Abstract
Background Promotion of increased physical activity is advocated for survivors of an intensive care unit (ICU) admission to improve physical function and health-related quality of life.
Objective The primary aims of this study were: (1) to measure free-living physical activity levels and (2) to correlate the measurements with scores on a self-reported activity questionnaire. A secondary aim was to explore factors associated with physical activity levels.
Design This was a prospective cohort study.
Methods Nested within a larger randomized controlled trial, participants were block randomized to measure free-living physical activity levels. Included participants wore an accelerometer for 7 days during waking hours at 2 months after ICU discharge. At completion of the 7 days of monitoring, participants were interviewed using the Physical Activity Scale for the Elderly (PASE) questionnaire. Factors associated with physical activity were explored using regression analysis.
Results The ICU survivors (median age=59 years, interquartile range=49–66; mean Acute Physiologic Chronic Health Evaluation [APACHE II] score=18, interquartile range=16–21) were inactive when quantitatively measured at 2 months after hospital discharge. Participants spent an average of 90% of the time inactive and only 3% of the time walking. Only 37% of the sample spent 30 minutes or more per day in the locomotion category (more than 20 steps in a row). Activity reported using the PASE questionnaire was lower than that reported in adults who were healthy. The PASE scores correlated only fairly with activity measured by steps per day. The presence of comorbidities explained one third of the variance in physical activity levels.
Limitations Accelerometer overreading, patient heterogeneity, selection bias, and sample size not reached were limitations of the study.
Conclusions Survivors of an ICU admission greater than 5 days demonstrated high levels of inactivity for prolonged periods at 2 months after ICU discharge, and the majority did not meet international recommendations regarding physical activity. Comorbidity appears to be a promising factor associated with activity levels.
Several different nations provide guidelines for physical activity.1–3 At least 150 minutes of moderate- to vigorous-intensity aerobic physical activity per week is recommended. This exercise can be accumulated in bouts of 10 minutes or more per day. There is a strong dose-response relationship between physical activity and health benefits, especially when comparing individuals who are sedentary to those undertaking even low and moderate exercise intensities.4 Current research highlights a clear association between physical activity levels and all-cause mortality and morbidity from chronic disease.5–8
Independent of physical activity levels, sedentary behaviors (particularly television watching) also pose a health risk. Salmon et al9 reported an average of 36.8 hours per week spent in sedentary behaviors among Australian adults. Higher levels of sitting time have been reported to lead to a progressively increased risk of all-cause and cardiovascular disease mortality.
Results from the 1999 National Physical Activity Survey showed that only 57% of Australians meet Australian physical activity guidelines, and 15% of Australians are considered inactive (do not participate in any leisure-time physical activity).10 The direct health care cost of physical inactivity in Australia has been estimated to be $377 million per annum.11
Given the increasing evidence for the benefits of physical activity, promotion and measurement of activity are the focus of a significant body of research in many different patient populations. Promotion of increased activity and exercise is advocated for survivors of an intensive care unit (ICU) admission to improve physical function levels and health-related quality of life.12 Survivors of an ICU admission were reported to have reduced physical function and health-related quality of life for as long as 5 years after discharge.13 This finding may be the result of a combination of premorbid disease and management in the ICU. These factors may lead to both physical and cognitive deficits that can be termed “post–intensive care syndrome.”14 This syndrome includes ICU-acquired weakness, as measured using the Medical Research Council (MRC) sum score of less than 48 out of 60.15 Although 3 trials demonstrated the benefits of early ICU rehabilitation16–18 and another trial demonstrated the benefits of an exercise diary after discharge,19 no previous studies have quantitatively measured daily physical activity levels of ICU survivors once discharged from hospital. The primary aims of this study were: (1) to measure physical activity levels with an accelerometer and (2) to correlate the measurements with scores on a self-reported activity questionnaire. A secondary aim was to explore whether any factors might predict activity levels at 2 months after ICU discharge.
