- Email: [email protected]
Outcomes in nursing home patients with traumatic brain injury
ARTICLE IN PRESS
JID: YMSY
[m5G;May 8, 2018;8:51]
Surgery 0 0 0 (2018) 1–5
Contents lists available at ScienceDirect
Surgery journal homepage: www.elsevier.com/locate/surg
Outcomes in nursing home patients with traumatic brain injury Stephanie N. Lueckel a,c,∗, Cyrus M. Kosar a, Joan M. Teno b, Sean F. Monaghan c, Daithi S. Heffernan a,d, William G. Cioffi c, Kali S. Thomas a,d a
Brown University School of Public Health, Providence, RI University of Washington School of Public Health, Seattle, WA c Warren Alpert Medical School at Brown University, Providence, RI d Department of Veterans Affairs Medical Center, Providence, RI b
a r t i c l e
i n f o
Article history: Accepted 22 February 2018 Available online xxx
a b s t r a c t Background: Traumatic brain injury is a leading cause of death and disability in the United States. In survivors, traumatic brain injury remains a leading contributor to long-term disability and results in many patients being admitted to skilled nursing facilities for postacute care. Despite this very large population of traumatic brain injury patients, very little is known about the long-term outcomes of traumatic brain injury survivors, including rates of discharge to home or risk of death in long-term nursing facilities. We hypothesized that patient demographics and functional status influence outcomes of patients with traumatic brain injury admitted to skilled nursing facilities. Methods: We conducted a retrospective cohort study of Medicare fee-for-service beneficiaries aged 65 and older discharged alive and directly from hospital to a skilled nursing facility between 2011 and 2014 using the prospectively maintained Federal Minimum Data Set combined with Medicare claims data and the Centers for Medicare and Medicaid Services Vital Status files. Records were reviewed for demographic and clinical characteristics at admission to the skilled nursing facility, including age, sex, cognitive function, ability to communicate, and motor function. Activities of daily living were reassessed at discharge to calculate functional improvement. We used robust Poisson regression with skilled nursing facility fixed effects to calculate relative risks and 99% confidence intervals for mortality and functional improvement associated with the demographic and clinical characteristics present at admission. Linear regression was used to calculate adjusted mean duration of stay. Results: Overall, 87,292 Medicare fee-for-service beneficiaries with traumatic brain injury were admitted to skilled nursing facilities. The mean age was 84 years, with 74% of patients older than age 80. Generally, older age, male sex, and poor cognitive or functional status at admission to a skilled nursing facility were associated with increased risk for poorer outcomes. Older patients (age ≥80 years) with traumatic brain injury had a 1.5 times greater risk of death within 30 days of admission compared with adults younger than 80 years (relative risk = 1.49, 99% confidence interval = 1.36, 1.64). Women were 37% less likely to die than men were (relative risk = 0.63, 99% confidence interval = 0.59, 0.68). The risk of death was greater for patients with poor cognitive function (relative risk = 2.55, 99% confidence interval = 2.32, 2.77), substantial motor impairment (relative risk = 2.44, 99% confidence interval = 2.16, 2.77), and patients with impairment in communication (relative risk = 2.58, 99% confidence interval = 2.32, 2.86) compared with those without the respective deficits. One year after admission, these risk factors continued to confer excess risk for mortality. Duration of stay was somewhat greater for older patients (30.1 compared with 27.5 average days) and patients with cognitive impairment (31.7 vs 27.5 average days). At discharge, patients with cognitive impairment (relative risk = 0.86, 99% confidence interval = 0.83, 0.88) and impairment in the ability to communicate (relative risk = 0.67, 99% confidence interval = 0.54, 0.82) were less likely to improve in physical function.
∗ Corresponding author: Brown University Surgery, 593 Eddy Street, APC 4th Floor Rhode Island Hospital, Providence, RI 02903. Tel.: 401-444-6461. E-mail address: [email protected] (S.N. Lueckel).
https://doi.org/10.1016/j.surg.2018.02.023 0039-6060/© 2018 Elsevier Inc. All rights reserved.
Please cite this article as: S.N. Lueckel et al., Outcomes in nursing home patients with traumatic brain injury, Surgery (2018), https://doi.org/10.1016/j.surg.2018.02.023
JID: YMSY 2
ARTICLE IN PRESS
[m5G;May 8, 2018;8:51]
S.N. Lueckel et al. / Surgery 000 (2018) 1–5
Conclusion: Our results suggest that among patients with traumatic brain injury admitted to skilled nursing facilities, the likelihood of adverse outcomes varies significantly by key demographic and clinical characteristics. These findings may facilitate setting expectations among patients and families as well as providers when these patients are admitted to skilled nursing facilities for rehabilitation after their acute episode. © 2018 Elsevier Inc. All rights reserved.
Introduction In 2010, 2.5 million people suffered a traumatic brain injury (TBI) at a cost of $76.5 billion. TBI is a leading cause of death and disability in the United States, accounting for 30% of all traumarelated deaths. Despite the high rate of mortality, it is estimated that 5.3 million individuals are living with disabilities resulting from TBI.1 Many of these patients suffer physical, emotional, cognitive, and behavioral disabilities that require treatment in postacute care facilities. The largest proportion of patients with TBI necessitating postacute care are discharged to skilled nursing facilities (SNFs).2,3 Little, however, is known about the TBI population admitted to SNFs and their outcomes. As such, postacute care for patients with TBI in SNFs remains unstandardized with no evidence-based guidelines or treatment protocols. Before such treatment protocols can be established, however, an understanding of the population and their trajectory is needed to guide the development of treatment strategies. The objective of this study was to describe the outcomes of patients with TBI who are admitted to SNFs for postacute care by key demographic and clinical characteristics. We hypothesized that within the population of TBI patients receiving postacute care, clinical and demographic characteristics, such as older age, male sex, and lower functional status at admission, confer excess risk for poor outcomes. Understanding this specific population and whether or not differences exist by readily available risk factors and assessed at admission to postacute care will assist in care planning, optimizing therapy (physical, occupational, psychological), and informing clinical expectations to families.
prior postacute SNF use within 1 year of the index hospitalization. TBI was identified if any one of the following International Classification of Diseases, Ninth Revision (ICD-9) codes was present on the hospitalization claim: 800–801.99, 803–804.99, 850–854.99, 310.2, 959.01, 907.0, 905.0, 873.0, 873.1.1 This study was approved by the Institutional Review Board at Brown University. Outcomes We examined mortality at 30 days, 90 days, and 1 year after admission to an SNF. For residents who were discharged from the SNF, we assessed the duration of SNF stay and functional improvement during the SNF stay. Functional improvement was measured using change in the 28point Activities of Daily Living (ADL) Scale score4 between admission and discharge assessments. The ADL score ranges from 0 (no impairment) to 28 (total dependence). Residents were considered to have improved in function if their discharge ADL score was less than their admission score. Because functional improvement may influence discharge, this data point was analyzed separately for patients with duration of stay less than 30 days. Demographics and admission characteristics
Data for this study come from several sources (Medicare Claims [Part A] and Enrollment Records, Minimum Data Set [MDS] resident assessments). Medicare Claims data were used to identify all hospitalizations and discharges to SNF among Medicare feefor-service beneficiaries. The Medicare enrollment records contain demographic information about each beneficiary ever entitled to Medicare including their date of death. The MDS is a federally mandated database that includes a comprehensive clinical assessment of individuals admitted to Medicare or Medicaid certified nursing facilities. The MDS contains items that measure physical, psychological, and psychosocial functioning and give a multidimensional view of the resident’s functional capacities. These data are required to be collected at admission, quarterly, annually, at discharge, and whenever there is a substantial change in a resident’s status.
