurrence or persistence of organ failure should lead to rapid surgical intervention. Koperna et al. reported that (n = 105) mortality was significantly lower if a relook on demand was performed within 48 h after initial emergency surgery, if necessary [59]. Both on-demand and planned relaparotomy present the risk of acute abdominal compartment syndrome (ACS) development: peritonitis itself on the one hand (primary ACS) and capillary leakage and fluid resuscitation (secondary ACS) on the other can lead to sustained intra-abdominal pressure > 20 mmHg with concomitant organ dysfunction [60]. As the diagnostic of choice, intra-abdominal pressure is typically measured indirectly through the bladder. Surveys revealed that, despite its hazardousness, ACS is often diagnosed too late. Only 47% of physicians could define ACS in that trial. Once suspected, the guidelines recommend monitoring the intra-abdominal pressure every 6 h in such cases [60].

Although it is one potential element of damage-control surgery, the current clinical guidelines do not recommend routine use of the open abdomen technique for secondary peritonitis [61, 62]. While open-abdomen surgery prevents ACS development and allows a rapid and easy second look, it presents the risk of enteroatmospheric fistulas or fascial deviation [63]. The increased morbidity is furthermore based on physiologic changes, which are in line with persistent opening of the peritoneal cavity: hypothermia, impaired immune function, fluid loss, and increased muscle proteolysis must lead to the modification and adaptation of intensive care therapy (passive rewarming/air warmers, pain control, tailored ventilator support, monitoring of pH and lactate, etc.). This complicated pathophysiological condition can result in increased mortality, which was reported recently [64]. Nevertheless, the open-abdomen approach is indicated for patients with secondary/persisting peritonitis, who face the risk of ACS development or in whom a second-look operation is expected. The prospective COOL trial (Closed or Open after Source Control Laparotomy for Severe Complicated Intra-abdominal Sepsis) is recruiting patients with severe intra-abdominal sepsis (defined as septic shock or a Predisposition, Infection, Response, Organ Dysfunction score > 3 or a World Society of Emergency Surgeons (WSES) sepsis severity score > 8) to analyze the influence of open versus fascial closure on mortality after source control [65].

Antibiotic therapy

Blood cultures should be collected prior to any antimicrobial treatment. Two to three pairs (aerobic and anaerobic) of blood culture samples should be collected from both the peripheral blood and from central venous catheters. Any antibiotic therapy dramatically reduces the detection rate of the blood culture technique [66].

The 2016 SSC recommendations [49] postulate that:

• 1. Initial administration of intravenous antimicrobials should be performed within 1 h after admission.

• 2. The first choice is a broad-spectrum antibiotic (or a combination of antibiotics).

• 3. The antibiotic spectrum should be narrowed after the microbes have been isolated.

• 4. Based on the clinical situation, de-escalation of antimicrobial pharmacotherapy should be considered as soon as possible.

As outlined in the International Guidelines for Management of Sepsis and Septic Shock: 2016, the initial administration of antibiotics is a key step in the early management of sepsis and septic shock (3-hour bundle, Fig. 3, Table 3) [49]. In hypotensive patients, every hour of delayed initiation of antimicrobial therapy leads to an increase in mortality [67]. These beneficial effects of rapid empirical antibiotics within the “golden hour of sepsis” could also be confirmed for patients with sepsis (with organ dysfunction!) [68]. The German Medical Education for Sepsis Source Control and Antibiotics (MEDUSA) showed an increase in mortality of 2% per hour of delayed antimicrobial therapy (1%/hour of delayed source control) [69].

