Sepsis - Causes and Diagnosis

DIFFERENTIAL DIAGNOSIS

               The differential diagnosis for SIRS is extremely broad and this is reflected in the fact that more than two-thirds of intensive care unit patients and a substantial number of patients on general medical units at some point during their hospitalization meet the criteria for this diagnosis {Marshall, 2000 #2; Vincent, 1997 #3}. Causes of SIRS include both infectious and noninfectious conditions such as trauma, autoimmune disease, drug reaction, myocardial infarction and pulmonary embolism. When a patient with SIRS develops concomitant shock it imperative that the hypotension be further defined as therapies for various types of shock differ greatly. There are four broad shock states: hypovolemic, distributive, cardiogenic, and extracardiac obstructive shock. The various shock states can be distinguished by characteristic hemodynamic profiles and clinical characteristics. However, as patients have increasing comorbidities, either confounding organ dysfunction or overlapping features can cloud this differential (Tables 2 and 3). Each type of shock will be briefly reviewed in the context of a differential diagnosis for a patient presenting in septic shock.

Table 2: Differential Diagnosis of Shock

Table 3: Hemodynamic Profiles of Shock States

               Hypovolemic shock is caused by a decrease in circulating blood volume. The decrement in blood volume can be either hemorrhagic in origin (e.g. related to trauma or GI bleeding) or non-hemorrhagic (e.g. due to vomiting, diarrhea, burns with increased insensible losses, or polyuric states such as diabetes insipidus or DKA). The patient in hypovolemic shock presents with tachycardia and hypotension. Extremities are cool to touch and cyanotic related to a low cardiac output state. Patients exhibit signs of hypovolemia such as collapsed neck veins and oliguria or anuria. Similar findings can be seen in early septic shock when a vasodilatory response to overwhelming infection leads to a relatively hypovolemic picture.

               Distributive shock is characterized by vasodilatation with either normal or elevated cardiac output. Classically, patients will present with warm extremities and tachycardia although decompensated distributive shock may be notable for extremities which are cool to touch. Organ injury is due to the inadequate cardiac preload, maldistribution of blood flow and may be further worsened by the inability to utilize oxygen adequately on a cellular level. Besides sepsis, anaphylaxis or neuralgic injury with loss of autonomic nervous system function can also cause distributive shock; therefore, allergen exposure or spinal cord injury should be considered when a patient presents with a clinical picture consistent with distributive shock.

               Cardiogenic shock results from inadequate tissue perfusion due to cardiac dysfunction. It is characterized by a low cardiac output and a high systemic vascular resistance in the setting of normo- or hypervolemia. Cardiac dysfunction can be driven by myocardial issues such as ischemia/infarction and cardiomyopathies, mechanical issues such as valvular disease or conduction issues such as arrhythmias. In the setting of the septic patient it is important to consider primary cardiac issues in the differential of the shock state since underlying heart disease can complicate the management of the septic patient. In addition, septic shock can have primary cardiac depressive effects which will effect the presentation of the patient.

               Extracardiac obstructive shock can be divided into two broad categories: increased intrathoracic pressure and intrinsic vascular flow obstruction. The primary effect of both categories can be either pump failure if cardiac output is compromised or loss of preload if blood return to the heart is compromised. As a result, extracardiac obstructive shock can present as a primarily cardiogenic shock type picture when the inciting event leads to primary cardiac failure or it can also present as a hypovolemic picture when preload is lost, or a mixture of the two. Increases in intrathoracic pressure leading to a shock state can be seen in the setting of a tension pneumothorax or as a result of positive pressure ventilation such as in obstructive lung disease with air trapping and the progressive rise of auto peep. Intrinsic vascular flow obstruction can be seen in massive pulmonary embolism, cardiac tamponade, and severe pulmonary hypertension. It is important to consider these issues in the differential diagnosis of the septic patient as some of these are mechanical issues which can be corrected expeditiously leading to rapid hemodynamic improvement.