The factors examined have been linked previously to outcomes of patients with TBI in other settings.5-7 Specifically, we examined differences in outcomes by age, sex, cognitive function, communication ability, and motor function. Age and sex were obtained from the Medicare enrollment records. Cognitive function, communication ability, and motor function were derived from the MDS admission assessment. Cognitive function was captured with the Cognitive Function Scale (CFS),8 derived from the Brief Interview for Mental Status and the Cognitive Performance Scale.9 Scores ranges from 1 (cognitively intact) to 4 (severely impaired). We dichotomized this variable to indicate whether or not patients were cognitively intact (CFS = 1) versus impaired (CFS <1). Ability to communicate was determined by staff rating of the level to which patients were able to make themselves understood (verbally and nonverbally), such as through expression of ideas and wants. Patients were considered to have severe impairment in communication if they were rarely or never understood. Motor function was ascertained from the transferring ADL score: a rating of the patient’s ability to move between surfaces (to or from) such as a bed, chair, wheelchair, or simply transition from sitting to standing. Patients were considered to have severe impairment in motor function if they were totally dependent on staff to transfer or transferring did not occur during the week of the assessment. Finally, we calculated the Deyo-Charlson Comorbidity Index from the diagnoses present on admission to the SNF as descriptors of patient acuity.10,11
Study population
Statistical analysis
The study was conducted on a cohort of hospitalized Medicare beneficiaries with an active diagnosis of TBI who were discharged directly to an SNF between January 1, 2011 and December 31, 2014. We excluded individuals who were <65 years of age and who had
Descriptive characteristics of the study population were summarized with standard measures, including means ± standard deviations, frequencies, and proportions. Using robust Poisson regression,12 we calculated relative risks (RRs) for mortality and
Methods Data
Please cite this article as: S.N. Lueckel et al., Outcomes in nursing home patients with traumatic brain injury, Surgery (2018), https://doi.org/10.1016/j.surg.2018.02.023
JID: YMSY
ARTICLE IN PRESS
[m5G;May 8, 2018;8:51]
S.N. Lueckel et al. / Surgery 000 (2018) 1–5 Table 1 Characteristics of TBI patients admitted to skilled nursing facilities after an inpatient hospitalization. Characteristic
N = 87292
Age, mean y (SD) Age ≥80 y, n (%) Female sex, n (%) Nonwhite race, n (%) ADL at baseline, mean score (SD) ADL ≥23, n (%) Deyo-Charlson Comorbidity Index, mean (SD) Deyo-Charlson Comorbidity Index Score ≥2, n (%) CFS Score, n (%) 1: Intact 2: Mild impairment 3: Moderate impairment 4: Severe impairment Terminal illness or on hospice, n (%)
83.9 (7.5) 64,541 (74) 55,418 (63) 7,455 (9) 18.1 (4.4) 8,674 (10) 1.2 (1.4) 28,106 (32) 37,714 (43) 22,043 (25) 20,282 (23) 5,144 (6) 753 (1)
Deyo-Charlson Comorbidity Index ranges from a score of 0 to 3 or more with greater scores indicating more comorbidities.11 The ADL Scale ranges from 0 to 28 with greater scores indicating more dependence. The CFS is a 4-point scale ranging from 1 (intact) to 4 (severely impaired). SD, standard deviation
functional improvement associated with older age, female sex, cognitive impairment, impairment in communication, and motor impairment for patients with TBI in SNFs. Linear regression was used to calculate the adjusted mean duration of stay associated with each factor present on admission (older age, female sex, cognitive impairment, impairment in communication, and impairment in motor function). To avoid potential type I errors from making multiple comparisons, the Bonferroni correction was applied, and point estimates with P < .008 were considered significant. Accordingly, results are presented with 99.2% confidence intervals (CIs). We also adjusted for factors that could influence a patient’s care (ie, the effect of the nursing home, whether or not the resident was receiving hospice services, and whether or not the resident had a documented prognosis of less than 6 months). Point estimates for each factor were obtained from a single adjusted model, which also included an indicator for terminal illness, hospice use, and SNF fixed effects. All statistical analysis was done in Stata (Version 15, Stata Corporation, College Station, TX) Results Between 2011 and 2014, 87,292 hospitalized Medicare beneficiaries aged 65 years and older without a nursing facility stay within 1 year of hospital admission were discharged to SNFs with an active TBI diagnosis (Table 1). The average age was 83.9 ± 7.5 years with 74% at or older than age 80. Females and nonwhites accounted for 63% and 9% of the population, respectively. The average Deyo-Charlson Comorbidity Index was 1.2 ± 1.4 with 32% having a score ≥2. The average baseline ADL score was 18.1 ± 4.4, and severe functional impairment (an ADL score of ≥23) was present in 10% of residents. Approximately 57% of residents had some level of cognitive impairment (CFS >1), and 29% had more severe cognitive deficits (CFS >2). Mortality Older age, cognitive impairment, impairments in communication, and severe motor deficits conferred greater risk for mortality among residents with TBI (Table 2). The 30-day mortality rate among residents aged <80 years was 5.9% compared with 9.1% in those aged 80 years and older (RR = 1.49, 99% CI = 1.36, 1.64); this trend was consistent at 1 year after discharge (RR = 1.42, 99% CI = 1.37, 1.48). Female sex was protective against mortality. The