Method
This was an observational cohort study nested within a larger randomized controlled trial (RCT) assessing physical function and health-related quality of life in ICU survivors. The study protocol was published previously.20 Trial recruitment was undertaken in a single tertiary referral center with a 20-bed ICU in Melbourne, Australia. To be eligible for enrollment, participants had to: (1) be older than 18 years; (2) have an ICU length of stay of more than 5 days; (3) be able to understand written and spoken English; (4) reside within a radius of 50 km from the hospital; (5) have their participation agreed upon by the attending intensivist; (6) have no neurological, spinal, or skeletal dysfunction preventing participation in physical rehabilitation; (7) be cognitively able to complete the self-report measures and perform physical tests; (8) be assessed by medical staff as not approaching imminent death or withdrawal of medical management; and (9) not have remained in the ICU because of unavailability of a ward bed.
Procedure
Participants were block randomized between October 2007 and October 2009 to participate in the measurement of physical activity levels. The sample size for this observational study was pragmatic and chosen based upon the number of accelerometers that could be used concurrently. We aimed to screen from participant number 35 in the RCT (the intended sample size was 200) in blocks of 20 with a gap of 5 participants between blocks of recruitment. Ethics approval was given, informed consent was obtained from the patient or his or her substitute decision maker, and the trial was registered with the Australian and New Zealand Clinical Trials Network (ACTRN12605000776606). Reporting follows the STROBE statement for the reporting of cohort studies.21
Included participants attended an initial appointment an average of 39 days after hospital discharge, where they were instructed by a physical therapist regarding correct application, positioning, and removal of the accelerometer, given an information sheet, and instructed to wear the accelerometer during waking hours for 7 days. Participants were informed that the accelerometer is waterproof but that they could remove it for bathing or other water activities if desired. At completion of the 7 days of monitoring, participants returned to the hospital outpatient clinic and were interviewed using the Physical Activity Scale for the Elderly (PASE) questionnaire.22 Scores on the Six-Minute Walk Test (6MWT) and the Timed “Up & Go” Test (TUG) were recorded for all participants at ICU discharge, at hospital discharge, and at 10 weeks, 6 months, and 12 months after ICU discharge as part of the larger trial. The 10-week measurement time point was used to define the exercise and functional capacity of this cohort.
Measures
AMP 331 Accelerometer.
The AMP 331 (Activity Monitoring Pod, Dynastream Innovations Inc, Cochrane, Alberta, Canada) is an accelerometer that uses a combination of inertial sensors to classify activity into 3 categories: inactive (no steps detected for at least 20 seconds), active (less than 20 steps recorded), and locomotion (at least 20 consecutive steps recorded). The AMP 331 accelerometer was chosen for this study for the following reasons: application ease, cost, ability to record for a 7- to 10-day period, and ability to directly download data. In addition, findings from previous research show that ankle-mounted devices, rather than hip-mounted devices, have better accuracy for measuring step count at slow walking speeds.23 The accuracy of the AMP 331 during overground walking has previously been reported as 94% for distance and 99% for step count in a population of healthy individuals.24 During our own preliminary analysis of the AMP 331 accelerometer, we recruited a sample of 12 participants who were healthy to complete a 30-minute choreographed routine to assess the accuracy of activity classification compared with video recording. We found no significant differences in recording of time spent in locomotion; however, the AMP 331 significantly overestimated time spent inactive by 20% (P<.001) and underestimated time spent active by 15% (P=.001) compared with video recordings. Previous findings regarding energy expenditure have shown the AMP 331 to be accurate during walking but to overestimate sedentary or light activities and to underestimate other activities.25
PASE Questionnaire.
