Evidence Report/Technology Assessment

Number 27

Prepared for:
Agency for Healthcare Research and Quality
U.S. Department of Health and Human Services
2101 East Jefferson Street
Rockville, MD 20852

http://www.ahrq.gov/

Contract No. 290-97-0014

Prepared by:
Duke Evidence-based Practice Center
Center for Clinical Health Policy Research
Douglas C. McCrory, MD, MHSc
Principal Investigator

Gregory P. Samsa, PhD
Byron B. Hamilton, MD, PhD
Joseph A. Govert, MD
David B. Matchar, MD
Ruth E. Goslin, MAT
Jane T. Kolimaga, MA
Investigators

AHRQ Publication No. 01-E014

September 2001

ISSN:1530-4396
ISBN:1-58763-067-2

On December 6, 1999, under Public Law 106-129, the Agency for Health Care Policy and Research (AHCPR) was reauthorized and renamed the Agency for Healthcare Research and Quality (AHRQ). The law authorizes AHRQ to continue its research on the cost, quality, and outcomes of health care and expands its role to improve patient safety and address medical errors.

This report may be used, in whole or in part, as the basis for development of clinical practice guidelines and other quality enhancement tools, or a basis for reimbursement and coverage policies. AHRQ or U.S. Department of Health and Human Services endorsement of such derivative products may not be stated or implied.

AHRQ is the lead Federal agency charged with supporting research designed to improve the quality of health care, reduce its cost, address patient safety and medical errors, and broaden access to essential services. AHRQ sponsors and conducts research that provides evidence-based information on health care outcomes; quality; and cost, use, and access. The information helps health care decisionmakers -- patients and clinicians, health system leaders, and policymakers -- make more informed decisions and improve the quality of health care services.

The Agency for Healthcare Research and Quality (AHRQ), through its Evidence-based Practice Centers (EPCs), sponsors the development of evidence reports and technology assessments to assist public- and private-sector organizations in their efforts to improve the quality of health care in the United States. The reports and assessments provide organizations with comprehensive, science-based information on common, costly medical conditions and new health care technologies. The EPCs systematically review the relevant scientific literature on topics assigned to them by AHRQ and conduct additional analyses when appropriate prior to developing their reports and assessments.

To bring the broadest range of experts into the development of evidence reports and health technology assessments, AHRQ encourages the EPCs to form partnerships and enter into collaborations with other medical and research organizations. The EPCs work with these partner organizations to ensure that the evidence reports and technology assessments they produce will become building blocks for health care quality improvement projects throughout the Nation. The reports undergo peer review prior to their release.

AHRQ expects that the EPC evidence reports and technology assessments will inform individual health plans, providers, and purchasers as well as the health care system as a whole by providing important information to help improve health care quality.

We welcome written comments on this evidence report. They may be sent to: Director, Center for Practice and Technology Assessment, Agency for Healthcare Research and Quality, 6010 Executive Blvd., Suite 300, Rockville, MD 20852.

This report describes evidence on the respiratory management of persons with acute or chronic cervical level spinal cord injury (SCI), including mechanical ventilation and other interventions aimed at preventing or treating respiratory diseases.

Databases searched were MEDLINE (1966-Feb 2000), HealthSTAR (1975-Feb 2000), Cumulative Index to Nursing & Allied Health Literature (CINAHL) (1983-Feb 2000), and EMBASE (1980-Feb 2000). The search strategies included the MeSH terms spinal cord injuries, paraplegia, and quadriplegia [exploded] and text words for tetraplegia, quadriplegia, and paraplegia with a pulmonary disease concept. The search was limited to articles pertaining to humans and published in the English language.

The population of interest is adults with traumatic cervical SCI. Interventions considered include intubation and airway management, mechanical ventilation initiation, and weaning as well as medications (bronchodilators, mucolytics) and respiratory therapy (noninvasive positive pressure ventilation [NPPV], assisted cough, postural drainage, humidification, spirometry, vital capacity assessment). Evidence was considered from controlled or uncontrolled studies.