30-day mortality rate for female residents was 6.6% compared with 11.1% among male residents (RR = 0.63, 9% CI = 0.59, 0.68). One year after admission, 26.2% of women had died compared with 38.4% of men (RR = 0.70, 9% CI = 0.68, 0.72). Among residents with cognitive impairment, 11.1% died within 30 days of admission to an SNF compared with 3.2% of those without impairment (RR = 2.55, 99% CI = 2.32, 2.80); this difference endured at 1 year after admission (RR = 1.74, 99% CI = 1.69, 1.79). We found that 26.4% of residents with severe motor impairment died within 30 days of admission to an SNF compared with 6.6% of those without impairment or with limited motor function (RR = 2.58, 99% CI = 2.32, 2.86); this difference persisted at 1 year (RR = 1.46, 99% CI = 1.40, 1.53). Among residents with severe impairment in communication, 34.1% died within 30 days of admission to an SNF compared with 7.0% of those without or with milder impairments (RR = 2.44, 9% CI = 2.16, 2.77); similar results were seen at 1 year (RR = 1.49, 9% CI = 1.41, 1.58). Duration of stay Among SNF residents discharged to the community, those who were older, female, and impaired in cognition, communication, or motor function had on average greater durations of stay (Table 3). The mean duration of SNF stay for patients <80 years of age was 27.5 days compared with 30.1 days for residents aged 80 and older (adjusted difference = 2.6 days, 99% CI = 1.6, 3.8). Male residents stayed in the SNF for 27.8 days on average, whereas female residents stayed for 30.2 days (adjusted difference = 2.4 days, 99% CI = 1.5, 3.4). Those with cognitive impairment had a mean duration of stay of 31.7 days compared with 27.5 days among those without cognitive impairment (adjusted difference = 4.2 days, 99% CI = 3.1, 5.2). Mean duration of stay for patients with severe impairment in communication was not significantly different. Mean duration of stay among patients with severe impairment in motor function was 47.9 days compared with 28.7 days among those without impairment or with limited motor function (adjusted difference = 19.2 days, 99% CI = 15.0, 23.4). Functional improvement at discharge Among residents discharged within 30 days of SNF admission, older age and female sex did not affect functional improvement, but the probability of functional improvement was significantly decreased for residents with cognitive impairment and severe impairment in the ability to communicate (Table 3). Among residents with cognitive impairment, 57.4% had functional improvement at discharge compared with 68.8% without impairment (RR = 0.86, 99% CI = 0.83, 0.88). Among residents with severe impairment in communication, 36.9% had functional improvement at discharge compared with 63.8% among residents without impairment or with milder deficits (RR =0 .67, 99% CI = 0.54, 0.82). Residents with severe motor impairment were actually more likely to improve their function (RR = 1.13, 9% CI = 1.02, 1.25). Point estimates were consistent in analyses on residents discharged after 30 days of SNF stay except that older age decreased the probability of functional improvement among these residents (RR = 0.96, 9% CI = 0.93, 0.99). Discussion In this study of more than 87,0 0 0 Medicare beneficiaries with TBI admitted to SNF, patients who were older or who had impairments in cognition, communication, or motor function at the time of admission had markedly greater rates of mortality and extended durations of stay. Impairments in cognition and
Please cite this article as: S.N. Lueckel et al., Outcomes in nursing home patients with traumatic brain injury, Surgery (2018), https://doi.org/10.1016/j.surg.2018.02.023
3
ARTICLE IN PRESS
JID: YMSY 4
[m5G;May 8, 2018;8:51]
S.N. Lueckel et al. / Surgery 000 (2018) 1–5 Table 2 Adjusted risk of mortality conferred by demographic and clinical characteristics among patients with TBI admitted to SNFs after an inpatient hospitalization. 30 Days
Age <80 y Age ≥80 y Male Female CFS <2 CFS ≥2 Active/limited communication Severely impaired communication Intact/limited motor function Severely impaired motor function
90 Days
1 Year
n/N (%)
RR (99.2% CI)
n/N (%)
RR (99.2% CI)
n/N (%)
RR (99.2% CI)
1,332/22,751 (5.9) 5,860/64,541 (9.1) 3,541/31,874 (11.1) 3,651/55,418 (6.6) 1,210/37,714 (3.2) 5,377/47,469 (11.3) 5,834/83,673 (7.0) 978/2,864 (34.1) 5,269/79,998 (6.6) 1,916/7,260 (26.4)
— 1.49 (1.36, 1.64) — 0.63 (0.59, 0.68) — 2.55 (2.32, 2.80) — 2.44 (2.16, 2.77) — 2.58 (2.32, 2.86)
2,638/22,751 (11.6) 11,066/64,541 (17.1) 6,578/31,874 (20.6) 7,126/55,418 (12.9) 2,939/37,714 (7.8) 9,997/47,469 (21.1) 11,936/83,673 (14.3) 1,340/2,864 (46.8) 10,952/79,998 (13.7) 2,741/7,260 (37.8)
— 1.44 (1.35, 1.53) — 0.65 (0.62, 0.69) — 2.16 (2.03, 2.29) — 1.83 (1.68, 2.00) — 1.90 (1.77, 2.04)
5,236/22,751 (23.0) 21,508/64,541 (33.3) 12,239/31,874 (38.4) 14,505/55,418 (26.2) 7,254/37,714 (19.2) 18,435/47,469 (38.8) 24,386/83,673 (29.1) 1,846/2,864 (64.5) 22,808/79,998 (28.5) 3,915/7,260 (53.9)
— 1.42 (1.37, 1.48) — 0.70 (0.68, 0.72) — 1.74 (1.68, 1.80) — 1.49 (1.41, 1.58) — 1.46 (1.40, 1.53)
The CFS is a 4-point scale ranging from 1 (intact) to 4 (severely impaired). Patients were considered to have severe impairment in communication if they were rarely or never understood. Patients were considered to have severe impairment in motor function if they were totally dependent on staff to transfer or transferring did not occur. Models were adjusted for SNF fixed effects, hospice services, and resident prognosis <6 months. Table 3 Adjusted length of stay and probability of functional improvement among patients with TBI admitted to skilled nursing facilities after an inpatient hospitalization, by demographic and clinical characteristics. Duration of stay
Age <80 y Age ≥80 y Male Female CFS <2 CFS ≥2 Active/limited communication Severely impaired communication Intact/limited motor function Severely Impaired Motor Function
Functional improvement: discharged within 30 days
Functional improvement: discharged after 30 days
Adjusted mean
Mean difference (99.2% CI)
n/N (%)
RR (99.2% CI)
n/N (%)
RR (99.2% CI)
27.5 30.1 27.8 30.2 27.5 31.7 29.3
— 2.6 (1.6, 3.8) — 2.4 (1.5, 3.4) — 4.2 (3.1, 5.2) —
6,397/10,024 (63.8) 15,369/24,229 (63.4) 7,363/11,859 (62.1) 14,403/22,394 (64.3) 13,609/19,791 (68.8) 8,089/14,083 (57.4) 21,634/33,838 (63.9)
— 1.01 (0.99, 1.04) — 1.02 (0.99, 1.04) — 0.86 (0.83, 0.88) —
3,376/3,992 (84.6) 9,988/12,498 (79.9) 4,227/5,270 (80.2) 9,137/11,220 (81.4) 6,782/8,018 (84.6) 6,537/8,405 (77.8) 13,186/16,223 (81.3)
— 0.96 (0.93, 0.99) — 1.00 (0.97, 1.03) — 0.92 (0.90, 0.95) —
32.7
3.4 (–2.9, 9.7)
123/333 (36.9)
0.67 (0.54, 0.82)
170/255 (66.7)
0.89 (0.76, 1.03)
28.7
—
21,353/33,492 (63.8)
—
12,347/15,375 (80.3)
—
47.9
19.2 (15.0, 23.4)
413/748 (55.2)
1.13 (1.02, 1.25)
1,017/1,115 (91.2)
1.16 (1.11, 1.21)
The CFS is a 4-point scale ranging from 1 (intact) to 4 (severely impaired). Patients were considered to have severe impairment in communication if they were rarely or never understood. Patients were considered to have severe impairment in motor function if they were totally dependent on staff to transfer or transferring did not occur. Models were adjusted for SNF fixed effects, hospice services, and resident prognosis <6 months.
