| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1998;66:1452-1465
© 1998 The Society of Thoracic Surgeons
Address reprint requests to Dr Pae, Division of Cardiothoracic Surgery, The Pennsylvania State University, College of Medicine, 500 University Dr, PO Box 850, Hershey, PA 17033
Staff
The Conference was staffed with representatives from both The Society of Thoracic Surgeons and American College of Cardiology. Special appreciation is given to Mr David Field of the ACC and Mr Donald Turney of the STS for their efforts in conducting the conference.
Because the pool of potential candidates for cardiac transplant continues to expand while donor organ availability remains critically scarce, the discrepancy between the number of patients who would benefit from heart transplant and the number of patients who actually receive a transplant continues to increase at an alarming pace. The clinical outcome for patients who receive devices as a "bridge" to transplant may well exceed medically treated status I patients and has allowed development and evaluation of devices that can be eventually used as a long-term alternative to transplant. The expectation that fully implantable ventricular assist devices (VAD) would be available by the early 1990s as an alternative to transplant has been precluded by technical regulatory difficulties in device development. In an effort to facilitate the process of bringing this remarkable therapeutic tool to widespread clinical application a cooperative effort was undertaken to bring investigators, industry, and regulatory agencies together in an attempt to prospectively define mutually agreeable and sound guidelines for clinical trials necessary to achieve pre-market approval of VADs for long-term use. The following Task Force Reports represent the culmination of these efforts and summarize the many months of dialog before and after this Bethesda Conference.
Task Force 1: Trial design
Eric A. Rose, MD, Chairman, D. Glenn Pennington, MD, Co-Chairman, John Wallwork, MD, Lynn Warner-Stevenson, MD, Lisa Parker, MD, Gene Blackstone, MD, Victor L. Poirier, MD, Peer Portner, MD, Arthur Ciarkowski, MD, Wolf Sapierstein, MD, Harvey Boroevetz, MD
General philosophy
This section outlines criteria for the design of clinical trials to obtain FDA pre-marketing approval for long-term use of mechanical circulatory assist devices, beyond their current approved indications of post-cardiotomy support and bridging to cardiac transplantation.
We firmly believe that clinical trials for this purpose should be scientifically sound, clinically meaningful, and achievable in a finite time frame at reasonable expense. It is anticipated that early trials will focus on well-defined patient populations, with subsequent trials including a wider spectrum of patients, once initial success is demonstrated.
To meet pre-marketing approval requirements, sponsors will need to demonstrate the efficacy and safety of mechanical circulatory assist devices in the treatment of end-stage heart failure. This section discusses the target population, endpoints, and the design of clinical studies for this purpose. We shall specifically define and justify the choices of population, endpoints, and hypotheses of benefit as follows:
Target population
Patients participating in these clinical trials must have demonstrated evidence of severe sustained congestive heart failure despite optimal medical therapy. All patients should be receiving the optimal medical treatment as currently provided under the supervision of a team experienced in the care of patients with severe heart failure. This heart failure may be related to coronary artery disease or non-ischemic cardiomyopathy. The population involved in these trials should include an appropriate representation of minorities and women and be chosen with a sufficient number of endpoints to ensure valid conclusions.
Considerations for exclusion from these studies may include:
Endpoints
The objective of implanting a device is to provide improvement in survival and quality of life when compared to alternative therapy. The primary endpoints that need to be measured are all cause mortality and impact severely on health-related quality of life. Secondary endpoints that need to be measured include cardiovascular mortality and functional cardiac status.
Device safety is also an important endpoint. Monitoring the safety of the device requires measurement of the incidence of adverse effects, including (in particular) infections, thromboembolic complications, bleeding, end organ dysfunction, right heart failure, emotional and psychiatric complications, and mechanical failure (see Task Force 2: Adverse Events).
Clinical trial design
The separation of transplantable from non-transplantable patients is an important consideration in minimizing the incidence of patient crossover from device therapy to transplantation or vice versa, which may confound the analysis of trial results and reduce the statistical power of the trial to detect effects. However, the competing risk of death may confound the analysis of non-fatal events.
