The programs listed are provided for reference purposes only. They were current when produced, but are no longer maintained and may now be outdated. Persons with disabilities having difficulty accessing information on this page may contact us for assistance. Please select the ODP's home page to access current information.
Critical Care Medicine

National Institutes of Health
Consensus Development Conference Statement
March 7-9, 1983

Conference artwork, a black and white photo of a bed in a critical care unit with the title above.

This statement is more than five years old and is provided solely for historical purposes. Due to the cumulative nature of medical research, new knowledge has inevitably accumulated in this subject area in the time since the statement was initially prepared. Thus some of the material is likely to be out of date, and at worst simply wrong. For reliable, current information on this and other health topics, we recommend consulting the National Institutes of Health's MedlinePlus

This statement was originally published as: Critical Care Medicine. NIH Consens Statement 1983 Mar 7-9;46):1-26.

For making bibliographic reference to the statement in the electronic form displayed here, it is recommended that the following format be used: Critical Care Medicine. NIH Consens Statement Online 1983 Mar 7-9 [cited year month day];46):1-26.


A Consensus Development Conference was held at the National Institutes of Health on March 7-9, 1983, to discuss issues related to the practice of critical care medicine.

NIH Consensus Development Conferences bring together investigators in the biomedical sciences, practicing physicians and other health professionals, and representatives of the public to provide a scientific assessment of medical technologies and to develop a consensus statement on their safety and effectiveness.

On the first two days of the meeting, a Consensus Development Panel and members of the audience heard scientific presentations from a distinguished group of medical scientists. The panel then considered the following questions:


  1. Is there empirical evidence that intensive care units (ICUs) cause a decrease in patient morbidity or mortality? Which patients are most likely to benefit from intensive care?
  2. What skills are essential for personnel in a critical care unit? How should this personnel be trained and organized to assure the best care for patients most in need?
  3. What special technology and therapeutic interventions should be routinely available for the most effective ICU function?
  4. How is a hospital's critical care delivery system best structured: one large multispecialty unit or multiple small subspecialty units?
  5. How has the development of ICUs affected the traditional functions of a hospital?
  6. What direction should critical care research follow?


Members of this panel included biomedical investigators, critical care physicians, other medical specialists, nurses, a biostatistician, and a jurist.

Critical Care Medicine (CCM) is a multidisciplinary and multiprofessional medical/nursing field concerned with patients who have sustained or are at risk of sustaining acutely life-threatening single or multiple organ system failure due to disease or injury. These conditions necessitate prolonged minute-to-minute therapy or observation in an intensive care unit (ICU) which is capable of providing a high level of intensive therapy in terms of quality and immediacy.

In its broadest meaning, CCM includes management at the scene of onset of critical illness or injury, during transportation, in the emergency department, during surgical intervention in the operating room, and finally in the ICU. However, this consensus report is limited to CCM as it relates to the adult patient's management in the ICU. The report focuses not only on units that offer the full range of CCM but also on units providing a narrower range of critical care services. However, pediatric and neonatal CCM, the care of patients with burns, and all forms of extra-hospital care are not addressed in this report. Although these units are excluded from this report, valuable similarities with other ICUs in their organization and structure should not be ignored.

While critical care may be considered a higher level of management than intensive care, the two terms are used synonymously in this report.

Is There Empirical Evidence That Intensive Care Units Cause a Decrease in Patient Morbidity or Mortality? Which Patients Are Most Likely To Benefit From Intensive Care?

The estimation of the efficacy of ICUs with respect to mortality and morbidity can be made only in the context of the patient populations admitted, the objectives for admission, the interventions which are being employed and the alternate patient management systems to which comparisons could be made. Mortality and morbidity may refer to events occurring in the ICU, hospital, or during a longer period of followup. Uniform goals and results for all ICUs are not attainable or desirable.

Historically, intensive care units were organized to serve specific purposes such as rhythm monitoring, the care of postoperative patients, airway maintenance and mechanical ventilation for patients with reversible neurological disease. The highly favorable outcomes derived from such specialized care served as the stimulus for establishing large numbers of such units. Over the past two decades, the availability of physical resources, nursing staff, and related specialized procedures, as well as patients' expectations, have resulted in an expansion of the original indications for admission to categories of patients for whom the achievable benefits are less clear.

There is empirical evidence that interventions commonly restricted to areas designated as intensive, coronary, or critical care units result in a decrease in mortality or morbidity. However, evidence for such a benefit can be considered unequivocal for only a portion of the heterogeneous patient populations currently admitted to such units. For a larger proportion of patients, evidence is equivocal but the weight of clinical opinion is that ICU care improves survival. It is recognized, however, that for some patients the risk of iatrogenic illnesses associated with ICU care may outweigh any potential benefit.

This spectrum of ICU effectiveness for different patients can be better understood by considering some typical patient categories.

First is the patient with acute reversible disease for whom the probability of survival without ICU intervention is low, but the survival probability with such interventions is high. Common clinical examples include the patient with acute reversible respiratory failure due to drug overdose, or with cardiac conduction disturbances resulting in cardiovascular collapse but amenable to pacemaker therapy. Because survival for many of these patients without such life-support interventions is uncommon, the observed high survival rates constitute unequivocal evidence of reduced mortality for this category of ICU patients. These patients clearly benefit from ICU care.