communication were also associated with a decreased probability of functional recovery. Though there is no direct comparison, our findings are similar to literature that describes the outcomes of patients with injuries or postoperative patients who are admitted to an SNF for postacute care. A study in 2010 examined patients who suffered from a subarachnoid hemorrhage admitted to an SNF. Of these patients, 49% were dead after a median of 1.1 years after admission and 69% did not have functional improvement at 2 years.13 Furthermore, Oresanya et al14 reported on studied patients who underwent lower extremity revascularization who were admitted to an SNF. At 1 year after admission to an SNF, 51% of the patients had died, 285 were nonambulatory, and 32% had experienced functional decline. Contributing factors to poorer outcomes included age, poor functional status on admission, decrease in ability to maintain ADLs, and cognitive impairment.14 The lack of long-term functional improvement in patients with TBI highlights the complexity of these injuries, but slow functional recovery in patients with TBI has also been reported previously.15 Though overall recovery appears poor in a majority of these patients, recovery does not occur at a steady, continuous rate. Literature supports that recovery tends to occur most quickly earlier in their SNF admission.16,17 In our study the majority of residents were discharged to the community within 30 days of admission, and this would suggest that the majority of functional improvement occurred during the earlier time period. This observa-
tion is consistent with literature that has reported a steep recovery curve in the first 5–6 months after TBI. It is further suggested that this recovery is not maintained.16,17 Additional studies suggest that these same patients went on to develop decreases in cognitive function as they moved further and further out from their injury.16,18 This observation lends support to the importance of early, appropriate, and directed therapies for these patients and may be used to guide future therapies in both SNFs and rehabilitative facilities dedicated to TBIs, as well as generate realistic expectations for both physicians and families. The findings of our study are important not only patients but also for the patients’ families in the acute setting. Knowing the types of patients who are less likely to regain the ability to live independently and who have an increased risk of dying can help guide physicians and families making clinical and long-term decisions. Consequently, these findings may affect the care and planning of the facilities that accept these patients for postacute care, such as SNFs, because they will need to adjust accordingly. Future work is needed to determine whether other factors such as the severity of the TBI drive the association of certain patient characteristics and outcomes.19 Additionally, therapies and treatments received at the time of admission to an SNF may also play an early and important clinical role in the outcomes of these patients. The overall strength of this study is reflected in the large, inclusive, national database we accessed as well as the evaluation of specific clinical outcomes. Knowing and understanding the
Please cite this article as: S.N. Lueckel et al., Outcomes in nursing home patients with traumatic brain injury, Surgery (2018), https://doi.org/10.1016/j.surg.2018.02.023
JID: YMSY
ARTICLE IN PRESS
[m5G;May 8, 2018;8:51]
S.N. Lueckel et al. / Surgery 000 (2018) 1–5
outcomes of patients with TBI newly discharged to SNF after an acute hospitalization can help guide the acute care of these unique patients. Limitations of this study: This study depended on the accurate diagnosis of TBI. Although we did attempt to be comprehensive in our definitions of TBI with ICD-9 codes, there is an element of subjectivity in assigning some of these codes; however, our per-year estimates of the number of SNF residents with TBI were consistent with prior reports.20,21 An additional limitation to this work is that we were unable to assess the severity, mechanism, or timing of the injury. Future work incorporating these characteristics of the injury will be an important next step in understanding the outcomes of patients with TBI. This study did not assess the effect of therapies received in the SNF on outcomes. Other research suggests that directed rehabilitation does improve both functional and cognitive ability; however, those with more severe impairments on admission are less likely to improve beyond their baseline status.22,23 In our study we were unable to determine if the lack of improvement was due to the underlying injury or rather the specific or nonspecific therapies they received. Understanding which therapies were offered and which therapies were received by patients with TBI in SNFs is an important area for future research and may guide future care of these patients. Finally, we are unable to characterize and analyze SNF residents who were not Medicare feefor-service beneficiaries, and we were unable to assess functional change among residents who were not discharged to the community, including those who died during their nursing facility stay. Conclusion Among a population with of patients with a TBI admitted to an SNF, the probability of poor outcomes varies significantly by key clinical and demographic characteristics. Understanding these differences and possibilities can help set expectations for patients and families as well as providers when these patients are admitted to SNFs for rehabilitation after their acute episode. References 1. Centers for Disease Control and Prevention (CDC). Traumatic brain injury & concussion [Internet]. Atlanta (GA): CDC. TBI: get the facts; [updated 2017 Apr 27], accessed on April 30, 2016. Available from: http://www.cdc.gov/ traumaticbraininjury/get_the_facts.html. 2. Tepas 3rd JJ, Pracht EE, Orban BL, Flint LM. High-volume trauma centers have better outcomes treating traumatic brain injury. J Trauma Acute Care Surg. 2013;74:143–147 discussion 7-8.
3. Hakkarainen TW, Ayoung-Chee P, Alfonso R, Arbabi S, Flum DR. Structure, process, and outcomes in skilled nursing facilities: understanding what happens to surgical patients when they cannot go home. A systematic review. J Surg Res. 2015;193:772–780. 4. Wysocki A, Thomas KS, Mor V. Functional improvement among short-stay nursing home residents in the MDS 3.0. J Am Med Directors Assoc. 2015;16:470–474. 5. Mosenthal AC, Lavery RF, Addis M, Kaul S, Ross S, Marburger R, et al. Isolated traumatic brain injury: age is an independent predictor of mortality and early outcome. J Trauma. 2002;52:907–911. 6. Leitgeb J, Mauritz W, Brazinova A, Janciak I, Majdan M, Wilbacher I, et al. Effects of gender on outcomes after traumatic brain injury. J Trauma. 2011;71:1620–1626. 7. Kouloulas EJ, Papadeas AG, Michail X, Sakas DE, Boviatsis EJ. Prognostic value of time-related Glasgow coma scale components in severe traumatic brain injury: a prospective evaluation with respect to 1-year survival and functional outcome. Int J Rehabil Res. 2013;36:260–267. 8. Thomas KS, Dosa D, Wysocki A, Mor V. The Minimum Data Set 3.0 Cognitive Function Scale. Med Care. 2015;55:e68–e72. 9. Morris JN, Fries BE, Mehr DR, Hawes C, Phillips C, Mor V, et al. MDS Cognitive Performance Scale. J Gerontol. 1994;49:M174–M182. 10. Klabunde CN, Potosky AL, Legler JM, Warren JL. Development of a comorbidity index using physician claims data. J Clin Epidemiol. 20 0 0;53:1258–1267. 11. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992;45:613–619. 12. Zou G. A modified poisson regression approach to prospective studies with binary data. Am J Epidemiol. 2004;159:702–706. 13. Greebe P, Rinkel GJ, Algra A. Long-term outcome of patients discharged to a nursing home after aneurysmal subarachnoid hemorrhage. Arch Phys Med Rehabil. 2010;91:247–251. 14. Oresanya L, Zhao S, Gan S, Fries BE, Goodney PP, Covinsky KE, et al. Functional outcomes after lower extremity revascularization in nursing home residents: a national cohort study. JAMA Intern Med. 2015;175:951–957. 15. Leon-Carrion J, Dominguez-Morales MR, Barroso y Martin JM, Leon– Dominguez U. Recovery of cognitive function during comprehensive rehabilitation after severe traumatic brain injury. J Rehabil Med. 2012;44:505–511. 16. Christensen BK, Colella B, Inness E, Hebert D, Monette G, Bayley M, et al. Recovery of cognitive function after traumatic brain injury: a multilevel modeling analysis of Canadian outcomes. Arch Phys Med Rehabil. 2008;89(12 Suppl):S3–15. 17. Green RE, Colella B, Hebert DA, Bayley M, Kang HS, Till C, et al. Prediction of return to productivity after severe traumatic brain injury: investigations of optimal neuropsychological tests and timing of assessment. Arch Phys Med Rehabil. 2008;89(12 Suppl):S51–S60. 18. Till C, Colella B, Verwegen J, Green RE. Postrecovery cognitive decline in adults with traumatic brain injury. Arch Phys Med Rehabil. 2008;89(12 Suppl):S25–S34. 19. Fuller GW, Ransom J, Mandrekar J, Brown AW. Long-term survival following traumatic brain injury: a population-based parametric survival analysis. Neuroepidemiology. 2016;47:1–10. 20. Gabella BA, Mangione EJ, Hedegaard H, Kelly JP. Comparison of nursing home residents with and without traumatic brain injury: use of the Minimum Data Set. J Head trauma Rehabil. 2007;22:368–376. 21. Buchanan RJ, Wang S, Huang C. Profiles of nursing home residents with traumatic brain injury using the Minimum Data Set. Brain Injury. 2003;17:507–523. 22. Sandhaug M, Andelic N, Vatne A, Seiler S, Mygland A. Functional level during sub-acute rehabilitation after traumatic brain injury: course and predictors of outcome. Brain Injury. 2010;24:740–747. 23. Park HY, Maitra K, Martinez KM. The effect of occupation-based cognitive rehabilitation for traumatic brain injury: a meta-analysis of randomized controlled trials. Occupat Ther Int. 2015;22:104–116.
