Abstract
Objective:
To describe patient, hospital, and geographic characteristics of older adult Medicare beneficiaries hospitalized with traumatic brain injury (TBI) and admitted to long-term acute care hospitals (LTACH).
Setting:
Acute hospital and LTACH facilities.
Participants:
15,148 Medicare beneficiaries 65 years and older with an acute TBI hospitalization who were discharged to an LTACH.
Design:
This retrospective cohort study used data from Centers for Medicare and Medicaid Services’ Medicare Enrollment and Provider Analysis and Review data files from 2011 to 2016.
Main Measures:
Patient variables (age, sex premorbid health burden, medical complications and procedure), hospital variables (for-profit status, bed size), and state/regional geographic variation associated with LTACH TBI admission.
Results:
Older adult Medicare beneficiaries admitted to LTACH facilities following TBI hospitalization averaged 77.1 years old and were predominantly white males. 94.6% of the sample had 2+ multimorbidities present during acute hospitalization. Average acute hospital length of stay of the sample was 19.4 days, and rates of acute mechanical ventilation of any duration and tracheostomy procedures were 56.6% and 40%, respectively. Only 4.1% of patients seen in LTACH were discharged home after LTACH stay; the primary discharge disposition was skilled nursing facilities (41.3%). Geographic analyses indicated selected Southern and Midwestern states had the greatest number of LTACH facilities and proportion of LTACH admissions.
Conclusions:
There has been limited characterization of the hospitalized TBI population admitted to LTACHs. Our findings among older adult Medicare beneficiaries suggest this population is highly medically-complex and are seldom discharged home after their LTACH stay. There are also notable geographic variations in LTACH TBI admissions across the United States. More research is warranted to understand long-term functional outcomes among this population.
Introduction:
Older adults have the highest rates of Emergency Department visits, hospitalizations, and deaths after traumatic brain injury (TBI).1 The rate of TBI among older adults has increased at rates faster than population growth,2 and recent data suggest rates of fall-related TBI deaths have increased 17% in the last decade.3 Among survivors of acute TBI, older age is associated with a lower likelihood of discharging directly home,2 greater disability,4 and reduced community participation post-injury.5
For patients who survive, but are unable to return home following the acute hospital stay, there are three inpatient post-acute care (PAC) settings where they may receive skilled, rehabilitative care: inpatient rehabilitation facilities (IRF), skilled nursing facilities (SNF), and long-term acute care hospitals (LTACH). The overall objective of all PAC settings is to address on-going medical needs, address functional impairments, and ideally, prepare patients to return home.6 However, there are differences in the care provided by each facility type. For example, IRFs must provide coordinated interdisciplinary care, with intensive rehabilitation services (15 hours/week/patient) and including follow up with a rehabilitation physician 3 times per week.7 SNFs must provide 24 hour/day nursing care, have a physician on-call 24 hours/day, and may provide rehabilitation services based on a patient’s plan of care. LTACHs treat patients with longer-term acute care needs due to medical complexity or illness severity, and may provide rehabilitation services.8 Older patients with TBI are most often admitted to SNFs if they are not expected to be able to tolerate therapy requirements, and also do not require hospital-level care.9 However, there is overlap in qualifying patients’ characteristics across PAC settings.10 Without a defined criteria guiding selection of the most appropriate inpatient PAC setting to maximize functional improvement, triage is often at the discretion of clinicians, case managers, and discharge planners. Overall, the population of patients in LTACHs do tend to have more complex medical presentation compared to patients in SNFs; however, non-patient factors, including geographic region and hospital characteristics (e.g., bed size, for-profit status) appear to play a significant role in referral patterns.10 In prior TBI research evaluating differences in PAC pathways, patients receiving care in LTACHs have been grouped together with patients admitted to SNFs.11 One study documented worse outcomes among patients with TBI admitted to either SNF or LTACH, compared to those in IRFs.11 Failure or inability to distinguish among the facilities where patients receive their long-term care precludes investigation into PAC-specific outcomes. As a result, the details of patient characteristics of those receiving care in LTACHs remain decidedly understudied.
Descriptive characterization of patients with TBI admitted to LTACHs will complement the existing literature on the more often-studied populations discharged to IRFs or SNFs to better understand PAC pathways after TBI. To that end, the objective of this study is to characterize patient-, hospital-, and regional-level factors associated with LTACH admissions among older Medicare beneficiaries hospitalized with TBI.
