Our favorite unproven ideas for future critical care

The future of critical care medicine will be shaped not only by the evidence-validated foundations of science, but also by innovations based on unproven and, in many cases, untested concepts and thoughtful visions of scientists and clinicians familiar with the complex problems actually faced in clinical practice. Clinical investigations and trials often lag behind collective experience and impressions, in a well-intentioned and necessary quest to determine the fallacy or validity of ongoing practice. Progress made in this way can be painfully slow, and imperfect theory may prove difficult to challenge. On occasion, an innovative paradigm shift fostered by a novel approach can reorient the forces of academic investigation toward generating an evidence base upon which such concepts and interpretations can find scientific justification. This discussion presents a selected set of ideas to improve the future practice of critical care - each having a defensible rationale, but unconfirmed validity.


Introduction
While technical innovations and scientifi c experiments often lead clinical practice, at times clinical science lags behind experience and simply documents what appears to be the fallacy or validity of ongoing practice. Progress made in this way can be slow, and imperfect theory may prove diffi cult to challenge. What follows is a wideranging set of unproven ideas, each having a defensible rationale but unverifi ed worth. Th e intent of this thoughtprovoking format is to stimulate thinking and interchange, and perhaps to point toward new directions with potential for productive research, conceptual modi fi cation, and eventual applications that could improve care of the critically ill. In what follows, a brief back ground outlines the rationale for each novel and unproven idea endorsed by the presenter.
We should retire the stethoscope (J-L Vincent)

Background
Dating from the days of Laennec, the stethoscope has proven itself a robust and helpful diagnostic instrument [1]. Chief among the instrument's attractive features are its portability, simplicity, ubiquity, low cost, and ability to complement the physical examination by opening an immediate acoustic window into the functional status of the lungs (for example, wheezes, rales) and heart (gallops, murmurs). It is a testimony to past contributions that this tool has not changed fundamentally in more than two centuries. Traditional stethoscopy does have drawbacks, however. Unrecorded sounds heard through the stethoscope are evanescent, the interpretation of the nature, timing, and amplitude of breath sounds is subjective, and high variability between observers is well documented [2]. Although acoustic information can be digitally processed as well as displayed [3], traditional stethoscopic information is descriptive rather than quantitative. Th e value of auscultation (similar to the unaided physical examination) depends on the skill of the individual and the completeness of reporting [4].
Th e stethoscope was particularly useful as an aid for physical examination when bedside imaging either was not available or was diffi cult to interpret. Not only do we now have better quality radiographs, but we also have echocardio graphy and ultrasound that can be quickly and repeatedly deployed to yield detailed and quantitative imaging infor mation [5]. Bed side computed tomography scanners and even magnetic resonance imagers are not far away from use. A continuous data-stream that yields dynamic functional monitoring may be aff orded by emerg ing technologies such as electrical impedance tomography and acoustic monitoring. By comparison, the venerable stethoscope seems dated and crude [6].
Nonetheless, the no-cost, instantly available stethoscope will continue to have indications, especially when immediate availability and feedback are at a premium. For example, auscultation aids in making sure that an endotracheal tube is in the right track or not placed too far down (decreased breath sounds in one lung). Another use is the classifi cation of very acute respiratory failure: Is the problem tube obstruction? Bronchospasm? Pneumothorax? Hemothorax? While not defi nitive, auscultation

Abstract
The future of critical care medicine will be shaped not only by the evidence-validated foundations of science, but also by innovations based on unproven and, in many cases, untested concepts and thoughtful visions of scientists and clinicians familiar with the complex problems actually faced in clinical practice. Clinical investigations and trials often lag behind collective experience and impressions, in a well-intentioned and necessary quest to determine the fallacy or validity of ongoing practice. Progress made in this way can be painfully slow, and imperfect theory may prove diffi cult to challenge. On occasion, an innovative paradigm shift fostered by a novel approach can reorient the forces of academic investigation toward generating an evidence base upon which such concepts and interpretations can fi nd scientifi c justifi cation. This discussion presents a selected set of ideas to improve the future practice of critical care -each having a defensible rationale, but unconfi rmed validity.
De-stressing therapy: blockade of the β-adrenergic system benefi ts mortality, metabolism, and the immune system (M Singer)