Please cite this article as: S.N. Lueckel et al., Outcomes in nursing home patients with traumatic brain injury, Surgery (2018), https://doi.org/10.1016/j.surg.2018.02.023
5
JID: YMSY
[m5G;May 8, 2018;8:51]
Surgery 0 0 0 (2018) 1–5
Contents lists available at ScienceDirect
Surgery journal homepage: www.elsevier.com/locate/surg
Outcomes in nursing home patients with traumatic brain injury Stephanie N. Lueckel a,c,∗, Cyrus M. Kosar a, Joan M. Teno b, Sean F. Monaghan c, Daithi S. Heffernan a,d, William G. Cioffi c, Kali S. Thomas a,d a
Brown University School of Public Health, Providence, RI University of Washington School of Public Health, Seattle, WA c Warren Alpert Medical School at Brown University, Providence, RI d Department of Veterans Affairs Medical Center, Providence, RI b
a r t i c l e
i n f o
Article history: Accepted 22 February 2018 Available online xxx
a b s t r a c t Background: Traumatic brain injury is a leading cause of death and disability in the United States. In survivors, traumatic brain injury remains a leading contributor to long-term disability and results in many patients being admitted to skilled nursing facilities for postacute care. Despite this very large population of traumatic brain injury patients, very little is known about the long-term outcomes of traumatic brain injury survivors, including rates of discharge to home or risk of death in long-term nursing facilities. We hypothesized that patient demographics and functional status influence outcomes of patients with traumatic brain injury admitted to skilled nursing facilities. Methods: We conducted a retrospective cohort study of Medicare fee-for-service beneficiaries aged 65 and older discharged alive and directly from hospital to a skilled nursing facility between 2011 and 2014 using the prospectively maintained Federal Minimum Data Set combined with Medicare claims data and the Centers for Medicare and Medicaid Services Vital Status files. Records were reviewed for demographic and clinical characteristics at admission to the skilled nursing facility, including age, sex, cognitive function, ability to communicate, and motor function. Activities of daily living were reassessed at discharge to calculate functional improvement. We used robust Poisson regression with skilled nursing facility fixed effects to calculate relative risks and 99% confidence intervals for mortality and functional improvement associated with the demographic and clinical characteristics present at admission. Linear regression was used to calculate adjusted mean duration of stay. Results: Overall, 87,292 Medicare fee-for-service beneficiaries with traumatic brain injury were admitted to skilled nursing facilities. The mean age was 84 years, with 74% of patients older than age 80. Generally, older age, male sex, and poor cognitive or functional status at admission to a skilled nursing facility were associated with increased risk for poorer outcomes. Older patients (age ≥80 years) with traumatic brain injury had a 1.5 times greater risk of death within 30 days of admission compared with adults younger than 80 years (relative risk = 1.49, 99% confidence interval = 1.36, 1.64). Women were 37% less likely to die than men were (relative risk = 0.63, 99% confidence interval = 0.59, 0.68). The risk of death was greater for patients with poor cognitive function (relative risk = 2.55, 99% confidence interval = 2.32, 2.77), substantial motor impairment (relative risk = 2.44, 99% confidence interval = 2.16, 2.77), and patients with impairment in communication (relative risk = 2.58, 99% confidence interval = 2.32, 2.86) compared with those without the respective deficits. One year after admission, these risk factors continued to confer excess risk for mortality. Duration of stay was somewhat greater for older patients (30.1 compared with 27.5 average days) and patients with cognitive impairment (31.7 vs 27.5 average days). At discharge, patients with cognitive impairment (relative risk = 0.86, 99% confidence interval = 0.83, 0.88) and impairment in the ability to communicate (relative risk = 0.67, 99% confidence interval = 0.54, 0.82) were less likely to improve in physical function.
∗ Corresponding author: Brown University Surgery, 593 Eddy Street, APC 4th Floor Rhode Island Hospital, Providence, RI 02903. Tel.: 401-444-6461. E-mail address: [email protected] (S.N. Lueckel).
https://doi.org/10.1016/j.surg.2018.02.023 0039-6060/© 2018 Elsevier Inc. All rights reserved.
Please cite this article as: S.N. Lueckel et al., Outcomes in nursing home patients with traumatic brain injury, Surgery (2018), https://doi.org/10.1016/j.surg.2018.02.023
JID: YMSY 2
ARTICLE IN PRESS
[m5G;May 8, 2018;8:51]
S.N. Lueckel et al. / Surgery 000 (2018) 1–5
Conclusion: Our results suggest that among patients with traumatic brain injury admitted to skilled nursing facilities, the likelihood of adverse outcomes varies significantly by key demographic and clinical characteristics. These findings may facilitate setting expectations among patients and families as well as providers when these patients are admitted to skilled nursing facilities for rehabilitation after their acute episode. © 2018 Elsevier Inc. All rights reserved.
Introduction In 2010, 2.5 million people suffered a traumatic brain injury (TBI) at a cost of $76.5 billion. TBI is a leading cause of death and disability in the United States, accounting for 30% of all traumarelated deaths. Despite the high rate of mortality, it is estimated that 5.3 million individuals are living with disabilities resulting from TBI.1 Many of these patients suffer physical, emotional, cognitive, and behavioral disabilities that require treatment in postacute care facilities. The largest proportion of patients with TBI necessitating postacute care are discharged to skilled nursing facilities (SNFs).2,3 Little, however, is known about the TBI population admitted to SNFs and their outcomes. As such, postacute care for patients with TBI in SNFs remains unstandardized with no evidence-based guidelines or treatment protocols. Before such treatment protocols can be established, however, an understanding of the population and their trajectory is needed to guide the development of treatment strategies. The objective of this study was to describe the outcomes of patients with TBI who are admitted to SNFs for postacute care by key demographic and clinical characteristics. We hypothesized that within the population of TBI patients receiving postacute care, clinical and demographic characteristics, such as older age, male sex, and lower functional status at admission, confer excess risk for poor outcomes. Understanding this specific population and whether or not differences exist by readily available risk factors and assessed at admission to postacute care will assist in care planning, optimizing therapy (physical, occupational, psychological), and informing clinical expectations to families.
prior postacute SNF use within 1 year of the index hospitalization. TBI was identified if any one of the following International Classification of Diseases, Ninth Revision (ICD-9) codes was present on the hospitalization claim: 800–801.99, 803–804.99, 850–854.99, 310.2, 959.01, 907.0, 905.0, 873.0, 873.1.1 This study was approved by the Institutional Review Board at Brown University. Outcomes We examined mortality at 30 days, 90 days, and 1 year after admission to an SNF. For residents who were discharged from the SNF, we assessed the duration of SNF stay and functional improvement during the SNF stay. Functional improvement was measured using change in the 28point Activities of Daily Living (ADL) Scale score4 between admission and discharge assessments. The ADL score ranges from 0 (no impairment) to 28 (total dependence). Residents were considered to have improved in function if their discharge ADL score was less than their admission score. Because functional improvement may influence discharge, this data point was analyzed separately for patients with duration of stay less than 30 days. Demographics and admission characteristics
Data for this study come from several sources (Medicare Claims [Part A] and Enrollment Records, Minimum Data Set [MDS] resident assessments). Medicare Claims data were used to identify all hospitalizations and discharges to SNF among Medicare feefor-service beneficiaries. The Medicare enrollment records contain demographic information about each beneficiary ever entitled to Medicare including their date of death. The MDS is a federally mandated database that includes a comprehensive clinical assessment of individuals admitted to Medicare or Medicaid certified nursing facilities. The MDS contains items that measure physical, psychological, and psychosocial functioning and give a multidimensional view of the resident’s functional capacities. These data are required to be collected at admission, quarterly, annually, at discharge, and whenever there is a substantial change in a resident’s status.
