Medicine is an ancient discipline, but the capacity to avert otherwise certain death is recent. Intravenous fluid therapy was first used in London during the cholera epidemic of 1832; refinements in an understanding of the role of intravascular volume in shock have altered the management of a spectrum of disorders, from multiple trauma to overwhelming infection. The development of dialysis techniques in the 1940s transformed renal failure from a rapidly lethal illness to a chronic condition. Similarly the development of techniques of mechanical ventilator support during the Scandinavian polio epidemic of the 1950s set the stage for intensive care units (ICUs) to become geographic locales capable of providing a spectrum of life-sustaining therapies —therapies whose target was not the specific cause of the illness but rather its life-threatening physiologic consequences.
The ability to avert, or at least delay, death through ICU intervention has fundamentally changed acute illness. It has allowed gravely ill or injured patients to survive conditions that in earlier times would have been lethal, but it has also created an entirely new spectrum of medical disorders that are only possible because death has been averted and whose roots lie solidly in the interventions used to accomplish that goal. Critical illness is a quintessentially iatrogenic disorder: it only arises in patients who in the absence of intervention would have died, but it is also shaped by the inadvertent consequences of that intervention. Understanding and mitigating the harmful effects of life-sustaining therapy have thus become paramount priorities because the consequences can be severe and prolonged.
Consider the following hypothetical, but uncomfortably familiar, scenario:
A previously healthy 72-year-old man is admitted to the ICU after a Hartman procedure for perforated diverticulitis. He remains intubated and paralyzed, and the plan is for overnight mechanical ventilation. He is noted to be tachycardic; 2 L of saline is given to correct a presumed fluid deficit, and the rate of his analgesic infusion is increased. The following morning he is thought to be too obtunded to consider extubation; further sedation is given to keep him comfortable, and a further fluid bolus administered when his blood pressure dips after a bolus of analgesic. Norepinephrine is administered, targeting a mean arterial pressure (MAP) of 65 and titrated up when the pressure drops but not down when it is increased. He continues taking vasopressors the next day, with a MAP of 78, but because of the use of vasopressors, it is deemed preferable to keep him sedated and ventilated for another day. Because he is receiving vasopressors and is still intubated, antibiotic coverage is broadened and a decision is made to perform a computed tomography (CT) scan to look for an intra-abdominal collection. The result of the scan is negative; he returns from the scanner on a controlled ventilator mode. Gastric residuals are high, perhaps because of ileus secondary to his illness and to the narcotics given for analgesia; bile is suctioned from the oropharynx. On the fourth day, purulent secretions are suctioned from his lung, and he has a low-grade temperature; Pseudomonas is cultured from the sputum. Review of his course so far shows that he is in 9 L positive fluid balance and still ventilator dependent; his creatinine level is twice the normal level. The clinicians decide that recovery will be slow and arrange a tracheostomy. After 32 days, several bouts of ventilator-associated pneumonia, and a short course of dialysis, he is discharged from the ICU. He is profoundly weak and has a sacral pressure ulcer. Two weeks later he is discharged from the hospital to a chronic care facility where he remains an additional 2 months.
Much of his course has been shaped by iatrogenic factors—well-meaning clinical decisions that unnecessarily prolonged his ICU stay—and set him up for new ICU complications. What if he had been extubated in the operating room? None of the individual decisions during his management was necessarily wrong, but in aggregate they increased his dependence on technologies that bring both benefit and harm; even more important, the fact that these technologies were used convinced the clinicians that they were needed. In the ICU, clinicians treat patients who are very ill, but clinicians also make them look ill and make them even more ill through the inadvertent consequences of their support.
Organ Dysfunction as an Iatrogenic Disorder
The establishment of the first ICUs in the 1950s and 1960s brought with it the emergence of new clinical syndromes whose development was only possible because patients who would otherwise have died were kept alive through the use of a spectrum of life support technologies. Initially described as syndromes reflecting derangements in a single organ system (e.g., acute respiratory distress syndrome [ARDS], septic shock, acute renal failure, disseminated intravascular coagulation), they came to be conceptualized as the manifestations of a common systemic process initially termed multiple systems organ failure, and more recently, the multiple organ dysfunction syndrome (MODS) . It is only relatively recently that clinicians have begun to realize that MODS is not only a descriptive term for the acute derangements that are the raison d’être for organ support in the ICU but also often a consequence of that support.
