Diagnostic challenges in postoperative intra-abdominal sepsis in critically ill patients: When to reoperate?

Sepsis is the systemic inflammatory response to an infection [40, 41, 42]. It can be defined as “life-threatening organ dysfunction caused by a dysregulated host response to infection.” The Sequential Organ Failure Assessment (SOFA) score is used to identify the organ dysfunction. Severe sepsis is the condition involving sepsis and dysfunction of at least one acute organ, hypoperfusion, or hypotension. The SIRS score, which was used in earlier definitions of sepsis, is now not used. The quick SOFA (qSOFA) has been introduced in the new definition of sepsis (Table 1) [43].

qSOFA is a specific attempt to identify high-risk patients with a suspected infection at the bedside. A several-fold increase in in-hospital mortality has been reported with an increase in two or higher points in qSOFA, reflecting the prognostic efficacy of the score [43]. However, the efficacy of qSOFA in the ED setting is not fully established. Dykes et al. reported a comparative study on qSOFA and SIRS criteria for early sepsis [17]. The study involved an elderly population, and the primary outcome was sepsis within 48 hours of admission. They reported that the qSOFA had lower sensitivity than SIRS (44.7% vs. 80.0%) but higher specificity (83.6% vs. 25.7%). Their results reflect low sensitivity of qSOFA; however, based on specificity and positive predictive value, the authors concluded that veterans with higher qSOFA had a high probability of having sepsis. The National Early Warning Score (NEWS) has also been reported to be useful in the ED setting; Lim et al. recently reported the effectiveness of NEWS in infection-related acute medical settings [44]. SIRS, SOFA, qSOFA, NEWS, and other scores are useful and have their own merits; however, none of them is specific enough to assist surgeons in deciding the feasibility of reoperation in patients with post-operative IAS.

Prophylactic antimicrobial therapy and other advances have reduced the incidences of surgery-bound infections. However, infections are still the leading causes of post-surgery morbidity and mortality [24, 45, 46, 47, 48]. Failure to remove infection at the initial surgery or anastomotic breakdown is the major contributor to postoperative IAS. Postoperative IAS is a challenging condition to diagnose. The outcome of patients with postoperative IAS correlates with the delay between the first and the corrective operation. With a rapid control of infection sources, the outcome is known to improve [7, 21].

The abdomen is the second most common source of sepsis [1, 49, 50]. Peritonitis, a common etiology of IAS, is categorized into primary, secondary, or tertiary. Around 11% of all patients with peritonitis are known to have complications of severe sepsis [51, 52, 53]. Early diagnosis and treatment of IAIs are important to sepsis control [2, 54, 55, 56, 57]. As many cases of complicated IAI involve the ICU setting, severity scoring systems for IAI can be disease-independent or disease-specific. Disease-independent scores include Acute Physiology and Chronic Health Evaluation (APACHE) II and Simplified Acute Physiology Score (SAPS II). Among disease-specific scores, Mannheim Peritonitis Index (MPI) is a valuable prognostic tool. However, some concerns have been raised regarding the prognostic significance of APACHE II in peritonitis, as the score does not include the role of interventions. The MPI has the advantage of being specific for peritonitis. However, these scoring systems or their combinations are not specific to prognosticate the reoperation decision-making process in postoperative IAS.

After abdominal surgery, reoperation presents severe morbidity and mortality risks, especially in IAS, in critically ill patients [22, 26, 48, 58, 59]. Currently, for patients with postoperative IAS, there are no accepted guidelines or scoring systems that can be used to select the patients for relaparotomy for sepsis control. Even with the recent advances in diagnostic technologies (radiological or laparoscopic), prognosis after laparotomy remains a challenge. A combination of diagnostic tests and clinical examinations is used in the clinical decision-making process. Different diagnostic methods, including laboratory, radiologic, and interventional techniques have been developed to assess the abdomen in postoperative IAS (Figure 1) [58].