CLINICAL MANIFESTATIONS

Neurologic Dysfunction

               A decline in a patients level of consciousness can be one of the first manifestations of sepsis. Focal neurologic deficits and seizures are rare and the brain appears anatomically normal on imaging unless patients have pre-existing disease that enhances the likelihood of ischemia induced focal pathology . Patients who have a decline in mental status related to sepsis have worse outcomes than those who do not. The etiology of this impaired level of consciousness is multifactorial and includes decreased cerebral perfusion and oxygen extraction.

Pulmonary Dysfunction

               Sepsis is one of the most common causes of Acute Lung Injury (ALI) and Acute Respiratory Distress Syndrome (ARDS). The sepsis syndrome accounts for up to 43% of cases of ARDS and leads to a significant increase in ICU mortality. How sepsis causes diffuse alveolar epithelial injury is unclear. As lung function and hypoxia worsen tissue perfusion and oxygenation are compromised. ARDS can readily be confused with diffuse pneumonia, or can be complicated by secondary pneumonia, and thus much ARDS is in fact a mixed picture with overlapping pathologies.

Cardiovascular Dysfunction

               Global cardiac depression is common in sepsis as described above. And there is no convincing evidence for ischemia as a cause of myocardial dysfunction in sepsis. Nonetheless, it is important that clinicians be cognizant of the fact that patients with preexisting coronary artery disease can suffer regional ischemia and infarction in the setting of sepsis.

Renal Dysfunction

               Oliguria and renal failure are common complications of sepsis. Renal dysfunction is caused by the shunting of blood away from the renal vascular bed to other vital organs. Oliguria may resolve with fluid resuscitation. Renal failure associated with sepsis is usually reversible. However, many other causes of renal dysfunction can complicate clinical presentation and outcome, such as contrast nephropathy, drug toxicity, diabetes, or hypertension.

Gastrointestinal Dysfunction

               Hypoperfusion of the GI tract in sepsis can increase intestinal permeability which is associated with the onset of multiple organ dysfunction syndrome. Mucosal hypoperfusion can lead to ulceration and increase the risk of GI bleeding. These consequences of sepsis can influence the efficacy of enteral nutrition, and the access of bowel flora to lymphatics and capillaries.

DIAGNOSIS

               Patients with sepsis need a careful history and physical examination. While laboratory assessment and monitoring provides invaluable and sometimes essential information for diagnosis and for therapeutic monitoring, failure to assess the history and failure to examine the patient carefully can lead to unnecessary therapies with resulting toxities, and to failure to initiate appropriate therapy promptly, thus reducing the likelihood of a successful outcome.

               History can direct diagnosis in many obvious ways. The environmental flora is important: specific pathogens might be suggested if a patient traveled to the Southwest United States (Coccidioides), to Sub-Saharan Africa (malaria or Yellow Fever) or to Martha’s Vineyard (Tularemia, Babesisia, Anaplasma). Specific pathogens might also be suggested by recent food ingestion (peanut butter and salmonella), travel on a cruise ship (Norovirus), or membership on a team with reported cases of MRSA skin disease. Pre existing medical conditions (ureteral calculi), immunosuppression (chronic corticosteroids, recent human stem cell transplant) are examples of the myriad of information that a skilled history can elicit.

               Physical examination is also vital to an informed management strategy. The presence of foreign bodies (intravascular catheters or urinary catheters), signs of pulmonary consolidation, skin lesions suggesting infection, or abdominal distention may direct the diagnostic and therapeutic strategies in more focused directions than an unfocused, empiric, “shotgun” approach.

               Imaging studies are increasingly important for localizing the source of infection. Computer assisted tomography (CAT scans) and magnetic resonance imaging (MRI) are more and more standardized for patients who can be transported and safely monitored for the procedure. Consideration must also be given to the advantage and toxicities of contrast agents. Such imaging is vital for providing diagnostic clues based on patterns or disease, or providing targets for aspiration or biopsy. Such imaging is also essential for recognizing obstructions of vital structures such as bronchi or ureters, for identifying collections of purulence or foreign bodies that must be removed.