The PASE questionnaire was developed to evaluate lifestyle physical activity in older adults over a 7-day period. It is a short, 10-item questionnaire, completed by telephone or self-administration. The range of possible scores is 0 to 400. Participants were asked to report on involvement in 3 activity subscales: leisure-time, household, and occupational. The leisure-time subscale includes: light, moderate, and strenuous sport and recreation activities; muscle strengthening/endurance exercises; and walking outside the home. Responses are recorded as frequencies per week: never, seldom (1–2 days per week), sometimes (3–4 days per week), or often (5–7 days per week). Duration of activity is recorded as: <1 hour, 1 to <2 hours, 2–4 hours, or >4 hours. The household activity subscale includes: light and heavy housework, home repairs, lawn work, outdoor gardening, and caring for another person. Responses for this subscale are recorded as a yes/no (scored as 1, 0) answer. The occupational activity subscale includes hours spent in paid or unpaid work (sedentary work is not included in the total PASE score).
The PASE score is calculated by first converting the reported frequency and duration of participation in the leisure-time subscale items into an average number of hours per day. For this calculation, and to calculate average sitting hours per day, we used the scoring chart provided by Washburn et al.26 Working hours per week are converted to average hours per day. All items then are multiplied by an empirically derived item weight and summed to give a total PASE score and 3 subscale scores. The PASE questionnaire has been validated previously,27 and validity and reliability are improved by interviewer rather than self-administered data collection. The mean PASE score in a sample of community-dwelling adults aged 65 years or over was 102.9 (SD=64.1).27 Recent studies have demonstrated mean PASE scores of 136.6 (SD=105.0) in residents of continuing care retirement communities28 and 228.2 (SD=102.1) 1 year after home-based exercise training following coronary artery bypass graft surgery.29 Previously reported correlations between PASE scores and accelerometer counts included a significant and moderate correlation (rho=.43, P<.01) in a sample of elderly participants who were healthy.30
6MWT.
The 6MWT is a commonly used practical test, is self-paced and submaximal, and reflects the functional exercise level for daily physical activities. It has been validated for populations with cardiorespiratory impairment.31 The 6MWT was performed according to reported guidelines,31 and the highest distance was recorded. In addition to raw data, the 6MWT score is reported as a percentage of predicted normative values.32,33 The TUG measures the time taken to stand from a chair, walk 3 m (10 ft), return, and sit down (in seconds).34 The best time for the 2 tests was recorded. Muscle strength was manually graded (0–5) in 6 muscle groups bilaterally, 7 days after awakening, as described by De Jonghe et al.15 A sum score out of 60 was calculated.
Data Analysis
Data from the AMP 331 accelerometer were automatically downloaded using the proprietary software (via a radiofrequency link) and displayed as a Microsoft Excel (Microsoft Corporation, Redmond, Washington) spreadsheet providing steps per day, total distance walked, and time spent in active, inactive, and locomotion categories. These data then were imported into SPSS Windows version 19.0 (SPSS Inc, Chicago, Illinois). It was determined a priori that a minimum of 3 days monitoring, 10 hours per day, was required35 for inclusion in the analyses of physical activity data. The mean data for individual participants across all of the days available were used in the analyses. Descriptive statistics and graphical displays were used to identify missing and out-of-range values and to assess the distribution of baseline characteristics and outcome variables.
Differences between the AMP 331 sample and the total sample were analyzed using t tests, chi-square tests, and analyses of variance for normally distributed data and Mann-Whitney U and Kruskal-Wallis tests for non–normally distributed data. Because the PASE scale measurement is ordinal and scores were not normally distributed, Spearman correlation coefficients were calculated to examine the relationship between PASE total and subscale scores and AMP 331 average daily step counts and average daily distance walked. Correlations between .25 and .50 are considered fair.36 Age was divided into 2 groups (20–64 and ≥65 years) based upon previous research using physical activity guidelines for adults.2,3
Multiple regression analyses were conducted to investigate factors associated with physical activity (average steps per day and distance). Univariate analyses were undertaken to identify potential factors to include in the linear regression model. Only those without collinearity or bivariate correlations were considered.