Fig. 3

Schematic overview of the antimicrobial therapy for patients with secondary and/or ongoing peritonitis according to the guidelines of the Paul-Ehrlich Society. The SSC guideline postulates the administration of a broad-spectrum antibiotic as soon as possible. Modern antibiotic stewardship programs involve the interdisciplinary, everyday re-evaluation of the critically ill patient, followed by either rapid de-escalation or modification of the antimicrobial therapy

Although it is a fact that early antimicrobial therapy is life-saving, the indiscriminate use of broad-spectrum antibiotics has promoted the development of antimicrobial resistance and is furthermore associated with adverse effects during the clinical course of the intensive care patient. To date, there is hardly any literature on this important field of pharmacotherapy in sepsis. The paramount importance of antimicrobial resistance is underlined by the foundation of several taskforces such as the World Alliance Against Antibiotic Resistance or the WSES AGORA (Antimicrobials: a global alliance for optimizing their rational use in intra-abdominal infections) initiative [70, 71].

Furthermore, the inadequate, non-specific use of antibiotics in the ICU is accompanied by increased rates of pulmonary (30%) and urinary (8%) infections [72]. As published recently [73], none of the 10 bacteria most frequently isolated from peritoneal sources of infection was sensitive to ampicillin/sulbactam. Table 6 summarizes the potential new-generation antimicrobials, which could be useful as second-line therapy in ongoing peritonitis caused by resistant bacteria.

Table 6 New-generation antibiotics and their potential indications

In cases of complicated ongoing peritonitis, antimicrobial therapy leads to selection pressure within the bacterial flora. This holds also true for Candida species, whose isolation from peritoneal fluid correlates with impaired clinical prognosis and persistent ongoing peritonitis [72]. Bassetti et al. reported up to 50% mortality of ICU patients with intra-abdominal candidiasis (!), while that of non-ICU patients was only half that. The European Society for Clinical Microbiology and Infectious Diseases (ESCMID) recommends echinocandins as first choice medication for Candida infections in ICU patients [84]. Fluconazole is a rational alternative for treating C. parapsilosis. It is important to know that medication must be administered up to a minimum duration of 14 days after a Candida-negative blood culture occurs. An inadequate therapeutic regimen of intra-abdominal candidiasis was proven to be an important negative prognostic survival parameter [72, 85].

Three key points should be kept in mind when discussing modern surgical antimicrobial therapy in ICU:

First, rapid detection of specific bacteria from the intra-abdominal source (e.g., from the peritoneal cavity) would shorten the period of empiric antimicrobial therapy to a tailored, more specific one. In contrast to blood culture, PCR-based techniques can allow the rapid identification of bacteria and associated antimicrobial resistances [22]. Particularly, surgical patients with ongoing, persisting peritonitis can be monitored by these modern approaches in the future.

Second, antibiotic stewardship should be implemented in any surgical ICU. The rapid, individualized de-escalation of the (broad-spectrum) antimicrobial therapy avoids bacterial in-hospital resistances and reduces pharmacological adverse effects such as renal and/or hepatic insufficiencies. A recent article on nosocomial pneumonia stated that de-escalation of antimicrobial therapy is postulated within 2–3 days after initiation [86]. On daily interdisciplinary rounds, multiple aspects such as clinical status and development; infectious parameters (e.g., PCT); metabolic, hepatic, and renal laboratory values; grade of source control, etc., have to be considered and result in flexible and individualized antimicrobial therapy. For uncomplicated intra-abdominal sepsis, an antimicrobial regimen should be finished after 7 days on average. A Staphylococcus aureus bloodstream infection (BSI) has to be treated for at least 14 days if uncomplicated; complicated Staph. aureus BSI require at least 4 weeks of antibiotic therapy.

For uncomplicated intra-abdominal sepsis, an antimicrobial regimen should be finished after 7 days on average, except for Staph. aureus infection, which requires up to 4 weeks of antibiotic therapy in the case of complicated infection.

