Evaluating Physical Outcomes in Acute Respiratory Distress Syndrome Survivors: Validity, Responsiveness, and Minimal Important Difference of 4-Meter Gait Speed Test. Chan KS, Friedman LA, Dinglas VD, Hough CL, Morris PE, Mendez-Tellez PA, Jackson JC, Ely WE, Hopkins RO, Needham DM. Critical Care Medicine. 2016 May;44(5):859-68.
This is a pair of secondary analyses from two large multicenter prospective ARDS studies (ARDSNet Long-Term Outcome Study – ALTOS – and Improving Care of Acute Lung Injury Patients – ICAP.) The studies followed more than 300 patients (all of whom had the 4-meter gait speed test done) for up to 60 months after discharge from the hospital. ALTOS participants were tested at 6 and 12 months, while ICAP patients were tested at 36, 48 and 60 months.
The 4-meter gait speed test is part of the Short Physical Performance Battery, and is recommended by the NIH (www.nihtoolbox.org) as a standardized physical performance measure for locomotion. Other measures of physical performance have also been validated previously including the 6-minute walk test and the SF-36 PF. The 4-meter gait speed test is particularly attractive for clinicians and researchers because it requires significantly less time and/or physical space to conduct the evaluation than the other measures.
The study group (including Drs. Jackson and Ely from Vanderbilt, and Dr. Needham from Johns Hopkins) validated the 4-meter gait speed test in this group of ARDS survivors with respect to other well-known functional tests and patient outcomes. With regard to construct validity, the gait speed test correlated with patient performance on the 6-minute walk test, the SF-36 physical function domain, and the EQ-5D mobility survey. The study group also tested the responsiveness of the tool: the results of the gait speed test were predictive of patient outcomes as well. Decreases in gait speed correlated with increasing risk of mortality and hospitalization; increases in gait speed correlated with increasing “risk” of remaining out of the hospital (ORs 0.78 and 0.77 for ALTOS and ICAP, respectively) alive at home (ORs 1.07 for ALTOS and 1.94 for ICAP, though only ICAP reached significance) and returning to normal activity (ORs 1.0 for ALTOS, but 1.34 for ICAP). In general, the ICAP study tended to have more significant ORs (in this case, meaning that the 95% confidence interval did not include 1.0). It is interesting to speculate that this is due to the longer follow up times in that study.
This paper was interesting not only because it touches on the hidden outcomes of ICU survivorship (return to work, return to normal activities of daily living, re-hospitalization, early mortality) but because of the thorough validation procedures followed by the study group (validating not only to convergent measures – grip strength, pulmonary function testing – but also to divergent measures – mental health, PTSD, depression and anxiety).
This is a useful tool not only in predicting risk of poor outcome in the ICU survivor population, but also in fall risk identification and prevention in the geriatric population. It is simple to use, intuitively easy to understand, and requires little time, equipment or testing space.
The following two articles (Article 2 from Journal of Critical Care and Article 3 from Critical Care Medicine) relate to the Critical Care Pain Observation Tool, which is one of only two tools recommended in the Pain, Agitation and Delirium CPG published in 2013 (reference follows after Article 2). The CPOT has not been previously validated in some fairly common subsets of critically ill patients, including patients with delirium and patients with brain injury.
The CPOT has 4 behavioral categories: 1) facial expression, 2) body movements, 3) muscle tension, and 4) ventilator compliance or vocalizations. Each is scored from 0-2, and the scores summed up. Any score of 2 or greater is consistent with patients experiencing pain. The CPOT has been validated in critically ill adult populations and demonstrates both high discriminant validity (CPOT scores are consistently higher during painful stimuli compared to non-painful stimuli) and good correlation with patient self-reporting.
Validation of the Critical-Care Pain Observation Tool in brain-injured critically ill adults. Joffe AM, McNulty B, Boitor M, Marsh R, Gelinas C. J Crit Care. 2016 Dec;36:76-80.
The CPOT was tested in the Neuroscience Intensive Care Unit (NSICU) at Harborview Medical Center in Seattle. This was a convenience sample of 80 patients admitted for traumatic brain injury (27%), aneurysm (29%), tumor resection (14%), or stroke (10%). APACHE II scores averaged 18.4. Exclusion criteria for this study involved causes of immobility of the face or body: severe obtundation (GCS >4); spinal cord injury; chemical paralysis; or deep sedation (RASS -5). Trained observer pairs (medical student, nurse) were used to measure the four behavioral domains of the CPOT before and during painful procedures (turning, ETT suction) as well as before and during non-painful stimulus (gentle touch). For the pre-procedural observations, the muscle tension domain of the CPOT was assessed by passive flexion and extension of the patient’s arm. In this study, 28 of the 80 patients were able to self-report; the observers were blinded to the patients’ reports. Notably, the ability to report pain included the mere capability of nodding the head; 52 of the patients observed could not even reliably do so little.
