1. Diuretics should be administered with caution in early ARF and in response to defined physiologic problems (hypervolemia, hyperkalemia).
2. Prophylactic administration of N-acetyl cysteine and sodium bicarbonate to patients with pre-existing renal disease appears to be beneficial in preventing contrast-induced nephropathy.
3. The weight of evidence supports an increased intensity of dialysis using daily standard hemodialysis, continuous renal replacement therapy (CRRT), or extended daily hemodialysis (“slow dialysis”).
VII. ENDOCRINE ASPECTS OF CRITICAL CARE MEDICINE
A. Glucose Management in Critical Illness. Hyperglycemia is associated with an increased risk of postoperative infection (wound and otherwise) and poor outcome in patients with stroke or TBI.
1. Strict glycemic control in critically ill patients has been advocated based on evidence from a single randomized trial in surgical patients in the ICU. A subsequent study by the same investigators in medical patients found no benefit of intensive insulin therapy on mortality rate.
2. The enthusiasm for intensive insulin therapy with tight glycemic control in ICUs has diminished.
B. Adrenal Function in Critical Illness. The stress response to injury includes an increase in serum cortisol levels in most critically ill patients. Adrenal insufficiency may also occur in critically ill patients, reflecting inhibition of adrenal stimulation or corticosteroid synthesis by drugs or cytokines and direct injury to or infection of the pituitary or adrenal glands. Until free cortisol assays are more widely available, the diagnosis of adrenal insufficiency in critical illness must be based on clinical suspicion and total cortisol levels. Adrenal insufficiency should be considered in all critically ill patients with pressor-dependent shock.
C. Thyroid Function in Critical Illness. Depression of triiodothyronine (T3) occurs within hours of injury or illness and can persist for weeks. Low thyroid hormone levels, particularly for T3, correlate with the severity of illness and are associated with an increased risk of death. Hypothyroidism (elevation of thyroid-stimulating hormone in the presence of a low thyroxine level) may be present in critically ill patients, particularly in the geriatric population, and should be considered in the face of refractory shock; adrenal insufficiency; unexplained coma; and prolonged, unexplained respiratory failure.
D. Somatotropic Function in Critical Illness. Growth hormone levels are low in patients with prolonged critical illness.
VIII. ANEMIA AND TRANSFUSION THERAPY IN CRITICAL ILLNESS. The vast majority of patients admitted to the ICU are anemic at some point in their hospital stay, and more than one-third of them will receive transfused blood. The cause of anemia in critical illness is multifactorial and is related to blood loss from the primary injury or illness, iatrogenic blood loss caused by daily blood sampling, and nutritional deficiencies (e.g., folate). It is assumed that critically ill patients have less efficient compensatory mechanisms and reduced physiologic reserve and thereby require a higher hemoglobin concentration than unstressed individuals. Data collected from ICUs at multiple centers in the United States suggest that the transfusion trigger is nearer 8.6 g/dL than the previously recommended 7 g/dL. Hemoglobin is an important determinant of oxygen delivery (DO2), and transfusion is an integral component of goal-directed therapeutic strategies that aim to optimize DO2 in early shock states.
IX. NUTRITION IN CRITICALLY ILL PATIENTS. Poor nutritional status is associated with increased mortality and morbidity rates among critically ill patients; adequate nutritional support should be considered a standard of care. Enteral nutrition is preferred over parenteral nutrition whenever possible because of its lower cost and less frequent complications.
A. Complications associated with enteral feedings include aspiration of gastric feeding, diarrhea, and fluid and electrolyte imbalance. To prevent aspiration with gastric feeding, the head of the patient’s bed should be raised 30 to 45 degrees during feeding; jejunal access can be considered in patients with recurrent tube feeding aspiration. To prevent or reduce diarrhea, all potential causes should be considered and corrected.
X. SEDATION OF CRITICALLY ILL PATIENTS. Critically ill patients are often deeply sedated because of the potential benefits afforded by a reduction in the sympathoadrenal response to injury. Additionally, complications associated with undersedation include ventilator dyssynchrony, patient injury, agitation, anxiety, stress disorders, and possibly unplanned extubation.
A. Recent studies have tempered the enthusiasm for deep sedation in the ICU. (Daily interruption of continuous sedative and analgesic drug infusions has been shown to be effective in reducing the length of mechanical ventilation and length of ICU stay.)
1. The depth of sedation may also play a role in long-term outcomes after discharge from the ICU. (The extent of ICU recall, including delusional memories, is a function of the extent of sedation.)