Background
Our innate responses may be ill adapted to lifethreatening illness. One clear possibility is that the native hormonal reaction to such challenges may be noxious, perhaps even accelerating injury processes in the wrong direction [8,9]. Physiological adaptations to catastrophic and therefore unsurvivable conditions are not likely to have evolved naturally. One could even argue that evolution would have selected against those genes that would have allowed the organism to experience trauma, infection, or serious acute organ dysfunction. Even the fi ght or fl ight response may be excessive.
Overly exuberant upregulation of the β-adrenergic system in response to stresses of various types (for example, cardiovascular crises, sepsis, massive wounding) has been shown to produce necrosis of vital tissues [10]. Published clinical trials demonstrate the potential for β-blockade to reduce myocardial damage following coronary occlusion. Sudden emotional stress may produce dramatic cardiac dysfunction, which persists for weeks afterwards (Takotsubo cardiomyopathy) [11]. βadrenergic blockade, while contraindicated during an acute attack, may reduce sensitivity to methacholine challenge in the setting of chronic asthmatic disease. Continuation of intense β-stimulation has the opposite eff ect. β-blockade has experimentally been shown to improve survival in rats challenged with intravenous lipopolysaccharide, and β-blockers are most eff ective if given before such a challenge [12]. Such observations suggest that interference with hardwired adrenergic responses to physiologic crises (such as severe sepsis) by modulation of their intensity could be a productive therapeutic avenue to pursue.

Idea
Rather than relying on high-dose catecholamine-based inotropic/vasoactive agents, using noncatechol inotropes (for example, milrinone) in conjunction with closely modulated β-blockade may be a rational approach to supporting the cardiovascular system during the stress of severe sepsis.

Background
Many patients survive the ICU only to face a protracted period of recovery, permanent disability, or eventual death from immobility. Recent data reveal more than 40% of the 6-month mortality following critical illness occurs after the patient has been discharged from the ICU [13]. Many of these deaths are believed to occur indirectly as a result of catabolism, loss of lean body mass, and lack of adequate physical activity, and to ultimately culminate in profound weakness and an inability to mobilize [14,15]. Many patients report very poor physical function for the entire fi rst year following ICU discharge [14] and these physical dis abilities have been shown to persist for 5 years or longer following the ICU discharge. Survivors of our advanced ICU care are thus often signifi cantly debilitated for months or years afterwards -a process known as ICU-acquired weakness. Many ICU survivors suff er rapid, dramatic loss of lean body mass, causing signifi cant debility that leaves them predisposed to aspiration, pneu monia, and other life-threatening compli cations [15]. Th e patient facing the abrupt stresses imposed by critical illness has not had time to evolve appropriate -and in this case adap tive -responses.
Th e ICU experience undergone by many normal individuals is not unlike the stresses encountered by elite athletes during training and competition. Critical illness progresses through acute, chronic, and recovery phases, each with their diff erent nutritional demands and suscepti bilities. Before competition, many of the most successful competitors take anabolic, performanceenhancing agents for extended periods that allow them to cope with the stresses of performing exceptionally at the time of their events. Elite strength-training athletes legally utilize high-quality protein (typically 2.0 g/kg/day), glutamine, fi sh oil (an anti-catabolic), hydroxy methyl butyrate, and creatine. As has been well described in the popular press in sports such as baseball and cycling, many athletes use nonsanctioned androgens (testosterone derivatives) and growth hormone, especially when trying to speed recovery after intense training and injury. Other anabolic hormones, such as insulin and insulinlike growth factor-I, are often stacked together with androgens and growth hormone. Despite risks of unsupervised use, performance enhancers are used because they work.
In recent years, common use of such agents has resulted in world and Olympic records being broken at an unprecedented pace by superhuman athletes. For athletic applications, eff ective dosing and timing have been worked out. Medical use of these agents, outside the burn unit [16,17], has been quite rare in the ICU setting. Further research is therefore needed to optimize the use of these agents to help our patients recover from the ultimate race -critical illness. In a sense, evolving to survive critical illness and to function again normally would be superhuman, and we should learn from athletes who perform superhuman feats daily to help our patients recover from what Mother Nature would have previously deemed unrecoverable.

