Published ahead of print on May 29, 2008, doi:10.1164/rccm.200712-1829OC
© 2008 American Thoracic Society doi: 10.1164/rccm.200712-1829OC
Acquired Weakness, Handgrip Strength, and Mortality in Critically Ill Patients1 Division of Pulmonary, Allergy, Critical Care and Sleep Medicine, 2 Department of Internal Medicine, Center for Biostatistics, Ohio State University, Columbus, Ohio; 3 Division of Critical Care, Department of Medicine, University of Manitoba, Winnipeg, Manitoba, Canada; 4 Division of Pulmonary and Critical Care Medicine, MetroHealth Medical Center, Cleveland, OH; 5 Division of Pulmonary, Critical Care and Sleep Medicine, Department of Internal Medicine, University of Cincinnati, Cincinnati, Ohio; 6 Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, University Hospitals Case Medical Center, Case-Western Reserve University, Cleveland, Ohio; 7 Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, Indiana University Medical Center, Indianapolis, Indiana; 8 College of Public Health, Ohio State University, Columbus, Ohio; and 9 Department of Internal Medicine, MetroHealth Medical Center, Columbus, Ohio Correspondence and requests for reprints should be addressed to Naeem A. Ali, M.D., 201G DHLRI, 473 W. 12th Avenue, Columbus, OH 43221. E-mail: Naeem.ali{at}osumc.edu
Rationale: ICU-acquired paresis (ICUAP) is common in survivors of critical illness. There is significant associated morbidity, including prolonged time on the ventilator and longer hospital stay. However, it is unclear whether ICUAP is independently associated with mortality, as sicker patients are more prone and existing studies have not adjusted for this. Objectives: To test the hypothesis that ICUAP is independently associated with increased mortality. Secondarily, to determine if handgrip dynamometry is a concise measure of global strength and is independently associated with mortality. Methods: A prospective multicenter cohort study was conducted in intensive care units (ICU) of five academic medical centers. Adults requiring at least 5 days of mechanical ventilation without evidence of preexisting neuromuscular disease were followed until awakening and were then examined for strength. Measurements and Main Results: We measured global strength and handgrip dynamometry. The primary outcome was in-hospital mortality and secondary outcomes were hospital and ICU-free days, ICU readmission, and recurrent respiratory failure. Subjects with ICUAP (average MRC score of < 4) had longer hospital stays and required mechanical ventilation longer. Handgrip strength was lower in subjects with ICUAP and had good test performance for diagnosing ICUAP. After adjustment for severity of illness, ICUAP was independently associated with hospital mortality (odds ratio [OR], 7.8; 95% confidence interval [CI], 2.4–25.3; P = 0.001). Separately, handgrip strength was independently associated with hospital mortality (OR, 4.5; 95% CI, 1.5–13.6; P = 0.007). Conclusions: ICUAP is independently associated with increased hospital mortality. Handgrip strength is also independently associated with poor hospital outcome and may serve as a simple test to identify ICUAP. Clinical trial registered with www.clinicaltrials.gov (NCT00106665).
Key Words: polyneuropathy, critical illness • muscle weakness • hand strength
The development of generalized weakness related to critical illness is a common and important complication for many patients in intensive care units (ICU) (1, 2). While most critically ill patients likely experience some weakness, it is only termed ICU-acquired paresis (ICUAP) (1, 3) when severe. Patients with ICUAP have a constellation of pathologic findings in peripheral nerve and skeletal muscle that has been described as critical illness polyneuromyopathy (CIPNM) (4–6). Recovery from this disorder may take months or years (7). Subclinical weakness may also contribute to the physical limitations commonly found in survivors of critical illness (8–11). Increased mortality is inconsistently observed in studies of subjects with either ICUAP or its physiologic surrogates critical illness polyneuropathy (CIP), myopathy (CIM), or neuromyopathy (CIPNM) (1, 12, 13). Several investigators suggest that the varied diagnostic approaches used in ICU-acquired weakness syndromes contributes to this inconsistency (3, 14). In prospective studies of subjects who survive to awakening, patients with ICUAP have increased observed mortality in unadjusted analyses. However, these same patients were also older and experienced more hyperglycemia and multiple organ failure. Any or all of these could have influenced outcome (1, 13, 15). To prevent (16, 17) or treat (18) ICUAP, the syndrome must first be recognized. Diagnostic strategies include specialized neurophysiologic testing and the bedside Medical Research Council physical strength exam. Neurophysiologic testing, such as electromyography, can be performed easily on most ICU patients; however, interpretation requires special expertise and is not universally available in all ICUs. In contrast, the bedside strength exam can be performed by most clinicians. However, to be performed properly the strength exam requires an awake and attentive patient with the mobility and stamina to participate in the evaluation of all twelve major limb muscle groups. A simpler assessment may result in improved recognition of ICUAP. One such measure is handgrip dynamometry, which has been used as a surrogate for global strength in other neuromuscular diseases (19, 20). We performed a multicenter study in critically ill, mechanically ventilated patients to test the hypothesis that ICUAP is associated with increased mortality independent of severity of illness or organ failure. Our secondary hypothesis was that handgrip strength as measured by dynamometry would provide a concise measure of global strength and would be independently associated with mortality. Some of the results of this study have been previously reported in abstract form (21).
