Functional status, age, and long-term survival after trauma

ARTICLE IN PRESS Functional status, age, and long-term survival after trauma Allan B. Peetz, MD,a Gabriel A. Brat, MD,b Jessica Rydingsward, PT,c Reza Askari, MD,b Olubode A. Olufajo, MD, MPH,b Kevin M. Elias, MD,d Kris M. Mogensen, MS, RD, LDN, CNSC,e Jessica L. Lesage, PT, DPT,c Clare M. Horkan, MB, BCh,f Ali Salim, MD,b and Kenneth B. Christopher, MD, SM,g Boston, MA Background. The association between functional status in trauma survivors and long-term outcomes is unknown. Methods. We performed an observational cohort study on adult trauma patients ($18 years), who required admission to the intensive care unit and who survived hospitalization between 1997 and 2011. The exposure of interest was a functional status defined as bed mobility, transfers, and gait level assessed at the time of hospital discharge. Adjusted odds ratios were estimated by multivariable logistic regression models. The primary outcome was all-cause, postdischarge mortality. Results. We analyzed 3,565 patients with a mean (standard deviation) age of 55 (12.4) years; 60% were male, and 78% were white. The 720-day postdischarge mortality was 22.8%. In a logistic regression model, the lowest functional status category at hospital discharge was associated with 4-fold increased odds of 720-day postdischarge mortality (adjusted odds ratio 4.06 (95% confidence interval, 2.65–6.20, P < .001) compared with patients with independent functional status. We compared the odds of 720-day postdischarge mortality in patients with independent functional status and in patients in the lowest functional status category at hospital discharge. The odds of 720-day postdischarge mortality were stronger in older adults ($65 years: adjusted odds ratio 3.34 [95% confidence interval, 1.72–6.50, P < .001]) than in younger adults (<65 years: adjusted odds ratio 2.53 [95% confidence interval, 1.39–4.60, P = .002]). Finally, improvement of functional status prior to discharge was associated with a 52% decrease in the odds of 720-day postdischarge mortality (adjusted odds ratio 0.48; 95% confidence interval, 0.30–0.75; P < .001) compared with patients without a change in functional status prior to discharge. Conclusion. In trauma intensive care unit survivors, functional status at hospital discharge is predictive of long-term mortality. (Surgery 2016;j:j-j.) From the Trauma, Surgical Critical Care & Acute Care Surgery,a Case Western Reserve University School of Medicine; Division of Trauma, Burns, and Surgical Critical Care, Department of Surgery,b Department of Rehabilitation,c Department of Obstetrics, Gynecology and Reproductive Biology,d Department of Nutrition,e Department of Medicine,f and The Nathan E. Hellman Memorial Laboratory, Renal Division, Department of Medicine,g Brigham and Women’s Hospital, Boston, MA AS SURVIVAL AFTER A CRITICAL ILLNESS has improved during the past 30 years, attention has pivoted toward the importance of long-term outcomes. Survivorship in the critically ill is complicated by substantial long-term mortality and morbidity, such as long-term physical impairments, profound neuromuscular weakness, exercise limitation, and Accepted for publication April 13, 2016. Reprint requests: Kenneth B. Christopher, MD, SM, The Nathan E. Hellman Memorial Laboratory, Division of Renal Medicine, Brigham and Women’s Hospital, Medical Research Building 418, 75 Francis Street, Boston, MA 02115. E-mail: kbchristopher@partners.org. 0039-6060/$ - see front matter Ó 2016 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.surg.2016.04.015 lower quality of life after hospital discharge,1-6 but long-term outcomes among trauma intensive care unit (ICU) survivors have not been studied. Physical therapy early in the ICU course has been shown to be safe.7-9 Early physical therapy is associated with improved functional independence in mechanically ventilated patients.7,8 Studies show that functional status may be modifiable in the ICU.8-12 The combination of interrupted sedation with physical and occupational therapy early in the ICU course is related to functional status improvements at hospital discharge.9 Although long-term functional independence is desirable among ICU patients, little information exists on critically ill trauma survivors’ functional status at hospital discharge or on the adverse outcomes they face after hospital discharge.13 SURGERY 1 ARTICLE IN PRESS 2 Peetz et al Because functional status may be an important driver of long-term outcomes in critically ill trauma patients, we performed this study to determine the relationship between critically ill trauma patients’ functional status at hospital discharge and all-cause 2-year post-hospital discharge mortality. We hypothesized that a decrease in functional status at discharge would be associated with adverse outcomes among critically ill and injured patients. METHODS Source population and data sources. We abstracted patient-level administrative and laboratory data from the Brigham and Women’s Hospital, a 793-bed urban