2. It is important to titrate medications according to established therapeutic goals and re-evaluate sedation requirements frequently (Ramsay Sedation Scale, Richmond Agitation Sedation Scale).
B. Confusion and agitation are common in ICU patients and may have unfavorable consequences on patient outcome. Agitation should be distinguished from delirium, which is relatively common in ICU patients and equally associated with increased length of stay, morbidity, and mortality. The distinguishing characteristics of delirium include acute onset and fluctuating course, inattention, disorganized thinking, and altered level of consciousness.
C. Nonpharmacologic and pharmacologic means can be used to provide comfort and safety to ICU patients. The former include communication, frequent reorientation, maintenance of a day–night cycle, noise reduction, and ensuring ventilation synchrony. Pharmacologic agents include hypnotics–anxiolytics, opioids, and antipsychotics.
1. The most commonly used hypnotics are propofol, midazolam, and lorazepam. Continuous infusion of propofol is associated with a shorter length of mechanical ventilation and ICU stay compared with lorazepam administration.
2. Dexmedetomidine has been effectively used as a single agent or in combination with other drugs in postsurgical and medical ICU patients.
3. Opioids. Morphine and fentanyl are the most commonly used opioids to provide analgesia in the ICU. Morphine should be avoided in patients with renal failure because of active metabolites that accumulate in the presence of impaired renal elimination.
4. Neuromuscular blockade may be occasionally indicated in ICU patients with severe TBI or respiratory failure, but routine use is discouraged because of concerns that this practice may predispose patients to critical illness, polyneuropathy, and myopathy and because of an increased risk of nosocomial pneumonia in patients receiving these agents.
XI. COMPLICATIONS IN THE INTENSIVE CARE UNIT: DETECTION, PREVENTION, AND THERAPY
A. Nosocomial infections are a major source of morbidity and mortality in critically ill patients. At some level, nosocomial infections are unavoidable and occur because of the nature of intensive care: patients are critically ill with altered host defenses; they require invasive devices (endotracheal tubes, intravascular catheters) for support, monitoring, and therapy that provide portals of entry for infectious organisms; and they receive therapies that increase the risk of infection (glucocorticoids, parenteral nutrition). On the other hand, many nosocomial infections are preventable with relatively simple interventions.
1. Sinusitis is common in critically ill patients with indwelling oral and nasal tubes. Prevention of sinusitis should focus on efforts to improve sinus drainage, including semirecumbent positioning and avoidance of nasal tubes. Bacterial sinusitis should be considered in patients with unexplained fever and leukocytosis in the ICU.
2. Ventilator-associated Pneumonia (VAP). Endotracheal intubation and mechanical ventilation increase the risk of VAP. Interventions that can reduce the incidence of VAP include strict hand washing before and after patient contact and semirecumbent positioning of the patient (head height at 30 degrees or greater from horizontal; level II evidence). These practices should be rigorously applied in all ICUs (granted that semirecumbent positioning is not possible in all patients).
a. Gastric and oropharyngeal colonization with resistant organisms appears to be a risk factor for the development of VAP. Oral decontamination with chlorhexidine (not nonabsorbable antimicrobial agents) appears to reduce VAP rates without leading to excess antibiotic resistance.
b. Acid-suppression therapies as prophylaxis against gastrointestinal (GI) bleeding have been associated with an increased risk of VAP because they allow bacterial overgrowth in the stomach. (Sucralfate should be considered as an alternative agent to acid-suppressive regimens despite its potentially reduced effectiveness.)
c. An important approach to reduce the overall mortality from VAP involves refinement of the diagnostic process and limitation of antibiotic therapy to avoid the development of antibiotic resistance. Antibiotics can then be narrowed in spectrum or discontinued depending on the results from quantitative cultures after 48 to 72 hours.
3. Catheter-related bloodstream infections are defined as clinical suspicion of catheter-related infection plus positive culture results for blood drawn from the catheter or of a segment of catheter and matching positive blood culture drawn from another site.
a. Catheter infection is more likely when placement occurs under emergency conditions and is reduced by the use of strict aseptic technique with full barrier precautions. Catheter-related infection and bacteremia increase with the duration of catheterization, particularly for durations >2 days. However, routine catheter replacement at 3 or 7 days does not reduce the incidence of infection and results in increased mechanical complications. Thus, routine guidewire change of catheters is not recommended.
b. Based on incidence of infection at the insertion site, the subclavian route should be used when possible if the duration of catheterization is predicted to be longer than 2 days.
c. Catheter-related venous thrombosis occurs commonly and is associated with an increased risk of infection. Routine flushing of catheter ports with heparin reduces the incidence of both thrombosis and infection.
d. When catheter-related bacteremia is confirmed, the offending catheter should be removed and appropriate antibiotics continued for a minimum of 7 days.