At least two reviewers independently screened titles and abstracts; references included by either rater were retained. Full reports were reevaluated according to the selection criteria and data describing study population, study design, interventions, and outcome data were recorded. Quality was assessed based on criteria related to external validity (characterization of the study population) and internal validity (strength of study design).

Patients with C4-level SCI have greater weaning success using progressive ventilatory-free breathing than synchronized intermittent mandatory ventilation techniques. In addition, high ventilator volume (more than 20 cc/kg) is associated with less atelectasis and faster weaning. Aggressive multimodal respiratory therapy interventions (including frequent turning, suctioning [and bronchial lavage], chest percussion and assisted coughing, inhaled bronchodilator treatments, deep breathing, and incentive spirometry) and rotating beds have been associated with reduced mortality, atelectasis, need for mechanical ventilation, or tracheostomy. Other secretion clearance modalities show evidence of improved cough (manual assisted cough, mechanical insufflator-exsufflator, glossopharyngeal breathing) but include no data on health outcomes. There is little evidence of an effect for other interventions, including active respiratory muscle exercise with incentive spirometry, inspiratory resistance training, and abdominal weight training.

Several alternatives to tracheostomy positive pressure ventilation (PPV) for long-term ventilatory support have been demonstrated, including electrophrenic respiration, noninvasive positive pressure ventilation, intermittent positive pressure breathing, pneumobelt, and glossopharyngeal breathing. Noninvasive ventilation may reduce the risk of pneumonia compared with tracheostomy PPV for patients requiring chronic ventilatory support.

Treatments aimed at improving ventilation, cough, and secretion clearance reduce atelectasis, pneumonia and the need for mechanical ventilation. Clinical research studies on pulmonary disease following cervical SCI cover only a small number of many important management decisions. Few studies use control groups (randomized or otherwise) or other designs to reduce bias.

This document is in the public domain and may be used and reprinted without permission except those copyrighted materials noted for which further reproduction is prohibited without the specific permission of copyright holders.

McCrory DC, Samsa GP, Hamilton BB, et al. Treatment of Pulmonary Disease Following Cervical Spinal Cord Injury. Evidence Report/Technology Assessment Number 27. (Prepared by the Duke Evidence-based Practice Center under Contract No. 290-97-0014.) AHRQ Publication No. 01-E014. Rockville, MD: Agency for Healthcare Research and Quality. September 2001.

The incidence of spinal cord injury (SCI) in the United States is approximately 10,000 new cases each year or 32 to 35 persons per million population. Because persons with SCI are surviving longer, the prevalence has been increasing and is now above 200,000 persons in the US. Despite being relatively uncommon, SCI is very expensive to treat incurring enormous costs for acute medical care, rehabilitation, medications and supplies, modification to home and vehicles, and personal assistance. SCI is the most expensive condition among all causes for hospitalization.

Injuries at the cervical level of the spinal cord, depending on the completeness of the lesion, can lead to tetraplegia, loss of sensory function, and paralysis of the respiratory muscles. Patients with paralyzed inspiratory muscles (principally the diaphragm) may not be able to breathe on their own and often must be placed on mechanical ventilators; patients with weak or paralyzed expiratory muscles (principally the abdominal and intercostal muscles) may have impaired ability to cough and clear mucoid secretions. The accumulation of retained secretions can lead to atelectasis and pneumonia.

Respiratory failure is the most common cause of death for patients with cervical injuries (particularly high-cervical injuries) during the acute phase of hospitalization, and one of the three most common causes of death subsequently; 30 percent during acute hospitalization and 20 percent subsequently. The rate of respiratory complications is highest in the weeks immediately following SCI; atelectasis and pneumonia are the most frequent respiratory complications.