level I trauma center. Data on all trauma patients $18 years old who were admitted to the intensive care unit between January 1, 1997, and December 31, 2011, and who survived to hospital discharge were obtained through the Research Patient Data Registry, a computerized registry that serves as a central data warehouse for all inpatient and outpatient records at Partners HealthCare sites.14,15 Approval for the study was granted by the Partners Human Research Committee (Institutional Review Board) Protocol Number: 2010P000645. Requirement for consent was waived, as the data were analyzed anonymously. Study population. During the study period, 7,450 unique patients met inclusion criteria. Exclusions included 3,885 patients who did not receive a formal structured evaluation from a physical therapist within 48 hours of hospital discharge. Thus, the analytic cohort comprised 3,565 patients. Exposure of interest and covariates. The exposure of interest was functional status at hospital discharge defined as physical function assessed at the time of hospital discharge. Determination of physical function was made by a licensed physical therapist and rated based on qualitative categories adapted from the functional mobility subscales of the Functional Independence Measure.16-18 The Functional Independence Measure mobility subscales incorporate transfers (including bed, chair, and wheelchair) as well as locomotion (including walking/wheelchair and stairs) and are scored on an ordinal scale based on percentage of active patient participation in the selected task.16 The scale scoring system grades patients on a scale of function for motor tasks assessed (independent, standby assist/supervision, minimal assist, moderate assist, maximal assist, and total assist) with a determination of “not applicable” used when a Surgery j 2016 patient was either incapable of progressing to the designated task or to indicate physical or medical limitations. Patients were assessed on bed mobility (roll side-to-side, supine-to-sit, sit-to-supine), transfers (sit-to-stand, stand-to-sit, bed-to-chair), and gait (level ambulation, stairs). A categorical risk prediction score was derived previously and validated based on a logistic regression model of the individual scale of function grades for each assessment.19 Definition and determination of the following covariates are outlined in the Supplementary Methods (online only version): Deyo-Charlson index,20 intensive care unit admission,21 race, sepsis,22 exposure to inotropes and vasopressors,23 acute kidney injury,24 noncardiogenic acute respiratory failure,25 emergency general surgery,26 packed red blood cells transfused,23 acute organ failure,21,27 malnutrition,28 and International Classification of Diseases, Ninth Revision, Clinical Modification, Ninth Edition (ICD-9-CM) derived injury severity score (ICISS).29-33 Malignant neoplasm history is defined by the presence of any of the following ICD-9-CM codes prior to the hospital discharge date: 140-209.34 For severity of illness risk adjustment, we used the acute organ failure score, an ICU risk-prediction score derived and validated from demographics (age, race), patient admission “type” as well as ICD-9-CM code based comorbidity, sepsis, and acute organ failure covariates, which has a similar discrimination for 30-day mortality as Acute Physiology and Chronic Health Evaluation (APACHE) II.35 The trauma-related ICD-9 diagnosis codes were grouped by body region of injury categories based on the Barell Injury Diagnosis Matrix.36 Endpoints. The primary endpoint was 720-day, all-cause mortality after hospital discharge. Secondary endpoints included 90- and 365-day allcause mortality after hospital discharge. Assessment of mortality. Vital status was obtained from the Death Master File of the Social Security Administration, which has high sensitivity and specificity for mortality37; we have validated the accuracy of this Death Master File for inhospital and out-of-hospital mortality in the Research Patient Data Registry database.21 In the study, 100% of the cohort had at least 720-day follow-up after hospital discharge. The censoring date was January 1, 2014. Power calculations and statistical analysis. Previously in a cohort of critically ill patients (n = 43,212), we studied postdischarge mortality in ICU survivors.23 From these data, we assumed that 720-day postdischarge mortality in ICU survivors was 18.8%. In our cohort of 1,224 patients ARTICLE IN PRESS Surgery Volume j, Number j with the lowest quartile of functional status and 340 patients with independent functional status assuming an alpha error level of 5% and a power of 80%, the smallest difference that we can detect between 720-day mortality rates is 7%. Categorical variables were described by frequency distribution and compared across outcome groups using contingency tables and v2 testing. Continuous variables were examined graphically and in terms of summary statistics and then compared across outcome groups using one-way analysis of variance or the Kruskal-Wallis test. For