The factors examined have been linked previously to outcomes of patients with TBI in other settings.5-7 Specifically, we examined differences in outcomes by age, sex, cognitive function, communication ability, and motor function. Age and sex were obtained from the Medicare enrollment records. Cognitive function, communication ability, and motor function were derived from the MDS admission assessment. Cognitive function was captured with the Cognitive Function Scale (CFS),8 derived from the Brief Interview for Mental Status and the Cognitive Performance Scale.9 Scores ranges from 1 (cognitively intact) to 4 (severely impaired). We dichotomized this variable to indicate whether or not patients were cognitively intact (CFS = 1) versus impaired (CFS <1). Ability to communicate was determined by staff rating of the level to which patients were able to make themselves understood (verbally and nonverbally), such as through expression of ideas and wants. Patients were considered to have severe impairment in communication if they were rarely or never understood. Motor function was ascertained from the transferring ADL score: a rating of the patient’s ability to move between surfaces (to or from) such as a bed, chair, wheelchair, or simply transition from sitting to standing. Patients were considered to have severe impairment in motor function if they were totally dependent on staff to transfer or transferring did not occur during the week of the assessment. Finally, we calculated the Deyo-Charlson Comorbidity Index from the diagnoses present on admission to the SNF as descriptors of patient acuity.10,11
Study population
Statistical analysis
The study was conducted on a cohort of hospitalized Medicare beneficiaries with an active diagnosis of TBI who were discharged directly to an SNF between January 1, 2011 and December 31, 2014. We excluded individuals who were <65 years of age and who had
Descriptive characteristics of the study population were summarized with standard measures, including means ± standard deviations, frequencies, and proportions. Using robust Poisson regression,12 we calculated relative risks (RRs) for mortality and
Methods Data
Please cite this article as: S.N. Lueckel et al., Outcomes in nursing home patients with traumatic brain injury, Surgery (2018), https://doi.org/10.1016/j.surg.2018.02.023
JID: YMSY
ARTICLE IN PRESS
[m5G;May 8, 2018;8:51]
S.N. Lueckel et al. / Surgery 000 (2018) 1–5 Table 1 Characteristics of TBI patients admitted to skilled nursing facilities after an inpatient hospitalization. Characteristic
N = 87292
Age, mean y (SD) Age ≥80 y, n (%) Female sex, n (%) Nonwhite race, n (%) ADL at baseline, mean score (SD) ADL ≥23, n (%) Deyo-Charlson Comorbidity Index, mean (SD) Deyo-Charlson Comorbidity Index Score ≥2, n (%) CFS Score, n (%) 1: Intact 2: Mild impairment 3: Moderate impairment 4: Severe impairment Terminal illness or on hospice, n (%)
83.9 (7.5) 64,541 (74) 55,418 (63) 7,455 (9) 18.1 (4.4) 8,674 (10) 1.2 (1.4) 28,106 (32) 37,714 (43) 22,043 (25) 20,282 (23) 5,144 (6) 753 (1)
Deyo-Charlson Comorbidity Index ranges from a score of 0 to 3 or more with greater scores indicating more comorbidities.11 The ADL Scale ranges from 0 to 28 with greater scores indicating more dependence. The CFS is a 4-point scale ranging from 1 (intact) to 4 (severely impaired). SD, standard deviation
functional improvement associated with older age, female sex, cognitive impairment, impairment in communication, and motor impairment for patients with TBI in SNFs. Linear regression was used to calculate the adjusted mean duration of stay associated with each factor present on admission (older age, female sex, cognitive impairment, impairment in communication, and impairment in motor function). To avoid potential type I errors from making multiple comparisons, the Bonferroni correction was applied, and point estimates with P < .008 were considered significant. Accordingly, results are presented with 99.2% confidence intervals (CIs). We also adjusted for factors that could influence a patient’s care (ie, the effect of the nursing home, whether or not the resident was receiving hospice services, and whether or not the resident had a documented prognosis of less than 6 months). Point estimates for each factor were obtained from a single adjusted model, which also included an indicator for terminal illness, hospice use, and SNF fixed effects. All statistical analysis was done in Stata (Version 15, Stata Corporation, College Station, TX) Results Between 2011 and 2014, 87,292 hospitalized Medicare beneficiaries aged 65 years and older without a nursing facility stay within 1 year of hospital admission were discharged to SNFs with an active TBI diagnosis (Table 1). The average age was 83.9 ± 7.5 years with 74% at or older than age 80. Females and nonwhites accounted for 63% and 9% of the population, respectively. The average Deyo-Charlson Comorbidity Index was 1.2 ± 1.4 with 32% having a score ≥2. The average baseline ADL score was 18.1 ± 4.4, and severe functional impairment (an ADL score of ≥23) was present in 10% of residents. Approximately 57% of residents had some level of cognitive impairment (CFS >1), and 29% had more severe cognitive deficits (CFS >2). Mortality Older age, cognitive impairment, impairments in communication, and severe motor deficits conferred greater risk for mortality among residents with TBI (Table 2). The 30-day mortality rate among residents aged <80 years was 5.9% compared with 9.1% in those aged 80 years and older (RR = 1.49, 99% CI = 1.36, 1.64); this trend was consistent at 1 year after discharge (RR = 1.42, 99% CI = 1.37, 1.48). Female sex was protective against mortality. The