Idea
Use well-timed nutritional and hormonal interventions to improve immunological and neuromuscular responses so as to aid healing and prevent debility. Use high-dose protein (1.5 to 2.0 g/kg/day), exercise (that is, early ambulation of ventilated patients), and hydroxymethyl butyrate during the acute phase, with no potent anabolics. In the later phases, use suffi cient calories and high-quality protein, β-blockers, and hydroxymethyl butyrate complemented by oxandrolone and creatine. In the recovery phase, add low doses of growth hormone to this prescription. (Following the acute phase of burns, approaches using β-blockers and oxandrolone have been quite success ful, provided that early and adequate protein/ calorie delivery is achieved [16,17].) Delivering suffi cient calories and protein in the chronic and recovery phases of critical illness is mandatory if oxandrolone, growth hormone, or other anabolic agents are to be both safe and eff ective.

Background
Critical care evolved from anesthetic approaches and postoperative practice. In elective cases, the sudden trauma of surgery interrupts resting physiology and is addressed by an attempt to restore a normal baseline. In the critical care setting, however, the price of maintaining normality may be injury infl icted by the therapy itself (ventilator-induced lung injury, drug toxicity, consequences of sedation, and so forth). Equally plausible is that our enforcing of normal indicators and protracted use of excessive support may impede the patient's potential to adapt to the abnormal physiology that characterizes chronic critical illness.
Health is characterized by gradual transitions between physiological states, by variation of cardiorespiratory indices, and by continual adaptation to biochemical, environ mental, and mechanical stressors. Acute disease, on the contrary, is often characterized by abrupt transitions, failure to adapt to the stressor, and monotonously infl exible physiological patterns [18]. Currently, our treatments rescue very eff ectively, but when sustained for too long they may frustrate the patient's innate adaptability to stressful but potentially survivable illness.
For example, early use of short-term muscle relaxants enables lung-protective strategies and may reduce mortality risk in acute respiratory distress syndrome [19], but when sustained for too long muscle relaxants can cause muscular atrophy and sustained weakness [20]. Because evolution did not provide for appropriate responses to severe acute injuries, an argument can be made that we must moderate the initial assault and response, and then build our capability to respond to lingering challenges over time.
During health, tolerable stresses above the resting baseline that are applied and released build strength, promote tolerance, and remodel tissues. Endurance athletes are a good example of what can be achieved with intermittent exposure to challenging stress. Stressinduced adaptations include enhanced pulmonary oxygen exchange, increased blood volume and hemo globin concentration, enriched blood fl ow to skeletal muscle, improved thermal regu lation, increased mitochondrial size and density, and increased capillarization of muscular beds. Th e healthy individual is also extremely adaptable when the physio logic milieu is gradually changed. Important lessons regarding adaptation to abnormal physiology have been learned from protracted exposure to hypoxic inspired gas, both in mountain climbing and in the experimental laboratory [21].
Functionally adaptive response needs some time to develop. For example, after exposure to a thermal stress, heat shock proteins require approximately 18 hours to be expressed to a level that protects against experimental ventilator-induced lung injury and sepsis [22]. Conversely, abrupt and synchronous exposure to the same high temperature may accelerate ventilator-induced lung damage, whereas cooling inhibits its expression [23]. Slowly achieving a given target is protective in other settings. For example, impressive depths of hypoxemia can be tolerated and the dangers of hyperoxia avoided by graded exposure to permissive hypoxemia [24]. Th e amplitude and timing of these conditioning challenges to the resting baseline and to the intermittent stresses are keys to tolerance and adaptation. Important unanswered questions related to conditioning of the critically ill patient relate to whether injured tissues are capable of stress conditioning and when in the course of the illness is it advisable to time the fl ip from the full support of rescue to a goal-directed conditioning pattern.