Another group consists of patients with a low probability of survival without intensive care whose probability of survival with intensive care may be higher--but the potential benefit is not as clear. Clinical examples include patients with septic or cardiogenic shock. The weight of clinical opinion is that ICUs reduce mortality for many of these patients, though this conviction is supported only by uncontrolled or poorly controlled studies. Often these studies do not allow one to distinguish between ICU effectiveness and the effects of patient selection and/or differences in cointerventions that do not require the ICU.

A third category is patients admitted to the ICU, not because they are critically ill, but because they are at risk of becoming critically ill. The purposes of intensive care in these instances are to prevent a serious complication or to allow a prompt response to any complication that may occur. It is presumed that the prompt response to a potentially fatal complication made possible by continuous monitoring plus the concentration of specialized personnel in the ICU increases the probability of a favorable outcome. The risk of complication may be high (as in the patient with an acute myocardial infarction and complex ventricular ectopy) or low (as in the patient with myocardial infarction suspected because of chest pain in the absence of electrocardiographic abnormalities). Also, the differences in probability of a favorable outcome following a complication inside rather than outside the ICU may be large (as in the patient with postcraniotomy intracranial bleeding) or small (as in the patient with gastrointestinal bleeding). The strength of evidence supporting the effectiveness of the ICU varies with the probability of a complication and with the difference in expected outcome inside and outside the ICU. When the risk of complication is high and the potential gain large, a decrease in mortality is likely. Similarly, when the risk is low and the potential gain small, an observable decrease in mortality is unlikely. These patients are not likely to benefit from ICU care.

The risk of iatrogenic morbidity and mortality for all patients must be included in consideration of ICU effectiveness. Complications associated with major interventions are not infrequent despite the concentration of skilled personnel in the intensive care unit. More subtle ill effects, including anxiety and psychiatric disturbances, are common and may increase the incidence of complications for which patients are monitored (e.g., ventricular arrhythmias in the coronary patient). Technical difficulties, errors in interpretation, increasing interventions induced by continuous monitoring, facilitated by immediate availability of personnel and equipment, are potential hazards for the monitored patient. Iatrogenic illness rates in the ICU are not known with any precision. This gap in our knowledge contributes substantially to our uncertainty about the effectiveness of ICU care.

A difficult clinical problem in the allocation of ICU resources is the disposition of patients with a very low probability of survival despite optimal ICU care.

It is not medically appropriate to devote limited ICU resources to patients without reasonable prospect of significant recovery when patients who need those services, and who have a significant prospect of recovery from acutely life-threatening disease or injury are being turned away due to lack of capacity. It is inappropriate to maintain ICU management of a patient whose prognosis has resolved to one of persistent vegetative state, and it is similarly inappropriate to employ ICU resources where no purpose will be served but a prolongation of the natural process of death.

Coronary care units (CCUs) deserve separate consideration as they represent the largest group of specialized intensive care units. These units appear to have improved survival of patients hospitalized with acute myocardial infarction. This change relates to the traditional function of CCUs; i.e., observations of cardiac rhythm, prompt recognition of life threatening arrhythmias, and appropriate drug and electrical cardioversion therapy. Since mortality within the first 24 hours following infarction is due primarily to ventricular fibrillation and since this arrhythmia (or "warning arrhythmia") is more likely to be treated successfully in monitored units than in general services, the assumption is made that the current relatively low hospital mortality from myocardial infarction is due to CCUs.

Considerable knowledge of the natural history and pathophysiology of acute myocardial infarction has accrued as a result of CCUs. Much of this relates to risk stratification based on clinical observations, demographic and hemodynamic factors, and characterizations of electrocardiographic phenomena, etc. Principles of therapy have evolved which reflect the heterogeneous nature of the disease; e.g., short-term, nonintervention care for low risk patients; aggressive, invasive medical/surgical therapy for high risk patients or for those who develop complications of myocardial infarction such as cardiogenic shock. The use of high risk procedures such as coronary angiography or intra-aortic balloon counterpulsation in patients with acute myocardial infarction requires a fine degree of judgment since the short-and long-term benefits have not yet been clarified.

In characterizing the effectiveness of CCU and ICU care it should be recognized that this is subject to ongoing change due to the interaction between trends in the natural history of those diseases leading to admission and improvements in available technology. For example, within two decades, potent antiarrhythmic agents, pulmonary artery catheters, computerized cardiac output measurements, and intraaortic balloon pumps have all been introduced to the coronary care unit. During the same period, a major national improvement has developed in the control of hypertension and interest in cardiac fitness. From another less complex perspective, the apparently "fixed" mortalities of respiratory failure, from poliomyelitis and idiopathic respiratory distress syndrome of the newborn, have been changed dramatically by the introduction of intermittent positive pressure ventilation and continuous positive airway pressure, respectively. In both instances, a single maneuver, not previously considered for use in those diseases, changed outcome almost overnight. For these reasons, decisions as to which patients may benefit must be subject to ongoing revision, based on continuously updated data. Considerations as to whom to exclude from access to intensive care should not ignore this changing picture.

What Skills Are Essential for Personnel in a Critical Care Unit? How Should This Personnel Be Trained and Organized to Assure the Best Care for Patients Most in Need?