We conducted a prospective multicenter cohort study from May 2005 through April 2007 at five medical ICUs in academic medical centers affiliated with the Midwest Critical Care Consortium (see APPENDIX A). Adult patients (age  18 yr) were eligible for enrollment if they required mechanical ventilation for at least 5 days. Exclusion criteria included patients with known diagnoses causing generalized weakness, mechanical ventilation for more than 24 hours before transfer from a referring hospital, surrogate or physician not committed to full support, inability to communicate with an examiner, and those without at least two limbs to examine. The complete list of exclusion criteria are listed in APPENDIX B. We collected baseline data including demographics (age, race, sex, and ethnicity), co-morbidities associated with weakness (diabetes mellitus, alcoholism, HIV, and a history of stroke), severity of illness (Acute Physiology and Chronic Health Evaluation [APACHE] III) (22) and organ failures (Sequential Organ Failure Assessment [SOFA]) (23) on the day of enrollment (Ventilator Day 5). We assessed reasons for admission through review of the physician notes in the medical record.
Screen for Awakening and Strength Exams ICUAP was defined as an average muscle strength score of less than 4 (anti-gravity strength) across all muscles tested, as described previously (1). Since grip strength is influenced by sex (29), we used sensitivity analysis to determine the optimal handgrip strength cutoff to identify ICUAP for males and females. A force value of less than 11 kg-force for males and less than 7 kg-force for females resulted in the maximum combination of sensitivity and specificity for the diagnosis of ICUAP.
Follow-up and Outcomes
Human Subjects Protection
Sample Size Estimates
Statistical Analysis
Cohort Development One-hundred seventy-four subjects were enrolled in the study and 136 (78%) survived to awakening, allowing strength measurement. All 38 subjects not examined for strength were deemed continuously ineligible from enrollment to death or hospital discharge. Subjects not examined had higher mortality (68.4% versus 12.6%, P < 0.001), were more often nonwhite (36.8% versus 19.8%, P = 0.03), less likely to have diabetes mellitus (13.2% versus 30.4%, P = 0.03), and had higher average APACHE III score at enrollment (102 versus 66, P < 0.001). The subjects undergoing strength examinations formed the final study cohort (Figure 1).
Strength Exams Figure 2 shows the average muscle strength for each muscle group assessed. The majority of patients (68.8%) were able to have all 12 muscle groups examined; 28.5% of the cohort had at least 9 muscle groups examined (11 groups, 13.2%; 10 groups, 10.2%; and 9 groups, 5.1%). One subject only had eight and two had only six groups examined. Consistent with previous reports (1), weakness was seen across all muscle groups tested. Inter-rater exams were performed in 8.8% (n = 12) of subjects by blinded examiners. There was complete agreement between examiners MRC designation of ICUAP and the average MRC scores were highly correlated (Pearson's correlation coefficient = 0.96, P < 0.001). One-hundred twenty-three subjects (90.4%) received serial exams and again there was excellent agreement in ICUAP assignment (Kendall's tau-b = 0.96, P < 0.001) and average MRC scores were highly correlated (Spearman's correlation coefficient = 0.90, P < 0.001). In subjects with ICUAP, the maximum handgrip strength was significantly lower. Using sex-specific thresholds (males, < 11 kg-force; females, < 7 kg-force; Figure 3), handgrip strength had good test performance (overall sensitivity 80.6%, specificity 83.2%, negative predictive value 92.3%, positive predictive value, 63.0%) when compared with an ICUAP diagnosis by MRC exam. Performance did not differ significantly by sex (males: sensitivity 78.6%, specificity 82.4%, negative predictive value 93.3%, positive predictive value, 55.0%; females: sensitivity 81.8%, specificity 84.0%, negative predictive value 91.3%, positive predictive value, 69.2%).
Cohort Characteristics Thirty-five subjects (25.7%) in the examined cohort had ICUAP. Demographics and severity of illness at enrollment (Day 5 of mechanical ventilation) are presented in Table 1. There were no major differences in the demographics of subjects with or without a diagnosis of ICUAP. Severity of illness was significantly higher in the group with ICUAP. Average glucose levels and intravenous insulin use were equivalent during follow-up between the groups.