The Lung: From Acute Respiratory Distress Syndrome to Ventilator-Induced Lung Injury
The earliest description of pulmonary dysfunction as a consequence of remote organ injury was by Moon who, in 1948, identified congestion and atelectasis in the lungs of a cohort of patients who had died of shock. Burke and colleagues described a condition they termed high output respiratory failure that complicated the course of some patients with peritonitis who had been supported by positive pressure mechanical ventilation. In their classic report, Ashbaugh et al. termed this disorder the adult (now acute ) respiratory distress syndrome (ARDS) , drawing attention to its cardinal features: severe arterial hypoxemia and diffuse bilateral fluffy infiltrates on the chest radiograph in the face of normal left atrial pressures and in association with autopsy findings of hyaline membranes. The cornerstone of the treatment of ARDS has been mechanical ventilation (although 5 of the 12 patients described by Ashbaugh et al. did not undergo ventilation); as this therapeutic modality expanded and evolved, it became not only the treatment for but also increasingly the cause of the clinical syndrome.
The earliest events in ARDS are increased pulmonary capillary permeability and an influx of innate immune cells—largely neutrophils—into the lung. Injury to the pulmonary parenchyma and the influx of inflammatory cells activate both the local microvascular coagulation and the processes of tissue repair, resulting, in conjunction with the debris from a loss of type I pulmonary epithelial cells, in hyaline membrane formation. Although external insults such as pneumonia or contusion or internal insults such as enhanced neutrophil recruitment in the face of an activated systemic inflammatory response play an important role in the evolution of ARDS, it has become evident that readily modifiable iatrogenic factors are equally culpable.
Computerized tomography of the lung of a patient with ARDS reveals that the apparently homogeneous diffuse fluffy infiltrates seen on chest radiograph are actually evidence of a more complex lesion, including dependent atelectasis and antidependent cystic degeneration of the lung ( Fig. 47-1 ). The former changes reflect atelectasis in dependent lung zones of a patient who has been nursed for an extended period of time in the supine position, whereas the latter reflect overdistention of the lung by excessively large tidal volumes. The terminology ventilator-induced lung injury was first used in the 1990s and shifted the focus of studies of ARDS from the biochemical mechanisms that underlie lung injury to the potentially modifiable iatrogenic factors that further aggravate the initial injury. Indeed, strategies to limit dependent atelectasis by prone positioning and to minimize lung overdistention by reducing tidal volumes have shown impressive effects on reducing the mortality of ARDS. Moreover, it is striking that, of all the clinical trials of interventions for patients with ARDS, only those where the intervention sought to minimize iatrogenic harm, rather than those that sought to modulate the pathologic processes mediating that harm, have significantly affected clinically important outcomes ( Table 47-1 ).
| Alter Outcomes | Ineffective |
|---|---|
| Reduced tidal volume | Antioxidants |
| Prone positioning | Beta-2 agonists |
| High frequency oscillation | G-CSF |
| Open lung ventilation | Nitric oxide |
| Neuromuscular blockade | Ketoconazole |
| Fluid restriction | Trophic feeding |
There is, moreover, evidence from human studies demonstrating that reducing tidal volumes attenuates the systemic inflammatory response, whereas animal studies indicate that injurious mechanical ventilation strategies can induce remote organ injury in the kidney.
ARDS is defined as an acute process, but its long-term consequences are profound. In a landmark study of ARDS survivors, Herridge and colleagues reported that disability, manifested as a reduced capacity for physical activity along with residual mild derangements in pulmonary function tests, persists even 5 years after the acute illness.
Fluids and Hemodynamic Support
The capacity to correct intravascular volume deficit through the administration of intravenous fluids and the ability to titrate resuscitation through the measurement of indices of intravascular filling and myocardial function represented major advances in the care of the acutely ill. Along with the provision of invasive respiratory support, advanced hemodynamic support and monitoring represent a major indication for ICU admission. This support also carries inadvertent iatrogenic consequences.