the 720-day postdischarge mortality model, specification of each continuous covariate (as a linear versus categorical term) was adjudicated by the empiric association with the primary outcome using Akaike information criterion; overall model fit was assessed using the Hosmer Lemeshow test. The discriminatory ability for 720-day mortality was quantified using the c-statistic. The improvement in model performance was evaluated via net reclassification improvement or integrated discrimination improvement.38 Adjusted odds ratios were estimated by multivariable logistic regression models with inclusion of covariate terms thought to associate plausibly with both functional status and 720-day mortality. We tested individually for effect modification by functional status by year of hospital admission, hospital, packed red blood cells transfusions, need for operation, vasopressors/inotropes, or age by adding an interaction term to the multivariate models. Furthermore, a multivariable Cox’s proportional hazards model was used to illustrate postdischarge survival related to functional status. To account for the differential outcomes that could be associated with highly debilitating illnesses, such as traumatic brain injury and spinal cord injuries, we performed different sensitivity analyses by including these variables in the models and by excluding patients with these injuries from the analyses. For the time to mortality, we estimated the survival curves according to functional status quartile with the Kaplan-Meier method and compared the results via the log-rank test. In all analyses, P values are 2-tailed. All analyses were performed using STATA 13.1MP statistical software (StataCorp LP, College Station, TX). RESULTS Table I shows characteristics of the study population. Most patients were men (60%), white (78%), with a mean age at hospital admission of 60.1 years (standard deviation, 20.1 years). Sixteen Peetz et al 3 percent of the cohort had sepsis, 7% had acute kidney injury, and 19% had noncardiac acute respiratory failure. In the study, 1,768 patients (50%) were transferred to rehabilitation; 515 (14%) were transferred to a nursing home; and 722 (20%) were discharged home with services. The crude 90-, 365-, and 720-day postdischarge mortality rates were 8.5%, 17.5%, and 22.8%. There were 1,126 patients who died subsequently with 15,940 person-years of follow-up, yielding a mortality rate of 70.6 per 1,000 person-years. Table I indicates that age, race, Deyo-Charlson index, acute organ failures, malignant neoplasm, malnutrition, acute kidney injury, sepsis, vasopressors/inotropes, and ICISS are associated significantly with 720-day mortality. Patient characteristics of the study cohort were stratified according to functional status categories (Table II). With the exception of race, all covariates differed significantly across functional status categories. Table III demonstrates the body region of injury. Head and neck injuries accounted for 22% of the study population. Functional status was associated with postdischarge mortality. Fig illustrates the different survival curves for the functional status groups. The log-rank test indicated that there is a significant difference in the overall survival distributions between the patient groups. In a logistic regression model after adjustment for ICISS, the acute organ failure score and sex, the lowest functional status category at hospital discharge was associated with 4-fold increased odds of 720-day postdischarge mortality (OR 4.06, 95% CI, 2.65–6.20; P < .001) compared with patients with independent functional status (Table IV). The adjusted functional status model showed good calibration (Hosmer-Lemeshow v2 8.73, P = .37) and good discrimination for 720-day postdischarge mortality (c-statistic = 0.77, 95% CI, 0.75–0.78). Additional individual adjustment for operative intervention, traumatic brain injury, or spinal cord injury did not alter materially the association between functional status and 720-day postdischarge mortality. Furthermore, exclusion of traumatic brain injury and spinal cord injury patients did not alter the functional status 720-day postdischarge mortality association. Differences in discrimination between the primary model and one including only sex, ICISS, and the acute organ failure score (area under the curve = 0.74 [95% CI, 0.73–0.76]) are significant (v2 27.67, P < .001). Furthermore, the net reclassification improvement was estimated at 6.3% (P < .001), and the integrated discrimination improvement was estimated at 2.9% (P < .001). ARTICLE IN PRESS 4 Peetz et al Surgery j 2016 Table I. Population characteristics of the derivation cohort and unadjusted association of potential prognostic determinants with 720-day postdischarge mortality* Age, mean ± SD Male, n (%) Non-white race, n (%) Deyo-Charlson index, n (%) 0–1 2–3 4–6 $7 Acute organ failure, n (%) 0 1 2 3 $4 Malignant neoplasm, n (%) Malnutrition, n (%)x Acute kidney injury, n (%)jj Sepsis, n (%) Noncardiogenic acute respiratory failure, n (%) Vasopressors/inotropes, n (%) ICISS, mean ± SD Acute organ failure score, mean ± SD{ Length of stay, median (IQR) Discharge to care facility, n (%) Alive N = 2,751 Dead* N = 814 Total N = 3,565 56.4 ± 20.1 1,672 (61) 671 (24) 72.9 ± 13.8 469 (58) 124 (15) 60.1 ± 20.1 2,141 (60) 795 (22) 1,374 844 463 70 (50) (31) (17) (3) 108 295 330 81 (13) (36) (41) (10) 1,482 1,139 793 151 (42) (32) (22) (4) 817 860 589 307 178 323 290 145 180 542 (30) (31) (21) (11) (6) (12) (16) (6) (7) (20) 137 247 203 139 88 293 129 87 94 125 (17) (30) (25) (17) (11) (36) (28) (14) (12) (15) 954 1,107 792 446 266 616 419 232 274 667 (27) (31) (22) (13) (7) (17) (18) (7) (8) (19) P value <.001y .11 <.001 <.001 Unadjusted OR (95% CI) for 720-day postdischarge mortality 1.05 (1.05, 1.06) 0.88 (0.75, 1.03) 0.56 (0.45, 0.69) 1.00 4.45 9.07 14.72 (Referent) (3.51, 5.63) (7.12, 11.54) (10.12, 21.42) <.001 <.001 <.001 <.001 <.001 .005 1.00 1.71 2.06 2.70 2.95 4.23 2.03 2.64 1.87 0.74 (Referent) (1.36, 2.16) (1.61, 2.62) (2.06, 3.54) (2.16, 4.03) (3.52, 5.08) (1.60, 2.57) (1.99, 3.50) (1.43, 2.43) (0.60, 0.91) 1,014 (37) 0.67 ± 0.21 8.4 ± 4.6 339 (42) 0.79 ± 0.13 11.9 ± 4.0 1,353 (38) 0.70 ± 0.20 9.2 ± 4.7 .013 <.001y <.001y 1.22 (1.04, 1.43) 1.04 (1.03, 1.05) 1.19 (1.16, 1.21) 12 (6, 22) 2,317 (84) 17 (9, 33) 753 (93) 13 (7, 24) 3,070 (86) .47z <.001 1.01 (1.01, 1.02) 2.31 (1.75, 3.06) *Died within 720 days of hospital discharge. xMalnutrition data is available for 2,301 patients. jjAcute kidney injury is a RIFLE (Risk, Injury, Failure, Loss, End-stage renal disease) class injury or failure, with data available for 3,113 patients. {Acute organ failure score is a severity of illness risk-prediction score ranging from 030 points, with 30 being the highest risk for mortality. Data presented as n (%) unless otherwise indicated. P values were determined by v2 except when determined by one-way analysis of variancey or by Kruskal-Wallis test.z OR, Odds ratio; CI, confidence interval; SD, standard deviation; ICSS, (ICD-9-CM) derived injury severity score; IQR, interquartile range. The net reclassification improvement and integrated discrimination improvement suggest that the addition of functional status results in a significant improvement in model performance. The relationships between functional status at hospital discharge and mortality were magnified at 90 and 365 days postdischarge (Table IV). Furthermore, the hazard ratio of mortality adjusted for sex, ICISS, and the acute organ failure score in patients with the lowest functional status category at hospital discharge was 3.58 (95% CI, 2.56–5.00) relative to patients with independent functional status. There was no effect modification of the functional status 720-day postdischarge mortality association on the basis of year of hospital admission, hospital, packed red blood cells transfusions, need for operation, or vasopressors/inotropes used (P interaction $ .11 each). Effect modification is present with age (P interaction <0.01). Crude allcause, 720-day postdischarge mortality rates were 10.2% and 37.4% in patients less than 65 years and greater than 65 years, respectively. In patients with the lowest functional status category at hospital discharge, the odds of 720-day postdischarge mortality compared with patients with independent functional status was stronger in older adults ($65: adjusted OR 3.34 [95% CI, 1.72–6.50, P < .001]) than in younger adults (<65: adjusted OR 2.53 [95% CI 1.39–4.60, P = .002]). In a subset of patients with physical function assessed at the time of hospital discharge and at least 1 week prior (n = 1,040), we studied the change in functional status and postdischarge mortality. In a logistic regression model adjusted for ARTICLE IN PRESS Peetz et al 5 Surgery Volume j, Number j Table II. Patient characteristics sorted by functional status risk category group Category 1 (highest function) N Age, mean ± SD Male, n (%) Non-white race, n (%) Deyo-Charlson index, n (%) 0–1 2–3 4–6 $7 Acute organ failure, n (%) 0 1 2 3 $4 Malignant neoplasm, n (%) Malnutrition, n (%)z Acute kidney injury, n (%)x Sepsis, n (%) Noncardiogenic acute respiratory failure, n (%) Vasopressors/inotropes, n (%) ICISS, mean ± SD Acute organ failure score, mean ± SD Operations, n (%) Emergency general operation, n (%) 365-day postdischarge mortality, n (%) 720-day postdischarge mortality, n (%) 1a (independent) 1b (less than independent) Category 2 Category 3 Category 4 (lowest function) 340 48.0 ± 17.3 244 (72) 89 (26) 319 50.5 ± 18.7 213 (67) 83 (26) 860 57.9 ± 19.0 551 (64) 184 (21) 824 65.2 ± 19.4 451 (55) 165 (20) 1,222 64.2 ± 19.9 682 (56) 274 (22) 208 84 42 6 (61) (25) (12) (2) 179 90 43 7 (56) (28) (13) (2) 371 271 188 30 (43) (32) (22) (3) 308 279 202 35 (37) (34) (25) (4) 416 415 318 73 (34) (34) (26) (6) 150 104 49 20 17 49 25 17 20 54 (44) (31) (14) (6) (5) (14) (10) (5) (6) (16) 119 101 57 27 15 47 26 14 10 45 (37) (32) (18) (8) (5) (15) (12) (5) (3) (14) 260 276 173 94 57 167 90 53 54 145 (30) (32) (20) (11) (7) (19) (16) (7) (6) (17) 210 271 190 93 60 162 102 55 62 130 (25) (33) (23) (11) (7) (20) (20) (8) (8) (16) 215 355 323 212 117 191 176 93 128 293 (18) (29) (26) (17) (10) (16) (23) (9) (10) (24) P value <.001* <.001 .079 <.001 <.001 .019 <.001 .015 <.001 <.001 108 (32) 0.71 ± 0.18 6.8 ± 4.0 98 (31) 0.70 ± 0.18 7.6 ± 4.3 302 (35) 0.72 ± 0.18 8.9 ± 4.4 303 (37) 0.71 ± 0.19 9.8 ± 4.8 542 (44) 0.67 ± 0.23 10.2 ± 4.8 <.001 <.001* <.001* 177 (52) 73 (22) 160 (50) 57 (18) 428 (50) 207 (24) 436 (53) 203 (25) 788 (65) 571 (47) <.001 <.001 16 (5) 25 (8) 105 (12) 169 (21) 309 (25) <.001 28 (8) 31 (10) 155 (18) 221 (27) 379 (31) <.001 zMalnutrition data is available for 2,301 patients. xAcute kidney injury is RIFLE (Risk, Injury, Failure, Loss, End-stage renal disease) class injury or failure, with data available for 3,133 patients. Data presented as n (%) unless otherwise indicated. P value was determined by v2 except when determined by one-way analysis of variance* SD, Standard deviation; ICISS, (ICD-9-CM) derived injury severity score. sex, ICISS, initial functional status, and the acute organ failure score, the highest quartile of functional status improvement at hospital discharge was associated with a 52% decrease in the odds of 720-day postdischarge mortality (adjusted OR 0.48 [95% CI, 0.30–0.75, P < .001]) compared with patients without a change in functional status prior to discharge. DISCUSSION Our study aimed to determine whether a decrease in functional status was associated with decreased mortality in critically ill trauma patients. In both adjusted and unadjusted analyses, we found an increase in the odds of 720-day postdischarge mortality relative to a decrease in functional status directly determined by a physical therapist at hospital discharge. This association seems to be stronger in patients $65 years. Patients with improvement in functional status prior to discharge seem to have improvement in 720-day postdischarge mortality. Because our study is observational and not interventional, however, a causal relationship between change in functional status and outcomes after trauma cannot be inferred from these data alone. Decreased long-term functional status is common in ICU survivors.39 Patients with functional disability in basic, instrumental, and mobility activities prior to an ICU stay are at high risk for 1-year ARTICLE IN PRESS 6 Peetz et al Surgery j 2016 Table III. Patient characteristics of cohort by nature and body region of injury Alive Dead* Total N = 2,751 N = 814 N = 3,565 Body region of injury, n (%) Extremities, n (%) 509 Head and neck, n (%) 610 Type 1 TBI, n (%) 514 Spine and 251 back, n (%) Torso, n (%) 354 Unclassifiable by 1,027 site, n (%) (19) (22) (19) (9) 117 191 162 29 (14) (23) (20) (4) 626 801 676 280 (18) (22) (19) (8) (13) 63 (8) 417 (12) (37) 414 (51) 1,441 (40) *Died within 720 days of hospital discharge. Note: Type 1 TBI are a subset of the head and neck and are present if there is recorded evidence of an intracranial injury, a moderate or a prolonged loss of consciousness, or injuries to the optic nerve pathways. Data are presented as n (%) unless otherwise indicated. TBI, Traumatic brain injury. Fig. Time-to-Event curves for mortality Unadjusted allcause post-discharge mortality rates were calculated with the use of the Kaplan-Meier methods and compared with the use of the log-rank test. Categorization of risk groups is per the primary analysis. The global comparison log rank p value is <0.001. mortality.40 Even in healthy subjects, atrophy of skeletal muscle is shown to occur with more than 3 days of immobilization.41 With prolonged immobility, such as bed rest, greater losses of strength and muscle mass are demonstrated in older compared with younger adults.42 Decreased strength and loss of muscle mass is common in critical illness.43-46 There is little information regarding long-term outcomes for trauma ICU survivors.47-49 Multiorgan failure in the critically ill trauma patient is associated with long-term increases in mortality and decreased functional status as determined by the Karnofsky Index and Glasgow Outcome Score.49 There has been a growing focus on increasing mobility in ICU patients, which improves short-term outcomes, such as delirium, ventilator-free days, physical function, peripheral and respiratory muscle strength, and duration of hospital and ICU stay.7-9,50 While studies suggest that functional status may be modifiable in the ICU, limited information exists in trauma ICU survivors regarding the improvement in functional status and outcomes after hospital discharge. We demonstrated an association between improvement in functional status and long-term outcomes in critically ill trauma patients. It remains to be seen if this association is due to unmeasured host factors, the physical therapy itself, or a combination. The present study has all the inherent limitations of a retrospective study. Because our study is observational, causality is limited. Selection bias may exist, as the patient cohort under study had functional status determined at discharge most likely for rehabilitation placement. These issues may