Weakness and Outcomes Outcomes were significantly different between the strength groups. In unadjusted analyses, mortality increased as average muscle strength or maximum handgrip strength declined (Figure 4). Hospital mortality was higher in patients with ICUAP than those without weakness (Table 2). After adjustment for severity of illness and organ failures, the odds of hospital mortality were significantly higher in subjects with ICUAP (odds ratio [OR], 7.8; 95% confidence interval [CI], 2.4–25.3; P = 0.001, Table 3) by MRC exam. We observed similar results when we used handgrip force as the risk factor (Table 3). An additional analysis using a propensity score to account for factors associated with ICUAP (including age, sex, ventilator days, and organ failures before MRC exam and severity of illness) yielded similar increases in mortality risk for subjects with ICUAP (OR, 5.2; 95% CI, 1.5–18.3; P = 0.01). After risk adjustment, the numbers of ICU- and hospital-free days were also significantly reduced in ICUAP subjects by MRC exam. There also was a strong association between handgrip strength and the number of ICU- and hospital-free days (Table 3). Finally, there was significantly greater morbidity in those subjects with ICUAP as measured by the secondary outcomes (Table 2).
We have prospectively shown that, among patients requiring at least 5 days of mechanical ventilation, ICUAP, assessed by MRC exam or handgrip dynamometry, is independently associated with hospital mortality, ICU, and hospital-free days. Handgrip dynamometry may provide a rapid, simple alternative to the comprehensive MRC examination for the diagnosis of ICUAP. Previous studies have reported that ICUAP is a morbid disease, but have not definitively demonstrated an independent association with mortality (1, 30–32). One study showed that mortality was independently associated with CIPNM (13) as diagnosed by electrophysiologic testing. It is unclear if these patients ever manifest weakness, as many may have died before awakening. It is possible that electrophysiologic abnormalities are more common than clinical weakness (ICUAP) and may have different implications. For example, one study found electromyographic abnormalities in 58% of patients who required 7 or more days of mechanical ventilation (12). But in a separate study of a similar population, ICUAP only occurred in 25.3% (30). While not a definitive comparison of the prevalence rates, the possibility exists that these techniques identify different patient groups. Such variation has been suggested as a possible limitation in interpreting studies of this syndrome (14). While it may be important to identify ICUAP or CIPNM, there are barriers to accomplishing this goal. The availability of electrophysiologic equipment or the expertise to interpret the results can limit the utility of this test. Since limb positioning is important for achieving valid results, the ICU environment can contain significant impediments to performing bedside strength assessments. As a result, some clinicians may perceive that the strength exam is difficult and time consuming to perform. While we did not find the MRC exam in our study to be burdensome, it was done by researchers who dedicated specific time for this exam, rather than busy clinicians. It is possible that the perceived barriers to performing the full exam may lead to delays in formally assessing muscle strength. In addition, the MRC scale has a subjective component (for example, MRC 5 = active movement against "full" resistance; APPENDIX C) (26), which may introduce some variability in the measure. However, our data indicate that when trained properly, physicians can generate highly reproducible results. Because handgrip strength identifies patients at increased risk of death, it might provide a reasonable alternative for diagnosing ICUAP and identifying patients at risk of poor outcomes and candidates for interventions to mitigate such risk. While we did not collect data about the effort required for each exam, all subjects able to participate with the MRC exam were also able to perform handgrip dynamometry. The time and precautions to perform handgrip dynamometry would likely be less than the strength exam as there is no need to reposition the patient for testing of multiple muscle groups. Because handgrip dynamometry does not require extensive repositioning and results in a more objective numeric value (28), it may be able to be performed more easily as a screening tool for global paresis in a busy ICU practice. Additional research is needed to understand the appropriate threshold of handgrip strength at which critically ill patients are at increased risk of death. Given that we attempted to determine the threshold of handgrip that optimized the diagnosis ICUAP, these values may not accurately describe the wide range of strength that may be "normal." Studies in healthy adults suggest that age and gender both affect normal handgrip strength (29). Our analysis suggests that using a force value cutoff for each sex (males, < 11 kg-force; females, < 7 kg-force) is adequate. However, our cutoff values are well below age- and sex-matched hospitalized patient's "normal" values (33), making it possible that different factors influence handgrip strength in critically ill patients than normal volunteers. Factors involved with testing hand strength, like handedness, bed position, and upper extremity entrapment syndromes, should be more formally examined as well. We also suggest that while most studies emphasize proximal muscle weakness in ICUAP (1, 34), the involvement of distal muscle groups may also be important. The observed association between handgrip weakness and ICUAP should lead to further exploration of the pathogenesis and treatment of the syndrome. There are limitations to our study. First, we did not employ neurologists in the performance of the strength exams. We chose