Large volumes of intravenous fluids are characteristically administered to unstable critically ill patients, not only during the initial phases of resuscitation but also over the course of the ICU stay. A cumulatively positive fluid balance is associated with increased ICU mortality and with an increased risk of complications such as abdominal compartment syndrome. Conversely there is evidence that more conservative fluid management strategies can attenuate organ dysfunction and improve outcomes. Iatrogenic edema contributes to organ dysfunction involving the brain, heart, lung, kidney, and gastrointestinal (GI) tract; it contributes as well to the development of pressure sores. The adverse effects of interstitial edema result from several factors, including a greater distance for oxygen diffusion to reach cells and a loss of tissue compliance.
Vasoactive agents increase blood pressure by virtue of their ability to increase peripheral vascular resistance and so, potentially, to reduce tissue blood flow. For example, it has been shown that vasopressor use is an independent risk factor for anastomotic leak after GI surgery. Similarly, although inotropic agents can increase cardiac output, in large doses they also may increase mortality.
Blood transfusion is similarly a double-edged sword: lifesaving in the face of massive hemorrhage but also potentially injurious. The Transfusion Requirements in Critical Care (TRICC) trial revealed that transfusion to an arbitrary threshold of 10 g/dL resulted in increased organ dysfunction, primarily pulmonary and cardiovascular.
Sedation and Analgesia
Alleviating pain and anxiety for gravely ill patients in an ICU is a clinical and humane priority, but doing so may result in further harm. In a landmark trial, Kress and colleagues showed that daily wakening of critically ill patients enhanced the probability of survival. Others have confirmed that prolonged early sedation is associated with an increased risk of ICU mortality, and a randomized clinical trial indicated that withholding sedation resulted in a shorter duration of mechanical ventilation and ICU stay. On the other hand, inadequate control of delirium is also a risk factor for adverse outcome in the ICU. The optimal balance between attenuation of anxiety and oversedation has yet to be defined; however, emerging evidence suggests that activity, rather than rest, is central to a successful outcome from critical care.
Anti-infective Strategies
The indigenous microbial flora of the GI tract plays a key role in normal development and immune homeostasis. Conversely, derangements in normal patterns of microbial colonization are both common in critical illness and associated with adverse outcome. The normal indigenous flora of the healthy individual comprises in excess of 1000 microbial species and remains remarkably constant over time. Loss of microbial diversity is characteristic of critical illness; the proximal gut in particular shows striking patterns of pathologic colonization with the same microbial flora that predominates in ICU-acquired infections.
The mechanisms underlying normal host-microbial homeostasis in the GI tract are enormously complex. Nonetheless, studies in animal models reveal that disruption of the normal gut flora by the administration of systemic antibiotics is sufficient to induce microbial translocation from the gut lumen into regional mesenteric lymph nodes, and that changes in the composition of the gut flora can induce alterations in systemic immune responsiveness. The extent to which antibiotic-induced changes in gut flora contribute to an increased risk of ICU-acquired infection or to other derangements of critical illness is unknown.
Stress Ulcer Prophylaxis
Acute GI bleeding from gastric erosions was one of the first life-threatening derangements reported in association with care in an ICU and also a risk factor for mortality for those patients in whom it developed. Rates of stress-induced bleeding have declined since these initial reports, likely as a consequence of better resuscitation, earlier initiation of enteral nutrition, and improved diagnosis and management of infection. A legitimate question in the contemporary ICU is whether strategies to prevent stress-induced GI bleeding yield clinical benefit or whether they expose the patient to greater harm by predisposing to nosocomial ICU-acquired pneumonia or Clostridium difficile colitis.
Bed Rest
Not only is prolonged bed rest unnatural, but it is also a risk factor for complications such as atelectasis, pneumonia, and deep venous thrombosis and pulmonary embolism. Prone positioning partially corrects some of the adverse effects of prolonged bed rest and improves survival in ARDS. Similarly, a program of physical therapy and mobilization has been shown to reduce the duration of delirium and mechanical ventilation and to improve functional outcomes at the time of hospital discharge.
Related posts:
What Lessons Have Intensivists Learned During the Evidence-Based Medicine Era?
What Is the Role of Invasive Hemodynamic Monitoring in Critical Care?
Should Fever Be Treated?
Are Computerized Algorithms Useful in Managing the Critically Ill Patient?
Inhaled Vasodilators in ARDS: Do They Make a Difference?
How Can We Monitor the Microcirculation in Sepsis? Does It Improve Outcome?
Full access? Get Clinical Tree