Non-transplant patients
In the case of patients who are not candidates for transplantation, it is feasible to conduct prospective, randomized, controlled clinical trials with a parallel study design. A non-randomized concurrent control group could also be considered, with adequate justification. It should be recognized that patients who are not transplant candidates may also be at greater risk of complications and death when receiving devices compared to transplant candidates. In the following paragraphs, we will specify the features that such trials should address.
Nature of comparison
At this time, pre-marketing approval of applicant devices should be based on a comparison to the survival and quality of life of patients who are randomly assigned to remain on medical therapy. The follow-up time should be sufficient to establish appropriate survival in the control group. This may be an inadequate amount of time to monitor device failure. Long-term safety and effectiveness data will be captured by post-marketing surveillance. The criteria for approving such devices should not be directly based on the absolute working life and failure rate of the device but rather on the relative improvement of quality of life or survival of the recipient population.
Blinding
It will be impossible for care-giving investigators or patients to be blinded to the nature of treatment received. However, to the extent that it is possible, outcome determinations should be made or at least confirmed by investigators that are blinded to the patients’ treatment allocations. Sham operations performed in support of the concept of blinded trials are surely unethical.
Randomization
Randomization is useful to ensure equitably constituted comparative groups and, when feasible, it would be employed. Because practical considerations dictate relatively small sample sizes, the efficacy of these devices should be high if they are to have a meaningful impact on survival and quality of life. Investigators may consider blocking by center to ensure approximately equivalent numbers of patients in each arm of the trial throughout its course.
Sample size
The determination of sample size to conduct any clinical trial is based upon the differences in the primary outcome event that the investigator is attempting to establish. For comparisons to standard medical therapy, sample size should be sufficient to establish clinical utility. For example, a study might be designed to establish a 33% survival gain and improved quality of life by two years with a significance level of 0.05 with 80% power. The goal of a 33% reduction in mortality is roughly equivalent to double the effect seen with angiotensin-converting enzyme inhibitors for heart failure and thrombolytic drugs for acute myocardial infarction.
Candidates for cardiac transplantation
It is widely recognized that donor hearts are becoming relatively more scarce due to the increasing population of potential recipients and a stagnation in the number of donors. Thus, the strategy of transplantation today often includes a prolonged waiting period and the inability to be transplanted before death from heart failure. In many areas of the country, this waiting period is in the range of 100 days for United Network for Organ Sharing (UNOS) Status I patients and much longer for UNOS Status II patients. An undefined number of potential transplant candidates are never listed for transplantation because of the severe shortage of donor hearts.
It may be feasible to conduct prospective randomized clinical trials of parallel study design; however, a non-randomized concurrent group could also be considered with adequate justification. A trial to demonstrate the utility of devices in ambulatory candidates might include all those candidates with a designated severity of disease, or it could be limited to patients with a low probability of being transplanted such as large type O patients, or patients with multiple antibodies. However, accruing enough patients of this type may be difficult and may necessitate a non-randomized, prospective cohort study. The device group might then consist of transplantation candidates with a low likelihood of being transplanted while the controls would be selected from the general group of candidates. Study design in this population requires careful consideration of the patients’ options for crossover to transplantation and devices as bridges and the impact such crossovers may have on analysis and interpretation of results. In addition, extensive multivariant analysis would be required to ensure the validity of group comparisons.
Hybrid trial: Transplant candidates and non-candidates
Hybrid trials, including patients who are and patients who are not candidates for transplantation, might also be considered. Design of such trials is confounded by issues of crossovers between groups and the necessity for a large number of patients to be enrolled. However, such trials may be justified if these obstacles can be overcome. The advantage of this hybrid trial would be to provide an increased number of patients, more generalizability, and the opportunity to test the devices simultaneously against medical therapy and the strategy of cardiac transplantation.