This report assesses the evidence currently available on the prevention and treatment of pulmonary disease following traumatic cervical SCI. The report focuses on empirical studies relating to two research questions about (1) management of ventilatory insufficiency in acute and chronic phases, including weaning from mechanical ventilation (MV), management of secretions, atelectasis, and pulmonary infection, and (2) the prevention of late respiratory failure in patients with traumatic cervical SCI. Excluded from consideration were non-pulmonary complications of SCI and venous thromboembolism/pulmonary embolus. The report does not cover patients with SCIs occurring below the cervical level or respiratory muscle weakness caused by neuromuscular or other spinal cord diseases such as Guillain-Barré syndrome and polio.

The key questions addressed in the report are:

We focused on patients with acute traumatic cervical SCI, regardless of the degree of completeness of injury. We are interested in treatment in the days to months following acute injury and also the long-term followup over years.

Databases searched for literature were MEDLINE (1966-Feb 2000), HealthSTAR (1975-Feb 2000), Cumulative Index to Nursing & Allied Health Literature (CINAHL) (1983-Feb 2000), and EMBASE (1980-Feb 2000). The search strategies combined an SCI concept (implemented using MeSH terms spinal cord injuries, paraplegia, and quadriplegia [exploded] and text words for tetraplegia, quadriplegia, and paraplegia) with a pulmonary disease concept. The search was limited to articles pertaining to humans and published in the English language.

Empirical studies or review articles were included after screening by the following criteria: (1) the study population includes traumatic cervical SCI; (2) study question relates to the research questions described above; (3) study includes data on health outcomes, health services utilization or economic outcomes, or physiological measures related to respiratory status; (4) study design is controlled trial, prospective trial with historical controls, prospective or retrospective cohort study, or medium- to large-sized (more than 20 subjects) case series.

Each article was independently reviewed by at least two investigators.

Patients with cervical SCI are at significant risk for ventilatory failure, and this risk differs by the level and completeness of injury. Ventilatory support is needed for a majority of patients with C5 and higher injuries and virtually all patients with C3 and higher injuries in the acute phase. Patients with lower-cervical SCI are less likely to require mechanical ventilation (MV), as are those patients with greater forced vital capacity (FVC) on presentation. The amount of secretions and serious pulmonary infection (pneumonia) are associated with the need for MV. This finding suggests that treatments aimed at improving ventilation, cough, and secretion clearance may reduce the need for MV.

MV for ventilatory failure is usually performed using invasive (endotracheal or nasotracheal) intubation. However, NPPV can be successful for short- and long-term ventilatory assistance when used properly, particularly when patients have spontaneous FVC approaching 1 liter. In patients with C4 level of injury (LOI), weaning from MV is often a protracted process that is sometimes not successful. However, patients with C4-level SCI appear to have greater success when weaned by progressive ventilatory-free breathing than by synchronized intermittent mandatory ventilation techniques, which are more commonly used in patients with other causes for ventilatory failure. Furthermore, high ventilator volume (more than 20 cc/kg) is associated with less atelectasis and faster weaning.

Evidence is available on a variety of therapeutic and clinical assessment strategies among patients with traumatic tetraplegia, including the following:

• Active respiratory muscle exercise with incentive spirometry, inspiratory resistance training, and abdominal weight training have all been associated with improvements in spontaneous vital capacity (VC) in prospective case series. However, in small randomized controlled trials (RCTs) these techniques have not been shown to improve ventilation, nor have the trials shown any difference between inspiratory resistance training and abdominal weight training.
• Deflating cuffed tracheostomies or switching to cuffless tracheostomies is necessary to permit speech, but can cause hypocapnia as a result of increasing tidal volume to compensate for variable upper-airway leak. Adding dead space can increase pCO₂ to near-normal levels, reducing hypocapnia during waking and sleeping periods. Speech valves are available to permit better speech.
• Electrophrenic respiration has been successfully used for chronic ventilatory support, most often for part-time ventilatory support. Technological advances in recent years appear to have improved the success of the technique, but reliability remains an important problem.
• The pneumobelt may be used for part-time ventilatory support or as an adjunct to full-time non-invasive ventilatory support.
• Noninvasive ventilation may reduce the risk of pneumonia compared with tracheostomy positive pressure ventilation for patients requiring chronic ventilatory support. This technique may also prevent the need for invasive ventilation in patients with acute ventilatory insufficiency.
• Glossopharyngeal breathing training can allow patients ventilator-free breathing for short periods and improve the effectiveness of cough and audibility of their voices.
• Cough is improved with manual assisted cough by 15 to 33 percent in two studies. Positive pressure insufflation can improve cough as well. These techniques used together provided near-normal cough peak flows. Although abdominal binding (corset) did not improve spontaneous cough when used alone, when used in conjunction with manual assisted cough, positive pressure insufflation, or both, a trend toward improvement was suggested by one study. It follows that cough can also be improved by mechanical insufflation-exsufflation, as this technique provides 10 liters/second of expiratory flow.
• Rotating beds have been associated with lower rates of pulmonary complications compared with less frequent turning with a wedge turning device in patients with cervical SCI in retrospective case series.
• Aggressive multimodal respiratory therapy interventions (including frequent turning, suctioning [and bronchial lavage], chest percussion and assisted coughing, inhaled bronchodilator treatments, deep breathing, and incentive spirometry) have been associated with improvements in atelectasis in a small case series. The interventions have also been associated with reduced mortality, atelectasis, need for MV, and tracheostomy in a larger historical cohort comparison.
• The reduction in VC associated with a change in position from supine to sitting is both statistically significant and clinically important. Hence, the seated position in which spirometric tests are usually performed is not optimal for patients with cervical SCI, in that FVC will be underestimated in this position.
• Periodic chest roentgenography often discloses lesions in patients with tetraplegia, even in the absence of symptoms. Conversely, normal upright chest roentgenograms are not sensitive enough to exclude pulmonary abnormalities, particularly pleural effusions (which can be seen on lateral decubitus views).
• In imaging the upper airway, tomography had better agreement with endoscopy for tracheal than glottic stenosis. Computed tomography was more accurate than tomography for diagnosing glottic stenosis, but was similarly accurate for grading the severity of tracheal stenosis.
• Studies of bronchodilator effects suggest potential therapeutic benefit from long-term prophylactic therapy with inhaled beta-agonists to reduce symptoms of breathlessness associated with airway reactivity in persons with tetraplegia. However, none of these studies evaluated the long-term efficacy of bronchodilator or other drug treatments on symptoms, pulmonary function, or the incidence of pulmonary complications.

We also conclude the following:

• Very little data are available with which to describe the risk of late ventilatory failure or the potential effectiveness of treatments to delay or prevent it.
• Despite the expectations of non-disabled persons, the long-term perceived quality of life and well-being are similar in ventilator-dependent and autonomously breathing patients with SCI.

Further research must be performed to answer definitively questions about the best care for patients with acute and chronic cervical SCI. The evidence base of clinical research on the management of pulmonary disease in this condition covers only a small number of many important management decisions. In addition, the amount and quality of the literature is relatively poor, with few studies using RCT or other designs to reduce bias. Clinical research on patients with cervical SCI has primarily been limited to observational studies, and primarily retrospective case series. Determining the influences of the interventions is problematic for several reasons, including imprecise characterization of treatments and lack of appropriate controls. Few of the extant studies have any type of comparison group. Case series can provide rates of outcome events (e.g., complications, successful weaning) that can be compared across series (benchmarks). However, case series provide biased and imprecise estimates of relative effectiveness and are less powerful than study designs that offer internal comparisons (e.g., RCTs, concurrent cohort comparisons, and historical cohort comparisons).

There is also a need for more precise and consistent characterization of the patients in terms of important prognostic features such as LOI, completeness, and time since injury. Most of the evidence comes from retrospective case series in which the intervention is difficult to characterize. In various studies, the method of clinical care is either not described or described individually for each patient in the series. In some studies that describe well the protocol for care, it is not possible to estimate the effects for individual components of care.