Methods:
Study Design and Cohort
We conducted a retrospective cohort study of Medicare beneficiaries admitted to LTACH after acute hospitalization for TBI using the Medicare Provider Analysis and Review (MedPAR) file linked to the Medicare Master Beneficiary Summary File (MBSF) available through Centers for Medicare and Medicaid (CMS). All Medicare beneficiaries were admitted to LTACH within 72 hours of TBI acute hospital discharge. The MBSF provides details of Medicare enrollment and demographic information, whereas MedPAR provides details of inpatient hospital encounters, for 100% of Medicare fee-for-service beneficiaries.
We used the linked MedPAR and MBSF files to identify Medicare beneficiaries age 65 years or older who were newly admitted to an LTACH after acute TBI hospitalization from 2011–2016. Consistent with a previous study,12 we used International Classification of Disease (ICD)-9th revision codes (800–801.99, 803–804.99, 850–854.99, 293.0, 310.2, 959.01, and 907.0) and ICD-10th revision codes (S01.0-S01.9, S02.0, S02.1, S02.3, S02.7-S02.9, S06.0-S06.9, S07.0, S07.1, S07.8, S07.9, and S09.7-S09.9) to capture TBI-related hospitalizations. To focus on incident LTACH admissions, we excluded individuals with an LTACH admission in the 2 years prior to index TBI hospitalization (n=1,573). We excluded Medicare Advantage beneficiaries, as their claims were unavailable (n=3,301).
Measures
We evaluated patient-level variables available in 2011–2016 Medicare files related to premorbid health status, index acute hospitalization, and LTACH stay. Study results were pooled across 2011–2016, and presented by year. Demographic variables included age, sex, race, and dual Medicare and Medicaid enrollment. Using the Chronic Conditions Warehouse (CCW) we evaluated the prevalence of seven complex chronic conditions (ischemic heart disease, congestive heart failure, atrial fibrillation, asthma, Alzheimer’s Disease/Related Dementias, chronic obstructive pulmonary disease, and chronic kidney disease) present at the time of TBI hospitalization. These selected conditions have been previously found to increase healthcare utilization and outcomes among Medicare beneficiaries (i.e., mortality, hospitalizations, costs).13 We also calculated frequency of multimorbidity (2+, 4+, or 6+ conditions) at the time of TBI hospitalization. For index acute hospitalization, we calculated length of stay (LOS), intensive care unit (ICU) use, as well as hospital procedures and complications. For LTACH stay, we documented the setting immediately preceding LTACH admission, LOS, complications, and discharge disposition after LTACH stay.
We also evaluated hospital, state, and regional variables that are known to be associated with LTACH admissions.10 Hospital variables included LTACH ownership (non-profit, for-profit, government) and bed count. We evaluated LTACH admission rates by state and geographic region, and also by state certificate of need (CON) law, which are state regulatory mechanisms requiring approval for establishment or expansion of new health care facilities. We gathered CON law information from the National Conference of State Legislatures,14 and all states (with the exception of New Hampshire) did not have any changes to CON law legislation during the study period. For New Hampshire, which revoked their CON law in 2016, we considered New Hampshire Medicare beneficiaries hospitalized with TBI from 2011–2015 to have CON laws present, and those hospitalized in 2016 to not have CON laws present at the time of LTACH stay. We obtained information on the number of acute hospitals that discharge to LTACH and number of LTACH facilities in each state from the Provider of Services (POS) file.
Statistical analysis
We described characteristics of patients with TBI admitted to LTACH using means and standard deviations or median and range, as appropriate, for continuous variables, and frequencies and percentages for categorical variables. We created a geospatial map of LTACH admissions by state using ArcGIS Version 10.4 (Environmental Systems Research Institute, Redlands, CA). To calculate proportion of LTACH admissions by state, we restricted to acute hospitals in a state that discharge to LTACHs. We analyzed all other data using SAS version 9.4 (SAS Institute, Cary, NC).
Results:
Description of the sample
We identified 15,148 Medicare beneficiaries with an acute TBI hospitalization who were discharged to an LTACH from 2011–2016. We provided the characteristics of the sample in Table 1. The average age was 77.2, and Medicare beneficiaries with TBI aged 65–74, 75–84, and 85+ made up 40.9%, 38.3%, and 20.8% of the sample, respectively. A majority of Medicare beneficiaries with TBI discharged to LTACH were white and male, and 21.1% were dually-enrolled in Medicare and Medicaid.