decrease the generalizability of our results to all trauma patients. Reliance on ICD-9 codes to determine the Deyo-Charlson Index, ICISS, and diagnosis of sepsis, and acute respiratory failure does not measure the true incidence, which is likely greater. Though not the gold standard, ICISS is comparable to the ISS alone as a predictor of survival.29,51 Furthermore, due to lack of physiologic and hemodynamic data in our dataset, we were unable to use a physiologic-based acuity of illness scores; however, for severity of illness adjustment, we used the acute organ failure score, a validated ICU risk-prediction score with similar discrimination for mortality as APACHE II.35 Although we adjusted for multiple potential confounders, there might be residual confounding of unmeasured variables, leading to observed differences in outcomes. Due to limitations of the dataset, we cannot determine with precision what treatment modality may have led to improvement in the functional status. Important information on duration of stay for rehabilitation or functional status at discharge from rehabilitation is not available in our study. Therefore, our data do not address the importance of functional status and quality of life after a rehabilitation stay. Such outcomes are likely more important to ICU survivors and to society than functional status at hospital discharge. Our study does, however, highlight the importance of optimizing functional status of trauma patients even prior to hospital discharge. The present study has several strengths and is unique in that it investigates the effect of functional status measured directly at hospital ARTICLE IN PRESS Peetz et al 7 Surgery Volume j, Number j Table IV. Unadjusted and adjusted associations between functional status and postdischarge mortality in trauma patients Category 1a (independent) Category 1b (less than independent) Category 2 Category 3 Category 4 (lowest function) OR (95% CI) P value OR (95% CI) P value OR (95% CI) P value OR (95% CI) P value OR (95% CI) P value 1.28 (0.39, 4.25) .68 1.20 (0.36, 4.01) .76 3.61 (1.42, 9.19) .007 2.84 (1.11, 7.27) .03 7.40 (2.97, 18.43) <.001 5.51 (2.19, 13.85) <.001 1.61 (4.32, 26.03) <.001 8.26 (3.33, 2.5) <.001 1.72 (0.90, 3.29) .10 1.62 (0.84, 3.13) .15 2.82 (1.64, 4.84) <.001 2.19 (1.26, 3.81) .005 5.23 (3.08, 8.87) <.001 3.92 (2.28, 6.75) <.001 6.85 (4.08, 11.51) <.001 5.49 (3.22, 9.35) <.001 1.20 (0.70, 2.05) .51 1.11 (0.64, 1.93) .71 2.45 (1.60, 3.74) <.001 1.90 (1.23, 2.94) .004 4.08 (2.69, 6.19) <.001 3.08 (1.99, 4.75) <.001 5.01 (3.34, 7.51) <.001 4.06 (2.65, 6.20) <.001 90-day mortality Crude 1.00 (Referent)* Adjustedy 1.00 (Referent)* 365-day mortality Crude 1.00 (Referent)* Adjustedy 1.00 (Referent)* 720-day mortality Crude 1.00 (Referent)* Adjustedy 1.00 (Referent)* *Referent in each case is independent functional status. yEstimates adjusted for the acute organ failure score, sex, and injury severity score. OR, Odds ratio; CI, confidence interval. discharge on long-term mortality in survivors of trauma critical care. The current study has ample statistical power to detect a clinically relevant difference in 720-day postdischarge mortality if one exists. In determining the ICISS, we included deaths that occurred outside hospital (as many as 30 days after ICU admission), which likely improve the accuracy of the resulting severity estimates.52 We used previously validated approaches to define Functional Independence Measure mobility subscales, Deyo-Charlson index, intensive care unit admission, sepsis, acute kidney injury, noncardiogenic acute respiratory failure, and the acute organ failure score.19-22,24,25,35 Although we do not have data on cause of death, determination of allcause mortality was based on the Death Master File of the Social Security Administration and validated in the administrative dataset.21 Our study shows that in trauma patients, functional status is an important driver of adverse outcomes after hospital discharge. Trauma patients with improvement in their functional status prior to hospital discharge had better outcomes. Our data underscore the fact that high-risk patients with low physical function can be identified at hospital discharge. To modify potentially longterm outcomes in such high-risk patients, we suggest implementation of targeted practices, such as prioritizing admission to higher-level rehabilitation centers, more extensive follow-up in multispecialty ICU survivor clinics, longitudinal nutritional supplementation, and longer-term physical therapy interventions.53 This article is dedicated to the memory of our dear friend and colleague Nathan Edward Hellman, MD, PhD. We thank Shawn Murphy, MD, PhD, and Henry Chueh, MD, and the Partners HealthCare Research Patient Data Registry group for facilitating use of their database. Kenneth B. Christopher, MD, SM, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. SUPPLEMENTARY DATA Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.surg.2016.04.015. REFERENCES 1. Dowdy DW, Eid MP, Sedrakyan A, Mendez-Tellez PA, Pronovost PJ, Herridge MS, et al. Quality of life in adult survivors of critical illness: A systematic review of the literature. Intensive Care Med 2005;31:611-20. 