Results
Ninety-eight ICU survivors were screened to participate in the trial. Fifty-three ICU survivors, a mean of 61 days (SD=31) after discharge from the ICU, agreed to participate and wore an AMP 331 accelerometer in an ankle pouch with Velcro fasteners (Velcro USA Inc, Manchester, New Hampshire) for a 7-day period. Four participants' data were excluded from the analysis. In one of these cases, data were not recorded due to a flat battery, and another participant reported not wearing the accelerometer due to ankle edema. Reasons for insufficient data collection were not recorded for the other 2 participants. The flow of participants through the trial is shown in Figure 1. Demographics of the remaining 49 participants included in this trial and in the larger trial are presented in Table 1.
Flow chart of participants through the trial. RCT=randomized controlled trial, PASE=Physical Activity Scale for the Elderly.
Participant Demographics and Six-Minute Walk Test (6MWT) and Timed “Up & Go” Test (TUG) Scores at 10 Weeks After Intensive Care Unit Dischargea
There were no significant differences in demographics between the participants who agreed to wear an accelerometer and other participants. However, comparing mean scores, participants who wore the accelerometer performed significantly better, both statistically and clinically, on the physical function tests (6MWT and TUG) 10 weeks after ICU discharge compared with other trial participants (P<.05) (Tab. 1).
Physical Activity Monitoring
On average, participants wore the accelerometer for a mean of 13 hours 40 minutes (SD=1 hour 47 minutes) per day for 6 days (SD=1.6). Data included in the analyses were collected from all seasons of the year (spring=31%, summer=20%, autumn=25%, and winter=22%). Season of wear was unable to be reported for one participant whose dates were incorrectly entered into the device. Data from 47 participants (96%) included at least one weekend day. The days of wear used in the analyses included an even distribution of weekdays and weekend days (Monday=12%, Tuesday=15%, Wednesday=15%, Thursday=13%, Friday=16%, Saturday=14%, and Sunday=15%).
On average, participants took 4,894 steps per day (SD=3,070) (Tab. 2). Eighty percent (39/49) of the sample took fewer than 7,500 steps per day, and only 6% (3/49) walked 10,000 or more steps per day. Only 54% of the steps were taken while in the locomotion category (more than 20 consecutive steps). Therefore, nearly half of the steps were taken in small bouts of fewer than 20 steps. Participants walked a median of 1.69 km per day (interquartile range=0.69–2.89) (1 km=0.62 mile) (Fig. 2). There were no statistically significant seasonal or age (20–64 and ≥65 years) differences in average number of steps per day or distance walked per day or between male and female participants (Tab. 2).
Steps Taken per Day, Distance Walked per Day, and Time Spent in Each Activity Classification by Age and Sexa
Frequency of average distance walked by participants (1 km=0.62 mile, 1 m=1.09 yd).
During the waking day, participants spent an average of only 3% of the time (X̅=26 minutes, SD=10–36) in the locomotion category. Sixty-three percent (31/49) of the cohort spent an average of less than 30 minutes per day in the locomotion category. An average of 90% of the time (X̅=12 hours 17 minutes, SD=1 hour 33 minutes) was spent in the inactive category (Tab. 2).
PASE Questionnaire
Forty-five participants completed the PASE questionnaire. The mean total PASE score was 97.4 (SD=56.7, range=5.0–242.7). There were no significant differences in either total PASE score or subscale scores between participants aged 20 to 64 years and those aged ≥65 years. There also were no significant differences between male and female participants for total PASE scores or subscale scores or by season of wear (Tab. 3).
Mean (SD) Physical Activity Scale for the Elderly (PASE) Scores for Total Sample and by Age and Sexa
The total PASE scores showed fair correlation with average steps per day (rho=.332) and average distance walked per day (rho=.313) at P=.05. The PASE occupation subscale scores also were fairly correlated with average steps per day (rho=.332, P=.031). However, only 7% (3/49) of the participants reported they had worked in the week prior to PASE completion (on average, more than 2 months after ICU discharge). Average hours worked per week for these participants was 16 hours. The other subscale scores did not correlate significantly with average steps per day.