Table 1:
Individual, Regional, and Hospital Characteristics of Patients with TBI newly admitted to LTACH (2011–2016)
| Total (N=15148) | |
|---|---|
| Demographic characteristics, n (%) | |
| 85+ | 3151 (20.8) |
| Age, Mean (SD) | 77.17 (8.05) |
| Female | 6309 (41.65) |
| Other | 268 (1.77) |
| Dual eligible Medicare/Medicaid at the time of admission | 3188 (21.05) |
| Prior Chronic Disease Diagnoses, n (%) | |
| ADRD | 4427 (29.22) |
| Asthma | 2901 (19.15) |
| Atrial Fibrillation | 4555 (30.07) |
| Heart Failure | 7757 (51.21) |
| Chronic Kidney Disease | 7128 (47.06) |
| Chronic Obstructive Pulmonary Disease | 6641 (43.84) |
| Ischemic Heart Disease | 10245 (67.63) |
| 6+ conditions | 12258 (80.92) |
| Hospital and regional level characteristics | |
| West | 2207 (14.57) |
| No | 4,527 (29.9) |
| Government | 367 (2.42) |
| >400 | 92 (0.61) |
Note: column percentages reported;
multimorbidity includes number of conditions present out of 22 reported in the Chronic Condition Warehouse: Alzheimer’s disease and related disorders or senile dementia, acute myocardial infarction, anemia, rheumatoid arthritis / osteoarthritis, asthma, atrial fibrillation, benign prostatic hyperplasia, cancer (breast cancer, colorectal cancer, endometrial cancer, lung cancer, and prostate cancer), cataract, heart failure, chronic kidney disease, chronic obstructive pulmonary disease, depression, diabetes, glaucoma, hip/pelvic fracture, hypertension, acquired hypothyroidism, hyperlipidemia, ischemic heart disease, osteoporosis, and stroke / transient ischemic attack.
Prior Comorbid Disease Diagnoses
The rates of multimorbidity of the sample were high; approximately 95% had 2+, 89.3% had 4+, and 80.1% had 6+ prior chronic disease diagnoses. The rates of complex chronic diseases were also high; 67.6% of the sample had Ischemic Heart Disease, 51.2% had Heart Failure, and 47.1% had Chronic Kidney Disease.
Clinical characteristics of Index Acute Hospitalization stay
We provided clinical characteristics during the index hospitalization of the sample in Table 2. The average length of acute hospital stay of the sample was 19.4 days, and 81% required ICU care. The rates of mechanical ventilation (any duration) and tracheostomy were 56.6% and 40.0%, respectively. Acute respiratory failure (16.6%) was the most common acute medical complication. Acute hospital characteristics of the sample were mostly stable over the study period (Supplemental Table S1).
Table 2:
Clinical Characteristics during Index Acute Hospitalization among Individuals with TBI Admitted to LTACH (2011–2016)
| Total (N=15148) | |
|---|---|
| Acute Hospitalization stay | |
| Length of stay (days), Median (range) | 17 (1–259) |
| ICU use, n (%) | 12249 (80.9) |
| Craniotomy or craniectomy | 338 (2.2) |
| Septic shock | 1646 (10.9) |
| LTACH stay | |
| 90+ days | 229 (1.5) |
| Septic shock | 451 (3.0) |
| Death | 2465 (16.3) |
Clinical characteristics of LTACH stay
We provided the clinical characteristics of the sample during LTACH stay in Table 2. Nearly all beneficiaries (99.5%) were transferred directly from the acute hospital to LTACH; however, a small percentage (0.5%) were in alternate settings (e.g., SNF, IPR, home) for less than a 72 hour window after TBI acute hospitalization prior to LTACH admission. The majority of the sample had a LOS during LTACH between 1–30 days (64.9%) and approximately 30% between 31–59 days. The most common discharge disposition after LTACH stay was SNF/nursing home (41.1%), while home (4.8%) and hospice (3.5%) disposition were least common. Additionally, 10.5% of Medicare beneficiaries were discharged back to acute care and 11.5% were discharged to inpatient rehabilitation after their LTACH stay. The mortality rate of the sample after admission to LTACH was 16.3%. Among the overall sample, acute respiratory failure (13.3%) and sepsis (12.6%) were the most common medical complications. The sample characteristics during LTACH stay were mostly stable over the study period (Supplemental Table S2).