2. van der Schaaf M, Beelen A, Dongelmans DA, Vroom MB, Nollet F. Poor functional recovery after a critical illness: A longitudinal study. J Rehabil Med 2009;41:1041-8. 3. Wunsch H, Guerra C, Barnato AE, Angus DC, Li G, Linde-Zwirble WT. Three-year outcomes for Medicare beneficiaries who survive intensive care. JAMA 2010;303: 849-56. ARTICLE IN PRESS 8 Peetz et al 4. Desai SV, Law TJ, Needham DM. Long-term complications of critical care. Crit Care Med 2011;39:371-9. 5. Herridge MS, Tansey CM, Matte A, Tomlinson G, DiazGranados N, Cooper A, et al. Functional disability 5 years after acute respiratory distress syndrome. N Engl J Med 2011;364:1293-304. 6. Brummel NE, Balas MC, Morandi A, Ferrante LE, Gill TM, Ely EW. Understanding and reducing disability in older adults following critical illness. Crit Care Med 2015;43:1265-75. 7. Bailey P, Thomsen GE, Spuhler VJ, Blair R, Jewkes J, Bezdjian L, et al. Early activity is feasible and safe in respiratory failure patients. Crit Care Med 2007;35:139-45. 8. Morris PE, Goad A, Thompson C, Taylor K, Harry B, Passmore L, et al. Early intensive care unit mobility therapy in the treatment of acute respiratory failure. Crit Care Med 2008;36:2238-43. 9. Schweickert WD, Pohlman MC, Pohlman AS, Nigos C, Pawlik AJ, Esbrook CL, et al. Early physical and occupational therapy in mechanically ventilated, critically ill patients: A randomised controlled trial. Lancet 2009;373: 1874-82. 10. Needham DM. Mobilizing patients in the intensive care unit: Improving neuromuscular weakness and physical function. JAMA 2008;300:1685-90. 11. Needham DM, Korupolu R, Zanni JM, Pradhan P, Colantuoni E, Palmer JB, et al. Early physical medicine and rehabilitation for patients with acute respiratory failure: A quality improvement project. Arch Phys Med Rehabil 2010;91:536-42. 12. Burtin C, Clerckx B, Robbeets C, Ferdinande P, Langer D, Troosters T, et al. Early exercise in critically ill patients enhances short-term functional recovery. Crit Care Med 2009;37:2499-505. 13. Engels PT, Beckett AN, Rubenfeld GD, Kreder H, Finkelstein JA, da Costa L, et al. Physical rehabilitation of the critically ill trauma patient in the ICU. Crit Care Med 2013;41:1790-801. 14. Murphy SN, Chueh HC. A security architecture for query tools used to access large biomedical databases. Proc AMIA Symp 2002:552-6. 15. Nalichowski R, Keogh D, Chueh HC, Murphy SN. Calculating the benefits of a research patient data repository. AMIA Annu Symp Proc 2006;2006:1044. 16. Granger CV, Hamilton BB, Zielezny M, Sherwin FS. Advances in functional assessment in medical rehabilitation. Topics Geriatric Rehabil 1986;1:59-74. 17. Dodds TA, Martin DP, Stolov WC, Deyo RA. A validation of the functional independence measurement and its performance among rehabilitation inpatients. Arch Phys Med Rehabil 1993;74:531-6. 18. About The FIM System. Uniform data system for medical rehabilitation: Uniform data system for medical rehabilitation; 2015 [cited 2015 Mar 13]. Available from: http:// www.udsmr.org/WebModules/FIM/Fim_About.aspx. 19. Rydingsward JE, Horkan CM, Mogensen KM, Quraishi SA, Amrein K, Christopher KB. Functional status in ICU survivors and out of hospital outcomes: A cohort study. Critical Care Med 2016;44:869-79. 20. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 1992;45:613-9. 21. Zager S, Mendu ML, Chang D, Bazick HS, Braun AB, Gibbons FK, et al. Neighborhood poverty rate and mortality in patients receiving critical care in the academic medical center setting. Chest 2011;139:1368-79. Surgery j 2016 22. Liu V, Escobar GJ, Greene JD, Soule J, Whippy A, Angus DC, et al. Hospital deaths in patients with sepsis from 2 independent cohorts. JAMA 2014;312:90-2. 23. Purtle SW, Moromizato T, McKane CK, Gibbons FK, Christopher KB. The association of red cell distribution width at hospital discharge and out-of-hospital mortality following critical illness. Crit Care Med 2014;42:918-29. 24. Braun AB, Litonjua AA, Moromizato T, Gibbons FK, Giovannucci E, Christopher KB. Association of low serum 25-hydroxyvitamin D levels and acute kidney injury in the critically ill. Crit Care Med 2012;40:3170-9. 25. Thickett DR, Moromizato T, Litonjua AA, Amrein K, Quraishi SA, Lee-Sarwar KA, et al. Association between prehospital vitamin D status and incident acute respiratory failure in critically ill patients: A retrospective cohort study. BMJ Open Respir Res 2015;2:e000074. 26. Shafi S, Aboutanos MB, Agarwal S Jr, Brown CV, Crandall M, Feliciano DV, et al. Emergency general surgery: Definition and estimated burden of disease. J Trauma Acute Care Surg 2013;74:1092-7. 27. Martin GS, Mannino DM, Eaton S, Moss M. The epidemiology of sepsis in the United States from 1979 through 2000. N Engl J Med 2003;348:1546-54. 28. Robinson MK, Mogensen KM, Casey JD, McKane CK, Moromizato T, Rawn JD, et al. The relationship among obesity, nutritional status, and mortality in the critically ill. Crit Care Med 2015;43:87-100. 