4. Urinary tract infection is the second most common source of infection in ICUs. Its incidence increases with the duration of bladder catheterization.
5. Invasive fungal infections in nonneutropenic patients are caused by Candida spp. in the vast majority of cases. These infections are increasingly common in the ICU population, accounting for 5% to 10% of all bloodstream infections in the ICU. A high level of suspicion for invasive Candida infection in critically ill patients is necessary, and “preemptive” therapy should be considered in patients with a high likelihood of invasive Candida infection while awaiting blood culture results. An ophthalmologic examination is warranted in patients with documented or suspected bloodstream infection because patients with endophthalmitis may require longer courses of therapy. Intravascular catheters that are potential sources of bloodstream infection should be removed. Organisms sensitive to the azole derivative fluconazole cause the majority of invasive Candida infections in the ICU, and fluconazole is the first-line treatment given its reasonable efficacy and limited toxicity.
B. Stress Ulceration and Gastrointestinal Hemorrhage. Gastric mucosal breakdown with resulting gastritis and ulceration (“stress ulceration”) can lead to GI bleeding in ICU patients. The major risk factors for stress-related GI bleeding are mechanical ventilation and coagulopathy; secondary risk factors among mechanically ventilated patients include renal failure, thermal injury, and possibly head injury. Enteral nutrition may protect against significant GI bleeding.
1. Prevention. Agents used to prevent stress ulceration and GI bleeding include methods to suppress acid production (H2 blockers and proton pump inhibitors) and cytoprotective agents (sucralfate). The agent of choice–and whether any prophylaxis is beneficial or indicated–is somewhat controversial.
2. It appears that stress ulcer prophylaxis is more widely used than necessary. Thus, although stress ulcer prophylaxis, predominantly with ranitidine, is commonly used in critically ill patients, the utility of this intervention is unclear.
C. Venous thromboembolism (VTE) occurs frequently in critically ill patients, with incidences of deep venous thrombosis (DVT) of 10% to 30% and of pulmonary embolism (PE) of 1.5% to 5%. In addition to classic lower extremity DVT, upper extremity DVT occurs with increased frequency in the ICU population. This is directly associated with the use of central venous catheters in the subclavian and internal jugular sites. Upper extremity DVT can result in pulmonary embolism in up to two-thirds of cases, with occasional fatalities.
1. Prophylaxis. The risks of VTE prophylaxis, including heparin-induced thrombocytopenia and bleeding, must be weighed when considering prophylaxis in the ICU population. Nonetheless, it is generally agreed that high-risk patients without contraindications should receive prophylaxis with low-molecular-weight heparin (LMWH) and that patients with low to moderate risk should receive low-dose unfractionated heparin (UFH) (Table 55-7). To reduce central venous catheter–associated thrombosis and infection, catheter tips should be positioned in the superior vena cava, and catheters should be flushed with a dilute heparin solution. Heparin bonding of catheters may also reduce local thrombosis.
2. Diagnosis. Despite the high incidence of DVT, routine screening studies for DVT do not appear to improve clinical outcomes in the ICU. VTE should be considered in critically ill patients in the face of relatively nonspecific findings, such as unexplained tachycardia, tachypnea, fever, asymmetric extremity edema, and gas exchange abnormalities (including high dead space ventilation). Compression Doppler ultrasonography is the most commonly used test for diagnosis of DVT. Ventilation–perfusion scanning and pulmonary angiography may have utility in specific circumstances, including in the presence of renal insufficiency (concerns about contrast-induced nephrotoxicity) or equivocal results on computed tomography scan. Pulmonary angiography may be the test of choice when the likelihood of PE is high and anticoagulation is contraindicated, necessitating immediate placement of a vena cava filter.
TABLE 55-7 RISK FACTORS FOR VENOUS THROMBOEMBOLISM
Strong Risk Factors
Fracture (hip or leg)
Hip or knee replacement
Major trauma
Spinal cord injury
Moderate Risk Factors
Arthroscopic knee surgery
Central venous lines
Chemotherapy
Congestive heart or respiratory failure
Hormone replacement therapy
Malignancy
Oral contraceptive therapy
Paralytic stroke
Pregnancy or postpartum period
Previous venous thromboembolism
Thrombophilia
Weak Risk Factors
Bed rest >3 days
Immobility caused by sitting (prolonged car or air travel)
Increasing age
Laparoscopic surgery (cholecystectomy)
Obesity
Pregnancy or antepartum period
Varicose veins