Essential Skills for an ICU

The skills essential for ICU personnel depend on the level of responsibilities and mission of that unit. For example, a CCU will require distinctly different skills from those required in a trauma unit. Skills required of a nurse or primary physician may be different from those required of the director of the ICU.

Once the mission of the ICU is established and the boundaries of expertise in the ICU are defined, certain specific skills may be omitted. For example, if the hospital lacks an active neurosurgical program, skills unique to a neurosurgery unit can be excluded and patients requiring this expertise will by necessity be triaged elsewhere.

Some personnel skills are generic to any ICU:

Decision-making Skills.

Clear lines of authority must be established for decision making. Certain decisions will be made as policy of the ICU with exceptions depending on the needs of the individual patient. Qualified personnel other than physicians may make life-saving decisions about interventions without prior consultation with the responsible physician. For each patient, medical management decisions should be made by a physician- coordinator who may delegate decision-making responsibility to other members of the ICU health care team and consultants. The physician-coordinator may be either an ICU-based physician or another physician credentialed by the hospital; but in the case of every patient the physician-coordinator will be explicitly designated.

Equipment Skills.

Qualified personnel must be available to inspect, maintain, and calibrate equipment used for monitoring and life support. All ICUs should be capable of arrhythmia monitoring and have equipment immediately available for cardiopulmonary resuscitation. Other equipment skills will differ depending on the mission of the ICU. For example, intracranial pressure monitoring may be important in a neurological unit, whereas such skills would not be required in a CCU.

Procedure Skills.

For a CCU and an intermediate CCU, arrhythmia detection and treatment and resuscitation skills are the basic procedural necessities. At the other end of the spectrum, a large medical/surgical ICU usually would have the procedural abilities of placing and maintaining arterial lines, placing and maintaining pulmonary artery catheters and central venous lines, inserting endotracheal tubes and managing mechanical ventilation, providing cardioversion and cardiopulmonary resuscitation, placing enteral tubes and hyperalimentation catheters, etc. Some procedures might be the function of personnel in the ICU while other procedural functions might be better provided by a specialized team (e.g., nutritional support team and hemodialysis team).

Administrative Skills.

In each ICU, there should be a policy for assuring continuity of patient care and the availability of both appropriate logistic support and experienced personnel for each shift.

Teaching and Training Skills.

A continuing teaching and training program is essential for each ICU to acquire and maintain skills and will be described in the following section.

Personnel Training

The general principle is that all involved personnel should be trained to do their jobs with emphasis on competence in their area of critical care medicine. The medical director and nursing director of the ICU should coordinate and/or participate in the various teaching programs for ICU personnel.

Depending on the categories of patients admitted and level of care provided in a specific ICU, there may be one or more physicians of one or several disciplines working in the unit. ICU experience should be part of most house staff training programs. In addition, those who seek special competence in critical care medicine should participate in structured ICU training programs under the direction of established professionals in appropriate disciplines. The ICU Director should be trained in a specialty of importance to the type of patients in the unit. Additional training in critical care medicine and/or another appropriate subspecialty is also desired.

Registered nurses must have significant postgraduate clinical experience prior to ICU training. Usually this should be at least one year in duration. This training and education should include a comprehensive orientation program, followed by on-the-job training with a preceptor. After significant ICU experience, each nurse should have the opportunity to participate in continuing education in critical care courses. If licensed practical nurses are used, adequate training must be provided that will be appropriate to the job performance required.

Depending upon the categories of patients, respiratory therapists should be integral members of the critical care team in order to provide respiratory care and related services. These personnel should have initial orientation followed by on-the-job training and, in addition, a regular lecture program on ICU patient problems and their respiratory care.

Other personnel categories that may be involved in the ICU but do not provide nursing interventions include paramedics, nurses aides, unit clerks, physiotherapists, laboratory technicians, biomedical electronics technicians, clinical engineers, pharmacists, dietitians, social workers, data managers, and clergy. They should all have access to appropriate teaching programs that will vary in content depending on the individual's profession and the goals and objectives of the ICU.

Organization of ICU Personnel

General Considerations.

An ICU combines the capacity to provide needed care and technology with a potential to do great harm. Any organizational structure must, therefore, match technology with the correct blend of personnel to guarantee safe application of invasive monitoring and insure that generated data are correct, interpretation of derived data appropriate, and therapy safely employed. The organizational structure will vary depending on the overall mission of the hospital. There should be risk stratification for patients leading to distribution among units-- matching the appropriate units with the necessity for intervention or monitoring. Within each ICU, organization should be structured to insure proper care of the total patient.

Physician Staff.

The ICU should be directed by a physician with demonstrated competence in the areas necessary for provision of critical care. These areas generally include a broad base in physiology, pharmacology, the continuum of disease, cardiopulmonary function, and the associated intervention skills. He or she must recognize the need for teamwork, specialty consultation, and have demonstrated administrative and leadership skills. Competence, availability, demonstrated interest and ability in promoting harmonious interaction of various members of the health care team rather than departmental affiliation should guide this selection.

Other physicians (ICU attending physicians), as credentialed by the hospital, may assume day-to-day unit and patient management responsibilities. All ICU physicians should demonstrate maintenance of clinical competence and skills through regular ICU practice. The role of the ICU physician may range, depending on unit and hospital type, from absolute control of patient management through coordinator to consultant. The medical director or designate should have final authority over admission and discharge of patients from the ICU.