Idea
A two-stage approach of initial rescue -that emphasizes minimizing demand, full support, gentle transitions, and assuming physiological control -followed by an adaptive phase -characterized by ongoing targeted reduction of vital supports and imposed intermittent stresses and rest periods, akin to the endurance and strength-building programs of athletic training -should be applied. (Such acclimatization and conditioning might be applied to the most common and unchallenged elements of acute care, such as the fraction of inspired oxygen, ventilating pressures, vaso pressors, and positioning.) Expressed poetically: 'First shelter the critically ill patient during the storm and then teach the patient to dance in the rain. '

Background
Awareness of the potential for ventilator-caused lung damage has grown over the past 25 years. Clinicians now appreciate that excessive airway pressures are dangerous, particularly when large tidal volumes are employed and insuffi cient positive end-expiratory pressure is used to keep the lung from experiencing high-pressure tidal opening and collapse cycles [25]. Th e healthy portions of the injured lungs have relatively normal mechanical properties [26]. A fi xed tidal volume crammed into a smaller lung will accentuate stress and strain. Th e clinically measurable correlates of resting lung volume and stretched lung volume are the functional residual capacity and the combination of the volume attributable to positive end-expiratory pressure and tidal volume [27].
Normal lungs begin to lose their integrity, however, when the gas volume of the given sector exceeds its resting value by a factor of 2.0 to 2.5. Such an extreme expansion of dimensional sizes is rarely seen in uniform lungs in the clinical setting [28]. However, inter dependence amplifi es local forces at the interface between open tissue and collapsed tissue, a concept fi rst brought to attention more than four decades ago in the work of Mead, Takashima, and Leith [29]. At these stress riser points, the alveolar-capillary membrane can be stressed suffi ciently to break down. When this membrane does break down, the rate of deterioration occurs with impressive speed -presumably because new stress risers are generated and the aerated lung size shrinks. Th e relatively poor correlation across diff erent patient populations between tidal volume and outcome, even when the latter is adjusted to predicted body weight, perhaps occurs for these reasons. If the number of stress riser points can be minimized, therefore, the lung will be better protected and catastrophic progression of lung injury may be avoided or slowed.

Idea
Promote uniformity and minimize the generation of stress rising points of heterogeneity by measures such as prone positioning (which reduces the gradient of transalveolar pressure) and recruiting the lung.
Monitoring of the injured lung should be conducted with methods that are sensitive to regional heterogeneitysuch as computed tomography and possibly electrical impedance tomography, ultrasound, and acoustic mapping -rather than exclusively by the global methods for monitoring lung mechanics and gas exchange in use today.

Background
Tissue health depends on adequacy of perfusion and oxygenation, not always well refl ected by global measures of cardiac output and oxygen extraction. In the setting of sepsis, the endothelium and vasculature are under assault by activated leukocytes, infl ammatory mediators, and reactive oxygen species that cause microcirculatory dysfunction in advance of organ failure [30]. A compromised microcirculation is no longer able to regu late blood fl ow distribution, resulting in functional shunting where the oxygen need of the parenchymal cells is not met by adequate delivery [31]. Th is condition describes distributive shock, one of the four states of shock defi ned by Weil and Shubin [32]. Early restoration of perfusion and oxygenation in addition to correction of the imbalances set into motion by endo thelial damage, inducible nitric oxide synthase, endo the lial nitric oxide synthase, and reactive oxygen species appear to improve outcomes [33]. Measures to reopen the microcirculation include boosting of hematocrit and oxygen transport, increasing the intravascular volume, and control of oxidative stress and infl ammatory load.
An extensive body of scientifi c data collected in an experimental setting documents such complex biochemical interplay. Clinical data demonstrate that early interventions targeting increased global perfusion improve outcomes and lower Sequential Organ Failure Assessment scores. Although the value of such an approach has been established for surgical patients, whether interventional targeting works as successfully for critically ill septic patients remains controversial. Th e ultimate objective, however, is to oxygenate the microcirculation by improved perfusion with blood that is optimally charged to off -load oxygen by maintaining adequate circulatory dimensions, fi ne-tuned vasoregulation, and improved plasmatic composition. Th e latter objective could conceivably be helped by using balanced starches to promote convection and correct hypovolemia. Blood transfusions, in addition to increasing oxygen-carrying capacity, also recruit previously unfi lled capillaries and shorten diff usion distances, thereby promoting oxygen transport to tissue [34]. To reverse the consequences of the shock cascade, optimal perfusion must be accomplished at the local microcirculatory level, not simply by improving the performance of the heart and charging the blood with high oxygen content.
In recent years, a variety of techniques that monitor microscopic blood fl ow and its consequences have been employed in the laboratory setting. Th ese have been brought closer to readiness for clinical implementation [35]. With such technology, the potential effi cacy or adverse consequences of drugs and fl uids can be monitored in representative tissue beds. Th e tools used to implement a new paradigm based on controlling infl ammation, resuscitating endothelial cells, and modulating reactive oxygen species will require such monitoring instruments.