29. Osler T, Rutledge R, Deis J, Bedrick E. ICISS: An international classification of disease-9 based injury severity score. J Trauma 1996;41:380-6; discussion 6–8. 30. Ciesla DJ, Tepas JJ 3rd, Pracht EE, Langland-Orban B, Cha JY, Flint LM. Fifteen-year trauma system performance analysis demonstrates optimal coverage for most severely injured patients and identifies a vulnerable population. J Am Coll Surg 2013;216:687-95; discussion 95–8. 31. Tepas JJ 3rd, Pracht EE, Orban BL, Flint LM. High-volume trauma centers have better outcomes treating traumatic brain injury. J Trauma Acute Care Surg 2013;74:143-7; discussion 7–8. 32. Stephenson S, Henley G, Harrison JE, Langley JD. Diagnosis based injury severity scaling: Investigation of a method using Australian and New Zealand hospitalisations. Inj Prev 2004;10:379-83. 33. Peetz A, Salim A, Askari R, De Moya MA, Simon TG, Gibbons FK, et al. A decrease in MELD score is associated with improved mortality in trauma patients with chronic liver disease. JAMA Surgery 2016;151:41-8. 34. Dreyfus B, Kawabata H, Gomez A. Selected adverse events in cancer patients treated with vascular endothelial growth factor inhibitors. Cancer Epidemiol 2013;37:191-6. 35. Elias KM, Moromizato T, Gibbons FK, Christopher KB. Derivation and validation of the acute organ failure score to predict outcome in critically ill patients: A cohort study. Crit Care Med 2015;43:856-64. 36. Barell V, Aharonson-Daniel L, Fingerhut LA, Mackenzie EJ, Ziv A, Boyko V, et al. An introduction to the Barell body region by nature of injury diagnosis matrix. Inj Prev 2002;8: 91-6. 37. Sohn MW, Arnold N, Maynard C, Hynes DM. Accuracy and completeness of mortality data in the Department of Veterans Affairs. Popul Health Metr 2006;4:2. 38. Pencina MJ, D’Agostino RB Sr, D’Agostino RB Jr, Vasan RS. Evaluating the added predictive ability of a ARTICLE IN PRESS Surgery Volume j, Number j 39. 40. 41. 42. 43. 44. 45. new marker: From area under the ROC curve to reclassification and beyond. Stat Med 2008;27:157-72; discussion 207–12. Iwashyna TJ, Ely EW, Smith DM, Langa KM. Long-term cognitive impairment and functional disability among survivors of severe sepsis. JAMA 2010;304:1787-94. Ferrante LE, Pisani MA, Murphy TE, Gahbauer EA, LeoSummers LS, Gill TM. Functional trajectories among older persons before and after critical illness. JAMA Intern Med 2015;175:523-9. Kortebein P, Ferrando A, Lombeida J, Wolfe R, Evans WJ. Effect of 10 days of bed rest on skeletal muscle in healthy older adults. JAMA 2007;297:1772-4. Iannuzzi-Sucich M, Prestwood KM, Kenny AM. Prevalence of sarcopenia and predictors of skeletal muscle mass in healthy, older men and women. J Gerontol A Biol Sci Med Sci 2002;57:M772-7. Batt J, dos Santos CC, Cameron JI, Herridge MS. Intensive care unit-acquired weakness: Clinical phenotypes and molecular mechanisms. Am J Respir Crit Care Med 2013;187: 238-46. Bagshaw SM, Stelfox HT, McDermid RC, Rolfson DB, Tsuyuki RT, Baig N, et al. Association between frailty and short- and long-term outcomes among critically ill patients: A multicentre prospective cohort study. CMAJ 2014;186: E95-102. Bagshaw SM, Stelfox HT, Johnson JA, McDermid RC, Rolfson DB, Tsuyuki RT, et al. Long-term association between frailty and health-related quality-of-life among survivors of critical illness: A prospective multicenter cohort study. Crit Care Med 2015;43:973-82. Peetz et al 9 46. Puthucheary ZA, Rawal J, McPhail M, Connolly B, Ratnayake G, Chan P, et al. Acute skeletal muscle wasting in critical illness. JAMA 2013;310:1591-600. 47. Jackson JC, Obremskey W, Bauer R, Greevy R, Cotton BA, Anderson V, et al. Long-term cognitive, emotional, and functional outcomes in trauma intensive care unit survivors without intracranial hemorrhage. J Trauma 2007;62:80-8. 48. Sluys K, Haggmark T, Iselius L. Outcome and quality of life 5 years after major trauma. J Trauma 2005;59:223-32. 49. Ulvik A, Kvale R, Wentzel-Larsen T, Flaatten H. Multiple organ failure after trauma affects even long-term survival and functional status. Crit Care 2007;11:R95. 50. Kayambu G, Boots R, Paratz J. Physical therapy for the critically ill in the ICU: A systematic review and meta-analysis. Crit Care Med 2013;41:1543-54. 51. Rutledge R, Osler T, Emery S, Kromhout-Schiro S. The end of the Injury Severity Score (ISS) and the Trauma and Injury Severity Score (TRISS): ICISS, an International Classification of Diseases, ninth revision-based prediction tool, outperforms both ISS and TRISS as predictors of trauma patient survival, hospital charges, and hospital length of stay. J Trauma 1998;44:41-9. 52. Henley G, Harrison JE. Injury severity scaling: A comparison of methods for measurement of injury severity. Injury Technical paper series number 10. Category number INJCAT 126. Canberra: Australian Institute of Health and Welfare; 2009. 53. Waters DL, Baumgartner RN, Garry PJ, Vellas B. Advantages of dietary, exercise-related, and therapeutic interventions to prevent and treat sarcopenia in adult patients: An update. Clin Interv Aging 2010;5:259-70.