this approach to improve the external validity of our observations for bedside intensivists. Our sample of inter-rater assessments suggested this still produced reliable exams. In addition, given that our cohort is very similar in age, sex, and co-morbidities to those in previously published studies (1, 34), we feel that our exams identified patients with true ICUAP. All ICUs participating in this study used ventilator liberation and sedation protocols that included daily wake-ups and self-breathing trials, but these practices were not standardized and could have influenced outcome. We were unable to standardize these practices for practical reasons, but analyses of mortality by site found no significant differences. Because we observed fewer deaths than anticipated in the examined cohort, our ability to adjust for some clinically important factors was limited. However, we remain confident in our results since adjustment for severity of illness and the use of a propensity score to "match" patients for factors associated with ICUAP did not eliminate this association. In addition, the results of our secondary outcome analyses support our mortality observation. Finally, we do not have information as to the causes of respiratory failure or death in our study participants, preventing us from drawing conclusions as to the mechanisms by which ICUAP could lead to mortality. Despite these limitations, we feel that this study has important implications. We believe handgrip dynamometry could be used to screen patients for ICUAP by anyone in the ICU, leading to earlier patient identification. Several strategies have been developed to prevent ICUAP (17, 35), but few studies have explored the treatment of clinically evident disease. Given that some of the known risk factors associated with ICUAP are nonmodifiable, it is likely that prevention alone will not eliminate the burden of ICUAP. Therefore, treatments should be studied to affect outcomes for those with established ICUAP while trying to reduce its incidence. Handgrip dynamometry could represent a useful tool to longitudinally assess strength recovery, allowing a way to monitor the effectiveness of these interventions. Our observations advance those made by other investigators (1, 13) and, for the first time, show that ICUAP is independently associated with increased odds of death after controlling for severity of illness and organ failures. In addition, we show that a simple measure of grip strength can serve as a useful surrogate for the MRC examination and is itself independently associated with hospital mortality. Further studies are needed to determine if handgrip strength can be used in a clinical prediction rule to help identify ICUAP patients as early as is feasible. Early identification of critically ill patients with possible ICUAP who are at increased risk of death might allow for more expeditious treatments that could reduce the morbidity and mortality of this disease.
Members of the Midwest Critical Care Consortium: Case Western Reserve University, University Hospital Case Medical Center, Cleveland, Ohio—Rana Hejal, Jeffrey Kern; Case Western Reserve University, MetroHealth Medical Center—Alfred F. Connors, Jr., James C.W. Finley, Allan Garland, Ted Warren; Indiana University—Karen M. Wolf; The Ohio State University Medical Center—Naeem A. Ali, Nitin Bhatt, Elliott Crouser, Stephen P. Hoffmann, Clay B. Marsh, John Mastronarde, James M. O'Brien, Jr.; University of Cincinnati—Khalid Almoosa, Frank McCormack, Mitch Rashkin. Data Coordinating Center (The Ohio State University)—Stanley Lemeshow and Gary Phillips.
Medical Research Council scale for evaluating peripheral muscle strength (26). 0: No muscular contraction detected; 1: barely detectable flicker or trace of contraction; 2: active movement with gravity eliminated; 3: active movement against gravity; 4: active movement against gravity and some resistance; 5: active movement against gravity and full resistance.
The Midwest Critical Care Consortium study of ICU-acquired weakness was performed at The Ohio State University Medical Center in Columbus, Ohio; MetroHealth Medical Center and University Hospitals Case Medical Center, both in Cleveland, Ohio; Indiana University Hospital in Indianapolis, Indiana; and the University of Cincinnati in Cincinnati, Ohio. The authors thank Wendy King, P.T., and Miriam Freimer, M.D., for their help in developing the exam training tools used for this study.
Originally Published in Press as DOI: 10.1164/rccm.200712-1829OC on May 29, 2008 Conflict of Interest Statement: N.A.A. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. J.M.O. has served as Principal Investigator of a clinical trial by Aerogen (8/05–6/06) but received no direct funds, served as a paid consultant to Medical Stimulation Corporation ($4,000, 2005–2006), co-authored a manuscript with Eli Lilly employees (2006), and currently serves as a unpaid consultant to Keimar, Inc. S.P.H. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. G.P. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. A.G. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. J.C.W.F. has been a co-principal investigator on two multicenter research grants from Boehringer Ingelheim, Inc. that included investigation of Tiotropium and Mirapex; K.A. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. R.H. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. K.M.W. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. S.L. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. A.F.C. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. C.B.M. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form December 14, 2007; accepted in final form May 27, 2008
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
AT A GLANCE COMMENTARY
level of 0.05. We planned to enroll 170 subjects to account for a 10% expected mortality before a patient being eligible for examination. 