Third, the differentiation between infection and colonization is of major importance for surgical patients with (ongoing) peritonitis. These patients are permanently threatened by hospital-acquired infections. While colonization by multidrug-resistant pathogens such as methicillin-resistant Staph. aureus (MRSA), vancomycin-resistant enterococci (VRE), or multidrug-resistant gram-negative bacteria (MRGN) is often diagnosed in surgical patients, it normally leads to isolation only. Whether the colonization of a patient with secondary and/or ongoing peritonitis should be treated with antibiotics should be investigated in future studies. The REDUCE (Randomized Evaluation of Decolonization versus Universal Clearance to Eliminate) MRSA trial changed the view on pharmacotherapy of colonized patients, showing a clear benefit for universal decolonization in comparison to screening plus isolation only [87]. Antibiotic treatment of colonized patients led to dramatically reduced rates of positive BSI in intensive care patients.

Give your patients a FAST-HUG

Every surgical patient with secondary peritonitis requires certain key elements of intensive care therapy such as prophylaxis of ulcers (e.g., proton pump inhibitor), lung protective ventilation (according to the ARDS (acute respiratory distress syndrome) network protocol), hemodynamic stabilization (MAP >65 mmHg, inotropics in cases of myocardial dysfunction, invasive hemodynamic monitoring, glomerular filtration rate > 0.5 ml/kg body weight, repetitive serum lactate measurement), blood glucose 110–180 mg/dl, prophylaxis of thrombosis, and enteral nutrition, if possible.

While these values and target parameters provide valuable assistance during everyday rounds, the exact doses, the kind of monitoring, etc., remain a matter of debate in the recent literature. As an example, adjunctive sepsis therapy with corticosteroids is still intensively discussed within the expert literature. While hydrocortisone did not reduce the development of cardiovascular instability/septic shock in the HYPRESS trial [88], recent literature reveals that at least some subgroups of patients with septic shock appear to benefit from continuous hydrocortisone administration (ADRENAL [89], APROCCHSS [90] trials).

As another example, the target parameters of (lung-protective) ventilation in the septic patient have shifted to lower PaO₂ (partial pressure of oxygen) values. In contrast to the conventional ventilator regimen (PaO₂ up to 150 mmHg, SpO₂ (blood oxygen saturation) = 97–100%), a more restrictive ventilation (PaO₂ = 70–100 mmHg, SpO₂ = 94–98%) seems to be beneficial [91].

For everyday rounds, the surgeon should remember to monitor the key aspects of modern intensive care medicine for critically ill surgical patients. As introduced by Vincent into clinical routine [92], every patient should get a FAST-HUG (Feeding, Analgesia, Sedation, Thromboembolic prophylaxis, Head-of-bed elevation, stress Ulcer prevention, Glucose control) at least once a day by both intensivists, anesthesiologists, and the abdominal surgeon.

Outcome and conclusions

Substantial improvement of sepsis survival is the main challenge of modern surgical research. In contrast to “historical” data reporting 40–60% hospital mortality for severe sepsis, more recent randomized trials that included strict implementation of protocol-based resuscitation therapy reported between 18% and 30% mortality [4, 93]. Nevertheless, both mortality and morbidity remain unacceptably high. Long-term morbidity leads to substantial functional disability, mental impairment, which finally is reflected in high rates of hospitalization in acute care or skilled nursing facilities. Permanent education, feedback, and audit initiatives are a new approach for monitoring the implementation of SSC sepsis measures. As stated by Levy et al., the strict transfer of the SSC guidelines into daily patient care led to a 9.6% decline in mortality [94].

Surgical source control is the obligatory treatment of every patient with secondary or ongoing peritonitis and is of both therapeutic and diagnostic importance. Future studies should evaluate the impact of damage-control surgery on the survival of patients with intra-abdominal sepsis. New diagnostic tools such as biomarker assays and PCR-based techniques for detecting microbes will accelerate the identification of complications after surgical treatment and will allow healthcare providers to initiate individualized antimicrobial therapy rapidly in the near future. The SSC constantly updates the guideline-based supportive care, but permanent medical education on sepsis diagnostics and therapy is required. Sepsis recognition and therapy require everyday re-evaluation of the patient during interdisciplinary rounds.