Fig. 1

Several factors need to be analyzed in the critical care setting before selecting an appropriate diagnostic modality for a patient with suspected postoperative IAS. The foremost consideration is the hemodynamic stability of the patient. For unstable patients, techniques that involve the transfer of patients from the ICU to another unit might not be feasible. Computerized tomography is among the most common modalities for the intra-abdominal examination for stable patients. The comparative efficacy of computed tomography (CT) and ultrasonography (US) for the diagnosis of IAIs have been reported in several studies [60]. It may be noted that though, in the critical care setting, portability of the sonography techniques offers an advantage over CT, the considerably low sensitivity and specificity of sonography make it of little help in surgical decision-making on patients with postoperative IAS. Magnetic resonance imaging and nuclear techniques are not routinely available in most hospitals in the ICU. Therefore, their uses are expected to have limited applications in postoperative IAS [61]. When imaging is not feasible or conclusive in identifying the causative pathology, diagnostic laparoscopy techniques or exploratory laparotomy may be used [62]. Abdominal Reoperation Predictive Index (ARPI) is the only predictive index developed for assessing the suitability of the surgical approach in postoperative care of IAS [23].

Serial intra-abdominal pressure is another technique that has been explored for the prognosis of the patient with suspected postoperative IAS. A detailed discussion of these techniques is presented in the following sections.

CT of the abdomen is extensively explored for IAI because of its high accuracy in diagnosing intra-abdominal abscess [36, 63, 64, 65, 66, 67]. However, a few factors restrict the use of CT in the critical care setting, including the risk associated with the movement of the patient from the safety of the ICU to the radiology room and the transfer- and optimization-associated time delays. Significant respiratory and hemodynamic deterioration can take place during transportation. Careful consideration of the patient's condition should be given before using CT in a patient with suspected postoperative IAS. CT scan has high sensitivity and specificity for intra-abdominal fluid collections and abscesses [68, 69, 70]. Velmahos et al. conducted a prospective case series of 85 critically injured patients. The reported sensitivity was around 98% in their work, and the reported specificity was around 62%. Around 70% of the total patients benefited from treatment changes due to the inputs provided from the CT scan [36]. Kumta et al. indicated that in around 80% of the patients in their study group, the management course was altered based on decisions made after CT [71]. As the patient's transfer to a CT scan is a severe limitation of the management of patients receiving critical care, one way to overcome this limitation is the use of portable CT [71]. However, portable CT is expensive, and not much work has been reported on its diagnostic efficacy for postoperative IAS. Maher et al. investigated the quality and clinical Impact of portable abdominal CT in 100 cases [72]. They reported the relatively low quality of the image obtained from portable CT compared to the image obtained from conventional CT. However, the authors concluded that, despite the inferior image quality, portable abdominal CT provides essential diagnostic information with the benefit of avoiding transport of the patient outside the ICU. Nonetheless, the interpretation of portable CT scans should be conducted with caution. Even the suitability of conventional CT scans for the patient with suspected postoperative IAS was questioned in some studies. Norwood et al. analyzed the influence of CT scans on the clinical decisions made on critically ill patients. They concluded 70% of scans were either of no help or were not used in the clinical decision process [73]. CT examinations should therefore be used according to proven clinical rationale; when case severity is high, alternative diagnostic techniques may be explored [74]. There is not much data on predicting the power of CT scans for reoperation for postoperative IAS. Bader et al. focused on elucidating early indications for relaparotomy in patients with diffuse secondary peritonitis. They reported a higher sensitivity in CT scans (97.2%) than in conventional radiography (66.2%) or in ultrasonography (44.3%) [75]. The use of multidetector-row CT (MDCT) scanners can further enhance the diagnostic efficacy for detecting the cause of abdominal sepsis. A sensitivity of 95% and a specificity of 91% was reported for abdominal sepsis [76].