Task Force 2: Adverse events
J. Donald Hill, MD, Chairman, O. Howard Frazier, MD, Co-Chairman, Kurt Daase, MD, Howard Leven, MD, James M. Anderson, MD, Rhona Shanker, MD, Benjamen H. Eidleman, MD, Maria Rosa Costanzo, MD
The following set of recommendations regarding adverse event definitions was written by the Task Force for inclusion in the Guidelines for the Design of Clinical Trials to Study Circulatory Support Devices for Chronic Heart Failure. The Task Force recognizes the dynamic nature of medicine and surgery and wishes to qualify these definitions:
|
Task Force 3: Cardiovascular function evaluation
Jack G. Copeland, MD, Chairman, Patrick M. McCarthy, MD, Co-Chairman, Robert L. Kormos, MD, David Farrar, MD, Donna Mancini, MD, Michael J. Domanski, MD, Ramiah Subramanian, MD
Introduction
It must be recognized at the outset that these guidelines are evolving as clinical experience accrues. In particular, predicting the need for additional right ventricular support, response of the right ventricle to implantable left ventricular assist device (LVAD) support, and the possibility for successfully removing the LVAD after cardiac recovery are areas that are still under study. These are areas of active clinical research and more specific answers should be forthcoming in the future.
Areas for review
Four areas of cardiovascular function will be reviewed:
Acceptable hemodynamic function with the device
The implantable LVAD is placed in patients with end-stage heart disease to reverse cardiogenic shock or to restore more effective hemodynamic parameters. As a minimum requirement, therefore, use of the device should result in a cardiac index greater than or equal to 2.2 L/min/m2 with a pulmonary capillary wedge pressure (or left atrial pressure, or pulmonary artery diastolic pressure) less than or equal to 18 mm Hg during periods of proper function.
In the early postoperative period, the effects of hypovolemia, transient right heart dysfunction, tamponade, or other problems related to surgery may interfere with the ability of the device to meet these demands. However, these are "patient-related" failures, and not failure of the device, per se. Inadequate function of the device itself has been defined by the task force on adverse events (see Task Force 2: Adverse Events).
Under optimal conditions, the design of the device should allow increased device output with exercise or other physiologic demands. Devices functioning in the automatic fill mode meet this requirement. [1] Devices in a fixed rate mode should allow external adjustment so that device output can be increased as needed. Also the system should have a monitor that displays pump output. This does not need to be a continuous display monitor, but such a monitor is useful in the early postoperative period. For patients who are chronically supported, periodic measurements of device output are important for troubleshooting, or determining proper device function.
Left ventricular assist device effects on left and right ventricular function
There was initial concern that the implantable LVAD would induce left ventricular atrophy. This has not been shown to be the case. Histologic findings in the bridge-to-transplant experience show resolution during support of acute changes from cardiac decompensation at implant. Eventually fibrosis and otherwise viable and healthy myocytes appear [2, 3]. Echocardiographic studies show a decrease in left ventricular chamber size immediately after the onset of LVAD function, while left ventricular dimensions remain stable during the period of support [3–5]. Left ventricular wall thickness increases.
Depending upon the device synchronization with left ventricular systole, pressures within the left ventricular cavity typically reach 40 mm Hg maximum on LVAD support. This continued left ventricular pressure (although greatly reduced from pre-LVAD levels) may help explain the absence of left ventricular myocyte atrophy. Because the histologic picture improves on support, successfully weaning and removing the LVAD following LV recovery may be possible. This concept will be dealt with in the section "Potential Cardiac Recovery While on Device Support."
The effect of LVAD support on the right ventricle and the right-sided circulation is a complex issue and not always predictable [6–11]. Impaired right ventricular function leads to decreased left atrial pressure, decreased LVAD filling, and decreased LVAD flow. It is an important early contributing factor to perioperative death [12, 13].
Factors thought to contribute to right-sided circulation dysfunction include right ventricular afterload (usually described in terms of pulmonary artery pressures, and pulmonary vascular or arteriolar resistance); right ventricular contractility (usually recorded as ejection fraction or fractional area of change by using automatic boundary detection with echocardiographic techniques or RV stroke work index); and cardiac rhythm (ventricular tachycardia, ventricular fibrillation, or asystole, which may impair LVAD filling) [1, 6–11, 14, 15].