The medical and nursing directors assume ultimate responsibility for the safety and appropriateness of services provided by the ICU. To this end, unstable clinical situations requiring minute-to-minute titration of therapy such as a delicate balance of multiorgan failure require that the ICU medical director or his/her designee direct patient care. This places even greater emphasis on the need to communicate regularly and frequently with primary physicians, family, and others involved in the current and future care of the patient.

Any hospital with a level II or III unit (level I units have in-unit physicians--see question 4) should have immediately available, in hospital, 24 hours a day, a physician credentialed by the hospital as competent in life support and airway and ventilator management. Where house staff programs exist, trainees must occupy appropriately supervised roles in patient care that add to the quality and clarity of care.

Nursing Personnel.

Nurses are the key element in critical care. They provide continuity while physicians and other health professionals come and go. The organizational structure must support rather than detract from this role. The same considerations in selecting a medical director apply to selection of the unit nursing director. Nursing management and nursing practice decisions should be made by the nursing director. The organizational structure should promote and require that nurses and physicians work together as colleagues at all levels-especially the medical director and nursing director. Clerical and administrative support are necessary to prevent distraction of the nurse from direct patient care functions. The nurse coordinates the activity of all other allied health personnel at the bedside. Staffing patterns should be keyed to level of patient illness rather than a fixed nurse-to-patient ratio. Nursing staff capabilities must match the spectrum of patients cared for in the unit.

What Special Technology and Therapeutic Interventions Should Be Routinely Available for the Most Effective ICU Function?

Effective intensive care units will have special technology, therapeutic capabilities, and personnel that are determined by the types of patients treated in the units. An ICU's technology and therapeutic capabilities must be based on an analysis of the reasons for which the unit was developed. This analysis should lead to clearly written guidelines for the care of patients in the ICU, defining clinical expertise needed, size of unit(s) needed, teaching versus nonteaching status, and commitment, if any, to research activities.

Incorporation of any technology and therapeutic capability must support the ultimate function of the ICU, which is to provide high quality care to patients who sustain or are at risk of developing potentially reversible severe illness. Thus, every ICU should have a well-trained team whose expertise matches the specific clinical problems that must be treated. This team must develop admission and continuing stay requirements, discharge criteria, and protocols that define the clinical scope of the ICU. In addition, policies for triage must be formally established and practiced.

An ICU, regardless of location, must have the following minimal technological capabilities:

  1. Cardiopulmonary resuscitation
  2. Airway management, including endotracheal intubation and assisted ventilation.
  3. Oxygen delivery systems and qualified respiratory therapists or registered nurses to deliver oxygen therapy
  4. Continual electrocardiographic monitoring
  5. Emergency temporary cardiac pacing
  6. Access to rapid and comprehensive laboratory services including but not limited to arterial blood gas analysis, electrolyte determinations, hemograms, measurement of cardiac enzymes, renal function studies, microbiologic studies, fluoroscopy, and other radiologic studies
  7. Access to nutritional support services to advise on both enteral and parenteral nutritional techniques
  8. Titrated therapeutic interventions with infusion pumps
  9. Based on determination of the ICU patient composition, technological capability must be available to support therapeutic interventions that are commonly accepted medical practice. For example, an ICU that manages shock syndromes needs hemodynamic monitoring capability techniques to allow for the rational diagnostic categorization and subsequent therapy of patients with shock syndromes
  10. Portable life-support equipment for use in patient transport, both within the hospital and for transfer


Should the above minimal capabilities not be available, patient stabilization and referral procedures should be implemented. Furthermore, health care providers must be conscious of their obligation not to use investigational therapy or technology without defined protocols and appropriate informed patient consent.

How Is a Hospital's Critical Care Delivery System Best Structured: One Large Multispecialty Unit or Multiple Small Subspecialty Intensive Care Units?

The nature of the hospital, the degree of subspecialty expertise available within the institution, the patient population being treated, and whether the institution is a referral center, a community hospital, or a teaching hospital, are all issues that must be taken into account to determine the best structure for its critical care delivery system.

In many hospitals a multispecialty ICU is most practical and allows better aggregation of ICU resources (technology, physicians, nurses, and allied health personnel). In other hospitals it is appropriate to combine one or more multispecialty units with one or more subspecialty units. Thus, the spectrum of ICUs today ranges from multispecialty to specific purpose ICUs.

In addition, there is diversity in the staffing and organizational structure in ICUs across institutions. These differences are identified below and grouped into levels representing current practices in hospitals. The levels are distinguished by differences in availability of certain human and technical resources and by the frequency of performance of certain interventions such as continuous ventilatory therapy. Suggested nurse-patient ratios are presented with some flexibility because the presence of other personnel such as respiratory therapists and temporal variations in patient mix may influence individual unit requirements. In many Level I and Level II ICUs, for example, continuous availability of on-site respiratory therapists will facilitate better patient care. The organization and scope of aggregated ICU resources integrated with patients who have special needs should occur in one or more of the following levels:

Level I.

This is a comprehensive and multisystem critical care unit. A physician-director or qualified designee is immediately available to the unit at all times. The nurse-patient ratio is 1:1 or greater depending on the severity of the patients' conditions. Measurement and derivation of all necessary invasive and noninvasive monitoring is accessible. There is a teaching and research obligation.

Level II.