Due to the risk associated with the transport of patients from ICU and in situations when routine radiological techniques do not yield a conclusive diagnosis, bedside diagnostic techniques are advised to identify causative intra-abdominal pathology [77, 78, 79, 80]. Furthermore, bedside techniques do not interrupt the ongoing treatment in the ICU and can also help in cost and time minimization. In the critical care setting, BDL is a feasible, safe, and efficient tool to examine the intra-abdominal pathology of hemodynamically unstable patients [78]. BDL has high accuracy for IAS diagnosis. BDL in appropriately selected patients can, therefore, be of great help in avoiding negative laparotomies [81]. Even in unstable patients with abdominal sepsis of unknown origin, DL has been found to have good accuracy with a high definitive diagnosis rate (86–100%) [77]. BDL can be successfully used for intubated as well as for non-intubated patients. It can be used when there are other signs of sepsis, but laparotomy indications are not clear. Diagnostic laparoscopy (DL) has very high sensitivity (98%) and specificity (96%) for unexplained severe abdominal pain [82]. In 36–95% of patients, DL has been reported to prevent laparotomy [83]. It is superior to computed tomography or ultrasound of the abdomen in terms of sensitivity and specificity. In BDL, morbidity risks are very low (<8%), and no mortality cases were directly attributed to DL. Due to better safety, DL is also preferred over diagnostic peritoneal lavage. However, since mortality rates among patients admitted in the ICU are generally very high, the diagnostic modality's safety cannot be the sole criterion for its use. The suitable diagnostic modality and the conditions of its use in the patients with suspected postoperative IAS should be case-specific and should not be generalized. The decision should be based on the patient's condition and vital signs and the surgical resources and expertise available at the hospital.

Before the availability of advanced CT techniques, DPL used to be a preferred method for diagnosing intra-abdominal pathology in the ICU [84, 85, 86, 87]. The process is not associated with mortality risk; however, it is reported to have risks such as small bowel/mesenteric injuries, intra-abdominal abscess, bladder punctures, abdominal wall infusions, and reoperation [88, 89]. Complication rates in DPL have been reported in the range of 0.8%–1.7%. This test is considered 100% sensitive and 88% specific; therefore, a negative diagnostic peritoneal lavage makes intra-abdominal surgical disease highly unlikely. On the other hand, a positive lavage may require further diagnostic examination [89, 90]. For critically ill ICU patients, Lee et al. recommended that DPL be a useful preoperative diagnostic tool for examining suspected intra-abdominal pathology [91]. A patient-specific risk-benefit analysis needs to be conducted before selecting DPL for critically ill patients with suspected postoperative IAS. One of the reasons the author has a low threshold for leaving intra-abdominal drains in high-risk patients is to utilize the drain body fluid analysis as a DPL.

Though surgeons can use the techniques mentioned above to better assess the severity of intra-abdomen sepsis and decide on reoperation, none of them is exclusively designed to assist in the decision-making process in postoperative IAS. ARPI is the only predictive index available exclusively for the abdominal re-operation decision-making process in postoperative IAS [27, 92]. A typical flow chart for ARPI is shown in Figure 2.

Fig. 2

ARPI has lower mortality, reduce the delay between the first operation and reoperation, and shorten the duration of ICU stay [27]. ARPI is calculated using the different parameters listed in Table 2 [27].

It has been claimed that with the proper use of ARPI, the surgical conundrum associated with the decision concerning re-operation of patients suspected with postoperative IAS can be resolved to a great extent [27]. ARPI is reported to be specifically useful in settings with limited resources [93]. In a cross-sectional, retrospective study of 94 patients diagnosed with abdominal sepsis, ARPI was used to assist in the decision process [23]. The authors reported reoperation in 75% of cases when ARPI was greater than 21 and 26% of cases when ARPI was less than 20. In another recent study, ARPI was found to be a helpful surgical decision aid regarding relaparotomy [23]. However, ARPI also has several limitations. The most important is that the accuracy of this scoring system has not been fully established. Furthermore, the variables that have high weight are related to postoperative IAI symptoms rather than peritonitis symptoms. The index also has subjective variables such as altered levels of consciousness and abdominal pain, which are difficult to measure in sedated and ventilated ICU patients [92].