Idea
Target the microcirculation as the primary objective. Directly visualize representative microcirculatory beds to monitor progress toward the goal of achieving adequate perfusion and oxygenation. Utilize the drugs and fl uids that are eff ective in oxygenating and perfusing the microcirculation (goal-directed fl uid therapy plus blood transfusion), opening the microcirculation (for example, sodium nitrite, a pure nitric oxide donor with potential for releasing nitric oxide from hypoxic red blood cells in compromised tissues), preserving the cellular glycocalyx (for example, vitamin C), and providing potent actions at the microcirculatory level.
Employ holistic care to take advantage of a missed opportunity (TJ Gan)

Background
As practiced in advanced industrial societies, western medicine leverages scientifi cally-based therapies that emphasize artifi cial procedures and drugs. ICU medicine has been an avid proponent of management that stresses measurement and correction. Traditional eastern practices, on the contrary, embrace a variety of treatments that integrate complementary and alternative measures (CAM) into a less methodically documented holistic approach that conceives of health as a tuned balance of body, mind, and spirit. In recent years, the importance of engaging these missing spheres has also been recognized in the outpatient practices of the advanced industrial world. Diet, exercise, relaxation, massage, prayer, music, and closer involvement of the family -methods that are outside established standards for ICU practice -have increasingly been integrated into intensive care management settings [36][37][38]. In the outpatient environment, CAM have been recognized as helpful in relieving stress, anxiety, discomfort, restlessness, and insomnia. Th e existing literature strongly backs the eff ectiveness of CAM for attenuating pain and anxiety; patients have generally been receptive and ultimately satisfi ed with their use [39][40][41]. Demands for inpatient application of CAM are increasingly strong, as they confer a measure of treatment control to the patient and family, and emphasize patient well-being and preferences during their hospital stay. Where implemented in the inpatient setting, CAM methodology has been viewed as clinically eff ective by patients and physicians alike [38].
Th e patient with critical illness faces major problems related to sleep, anxiety, and pain [40,41]. Th ese challenges are severe enough to warrant sedation and analgesia in doses that may ultimately compromise progress or cause complications. Recovery may be slow and arduous. CAM methodology is generally free of lasting consequences other than lingering relaxation and reduced pain. Acknowledging a strong recent emphasis on avoiding delirium, a holistic approach is an attractive option to complement our current measures for intensive care [42].

Idea
Wake patients earlier, use fewer drugs, and incorporate CAM into ICU practice of the alert and semiconscious patient. Use these techniques routinely to accomplish eff ective sleep, pain relief, and the sense of well-being that will accelerate rehabilitative progress and reduce the consequences of drug-based management.

Conclusion
Th e future of critical care medicine will be shaped not only by the evidence-validated foundations of science, but by innovations based on unproven and, in many cases, untested concepts and thoughtful visions of scientists and clinicians familiar with the complex problems actually faced in clinical practice. On occasion, an innovative paradigm shift fostered by a novel idea can reorient the forces of academic investigation to generate an evidence base upon which such concepts and interpretations can fi nd scientifi c justifi cation. Intriguing as this out-of-the-box thinking may be, however, such visions should not be implemented until supported by a substantial body of basic and clinical research.