This unit is a multipurpose or a specific purpose unit. Examples include respiratory ICU, neuro ICU, coronary care unit, mixed medical/surgical unit, etc. A physician-director or qualified designee is available in the hospital. This unit can perform therapeutic interventions based on invasive and noninvasive monitoring and has a nurse-patient ratio of 1:1 or 1:2 or 1:3 depending on the severity of the patients' conditions.

Level III.

This unit provides for limited use of invasive monitoring and therapeutic interventions such as assisted ventilation. A physician-director or his/her designee must be readily available. In- hospital coverage must be available from a physician who is credentialed by the hospital in life support and airway and ventilator management. Nurse-patient ratio is 1:2 or 1:3 or 1:4 depending on the severity of the patients' conditions.

Level IV.

This is a specialty care unit but does not meet the definition of an ICU. This unit provides noninvasive monitoring for those patients who may have arrhythmias and the potential for complications. A unit director or designee is provided and will respond as needed. Arrhythmia monitoring and basic CPR skills are provided. The nurse-patient ratio is 1:4 or 1:5 depending on the severity of the patients' conditions. This ratio does not apply if the number of patients is less than four.

Larger hospitals might have several units (Level I or II and Level IV as intermediate care, for example). Transfer among units and to the general service adds additional risk; the level of care during transfer must not be less than that provided prior to initiation of transfer.

The ability to administer each level effectively becomes increasingly difficult if the unit is too large. Ordinarily, intensive care units should not be larger than 12 beds. If there is a need for more intensive care beds, development of additional ICUs must be considered, either general or specific purpose units. Where possible, institutions with more than one unit should place them in close proximity and adjacent to needed support services.

Admission criteria must be based on the sound judgment and clinical expertise of practicing physicians who feel there is a reasonable probability that admission will benefit the patient. Once in the ICU, the medical and nursing directors or their designees are responsible for evaluating admission and determining a patient's need to stay in the unit.

How Has the Development of Intensive Care Units Affected the Traditional Functions of a Hospital?

The primary impact of the development of intensive care units in a hospital is that such units have enabled the hospital to provide more complete and higher quality services to the community. It has also enabled the hospital to more economically group its resources so that similar types of seriously ill patients can benefit equally.

The delivery of high quality service from ICUs has made it necessary for hospitals to evaluate and expand many of the ancillary areas that support these services. Specifically, hospitals that have developed such units have added more comprehensive laboratory, radiology, nutrition, biomedical engineering, respiratory therapy, psychological, and social work services, among others. Such expansions of support operations have added complexity to the day-to-day functions of institutions, resulting in the need for more refined organizational structures throughout hospitals. This in turn has necessitated the development of more controlled communication systems to ensure availability of needed medical information and continuity of patient care services.

ICUs have also added prestige and enhanced the communities' good will toward the institutions in which they are based. This good will has often taken the form of financial and volunteer support as well as general moral support.

But while the ICUs have improved patient care, day-to-day operations and institutional prestige, they also have added pressures and stresses to hospital personnel both within and outside of the ICUs. For example, movement of patients into ICUs has had undesirable effects upon the general medical surgical units of the hospital.

First, there has been, in many hospitals, a reduction in appreciation for the continued care given and the possibilities for effective patient care on the general floors. The aura of urgency associated with critically ill patients tends to reinforce the perception of personnel from the general services that they are not as important in the care chain.

A more serious result of the formation of ICUs has been the removal from general services of the more acutely ill, thus reducing the experience of the general staff in caring for such patients.

Nursing care in the ICU has an emphasis opposite from such care on general services. The effectiveness of the ICU nurse is his/her knowledge of all the details necessary to care for one or two patients while the effectiveness of the general service nurse rests upon his or her ability to direct care delivery by others to numerous patients. This contradiction in nursing practice techniques may not be fully understood by the physicians who tend to equate the quantity of nursing care with quality. As the physician becomes more reluctant to use the general services for patients who are acutely ill, but not critically ill, the general services staff is deprived of the opportunity to gain and maintain skills which ultimately must be exercised in behalf of the post-ICU patient. These problems are recognized by hospitals and require ongoing programs to deal with them.

Beyond these issues the development of ICUs has motivated hospitals to better focus on their mission in the community and to work with other institutions to insure a more rational approach to patient care services. Further, the nature of ICU medicine has precipitated the need for hospitals to develop protocols and statements on ethical issues.

Advancing technology requires ICU personnel to become more expert in evaluation of medical devices and much more active in staff training to insure that relevant medical practices and capabilities are provided in the highest quality manner.

In general, ICUs have had many positive effects on hospitals but cause pressures and stresses which must be dealt with continuously.

What Direction Should Critical Care Research Follow?

Directions for Critical Care Research

The national cost of ICU care is in excess of 15 percent of hospital costs, and is between $10 and $20 billion annually. Although the cost of ICU-related research is high, the panel believes that substantial savings--greatly in excess of research costs--can be gained by better defining indications for ICU admission and the use of monitoring technology and interventions.

Intensive care related research readily falls into one of four groups:

  1. Natural history, risk, and outcome research, directed at broad questions of utilization and efficacy
  2. Research focused on specific ICU technology and application, and therapeutic interventions
  3. Human resource research
  4. Disease entity research, ranging from basic studies of etiologic mechanisms to specific clinical therapeutic trials


Natural History, Risk, and Outcome Research

The combination of life-threatening diseases, finite resources, invasive therapeutic and monitoring techniques, and high costs makes the need for adequate data on which to base decisions a high priority.