Twenty-nine percent (13/45) of the participants reported that they spent less than 30 minutes per day walking outside the home or yard. Again, correlations between this question on the PASE and average steps and distance walked per day while the accelerometer was in locomotion mode were fair but significant (rho=.345 and .344, respectively, P<.022). Twenty-four percent (11/45) of the participants reported they had spent more than an average of 4 hours per day performing sitting activities.
Association
Five independent variables that were significant in univariate analyses (P<.05) were entered into a linear regression model using both steps per day and walking time as dependant variables. They were 6-minute walk distance at ICU discharge and discharge home, acute hospital length of stay, 36-Item Short-Form Health Survey questionnaire version 2 (SF-36v2) physical function score at baseline, presence of chronic disease (yes/no), and hospital readmission within 12 months of index admission (yes/no). Absence of chronic disease was significantly associated with an increased number of steps taken (beta=−0.537, P=.000). In this model, chronic disease explained 28.8% of the variance of average steps per day (regression equation for average steps per day: 7,757.802 − 3,792.934 × chronic disease). Absence of chronic disease was significantly associated with increased distance walked (beta=−0.579, P=.000). In this model, chronic disease explained 33.5% of the variance of average distance walked per day (regression equation for average distance walked per day: 3,667.221 − 2,166.325 × chronic disease).
Discussion
This cohort of predominantly male ICU survivors were moderately unwell upon admission to the ICU, with a mean APACHE II score of 18. It represents a heterogeneous sample of participants who had been in the ICU for a median of 7 days with either a medical or surgical diagnosis. The population demographics compare with those of other published samples in published ICU trials,18,37 but the cohort was less sick compared with trials including only participants with acute lung injury or those with a longer ICU stay.16 A large proportion of the cohort had premorbid comorbidities upon admission. These comorbidities were grouped as respiratory, cardviovascular, diabetes, renal, and “other,” which included arthritidies. Thirty-three percent of the participants in this study had 2 or more comorbidities (Tab. 1). Their average muscle strength (MRC sum score of 51 out of 60), as measured with the MRC scale at 7 days after awakening,15 was indicative of strength able to overcome applied resistance. The MRC scale grades muscle strength from 0 to 5; a sum score less than 48 is reported as a clinical diagnosis of ICU-acquired weakness.15 However, this measurement is subjective and grade 4 scores are likely to represent a large range of strengths.38 Twenty-four percent of this cohort (8 out of 34 participants; 15 participants had missing data) had muscle strength (MRC sum score) of less than 48 out of 60, defining ICU-acquired weakness.
The 6MWT is a test of functional exercise capacity, and in this cohort the mean 6MWT score was 64% of predicted norms for Australians of similar age, sex, and body mass index at 10 weeks after ICU discharge. At the time of measurement of the physical activity levels, this sample was clearly deconditioned. However, as it is not possible to measure strength and exercise capacity prior to ICU admission, it is difficult to know whether these values were at normal levels at baseline. The range and number of comorbidities suggest it is possible that, on average, participants were not at normal levels of exercise capacity premorbidly. This finding is consistent with other research in populations of ICU survivors.39,40
The ICU survivors were inactive when quantitatively measured at 2 months after hospital discharge, although there was wide variation among participants in distance walked and steps taken. The mean daily step count in this cohort of survivors (4,894) was low compared with the mean step count derived from a meta-analysis (6,565 in participants aged ≥65 years).41 Eighty percent of our sample had step counts of less than 7,500 per day, activity levels that are considered low.42 Fifty-five percent of the participants took an average of less than 5,000 steps per day. This finding places 80% of our sample at risk of further morbidity from chronic disease due to physical inactivity.11,43 Significantly, only 6% of the participants were considered “active” or “highly active” with 10,000 or more steps per day, meeting recommended activity guidelines.42 Previous research using the AMP 331 accelerometer demonstrated average daily step counts of 6,384 in community-dwelling older adults with self-reported mild to moderate functional limitations44 and 8,21945 and 10,26923 in adults who were healthy. Different brands of accelerometers have been used to record average daily step counts in different populations, including cystic fibrosis (9,398),46 chronic obstructive pulmonary disease (COPD) and obesity (6,991),47 type 2 diabetes (median=9,150),48 and recipients of lung transplants (4,977).49 The number of steps taken per day in this ICU cohort compares most closely with that of recipients of lung transplants but is less than that of other populations reported, including those with similar comorbidities.