               For patients with sepsis, severe sepsis, or septic shock, any potential source of infection should be carefully assessed. Bronchoscopy, lumbar punctures, sinus aspirations, soft tissue aspirations or biopsies, and lymph node biopsies are among the many invasive techniques that are a standard component of an effective search for the causative process and identification of the causative organism. Any purulent discharge needs to be examined by direct microscopy and culture. Similarly, suspicious body fluids, secretions, and excretions should be similarly examined, such as sputum, urine, stool, or CSF. Tissue biopsies of skin, lung, lymph nodes, or aspiration of collections are clearly important. Microbiology laboratories are more and more likely to be distant from the area of patient care. Despite this, clinicians must communicate with the laboratory to assure that specimens are collected and transported properly, and that the optimal stains and cultures are performed. As diagnostic and therapeutic interventions become more feasible for viruses, fungi, fastidious bacteria, optimal management requires increasingly sophisticated diagnostic approaches.

               Blood cultures are an important part of a diagnostic evaluation. The likelihood that useful information will be derived from blood cultures depends on selecting patients with a reasonable likelihood of having organisms in their blood stream. A patient with a likely viral infection, or a patient who is stable with a bacterial infection and doing well on antimicrobial therapy, by definition, is unlikely to be bacteremic or fungemic. Obtaining a blood culture on such a patient is more likely to yield a contaminant than a true pathogen, potentially causing unnecessary drugs, unnecessary expense, and unnecessary toxicity. Thus, care must be exercised in choosing which patients to draw blood cultures on, when to draw cultures, and what types of cultures to obtain.

               Blood cultures should ideally be obtained before new antimicrobial therapy is started. Obtaining blood cultures prior to starting or switching antibiotics will increase the likelihood that such cultures will identify the causative agent. However, obtaining cultures show not substantially delay the initiation of potentially life saving therapy.

               A blood culture is defined as a single blood draw with inoculation of all bottles used by the laboratory’s culture system. Some blood culture systems will require inoculation of two bottles, but others require one or three bottles for routine use. The suggested quantity of blood should be drawn: suboptimal ratio of media to blood, or suboptimal volume of blood will reduce the yield of cultures. Extra bottles must be inoculated for special pathogens such as viruses, certain fungi, certain zoonoses and mycobacteria. The yield of blood cultures, once a high risk patient and an appropriate time (e.g. before antimicrobial agents are started) are identified, depends on the volume of blood drawn and the processing technique. Current guidelines for evaluating fever in the intensive care unit now recommend three cultures be drawn rather than two cultures based on the relationship between blood culture yield and volume of blood. When blood cultures are drawn, at least two sets should be drawn. Having two sets offers several advantages: 1) the volume of blood cultured is double that of a single culture, increasing the yield; 2) if an organism is identified that could be a contaminant, having two positive cultures provides more convincing evidence than a single positive culture; 3) if there is an intravascular catheter in place, drawing one culture percutaneously and one through the catheter (or through the catheter most likely to be infected due to chronology or local signs or symptoms) permits an assessment of “time to positivity” which can provide suggestive information as to whether or not the catheter is the source of any bacteria or yeast cultured from the blood. Molecular tests on blood and body fluids are an increasingly important part of the diagnostic armamentarium, especially for viruses and fungi. Gene probes for tuberculosis, rapid tests for influenza or RSV virus, or PCR for hepatitis B or C or for HIV are examples of molecular techniques that are becoming routine aspects of patient evaluation. Serial serologies are useful diagnostically, but results rarely come back in time to influence acute management. Some serologies can be useful for patients with some form of sepsis, such as West Nile IgG and IgM, or Lyme disease ELISA and Western Blot. There has also been interest in surrogate markers for the presence of sepsis. While markers such as CRP, protein C, leptin, tumor necrosis factor and procalcitonin have been studied with intriguing results, there is no clearly established role for such biomarkers in the diagnosis or management of sepsis.