In addition, overall clinical condition of the patient at the time of implantation also correlates with the need for early right ventricular assist device (RVAD) support [13]. Patients who are the most severely decompensated in the bridge-to-transplant experience (that is, those with pulmonary edema, adult respiratory distress syndrome, and very severe hemodynamic compromise) appear to be at most risk for RVAD use.
A variety of factors including clinical state, hemodynamic factors, and experience of the implanting surgeon have led to a wide range of reported "need" for an RVAD in the early postoperative period. Recent experience (since 1994) with the implantable LVAD devices has shown the need for RVAD support to be approximately 15% to 20% (personal communication: P. Portner, Oakland, CA; K. Dasse, Woburn, MA) [1, 12, 13, 15, 16]. External devices that allow for the capability for biventricular support (such as the Thoratec device) have reported use of an RVAD in up to 50% of patients, although this seems to be decreasing with greater clinical experience (personal communication: David Farrar, PhD, San Francisco, CA). Because use of an RVAD is surgeon-dependent, and currently a clinical "art" rather than science, this Task Force has not set forth guidelines for RVAD use in implantable LVAD patients.
In most patients, who do not require RVAD support, improvement in RV function can be expected [5, 13, 16]. For most patients the left atrial pressure drops immediately after device insertion, and therefore pulmonary artery pressures drop rapidly. With this decrease in pulmonary artery pressures and pulmonary vascular resistance, RV function improves. Right atrial pressure also drops, but more slowly than the drop in left atrial pressure. Although right ventricular contractility improves, it rarely returns to normal levels of function, despite near-normal pulmonary artery pressures and pulmonary vascular resistance.
Device effects on exercise capacity and physiology
Patients in cardiogenic shock suffer from impending multiple organ failure, and many early deaths are due to progression to multiple organ failure, despite adequate pump outputs. For patients with a bridge-to-transplant, many investigators have documented return to normal or near-normal levels of hepatic, renal, hematologic, and pulmonary function [12, 16–19].
An important goal of chronic circulatory support is to provide the patient with a good quality-of-life. Exercise capacity has been one objective measurement to show that the patients have improved physical capacity and therefore quality-of-life. In bridge-to-transplantation exercise studies, it has been shown that, when the implantable LVAD reaches maximum pump output, the patient’s native heart may eject blood through the aortic valve and further contribute to total cardiac output [20, 21]. Therefore, total forward output of the heart may be significantly higher than LVAD output. However, correct interpretation of bridge-to-transplant studies may be difficult because the exercise studies and oxygen consumption studies are usually obtained early after the LVAD implantation, and the patient may still be recovering from surgery. Sequential studies of patients on longer support have shown improvement in exercise capacity beyond the initial studies (personal communication: OH Frazier, Houston, TX).
Exercise studies that may be useful in evaluating LVAD patients would include maximal or submaximal exercise studies, either with or without oxygen consumption, and a 6-minute walk test. These studies can provide objective measurements, but they do have limitations. Even after the patient has recovered from surgery, exercise capacity may be limited by noncardiovascular factors such as physical training, pulmonary dysfunction, or infection. It is too early in our clinical experience to define "acceptable" exercise capacity of patients recovered from surgery on chronic LVAD support.
Potential cardiac recovery while on device support
In the clinical bridge-to-transplant experience it has been noted that the cardiothoracic ratio decreased on chest film, left and right ventricular function appeared to improve with the device functioning, and histology showed a return to a more normal pattern in some patients [2]. In addition, it was noted that during brief periods with the device flow turned down after months of LVAD support, the patient’s native heart was able to provide improved hemodynamic function when compared to the pre-LVAD state. Therefore, it was reasoned that some patients may have recovered sufficiently so that the implantable LVAD can be "weaned" and subsequently surgically removed [2, 3, 22]. This in fact has occurred in a few clinical cases (personal communication: Peer Portner, PhD, Oakland, CA; Kurt Dasse, PhD, Woburn, MA).