The median CPOT scores for all pre-procedural assessments were 0, indicating absence of pain. Median CPOT scores in this brain-injured population during painful procedures ranged from 2.5 to 3.0. The most common source of behavioral scoring was facial expression (93% of patients who were turned as painful stimulus had facial indicators of pain). The least common behavioral domain scored was muscle tension, with only 22% of patients demonstrating muscle rigidity during painful stimulus. Patients who were able to report that they had pain also had higher CPOT scores (median 3.5) than those who were able to report they had no pain (median 1.5). Spearman correlation with patient self-report was 0.82 for gentle touch, 0.64 for turning.
The authors make the point that in their brain-injured population, more than half (48) were deemed conscious, but only 28 of those could effectively self-report their pain (and this included patients whose only reporting ability was nodding the head). In patients who are not able to express pain, behavioral tools like the CPOT become incredibly important. This is true not only in the ICU, but in the trauma bay as well. How many of us have watched our patients receive cares that ranged from mildly uncomfortable to very painful without really registering what we were seeing written on their faces and their bodies?
Additional References for Articles 2 and 3:
Clinical practice guidelines for the management of pain, agitation, and delirium in adult patients in the intensive care unit. Barr J, Fraser GL, Puntillo K, Ely EW, Gélinas C, Dasta JF, Davidson JE, Devlin JW, Kress JP, Joffe AM, Coursin DB, Herr DL, Tung A, Robinson BR, Fontaine DK, Ramsay MA, Riker RR, Sessler CN, Pun B, Skrobik Y, Jaeschke R. Crit Care Med. 2013 Jan; 41(1):263-306.
Validation of the critical-care pain observation tool in adult patients. Gélinas C, Fillion L, Puntillo KA, Viens C, Fortier M. Am J Crit Care. 2006 Jul;15(4):420-7.
Validation of the Critical Care Pain Observation Tool in Critically Ill Patients With Delirium: A Prospective Cohort Study. Kanji S, MacPhee H, Singh A, Johanson C, Fairbairn J, Lloyd T, MacLean R, Rosenberg E. Critical Care Medicine. 2016 May;44(5):943-7.
The CPOT was tested in two tertiary care hospitals in Ottawa. Study patients were identified as being delirious with administration of the CAM-ICU. The RASS was used to measure sedation in these ICUs. In this study, patients with traumatic brain injury or stroke were excluded, as were patients who were too deeply sedated to participate (RASS -4 or -5), chemically paralyzed, quadriplegic, or otherwise unable to move. Sample size was based on power analysis designed to be able to discriminate between CPOT scores by one point at a power of 80% and alpha of 0.05.
Forty patients admitted to the ICU between March and June of 2014 who were identified as being delirious (and who were still CAM-positive at the time of study administration) were then observed by trained pairs (pharmacist, nurse) at three points: baseline, non-painful stimulus, and painful stimulus (most commonly turning (88%) or ETT suction (10%). Three-quarters of enrollees had received an antipsychotic in the 24 hours prior to enrollment. Interestingly, this was higher than the percentage of patients who had received narcotics (72%). Ninety percent were mechanically ventilated, and the mean APACHE II score was 19.1. Sepsis was the most common reason for ICU admission (43%), and 78% of the patients enrolled were cared for by a medical ICU service. On average, there was a three-point difference between baseline CPOT (mean score 0.5) and CPOT with painful stimulus (mean score 3.63). These scores are similar to CPOT validations in other patient subgroups. Inter-rater reliability was uniformly high with kappa of 0.67 – 0.89.
The authors do acknowledge that patient self-report remains the gold standard, but that the presence of delirium can make the determination of whether a patient does not have pain difficult. Additional limitations included the use of a wide range of narcotics, anti-psychotics, hypnotics, and other psychoactive medications, which are all typical in a critical care unit. As the study group was able to demonstrate good validity with high inter-rater reliability despite these limitations, I agree with their recommendation that the CPOT be considered the tool of choice for pain assessment in critically ill patients. It was noteworthy that more patients (who were mostly ventilated) in this study received antipsychotics than had received narcotics prior to enrollment, suggesting that pain remains undertreated in our critical care units.