30-day mortality rate for female residents was 6.6% compared with 11.1% among male residents (RR = 0.63, 9% CI = 0.59, 0.68). One year after admission, 26.2% of women had died compared with 38.4% of men (RR = 0.70, 9% CI = 0.68, 0.72). Among residents with cognitive impairment, 11.1% died within 30 days of admission to an SNF compared with 3.2% of those without impairment (RR = 2.55, 99% CI = 2.32, 2.80); this difference endured at 1 year after admission (RR = 1.74, 99% CI = 1.69, 1.79). We found that 26.4% of residents with severe motor impairment died within 30 days of admission to an SNF compared with 6.6% of those without impairment or with limited motor function (RR = 2.58, 99% CI = 2.32, 2.86); this difference persisted at 1 year (RR = 1.46, 99% CI = 1.40, 1.53). Among residents with severe impairment in communication, 34.1% died within 30 days of admission to an SNF compared with 7.0% of those without or with milder impairments (RR = 2.44, 9% CI = 2.16, 2.77); similar results were seen at 1 year (RR = 1.49, 9% CI = 1.41, 1.58). Duration of stay Among SNF residents discharged to the community, those who were older, female, and impaired in cognition, communication, or motor function had on average greater durations of stay (Table 3). The mean duration of SNF stay for patients <80 years of age was 27.5 days compared with 30.1 days for residents aged 80 and older (adjusted difference = 2.6 days, 99% CI = 1.6, 3.8). Male residents stayed in the SNF for 27.8 days on average, whereas female residents stayed for 30.2 days (adjusted difference = 2.4 days, 99% CI = 1.5, 3.4). Those with cognitive impairment had a mean duration of stay of 31.7 days compared with 27.5 days among those without cognitive impairment (adjusted difference = 4.2 days, 99% CI = 3.1, 5.2). Mean duration of stay for patients with severe impairment in communication was not significantly different. Mean duration of stay among patients with severe impairment in motor function was 47.9 days compared with 28.7 days among those without impairment or with limited motor function (adjusted difference = 19.2 days, 99% CI = 15.0, 23.4). Functional improvement at discharge Among residents discharged within 30 days of SNF admission, older age and female sex did not affect functional improvement, but the probability of functional improvement was significantly decreased for residents with cognitive impairment and severe impairment in the ability to communicate (Table 3). Among residents with cognitive impairment, 57.4% had functional improvement at discharge compared with 68.8% without impairment (RR = 0.86, 99% CI = 0.83, 0.88). Among residents with severe impairment in communication, 36.9% had functional improvement at discharge compared with 63.8% among residents without impairment or with milder deficits (RR =0 .67, 99% CI = 0.54, 0.82). Residents with severe motor impairment were actually more likely to improve their function (RR = 1.13, 9% CI = 1.02, 1.25). Point estimates were consistent in analyses on residents discharged after 30 days of SNF stay except that older age decreased the probability of functional improvement among these residents (RR = 0.96, 9% CI = 0.93, 0.99). Discussion In this study of more than 87,0 0 0 Medicare beneficiaries with TBI admitted to SNF, patients who were older or who had impairments in cognition, communication, or motor function at the time of admission had markedly greater rates of mortality and extended durations of stay. Impairments in cognition and
Please cite this article as: S.N. Lueckel et al., Outcomes in nursing home patients with traumatic brain injury, Surgery (2018), https://doi.org/10.1016/j.surg.2018.02.023
3
ARTICLE IN PRESS
JID: YMSY 4
[m5G;May 8, 2018;8:51]
S.N. Lueckel et al. / Surgery 000 (2018) 1–5 Table 2 Adjusted risk of mortality conferred by demographic and clinical characteristics among patients with TBI admitted to SNFs after an inpatient hospitalization. 30 Days
Age <80 y Age ≥80 y Male Female CFS <2 CFS ≥2 Active/limited communication Severely impaired communication Intact/limited motor function Severely impaired motor function
90 Days
1 Year
n/N (%)
RR (99.2% CI)
n/N (%)
RR (99.2% CI)
n/N (%)
RR (99.2% CI)
1,332/22,751 (5.9) 5,860/64,541 (9.1) 3,541/31,874 (11.1) 3,651/55,418 (6.6) 1,210/37,714 (3.2) 5,377/47,469 (11.3) 5,834/83,673 (7.0) 978/2,864 (34.1) 5,269/79,998 (6.6) 1,916/7,260 (26.4)
— 1.49 (1.36, 1.64) — 0.63 (0.59, 0.68) — 2.55 (2.32, 2.80) — 2.44 (2.16, 2.77) — 2.58 (2.32, 2.86)
2,638/22,751 (11.6) 11,066/64,541 (17.1) 6,578/31,874 (20.6) 7,126/55,418 (12.9) 2,939/37,714 (7.8) 9,997/47,469 (21.1) 11,936/83,673 (14.3) 1,340/2,864 (46.8) 10,952/79,998 (13.7) 2,741/7,260 (37.8)
— 1.44 (1.35, 1.53) — 0.65 (0.62, 0.69) — 2.16 (2.03, 2.29) — 1.83 (1.68, 2.00) — 1.90 (1.77, 2.04)
5,236/22,751 (23.0) 21,508/64,541 (33.3) 12,239/31,874 (38.4) 14,505/55,418 (26.2) 7,254/37,714 (19.2) 18,435/47,469 (38.8) 24,386/83,673 (29.1) 1,846/2,864 (64.5) 22,808/79,998 (28.5) 3,915/7,260 (53.9)
— 1.42 (1.37, 1.48) — 0.70 (0.68, 0.72) — 1.74 (1.68, 1.80) — 1.49 (1.41, 1.58) — 1.46 (1.40, 1.53)
The CFS is a 4-point scale ranging from 1 (intact) to 4 (severely impaired). Patients were considered to have severe impairment in communication if they were rarely or never understood. Patients were considered to have severe impairment in motor function if they were totally dependent on staff to transfer or transferring did not occur. Models were adjusted for SNF fixed effects, hospice services, and resident prognosis <6 months. Table 3 Adjusted length of stay and probability of functional improvement among patients with TBI admitted to skilled nursing facilities after an inpatient hospitalization, by demographic and clinical characteristics. Duration of stay
Age <80 y Age ≥80 y Male Female CFS <2 CFS ≥2 Active/limited communication Severely impaired communication Intact/limited motor function Severely Impaired Motor Function
Functional improvement: discharged within 30 days
Functional improvement: discharged after 30 days
Adjusted mean
Mean difference (99.2% CI)
n/N (%)
RR (99.2% CI)
n/N (%)
RR (99.2% CI)
27.5 30.1 27.8 30.2 27.5 31.7 29.3
— 2.6 (1.6, 3.8) — 2.4 (1.5, 3.4) — 4.2 (3.1, 5.2) —
6,397/10,024 (63.8) 15,369/24,229 (63.4) 7,363/11,859 (62.1) 14,403/22,394 (64.3) 13,609/19,791 (68.8) 8,089/14,083 (57.4) 21,634/33,838 (63.9)
— 1.01 (0.99, 1.04) — 1.02 (0.99, 1.04) — 0.86 (0.83, 0.88) —
3,376/3,992 (84.6) 9,988/12,498 (79.9) 4,227/5,270 (80.2) 9,137/11,220 (81.4) 6,782/8,018 (84.6) 6,537/8,405 (77.8) 13,186/16,223 (81.3)
— 0.96 (0.93, 0.99) — 1.00 (0.97, 1.03) — 0.92 (0.90, 0.95) —
32.7
3.4 (–2.9, 9.7)
123/333 (36.9)
0.67 (0.54, 0.82)
170/255 (66.7)
0.89 (0.76, 1.03)
28.7
—
21,353/33,492 (63.8)
—
12,347/15,375 (80.3)
—
47.9
19.2 (15.0, 23.4)
413/748 (55.2)
1.13 (1.02, 1.25)
1,017/1,115 (91.2)
1.16 (1.11, 1.21)
The CFS is a 4-point scale ranging from 1 (intact) to 4 (severely impaired). Patients were considered to have severe impairment in communication if they were rarely or never understood. Patients were considered to have severe impairment in motor function if they were totally dependent on staff to transfer or transferring did not occur. Models were adjusted for SNF fixed effects, hospice services, and resident prognosis <6 months.