Similarly, our ICU survivors walked for less time than other patient populations. Compared with patients with stable COPD, who spend an average of 44 minutes per day walking,50 accelerometer data showed our ICU survivors spent only 26 minutes per day walking. Sixty-three percent did not reach the minimum activity guidelines of at least 30 minutes of moderate-intensity physical activity per day compared with 48% and 23% in stable COPD populations.51 Although nearing Australian guidelines for minimum requirements of daily moderate-intensity exercise, in the majority of participants these bouts of locomotion were of short duration, accumulated throughout the day rather than as recommended.1–3 In contrast to the accelerometer data, participants reported in the PASE a significantly higher daily frequency of walking over the same time period, with only 29% reporting they had walked for an average of less than 30 minutes per day outside of their home or yard. This finding, in part, may reflect the previously reported inaccuracies associated with self-report measures, with patients tending to overestimate activity and underestimate inactivity.52–54
It has been reported previously that Australian adults who are healthy spend 57% of their waking time sedentary.55 Accelerometer data showed our sample spent 90% of their time inactive. This finding may have been due, in part, to the large number of participants who had significant comorbidities, along with new impairments resulting from the ICU admission and varying recovery times. Additionally, lack of motivation to exercise or depression and cognitive impairments56 may affect activity levels, and the motivation of family and caregivers to be active may affect the ICU survivor's motivation to exercise.57 As discussed, our preliminary analyses of the AMP 331 accelerometer data revealed a significant overreporting of time spent inactive and underreporting of time spent active compared with video recordings, which may have contributed to this finding of inactivity in our ICU survivors.
Total PASE scores were well below mean scores previously obtained for adults who were healthy (X̅=125.2, SD=79.9).22 However, consistent with our study sample, participants with chronic diseases in that study had lower PASE scores. Scores also were well below those obtained in residents of community care retirement communities28 and those 1 year after a home-based exercise program following coronary artery bypass graft surgery.29 Further research is needed regarding physical activity guidelines to meet the needs of adults with chronic disease.3
Household activities contributed 67% of the total PASE score, which is consistent with previous findings. The leisure subscale scores were higher and the occupational subscale scores were lower than previously reported.22,27 At least 20% of our participants attended rehabilitation programs following their hospital admission, which may explain the higher leisure subscale scores. In contrast to the validation study,27 there were no significant differences in PASE scores with respect to season of wear. This finding may have been due to the fact that seasonal variations in outdoor temperature are not as significant in Melbourne, Australia.
The total PASE scores correlated fairly with average daily step counts and distance walked per day. Specifically, the question relating to walking outside the home or yard also correlated fairly with average step counts and distance walked while in the accelerometer “locomotion” phase. These correlations are lower than those previously reported,22 perhaps due to the period of monitoring being only 3 days and the use of a smaller sample size of 20 in that study. Also, it has been previously reported that accelerometers and the PASE do not measure the same construct. Accelerometers measure physical activity duration and intensity, whereas the PASE also measures activity type (more strenuous items being given higher weighting).58 Choice of measurement instrument (PASE or accelerometer) may be dependent upon the construct being measured, as measurement biases exist in both instruments.