Hospital and Regional Characteristics of TBI LTACH Admissions
The majority (63.8%) of LTACH ownership was for-profit, and total bed counts were predominantly <200 beds. States with CON laws had significantly lower proportion of TBI LTACH admissions (1.5%) than states without CON laws (2.1%) (p<0.0001). There were state and regional differences in propensity of TBI LTACH admissions among acute hospitals that discharge to LTACHs (see Figure 1). In Supplemental Table S3, we provided detailed state-level information of acute hospitals that discharge to LTACH. Louisiana and Texas had the greatest absolute number of LTACH facilities, and relative proportion (Louisiana: 5.7%; Texas: 3.7%) of TBI LTACH admissions. Other states (e.g., Massachusetts, Mississippi, Oklahoma, Idaho) had fewer absolute number of LTACH facilities, but relatively higher proportions of TBI LTACH admissions.
Figure 1. Percent of TBI-related acute hospitalizations with discharge to LTACH, by state.
Among TBI admissions to acute hospitals that discharge to LTACH facilities, we calculated the percent TBI LTACH admissions, by state. The state-level data are based upon location of acute hospital.
Discussion:
The current study evaluated Medicare beneficiaries over 65 years of age hospitalized with TBI and transferred within 72 hours to an LTACH from 2011–2016. The population of older adults with TBI discharged to LTACHs for their PAC are an understudied subgroup. Our findings indicated this sample was highly medically complex, had a high burden of tracheostomy and ventilator support during acute care, and were most commonly discharged to SNFs after their LTACH stay. Consistent with other research,10 we also documented variability in number of LTACH facilities and usage by state and geographic region.
Medicare is the payer for approximately 70% of LTACH stays (irrespective of diagnosis).8 Prior studies have indicated a rise in the number of LTACH admissions over time following hospitalization for a critical illness.15 Despite this, LTACHs historically have only accounted for approximately 1% of all PAC use.8 More medically-complex Medicare beneficiaries are disproportionately admitted to LTACHs.10 According to a Medicare Payment Advisory Commission report,8 about 60% of beneficiaries requiring tracheostomy with more than 96 hours of ventilator support during acute hospitalization were admitted to LTACHs, as are 15% of beneficiaries who experienced septicemia or respiratory failure requiring mechanical ventilation over 96 hours during acute care. Our data indicated Medicare beneficiaries with TBI admitted to LTACHs had high rates of complex medical conditions, ICU use, mechanical ventilation, and tracheostomy procedures during their acute care. We also documented that Medicare beneficiaries admitted to LTACH after TBI hospitalization had nearly a 10 times greater proportion of discharges to SNFs compared to home after their LTACH stay, further suggesting these patients required a continued high level of care. Comparatively, 67% of Medicare beneficiaries over 65 admitted to IRF for TBI are discharged home after inpatient rehabilitation.16
Our findings parallel prior work10 documenting geographic differences in LTACH admissions. State-level contributors to variability in LTACH admissions include lower supply of LTACH facilities, and also presence of CON laws, which add regulatory steps for the establishment of new health care facilities based on documented unmet needs. We found states with CON laws had significantly lower proportion TBI LTACH admissions compared to states without these laws. We also found variability in the rates of LTACH admissions by state; Louisiana (5.7%) and Texas (3.7%) had the highest proportion of LTACH admissions compared to all other states.
There are limitations to the current study. It is possible we had ascertainment bias due to differences in administrative codes before and after the ICD-10 transition period in October 2015. As a part of the MedPAR files, we did not have access to information on functional status of patients; therefore, were not able to track functional improvements during LTACH stay, or long-term outcomes after LTACH discharge. We did not exclude Medicare beneficiaries with extra-cerebral injuries (e.g., significant injuries outside of the brain). Future studies would benefit from studying the relationship between polytrauma plus TBI on post-acute care discharge destination. In the present study, we focused on incident LTACH admissions after TBI hospitalization and excluded persons with prior LTACH stays. Therefore, our estimates of total LTACH admissions after TBI hospitalization may be an underestimate of the overall number of Medicare beneficiaries admitted to LTACH after TBI hospitalization. This study was conducted among Medicare beneficiaries 65 years or older; observed characteristics associated with LTACH admission may not be generalizable to non-Medicare beneficiaries, beneficiaries with Medicare Advantage insurance coverage, or Medicare enrollees with TBI younger than 65.
To our knowledge, this is the first study to systematically characterize hospitalized Medicare beneficiaries with TBI admitted to LTACHs. The overall proportion of LTACH admissions after acute hospitalization for TBI was low, but we found this subgroup was medically complex, with high rates of multimorbidities and acute care complications. These are independent risk factors for poor long-term outcomes after TBI, suggesting this subgroup is at particularly high risk for poor outcomes. Future research is needed to directly compare characteristics associated with admission, and long-term outcomes, between patients admitted to different PAC settings after TBI (e.g. IRF vs. SNF vs. LTACH).