Such research is aimed at determining how ICUs can be used for the maximum benefit of the ICU population. This research should include procedures for "triaging" patients so that admission is not denied to patients who can most benefit from an ICU as well as excluding patients who have no reasonable chance to benefit. Research aimed at developing accurate outcome predictors as a function of initial presenting condition, diagnosis, and other ongoing prognostic variables should be encouraged.

Patients being monitored require sensitive predictors of potential complications, whereas patients being considered for intervention need specific predictors of success for the range of possible interventions.

ICU Technology and Applications and Therapeutic Interventions

A highly expensive component of intensive care is the monitoring technology. It is easy to assume a proportional relationship between quantity of information and quality of care, but this is not necessarily true. Indeed it has been suggested that, since there is a natural tendency to respond to abnormal data with a therapeutic maneuver, the net effect of monitoring a variable, which is not important in decision making, would be the sum total of the complications of the resulting therapy plus any direct complications from the mode of monitoring itself. Included in this negative effect would be the consequences of errors in the monitored data. The panel recommends encouraging research directed at detecting the outcome gain to be derived from the variables commonly monitored in ICUs. Similarly, the efficacy of specific interventions requires careful evaluation (see below). These two types of study (efficacy of monitoring and interventions) should include a careful assessment of the balance between gain and outcome, and iatrogenic complications.

For many of the disease states presently resulting in ICU admission the benefit is obvious (e.g., ventilation for apnea secondary to temporary neuromuscular paralysis). In such cases, the panel considers it inappropriate to conduct efficacy studies in the light of accumulated positive experience. The role of specific ICU protocols in influencing outcome is, for many diseases, not clear. There are wide variations among institutions regarding the interventions used for patients with the same disease. It is recommended that a series of protocol studies be carried out to determine optimal interventions for specific diagnoses. The term "protocol study" denotes a research plan where the study population, severity of disease, treatment intervention, and the objective measures of benefit are clearly stated. Ideally, such studies would include random allocation to treatment and control groups. Randomized studies may be difficult to carry out within a single unit or institution for many diagnoses because of the ethics and logistics of patient consent and the need to act quickly. Where this is the case, one may carry out a randomized study by enrolling more than one unit or institution to participate by randomizing units or institutions--not patients. Units or institutions would be randomized to a single treatment policy. They would treat a specified number of patients as dictated by the randomization allocation. After treating a specified number, a unit or institution would switch to the other treatment policy (in the case of studying two treatments). In this study plan, half the units or institutions would be first randomized to each treatment program and then crossed over. Adoption of such multicenter study plans will allow randomized prospective trials to be carried out which may lead to finding optimal intervention programs in intensive care units.

Human Resource Research

The complexity and intensity of care in the ICU requires the accumulation of information related to issues such as (a) the relationship between training levels and distribution of the various categories of staff, and patient care outcome, (b) consideration of factors leading to conflict and stress in staff, and (c) an evaluation of the efficacy of orientation and continuing education programs in the ICU.

Disease Entity Research

Numerous areas for research related to the various disease entities which result in ICU admission merit high priority. Among the more common causes of death are head injuries, sepsis, and multiple organ failure. While research in each of these areas has wider application than CCM, the panel recognizes that the teaching hospital ICU lends itself well to such clinical research. This should be encouraged as should more basic research at the level of mechanisms of the pathophysiology of brain damage following head injury, factors contributing to serious infections, and the mechanisms that lead to the adult respiratory distress syndrome.

In carrying out research in the ICU, certain difficulties must be recognized. In relation to natural history characterization, the existence of multiple subsets of any disease group and the very nonstandard way in which data are collected in the clinical setting pose a special problem. The relatively small number of patients in any one ICU and the multiple subsets suggest multicenter studies. The current nonstandard approach to data acquisition should be replaced with a uniform planned program of data collection which could be adopted by all ICUs. A minimum data set with common definitions would enable individual ICUs to evaluate the impact of their care over time and allow inter-ICU comparisons.

In conducting clinical studies in the ICU setting, special problems arise in obtaining informed consent. The unique dependency of the patient, the concern of the family, and the implications of the interventions under study all complicate anunemotional consideration of risks and benefits. This problem is compounded when the patient is unresponsive, and the net effect is to create a substantial difficulty in conducting research in this area. However, many of the questions to be addressed are of great import, both because of the enormous costs involved and the life-threatening nature of the factors requiring study.