Many areas need further exploration in this realm before guidelines can be written to specifically address them. These include indications for weaning and removing the device, proper technique for weaning, and long-term outcome following device removal. Clinical situations that may precipitate device removal could include infections, device failure, or patient dissatisfaction with the device. These clinical situations may occur in patients in whom there has been recovery, but are different than removal because the physician decided that cardiac recovery has occurred, and therefore device removal is warranted. Judgements of cardiac recovery would have to be based on the pathology of the underlying cardiac disorder, sequential echocardiographic observations of left and right ventricular function with the device functioning, and changes observed during device weaning. Ideally, there should be an evaluation of left and right ventricular function and hemodynamic parameters with the device turned off [23]. As a practicality, information with the LVAD "off" may be difficult to obtain because this could lead to blood stasis within the pump and the threat of thrombus formation and embolization when the pump is restarted. The most difficult problem will be predicting in which patients recovery will be sustained, so that the patient does not return to his previous condition after weeks or months of device removal.
Task Force 4: Assessment of quality of life
Bartley P. Griffith, MD, Chairman, Mary Amanda Dew, PhD, Co-Chairman, Roger Evans, PhD, Sharon Hunt, MD, Eleanor Schron, Fr Kevin O’Rourke, Charles-Julien Hahn, MD, Kay Kendell, RN, Dennis McNamara, MD
This section provides the rationale for and description of a standard protocol to evaluate health-related quality of life (HQL) in persons receiving cardiac support devices. The goal of the proposed HQL protocol is to enable sponsors to gather systematic, empirical data that can be used to aid in determinations of the safety and effectiveness of such devices.
Health-related quality of life has become a recognized and accepted endpoint in biomedical and clinical research. In fact, outcomes related to HQL may be the primary endpoints in studies where mortality differences are not anticipated [24–31]. The development of valid and reliable measures of HQL, plus growing evidence that HQL is sensitive and responsive to important biological and clinical changes, have led to its inclusion in clinical trials and quality-of-care evaluations across a spectrum of medical illnesses and conditions [24–27, 30, 32–42] (see also reviews by Testa and Nackley [43], and Wilson and Cleary [31]). The National Heart, Lung, and Blood Institute now routinely includes HQL assessments in clinical trials and efficacy studies of cardiovascular, lung, and blood therapies [44–47].
This emphasis is also growing in clinical practice settings and in health policy decision-making, where information about patients’ HQL after they have been given different therapies is increasingly being used by physicians and other health care providers for treatment decision-making and to plan allocation of resources [28, 36–39, 44, 48–50]. For example, HQL was a key consideration in the Health Care Financing Administration’s decision to extend Medicare coverage of recombinant human erythropoietin for dialysis-dependent end-stage renal disease [51, 52]. The data necessary for this decision were collected as part of the Phase III trial [35].
From a conceptual and medical ethics standpoint, the use of HQL as an indicator of treatment safety and efficacy is mandated by the World Health Organization’s definition of health as "a state of complete physical, mental, and social well-being and not merely absence of disease or infirmity" [53]. Indeed, the optimal assessments of HQL are multidimensional (as opposed to single, global indicators), which incorporate the measurement of patients’ functional status and well-being in each of the dimensions in the WHO definition [54, 55]. A multidimensional approach is essential in order to collect data adequate for the evaluation of the benefit to burden ratio associated with medical treatments [56].
Measurement approaches to health-related quality of life
In general, two major categories of instruments have been designed to assess HQL [29]. First, a number of generic measures have been created that assess health status in broad terms. They are useful for overall evaluations of HQL domains that are relevant to every individual, regardless of specific illness. They are highly useful because they allow comparisons across illness groups and across persons at different stages of the same illness. Second, disease- or illness-specific measures have been created to address unique symptoms and limitations related to particular health conditions. Such measures are useful for understanding finer-grained differences between patients with similar conditions and are very important in clinical trials designed to improve certain symptoms or illness states that are related to specific diseases or interventions. Also in the category of specific measures are instruments designed to focus on particular areas of functioning, such as psychiatric status or cognitive status.