Supplementary Material
Acknowledgements:
Sources of funding:
This research was supported by funding from the National Institute on Aging (R21 AG059120), the Agency for Healthcare Research and Quality (5T32HS000011–33, EE), the National Institute on Disability Independent Living and Rehabilitation Research (90DPTB0009), and a VA HSR&D Career Development Award (CDA-14–422), and the Foundation for Physical Therapy Research funded Center on Health Services Training and Research (CoHSTAR)
Footnotes
Conflicts of interest:
The authors have no conflicts of interest to report.
References:
- 1.Taylor CA, Bell JM, Breiding MJ, Xu L. Traumatic brain injury–related emergency department visits, hospitalizations, and deaths—United States, 2007 and 2013. MMWR Surveillance Summaries. 2017;66(9):1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Dams-O’Connor K, Cuthbert JP, Whyte J, Corrigan JD, Faul M, Harrison-Felix C. Traumatic brain injury among older adults at level I and II trauma centers. Journal of neurotrauma. 2013;30(24):2001–2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Peterson AB. Deaths from Fall-Related Traumatic Brain Injury—United States, 2008–2017. MMWR Morbidity and Mortality Weekly Report. 2020;69. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Mushkudiani NA, Engel DC, Steyerberg EW, et al. Prognostic value of demographic characteristics in traumatic brain injury: results from the IMPACT study. Journal of Neurotrauma. 2007;24(2):259–269. [DOI] [PubMed] [Google Scholar]
- 5.Erler KS, Whiteneck GG, Juengst SB, et al. Predicting the trajectory of participation after traumatic brain injury: a longitudinal analysis. The Journal of head trauma rehabilitation. 2018;33(4):257–265. [DOI] [PubMed] [Google Scholar]
- 6.Buntin MB. Access to postacute rehabilitation. Archives of physical medicine and rehabilitation. 2007;88(11):1488–1493. [DOI] [PubMed] [Google Scholar]
- 7.Commission MPA. Report to the Congress: Medicare Payment Policy. Chapter 10: Inpatient rehabilitation facility services. In: MedPAC, Washington DC; 2019. [Google Scholar]
- 8.Commission MPA. Report to the Congress: Medicare Payment Policy. Chapter 11: Long term care hospital services. 2019:281–300. [Google Scholar]
- 9.Commission MPA. Report to the Congress: Medicare Payment Policy. Chapter 8: Skilled Nursing Facility Services. In: MedPAC, Washington DC; 2019. [Google Scholar]
- 10.Makam AN, Nguyen OK, Xuan L, Miller ME, Goodwin JS, Halm EA. Factors associated with variation in long-term acute care hospital vs skilled nursing facility use among hospitalized older adults. JAMA internal medicine. 2018;178(3):399–405. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Mellick D, Gerhart KA, Whiteneck GG. Understanding outcomes based on the post-acute hospitalization pathways followed by persons with traumatic brain injury. Brain Injury. 2003;17(1):55–71. [DOI] [PubMed] [Google Scholar]
- 12.Lueckel SN, Teno JM, Stephen AH, et al. The Population of Patients with Traumatic Brain Injury (TBI) in Skilled Nursing Facilities: A Decade of Change. The Journal of head trauma rehabilitation. 2019;34(1):E39. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Keeney T, Belanger E, Jones RN, Joyce NR, Meyers DJ, Mor V. High‐Need Phenotypes in Medicare Beneficiaries: Drivers of Variation in Utilization and Outcomes. Journal of the American Geriatrics Society. 2020;68(1):70–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Legislatures NCoS. CON-Certificate of Need State Laws. http://www.ncsl.org/research/health/con-certificate-of-need-state-laws.aspx. Published 2019. Accessed 12/19/2019, 2019.
- 15.Kahn JM, Benson NM, Appleby D, Carson SS, Iwashyna TJ. Long-term acute care hospital utilization after critical illness. Jama. 2010;303(22):2253–2259. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Krebill C, Evans E, Devone F, Zhang W, Gutman R, Thomas K. Factors Associated with Successful Community Discharge from Inpatient Rehabilitation among Older Adults with Traumatic Brain Injury. Archives of Physical Medicine and Rehabilitation. 2020;101(11):e105. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.