communication were also associated with a decreased probability of functional recovery. Though there is no direct comparison, our findings are similar to literature that describes the outcomes of patients with injuries or postoperative patients who are admitted to an SNF for postacute care. A study in 2010 examined patients who suffered from a subarachnoid hemorrhage admitted to an SNF. Of these patients, 49% were dead after a median of 1.1 years after admission and 69% did not have functional improvement at 2 years.13 Furthermore, Oresanya et al14 reported on studied patients who underwent lower extremity revascularization who were admitted to an SNF. At 1 year after admission to an SNF, 51% of the patients had died, 285 were nonambulatory, and 32% had experienced functional decline. Contributing factors to poorer outcomes included age, poor functional status on admission, decrease in ability to maintain ADLs, and cognitive impairment.14 The lack of long-term functional improvement in patients with TBI highlights the complexity of these injuries, but slow functional recovery in patients with TBI has also been reported previously.15 Though overall recovery appears poor in a majority of these patients, recovery does not occur at a steady, continuous rate. Literature supports that recovery tends to occur most quickly earlier in their SNF admission.16,17 In our study the majority of residents were discharged to the community within 30 days of admission, and this would suggest that the majority of functional improvement occurred during the earlier time period. This observa-
tion is consistent with literature that has reported a steep recovery curve in the first 5–6 months after TBI. It is further suggested that this recovery is not maintained.16,17 Additional studies suggest that these same patients went on to develop decreases in cognitive function as they moved further and further out from their injury.16,18 This observation lends support to the importance of early, appropriate, and directed therapies for these patients and may be used to guide future therapies in both SNFs and rehabilitative facilities dedicated to TBIs, as well as generate realistic expectations for both physicians and families. The findings of our study are important not only patients but also for the patients’ families in the acute setting. Knowing the types of patients who are less likely to regain the ability to live independently and who have an increased risk of dying can help guide physicians and families making clinical and long-term decisions. Consequently, these findings may affect the care and planning of the facilities that accept these patients for postacute care, such as SNFs, because they will need to adjust accordingly. Future work is needed to determine whether other factors such as the severity of the TBI drive the association of certain patient characteristics and outcomes.19 Additionally, therapies and treatments received at the time of admission to an SNF may also play an early and important clinical role in the outcomes of these patients. The overall strength of this study is reflected in the large, inclusive, national database we accessed as well as the evaluation of specific clinical outcomes. Knowing and understanding the
Please cite this article as: S.N. Lueckel et al., Outcomes in nursing home patients with traumatic brain injury, Surgery (2018), https://doi.org/10.1016/j.surg.2018.02.023
JID: YMSY
ARTICLE IN PRESS
[m5G;May 8, 2018;8:51]
S.N. Lueckel et al. / Surgery 000 (2018) 1–5
outcomes of patients with TBI newly discharged to SNF after an acute hospitalization can help guide the acute care of these unique patients. Limitations of this study: This study depended on the accurate diagnosis of TBI. Although we did attempt to be comprehensive in our definitions of TBI with ICD-9 codes, there is an element of subjectivity in assigning some of these codes; however, our per-year estimates of the number of SNF residents with TBI were consistent with prior reports.20,21 An additional limitation to this work is that we were unable to assess the severity, mechanism, or timing of the injury. Future work incorporating these characteristics of the injury will be an important next step in understanding the outcomes of patients with TBI. This study did not assess the effect of therapies received in the SNF on outcomes. Other research suggests that directed rehabilitation does improve both functional and cognitive ability; however, those with more severe impairments on admission are less likely to improve beyond their baseline status.22,23 In our study we were unable to determine if the lack of improvement was due to the underlying injury or rather the specific or nonspecific therapies they received. Understanding which therapies were offered and which therapies were received by patients with TBI in SNFs is an important area for future research and may guide future care of these patients. Finally, we are unable to characterize and analyze SNF residents who were not Medicare feefor-service beneficiaries, and we were unable to assess functional change among residents who were not discharged to the community, including those who died during their nursing facility stay. Conclusion Among a population with of patients with a TBI admitted to an SNF, the probability of poor outcomes varies significantly by key clinical and demographic characteristics. Understanding these differences and possibilities can help set expectations for patients and families as well as providers when these patients are admitted to SNFs for rehabilitation after their acute episode. References 1. Centers for Disease Control and Prevention (CDC). Traumatic brain injury & concussion [Internet]. Atlanta (GA): CDC. TBI: get the facts; [updated 2017 Apr 27], accessed on April 30, 2016. Available from: http://www.cdc.gov/ traumaticbraininjury/get_the_facts.html. 2. Tepas 3rd JJ, Pracht EE, Orban BL, Flint LM. High-volume trauma centers have better outcomes treating traumatic brain injury. J Trauma Acute Care Surg. 2013;74:143–147 discussion 7-8.
3. Hakkarainen TW, Ayoung-Chee P, Alfonso R, Arbabi S, Flum DR. Structure, process, and outcomes in skilled nursing facilities: understanding what happens to surgical patients when they cannot go home. A systematic review. J Surg Res. 2015;193:772–780. 4. Wysocki A, Thomas KS, Mor V. Functional improvement among short-stay nursing home residents in the MDS 3.0. J Am Med Directors Assoc. 2015;16:470–474. 5. Mosenthal AC, Lavery RF, Addis M, Kaul S, Ross S, Marburger R, et al. Isolated traumatic brain injury: age is an independent predictor of mortality and early outcome. J Trauma. 2002;52:907–911. 6. Leitgeb J, Mauritz W, Brazinova A, Janciak I, Majdan M, Wilbacher I, et al. Effects of gender on outcomes after traumatic brain injury. J Trauma. 2011;71:1620–1626. 7. Kouloulas EJ, Papadeas AG, Michail X, Sakas DE, Boviatsis EJ. Prognostic value of time-related Glasgow coma scale components in severe traumatic brain injury: a prospective evaluation with respect to 1-year survival and functional outcome. Int J Rehabil Res. 2013;36:260–267. 8. Thomas KS, Dosa D, Wysocki A, Mor V. The Minimum Data Set 3.0 Cognitive Function Scale. Med Care. 2015;55:e68–e72. 9. Morris JN, Fries BE, Mehr DR, Hawes C, Phillips C, Mor V, et al. MDS Cognitive Performance Scale. J Gerontol. 1994;49:M174–M182. 10. Klabunde CN, Potosky AL, Legler JM, Warren JL. Development of a comorbidity index using physician claims data. J Clin Epidemiol. 20 0 0;53:1258–1267. 11. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992;45:613–619. 12. Zou G. A modified poisson regression approach to prospective studies with binary data. Am J Epidemiol. 2004;159:702–706. 13. Greebe P, Rinkel GJ, Algra A. Long-term outcome of patients discharged to a nursing home after aneurysmal subarachnoid hemorrhage. Arch Phys Med Rehabil. 2010;91:247–251. 14. Oresanya L, Zhao S, Gan S, Fries BE, Goodney PP, Covinsky KE, et al. Functional outcomes after lower extremity revascularization in nursing home residents: a national cohort study. JAMA Intern Med. 2015;175:951–957. 15. Leon-Carrion J, Dominguez-Morales MR, Barroso y Martin JM, Leon– Dominguez U. Recovery of cognitive function during comprehensive rehabilitation after severe traumatic brain injury. J Rehabil Med. 2012;44:505–511. 16. Christensen BK, Colella B, Inness E, Hebert D, Monette G, Bayley M, et al. Recovery of cognitive function after traumatic brain injury: a multilevel modeling analysis of Canadian outcomes. Arch Phys Med Rehabil. 2008;89(12 Suppl):S3–15. 17. Green RE, Colella B, Hebert DA, Bayley M, Kang HS, Till C, et al. Prediction of return to productivity after severe traumatic brain injury: investigations of optimal neuropsychological tests and timing of assessment. Arch Phys Med Rehabil. 2008;89(12 Suppl):S51–S60. 18. Till C, Colella B, Verwegen J, Green RE. Postrecovery cognitive decline in adults with traumatic brain injury. Arch Phys Med Rehabil. 2008;89(12 Suppl):S25–S34. 19. Fuller GW, Ransom J, Mandrekar J, Brown AW. Long-term survival following traumatic brain injury: a population-based parametric survival analysis. Neuroepidemiology. 2016;47:1–10. 20. Gabella BA, Mangione EJ, Hedegaard H, Kelly JP. Comparison of nursing home residents with and without traumatic brain injury: use of the Minimum Data Set. J Head trauma Rehabil. 2007;22:368–376. 21. Buchanan RJ, Wang S, Huang C. Profiles of nursing home residents with traumatic brain injury using the Minimum Data Set. Brain Injury. 2003;17:507–523. 22. Sandhaug M, Andelic N, Vatne A, Seiler S, Mygland A. Functional level during sub-acute rehabilitation after traumatic brain injury: course and predictors of outcome. Brain Injury. 2010;24:740–747. 23. Park HY, Maitra K, Martinez KM. The effect of occupation-based cognitive rehabilitation for traumatic brain injury: a meta-analysis of randomized controlled trials. Occupat Ther Int. 2015;22:104–116.
Please cite this article as: S.N. Lueckel et al., Outcomes in nursing home patients with traumatic brain injury, Surgery (2018), https://doi.org/10.1016/j.surg.2018.02.023
5