Currently, the approach that is considered optimal for HQL assessment [29, 43, 57] is first to select a generic measure to serve as the core instrument, and then to supplement the core instrument with other measures to assess specific areas of particular concern to a given patient population. The protocol proposed here for patients receiving experimental cardiac support devices is based on this approach, yet was designed to be brief enough to be easily administered in the context of other clinical and physiological assessments of patients’ status. In sum—and as detailed below—the optimal HQL battery incorporates the following features:
Protocol for patients receiving cardiac support devices: Rationale for selection of relevant domains
Previous research and clinical experience with patients undergoing cardiac support as a "bridge" to transplantation, as well as with non-bridged heart transplant candidates and recipients, indicate the importance of considering patients’ overall well-being along each of the domains of quality of life defined above, that is, physical, functional, emotional, and social. Dew et al [58] provide a full review of this evidence. For example, it is well-established that HQL in these broad domains shows marked improvement, on average, after heart transplantation [36, 59–65]. There is also some early, primarily small case-series evidence that, among transplant candidates receiving support devices, HQL in these domains can be improved by certain interventions, such as intensive rehabilitative exercise programs [66, 67] or psychosocial programs whereby device patients are discharged (with device in place) to outpatient settings [54, 58, 68]. The fact that broad HQL measures are sensitive to changes in patients’ clinical status (ie, changes from pre- to post-transplant or changes from inpatient to outpatient status) makes them critical to include as outcomes of new medical procedures and treatments for end-stage heart disease.
Extensive clinical experience, and growing research data, have indicated that in addition to the broad domains discussed above, it is critical to consider three HQL subcomponents that are of special concern for cardiac support device patients. They include: (1) specific physical functional limitations in the areas of mobility and ambulation, sleep, and body care; (2) cognitive functioning; and (3) patients’ concerns and worries about the device itself. Device patients frequently report problems in engaging in specific physical functional activities; mobility limitations and insomnia, for example, are common complaints [58, 69]. Patients’ cognitive status is a well-recognized concern, given the possibility of device-related cerebrovascular accidents [44, 70]. Finally, cardiac support devices have distinct mechanical features that appear to affect quality of life, including for example, noise of the system, and potential anxiety about device failure [44, 71, 72].
A measurement protocol has therefore been proposed which incorporates both a core, generic measure of HQL, as well as supplemental assessments of HQL areas of particular importance for cardiac support device patients. The domains of measures are summarized in Table 2, which (1) lists and defines each HQL domain to be assessed, (2) provides examples of suitable instruments that could be used to assess these domains, and (3) gives estimated completion time for those instruments.
|
In addition, an item has been included in the recommended core assessment to reflect patients’ overall perceptions of the quality of their lives. The Evans item, which constitutes one example of a suitable measure for overall perceptions, has been employed in numerous patient populations [37–39].
The additional subcomponents to be assessed are: (1) specific physical functional limitations, (2) a cognitive functioning screen, and (3) patients’ concerns and worries about the device. These areas are to be evaluated with brief supplemental measures to be administered in conjunction with the core assessment. Examples of measures appropriate to each supplemental area are provided here.
A suitable measure to assess specific physical functional limitations is the Sickness Impact Profile [75] physical subscales. These have already been used extensively in cardiac support device recipients [54, 58]. Cognitive functioning can be screened for attention, psychomotor speed, visuospatial processing, and memory deficits with two standard screens, the Trail Making test and the Digit-Symbol Substitution Task [76]. Device-specific concerns can be assessed with a brief instrument adapted from measures that have been used with LVAD patients in St. Louis and Pittsburgh [44] and Cleveland [69].
Administration, formulation of indices, and analysis issues
The HQL battery summarized in Table 2 should be administered to patients longitudinally over time as patients continue with the device in place. It should be noted that the cognitive screen will show marked learning effects initially. However, after patients reach their "ceiling" performance level, which occurs rapidly with repeated administration, the screen provides a sensitive indication of cognitive change. All instruments should be administered by interviewers, rather than being completed by patients on their own. Data collection in an interview format is necessary to maximize the quality and integrity of the data over that which would be expected from self-administered paper-and-pencil assessments. The interviewers can be lay personnel (ie, none of the examples of instruments provided here require interviewers with advanced specialty degrees), and only brief training would be required to standardize the interviewers’ administration of these instruments. With the exception of the cognitive screen, the examples of instruments described above can be completed equally well in person or by telephone.