Finally, many of the clinical studies had small sample sizes. A small size limits the generalizability of a case series. For more well-designed studies, the small sample sizes limit the statistical power; for example, the few RCTs identified in this study were negative and lacked the statistical power to show clinically important differences. Future studies should seek to improve both the generalizability and the statistical power by including more patients. Because of the relative rarity of the condition, it will be necessary to aggregate patients through merging standardized data sets from many institutions or developing multisite research networks.

The US Model Spinal Cord Injury Systems program is a network of (currently) 18 centers that work together to maintain a national database, provide continuing education, and participate in independent and collaborative research relating to SCI. The "Model Systems" database has contributed greatly to our knowledge of the prevalence, natural history, cost, and sequelae of SCI. The Model Systems data have clearly demonstrated improvement in survival of patients with SCI over the last few decades. Furthermore, the database, through measurement of incidence of complications, ventilator use at discharge, and other parameters, has been used to create benchmarks to guide the care of persons with SCI.

Currently the Model Systems database does not include detailed clinical data of the type needed to examine how differences in clinical practice (e.g., ventilator settings, frequency and type of respiratory therapy) might influence clinical outcomes such as weaning success, time on MV, incidence of atelectasis, or mortality. Agreeing on specific detailed clinical data to collect on day-to-day management practices for persons with SCI, and expanding the Model Systems database to include such data, would make it possible to correlate clinical practices with the health outcomes already obtained in Model Systems data. This would seem a natural extension of the current effort and would facilitate inter-site collaboration in the design and conduct of prospective clinical trials, which would be necessary to definitively answer questions of clinical management.

The Department of Veterans Affairs (VA) network of 23 designated SCI centers handles approximately two-thirds of VA inpatients visits for persons with SCI. The Department has recently developed a national registry of veterans with spinal cord dysfunction to allow more effective planning and administration for SCI care, both in the specialized centers and throughout the network of VA hospitals and clinics. Clinical data, obtained by linking data from the Department's patient treatment file, is, like the Model Systems database, insufficiently detailed for effective outcomes research.

Further queries of the Model Systems and VA databases are warranted to describe the occurrence of respiratory illnesses and complications in the course of care following SCI. Few of the currently available reports of Model Systems data have focused primarily on respiratory complications. Despite the lack of detailed data on treatment, the database does provide an opportunity to better describe the epidemiology of respiratory illness and complications. This more-detailed information could be useful for estimating the burden of illness, for modeling, or for cost-effectiveness analysis. These type of data, in turn, will provide motivation for funders to support collaborative clinical research efforts such as those described above.

An active clinical research network would offer the opportunity to implement protocols across multiple institutions, and to test for differences in outcomes associated with different protocols. A variety of management algorithms have been either described in the literature or put to use within systems or single institutions. None of these guidelines was developed using an explicit evidence-based process and none has been rigorously tested in clinical practice. The evidence compiled in this report will be used by the Consortium for Spinal Cord Medicine in their efforts to develop a new guideline using a more explicit process linking recommendations to evidence. Much of the justification for the current respiratory care of patients with tetraplegia comes from data on or experience with patients with other illnesses, thus guidelines on this topic will need to be based to a large extent on interpretation of these data. Explicit methods are needed for integrating data on related populations with our understanding of the pathophysiology of cervical SCI and other diseases. The Consortium for Spinal Cord Medicine, professional associations and similar groups, after examining the current state of research in this area, would be in a good position to outline an agenda for future research by prioritizing questions that would be most useful in guiding clinical care.

Most of the published research on pulmonary disease in SCI focuses on care of acutely injured subjects. As acute SCI is the most expensive condition among all causes for hospitalization, the acute management phase should be a high priority for research. However, as long-term SCI survival has increased, the prevalence of patients with chronic SCI grows ever larger. Pulmonary complications of chronic SCI will become of greater clinical importance. Currently, little data is available regarding the incidence of late respiratory failure, and even less is available on the effectiveness of therapies that may be effective to prevent it. In this instance, venues such as the VA Spinal Cord Injury Centers and the Model Systems institutions (in which cohorts of patients with chronic cervical SCI are followed) might decide that the systematic collection of data about their patients' clinical care could be valuable.