The only significant factor associated with activity levels at 2 months from ICU discharge was comorbidity. This finding offers some promise for future research and clinical practice to identify those individuals who may benefit most from intervention to improve physical activity levels. The Charlson Comorbidity Index,59 in which 3 levels of comorbidity are defined (low=score of 0 or 1), medium=score of 2–4, and high=score of 5 or higher), has been reported to predict readmission to ICU and death.17 Further research also demonstrated the ability of the Functional Comorbidity Index,60 which measures comorbidities prior to ICU admission, to predict post–ICU physical function.61
Limitations
The basis for the sample size was pragmatic and determined by the number of participants recruited to the RCT, the number of those who were still alive at the testing time point, and the number of accelerometers concurrently available. We did not reach our intended sample size in the RCT and, therefore, were able to screen only 98 participants. As shown in our preliminary reliability work and other published research, it is likely that the AMP 331 accelerometer may have overestimated time spent in the inactive classification and underestimated time spent in the active classification (<20 consecutive steps).23,25 In addition, participants did not keep an activity diary, making it difficult to determine periods of not wearing the AMP 331 accelerometer from periods of inactivity. Although the manufacturers report the AMP 331 accelerometer to be 99% accurate for step count,24 other studies have reported an underestimation of steps compared with video recordings62 and the StepWatch 3 (SW-3) accelerometer.23 The SW-3 accelerometer has been reported to record a higher number of steps taken during slow walking and lifestyle activities compared with the AMP 331 accelerometer.23 The comparisons between the accelerometer sample and the larger RCT sample suggest that some selection bias may have been present. The accelerometer group performed significantly better (both statistically and clinically) on physical function tests (6MWT and TUG) at 10 weeks following ICU discharge compared with the total sample. This finding suggests that perhaps it was the less severely involved, more active, and more motivated participants from the total sample who agreed to take part in the accelerometer trial. Our results may indeed be an overestimation of true activity levels of the general population of ICU survivors. Our sample size was small, and future research in this area should report on larger samples.
Conclusions and Clinical Implications
Survivors of an ICU admission greater than 5 days demonstrated high levels of inactivity for prolonged periods at 2 months after hospital discharge, and the majority did not meet recommendations regarding levels of physical activity. Providing a continuum of rehabilitation that includes post-ICU discharge is recommended to increase physical activity, improve performance on functional walking tests, and reduce all-cause mortality in survivors of ICU. We expect that, given the high levels of comorbidity and the possibility of cognitive impairments in these individuals, this treatment approach needs to be individualized and include training in lifestyle change regarding exercise such as cognitive behavioral strategies to increase and sustain participation in physical activity. This approach is similar to the need to promote continued activity to maintain improved patient-centered outcomes after pulmonary rehabilitation. Physical therapists can actively promote the continuum of care for survivors of ICU by working with health care providers and other members of a multidisciplinary team to investigate strategies to promote adherence to physical activity that meets international guidelines.
Footnotes
-
Dr Denehy, Dr Berney, and Ms Whitburn provided concept/idea/research design. Dr Denehy, Dr Berney, and Ms Edbrooke provided writing. Ms Whitburn provided data collection. Dr Denehy, Ms Whitburn, and Ms Edbrooke provided data analysis. Ms Edbrooke provided project management. Dr Denehy provided fund procurement.
-
This study was approved by the Austin Health Human Research Ethics Committee.
-
An abstract of this research was presented at the American Thoracic Society Conference; May 18–23, 2012; San Francisco, California.
-
This research was funded by a National Health and Medical Research Council of Australia project grant (454717), the Austin Health Medical Research Fund, the Australian New Zealand Society of Anesthetists, and the Physiotherapy Research Foundation.
- Received November 15, 2011.
- Accepted April 30, 2012.
- © 2012 American Physical Therapy Association
References
Issue highlights