The instruments all yield standard indices that can be used as specific outcomes in any given clinical evaluation. The computation of these indices is described in the publications referenced for each instrument. These indices can be computed for data gathered on each occasion that the patient is interviewed, and statistical analyses can then be employed to make three major types of comparisons (depending on the sponsor’s key research questions):
The decision as to which of these three categories of comparisons would be most appropriate to any specific protocol would depend on the sponsor’s research questions. In general, any proposed HQL evaluation of device patients should include a list of specific aims and research hypotheses. The analytic plan should address issues of sample size and statistical power to detect hypothesized HQL effects. As with any clinical trial outcome, the analytic plan should propose statistical techniques suitable for analyzing repeated measures data, especially when those data are likely to involve censoring (eg, due to patient death) or missing data at some timepoints but not others. There is now widespread availability of appropriate statistical tools for these situations [77, 78].
Acknowledgments
The conference was conducted with financial and other support from the following organizations: American College of Cardiology, The Society of Thoracic Surgeons, and American Society of Artificial Internal Organs. Additionally, participants included members of these United States agencies: Food and Drug Administration and National Institutes of Health.
Footnotes
1 This Conference, sponsored by The Society of Thoracic Surgeons, Food and Drug Administration/US Department of Health and Human Services, was held at Heart House, Bethesda, MD, Oct 14–15, 1995. 
2 The recommendations set forth in this report are those of the Conference participants and do not necessarily reflect the official position of The Society of Thoracic Surgeons. This report was created through the efforts of the selected personnel and financial support of the sponsoring and participating organizations including the FDA and the NIH and may not reflect the official position of participating or sponsoring organizations.
Appendix 1. Additional information about suitable instruments for HQL battery
Core assessment of HQL
SF-36
The SF-36 is a 36-item measure that represents the state-of-the-art in brief health assessment methods [57, 74]. It was designed to focus on critical elements of general health and well-being and on the most efficient measurement of those elements. As a "generic" indicator of health status, it can be used in combination with disease- or illness-specific measures as an outcome in a variety of settings. Eight multi-item scales can be formed from the 36 items in the measure. These include physical functioning; general mental health; role limitations due to physical health problems; role limitations due to emotional problems; social functioning; vitality, energy or fatigue; bodily pain; and general health perceptions. A published manual provides extensive information on administration of the measure and on its reliability and validity [74]. The manual also includes norms for the general US population by seven age groups and by gender, and norms for a variety of acute and chronic medical conditions.
A reprint of the SF-36, plus a useful commentary on it and its applications may be found in the text by McDowell and Newell [79]. The SF-36 is in the public domain and no royalties are required for using it. However, permission to use it should be obtained from the Medical Outcomes Trust, a nonprofit organization that supports the development and use of standardized health outcome measures. For additional information, write to The Health Institute, New England Medical Center Hospitals, Box 345, 750 Washington St, Boston, MA 02111, USA.
Evans quality of life item
This item is reproduced in Figure 1. It provides a quick, global assessment of the quality of the respondent’s life, including both health-related and non–health-related elements. It was created by Roger Evans, PhD, for use in a variety of populations with chronic physical illnesses as well as in physically healthy cohorts [37–39].
|
Cognitive status: Trail making and digit-symbol substitution tasks
Both of these tests are well-known neuropsychological measures that are easily administered and scored. They serve as a quick, portable screen for deficits in attention, psychomotor speed, visuospatial processing, and memory. Lezak [77] provides a description, a review, and examples of forms or items for each test. Trail Making is in the public domain; copies may be obtained from Mary Amanda Dew, PhD, at the address given above. The Digit-Symbol Substitution Task is part of the Wechsler Adult Intelligence Scale [81] and may be ordered from The Psychological Corporation, Order Service Center, PO Box 839954, San Antonio, TX 78283-3954 (or contact Dr Dew for more information).
Patients’ device-specific concerns
This measure has not been previously published but has been in use in the Artificial Heart Program, University of Pittsburgh Medical Center for several years [82]. It is reproduced in Figure 2. For additional information about administering the measure, contact Mary Amanda Dew, PhD, at the address given above.
|