Consensus Development Panel

Stephen M. Ayres, M.D. (Panel Chairman)
Professor and Chairman
Department of Internal Medicine
St. Louis University School of Medicine
Medical Director
St. Louis University Hospital
St. Louis, Missouri
Stephen C. Achuff, M.D.
Associate Professor of Medicine
Johns Hopkins University
School of Medicine
Director of Clinical Cardiology
The Johns Hopkins Hospital
Baltimore, Maryland
Hon. Christopher J. Armstrong
Massachusetts Appeals Court
Boston, Massachusetts
Donna Lee Bertram, R.N., B.S.N.
Department of Medical/Surgical Nursing
Harris Hospital-Methodist
Fort Worth, Texas
Roger C. Bone, M.D.
Professor of Medicine
University of Arkansas for Health Sciences
Chief, Division of Pulmonary and Critical Care Medicine
University of Arkansas Medical Center and Veterans Administration Complex
Little Rock, Arkansas
Joseph M. Civetta, M.D., F.A.C.S.
Professor of Surgery, Anesthesiology, Medicine, and Pathology
University of Miami School of Medicine
Chief, Division of Emergency Surgical Services
Jackson Memorial Medical Center
Miami, Florida
H. Barrie Fairley, M.D.
Professor of Anesthesia
University of California, San Francisco
San Francisco General Hospital
San Francisco, California
Frank H. Gafford IV, M.D.
Chairman, Department of Critical Care Medicine
Director, Intensive Care Units
St. Johns Mercy Medical Center
Co-Director, St. Louis University-St. John's Mercy Medical Center
Critical Care Medicine Fellowship Program
St. Louis, Missouri
David Ross Garr, M.D.
Clinical Assistant Professor of Family Medicine
University of Colorado Health Sciences Center
Director of Learning Resources
Family Medicine Residency Program
Mercy Medical Center
Denver, Colorado
Ake Grenvik, M.D., Ph.D.
Professor of Anesthesiology and Surgery
Director, Critical Care Medicine Training Program
University Health Center of Pittsburgh
Presbyterian-University Hospital
Pittsburgh, Pennsylvania
Thomas E. Macnamara M.B., Ch.B.
Professor and Chairman
Department of Anesthesia
Georgetown University
Chief, Anesthesia Section
Clinical Center
National Institutes of Health
Washington, D.C.
Albert G. Mulley, M.D.
Associate Director
Medical Practices Evaluation Unit
Massachusetts General Hospital
Boston, Massachusetts
Thomas G. Rainey, M.D.
Assistant Professor of Medicine
Uniformed Services University of the Health Sciences
Chairman, Department of Critical Care Medicine
Director, Intensive Care Unit
National Naval Medical Center
Bethesda, Maryland
Norma J. Shoemaker, R.N., M.N.
Executive Director
Society of Critical Care Medicine
Fullerton, California
W. Vickery Stoughton
Toronto General Hospital
Toronto, Ontario
Marvin Zelen, Ph.D.
Professor and Chairman
Department of Biostatistics
Harvard School of Public Health
Head, Division of Biostatistics and Epidemiology
Dana-Farber Cancer Institute
Harvard Medical School
Boston, Massachusetts