Abdominal exploration can also be used as a highly effective diagnostic modality in acute cases of IAS [94, 95, 96]. It is a rapid, cost-effective, and accurate method. However, the long-term mortality and morbidity risks associated with it dictate its use only in cases with apparent anticipation of the diagnosis [97, 98, 99, 100, 101, 102]. Advances in laparoscopic and imaging techniques for abdomen inspection have further reduced the necessity of exploratory laparotomy. Before exploratory laparotomy, it is essential to ensure that necessary expertise and resources are available to perform the adequate and timely therapeutic procedure if the suspected IAS is confirmed on exploration [59]. DPL can be used as the first-line diagnostic modality. If the results are positive, exploratory laparotomy can be used; in case of negative DPL findings, the patient can be closely monitored till the indications for exploratory laparotomy are more evident [103]. In a critical study on exploratory laparotomy, BSL (Bedside Exploratory Laparotomy) in the ICU was compared with emergency laparotomy in the operating room. The study reported 77.8% mortality in BSL and 45.5% in laparotomy [104]. This result does not suggest that BSL has higher mortality risks than laparotomy. This result underscores the ICU setting's mortality risk and the importance of timely and adequate surgical intervention. In cases where there are no rapid ways available to support the surgical decision process, emergent laparotomy may be used to identify surgically correctable causes, despite the anticipated high mortality with laparotomy [62, 104, 105, 106].

Further stressing the importance of timely interventions, some studies recommend on-demand relaparotomy within 48 hours of initial surgery or before the emergence of multiple organ dysfunction syndromes [25, 107, 108, 109]. In terms of deciding for relaparotomy, van Ruler et al. suggested that findings at emergency laparotomy for peritonitis are not adequate, and the monitoring of progressive or persistent organ failure is the best indicator for the need for reoperation [110]. In the absence of extensive studies, it is difficult to make a conclusive argument. However, all studies recognized the benefits of sensitive detection techniques that permit directed laparotomy before septic deterioration [13, 101, 111, 112].

IAS is reported to be associated with higher intra-abdominal pressure (IAP). IAP increase is ascribed to excessive bowel edema due to increased vascular permeability and fluid sequestration [113, 114]. IAP has been classified into four grades, and higher IAP commonly results in reduced regional and global perfusion, posing a severe risk for vital organ failure. After a comprehensive literature search, Plantefeve et al. concluded that IAP could be a useful tool for monitoring critically ill patients with a suspicion of postoperative IAS [115]. Basu et al. investigated the predictive potential of the increased IAP for relaparotomy. A study of 102 patients found that postoperative elevation in IAP is associated with the increased risk of peritonitis, concluding that the elevated IAP can be a predictor of early relaparotomy [116]. It is worth highlighting here that the IAP measurements conducted in the immediate postoperative period were used for predicting the need for relaparotomy. This will be particularly helpful in starting early intervention for the surgical correction of sepsis.

Using preoperative and intraoperative parameters, Kong et al. also reported developing a clinical model with a predictive value of more than 90% for the need for relaparotomy. In a study involving 1000 patients, four factors were used to accurately predict the need for subsequent relaparotomy. These factors were: patients referred from any rural center, duration of illness greater than 5 days, heart rate higher than 120 bpm, and perforation associated with generalized IAS [117]. Van Ruler et al. used a 10-point visual analog scale (VAS), and VAS score >5.0 was considered “important.” However, only three out of ten “important” variables were found to be independently predictive [118]. In another interesting study, postoperative procalcitonin (PCT) serum levels were monitored. The authors claimed sensitivity of 93% and a specificity of 71% for eliminating septic abdominal focus [119]. Sileikis et al. devised a scoring system using six factors: age, sex, leukocyte count, C reactive protein, time of symptoms to index operation, and MPI. A score ≥37 indicates a planned relaparotomy, and ≤24 indicates other milder therapeutic interventions rather than relaparotomy [120].