Diane C. Adler, M.A., C.C.R.N., R.R.T
"Hospital Management of Critical Care"
Clinical Director of Critical Care Nursing
Hospital of the University of Pennsylvania
Philadelphia, Pennsylvania
Henrik H. Bendixen, M.D.
"Respiratory Failure: Definition, Diagnosis and Physiology"
Professor and Chairman Department of Anesthesiology
Columbia University College of Physicians and Surgeons
New York, New York
Bernadine Healy Bulkley, M.D.
"The Coronary Care Unit"
Professor of Medicine
Johns Hopkins Hospital
Baltimore, Maryland
P. J. Burau, R.N., M.S.N.
"Problems in Attracting and Retaining Nurses for a Critical Care Environment"
Chief Critical Care and Heart and Lung Nursing Services
Clinical Center
National Institutes of Health
Bethesda, Maryland
John J. Caronna, M.D.
"The Neurological Intensive Care Unit"
Vice Chairman Department of Neurology
New York Hospital--Cornell Medical Center
New York, New York
David J. Cullen, M.D.
"The Anesthesia Intensive Care Unit"
Associate Professor of Anesthesia
Director of Recovery Room
Harvard Medical School Department of Anesthesia
Massachusetts General Hospital
Boston, Massachusetts
John B. Downs, M.D.
"Adult Respiratory Distress Syndrome (Shock Lung)"
Associate Professor of Anesthesia
Northwestern University
Medical Director, Anesthesiology
Mercy Hospital Urbana, Illinois
Paul F. Griner, M.D.
"Pulmonary Edema and the ICU"
Professor and Associate Chairman
Department of Medicine
University of Rochester School of Medicine and Dentistry
Rochester, New York
Bryan Jennett, M.D., F.R.C.S.
"Head Injury Management and Outcome in a Critical Care Unit"
Professor Department of Neurosurgery
Institute of Neurological Sciences
Dean of Medicine
University of Glasgow
Thomas Killip, M.D.
"Cardiac Rhythm Monitoring in Critical Care Units"
Chairman Department of Medicine
Henry Ford Hospital
Detroit, Michigan
Robert R. Kirby, M.D.
"Weaning from Mechanical Ventilation"
Chairman, Department of Anesthesiology
Wilford Hall U.S.A.F. Medical Center, SGHSA
Lackland Air Force Base, Texas
William A. Knaus, M.D.
"The Combined Medical-Surgical Intensive Care Unit"
Associate Professor of Anesthesia and Medicine
Director, ICU Research
George Washington University Medical Center
Washington, D.C.
James J. Leonard, M.D.
"Staffing of the Critical Care Unit: The Education and Training of Critical
Care Physicians"
Professor and Chairman, Department of Medicine
Uniformed Services University of the Health Sciences
Bethesda, Maryland
Byron D. McLees, M.D., Ph.D.
"Pulmonary Diagnostic Procedures in Critical Care Units"
Chief, Pulmonary Section
Bowman Gray School of Medicine
Winston-Salem, North Carolina
Sally Millar, R.N., C.C.R.N.
"Staffing of a Critical Care Unit: Nursing"
Clinical Nurse Leader
Intensive Care Nursing Service
Massachusetts General Hospital
Boston, Massachusetts
Hiltrud S. Mueller, M.D.
"Cardiogenic Shock"
Professor of Medicine
Albert Einstein College of Medicine of Yeshiva University
Associate Director of Cardiology
Montefiore Medical Center
Bronx, New York
Joseph E. Parrillo, M.D.
"Septic Shock"
Chief, Critical Care Medicine Department
Clinical Center
National Institutes of Health
Bethesda, Maryland
Henning Pontoppidan, M.D.
"Respiratory Failure: Management and Outcome"
Professor of Anesthesia
Harvard Medical School Department of Anesthesia
Massachusetts General Hospital
Boston, Massachusetts
Peter Safar, M.D.
"The Critical Care Medicine Continuum from Scene to Outcome"
University Professor
Director Resuscitation Research Center
University of Pittsburgh
Pittsburgh, Pennsylvania
Charles A. Sanders, M.D.
"Hospital Management of Critical Care I"
Executive Vice President Science and Administration
E. R. Squibb and Sons, Inc.
Princeton, New Jersey
Barry A. Shapiro, M.D.
"Use of Mechanical Ventilation in Critical Care Units"
Professor of Clinical Anesthesia
Director of Respiratory/Critical Care
Department of Anesthesia
Northwestern University Medical School
Chicago, Illinois
William C. Shoemaker, M.D.
"Accidental, Traumatic, and Hemorrhagic Shock"
Professor of Surgery
Department of Surgery
Harbor-UCLA Medical Center
Torrance, California
Robert N. Sladen, M.B., M.R.C.P.(U.K.), F.R.C.P.(C)
"Acute Renal Failure in Critically Ill Patients"
Associate Medical Director, Intensive Care
Assistant Professor in Anesthesia
Stanford University Medical Center
Stanford, California
H. J. C. Swan, M.D., Ph.D.
"Hemodynamic Monitoring in Critical Care Units"
Professor of Medicine
University of California, Los Angeles School of Medicine
Director of Cardiology
Cedars-Sinai Medical Center
Los Angeles, California
George E. Thibault, M.D.
"The Medical Intensive Care Unit"
Assistant Chief Department of Medicine
Harvard Medical School
Massachusetts General Hospital
Boston, Massachusetts
W. Leigh Thompson, Ph.D., M.D., F.A.C.P.
"Structure of a Critical Care Unit: An Overview"
Executive Director
Lilly Research Laboratories
Indianapolis, Indiana
John G. Weg, M.D.
"The Respiratory Intensive Care Unit"
Professor of Internal Medicine
Physician In Charge
Pulmonary and Critical Care Medicine Division
University of Michigan Medical School and Hospitals
University Hospital
Ann Arbor, Michigan
Robert F. Wilson, M.D., F.A.C.S.
"The Surgical Intensive Care Unit"
Professor of Surgery
Director of Thoracic and Cardiovascular Surgery
Wayne State University
Chief of Surgery
Detroit Receiving Hospital
Detroit, Michigan
Jack E. Zimmerman, M.D., F.A.C.P.
"Administrative Structure of a Critical Care Unit"
Professor of Anesthesiology
George Washington University
Washington, D.C.

Planning Committee

Joseph E. Parrillo, M.D. (Chairman)
Chief, Critical Care Medicine Department
Clinical Center
National Institutes of Health
Bethesda, Maryland
Stephen M. Ayres, M.D.
Professor and Chairman
Department of Internal Medicine
St. Louis University School of Medicine
Medical Director
St. Louis University Hospital
St. Louis, Missouri
Itzhak Jacoby, Ph.D.
Deputy Director
Office of Medical Applications of Research
National Institutes of Health
Bethesda, Maryland
James J. Leonard, M.D.
Professor and Chairman
Department of Medicine
Uniformed Services University of the Health Sciences
Bethesda, Maryland
Thomas E. Macnamara, M.B., Ch.B.
Professor and Chairman
Department of Anesthesia
Georgetown University
Chief, Anesthesia Section
Clinical Center
National Institutes of Health
Washington, D.C.
Thomas G. Rainey, M.D.
Assistant Professor of Medicine
Uniformed Services University of the Health Sciences
Chairman, Department of Critical Care Medicine
Director, Intensive Care Unit
National Naval Medical Center
Bethesda, Maryland
Jack E. Zimmerman, M.D., F.A.C.P.
Professor of Anesthesiology
George Washington University
Washington, D.C.

Conference Sponsors

The Warren Grant Magnuson Clinical Center
National Institutes of Health
Jay Shapiro, M.D.
Acting Director

Office of Medical Applications of Research
National Institutes of Health
J. Richard Crout, M.D.

Privacy Notice | Disclaimer | Accessibility | Contact NIH
Department of Health and Human Services HomepageNational Institutes of Health Home PageFirstGov Home Page - The U.S. Government's Official Web Portal  National Guideline ClearingHouse - a public resource for evidence-based clinical practice guidelines