The Trauma Dyad: the Role of Informal Caregivers for Older Adults following Traumatic Injury

Abstract

Objective:

To investigate the association between higher injury severity and increased informal caregiving received by injured older adults.

Summary of Background Data:

Injured older adults experience high rates of functional decline and disability after hospitalization. Little is known about the scope of caregiving received post-discharge, particularly from informal caregivers such as family.

Methods:

We used the National Health and Aging Trends Study (NHATS) 2011–2018 linked to Medicare claims to identify adults ≥65 with hospital admission for traumatic injury and an NHATS interview within 12 months pre- and post-trauma. Injury severity was assessed using the injury severity score (ISS, low 0–9; moderate 10–15; severe 16–75). Patients reported the types and hours of formal and informal help received and any unmet care needs. Multivariable logistic regression models examined the association between ISS and increase in informal caregiving hours after discharge.

Results:

We identified 430 trauma patients. Most were female (67.7%), non-Hispanic White (83.4%) and half were frail. The most common mechanism of injury was fall (80.8%) and median injury severity was low (ISS=9). Those reporting receiving help with any activity increased post-trauma (49.0% to 72.4%, p<0.01), and unmet needs nearly doubled (22.8% to 43.0%, p<0.01). Patients had a median of 2 caregivers and most (75.6%) were informal, often family members. Median weekly hours of care received pre vs post-injury increased from 8 to 14 (p<0.01). ISS did not independently predict increase in caregiving hours; pre-trauma frailty predicted an increase in hours ≥8/week.

Conclusions:

Injured older adults reported high baseline care needs which increased significantly after hospital discharge and were mostly met by informal caregivers. Injury was associated with increased need for assistance and unmet needs regardless of injury severity. These results can help set expectations for caregivers and facilitate post-acute care transitions.

Mini-Abstract:

Injured older adults experience high rates of functional decline and disability after hospitalization. We conducted a retrospective cohort study to identify older adult trauma patients and understand their care needs after discharge. We found that regardless of injury severity, care needs increased significantly after discharge; these were primarily met by informal caregivers (i.e. family), although many had unmet needs.

INTRODUCTION

Rates of traumatic injury in older adults (≥65 years) are rising rapidly: by 2030, older adults will comprise 19% of the U.S. population but incur almost 40% of trauma admissions.¹ Among the injured, 40% are frail and the most common mechanism of injury is fall.^2,3 Compared to younger counterparts, injured older adults experience higher rates of disability and death exacerbated by pre-existing comorbid conditions and frailty.^2,4 One-year morbidity is considerable and even robust patients experience prolonged recovery and functional decline. Among more frail patients, their “life space” shrinks post-trauma, restricting autonomy and exacerbating dependence on caregivers to perform daily activities.^5,6

Despite significant increases in the numbers of older trauma patients and expected complex recovery, post-acute care needs are incompletely understood. Post-acute care is often measured in terms of healthcare resource utilization based on billing data, including use of rehabilitation services, paid home healthcare, and readmission.⁷ However, this fails to capture personal and familial sources of support, particularly unpaid or “informal” care provided by friends and family.⁸ Over 50 million adults in the U.S. identify as caregivers for recipients >50 years old, providing unpaid assistance with daily activities, financial and social support, and making healthcare decisions, spending on average 24 hours/week providing care.^8–10 Increased magnitude of disease and disability among patients is associated with increased care needs from informal caregivers, which is subsequently associated with higher rates of caregiver burden.^11,12

The majority of patients who fall have pre-existing disability,¹³ and the additive effect of loss of independence, high healthcare usage and persistent disability make understanding the long-term needs of this population and the role of informal caregivers of critical public health importance. We sought to develop a more complete picture of how trauma impacts caregiving needs for older adults by characterizing the formal and informal care received by older adults before and after trauma. Further, we assessed for factors associated with increased informal caregiving intensity post-trauma. We hypothesized that higher injury severity would be independently associated with increased informal care needs and caregiving hours received after trauma.

METHODS

This study used data from the National Health and Aging Trends Study (NHATS) (2011–2018) linked to Medicare claims. NHATS is a publicly available national population-based survey which examines disability and health trajectories of adults ≥65 years who reside in the United States,¹⁴ comprised of Medicare enrollees and over-samples non-Hispanic Blacks and adults 90 years or older. Four categories are utilized for race/ethnicity: White non-Hispanic, Black non-Hispanic, Other non-Hispanic, and Hispanic. The survey includes details of physical and cognitive ability, emotional well-being and daily activities. Subjects report caregivers who helped them with mobility, personal care, household activities, medical activities, and transportation and hours of care provided.

Study Population

The study population included subjects 65 years and older who had a hospital admission associated with an injury diagnosis code (see Appendix 1) and who had participated in NHATS interviews within 12 months before and after the injury event.^15,16 Subjects with multiple qualifying traumas were enrolled at the first observed event.

Variables

Subject demographics, comorbidities, injury characteristics, hospital stay characteristics, and post-discharge healthcare utilization were derived from Medicare claims and NHATS interviews. Patient demographics include education (≥ high school), living arrangement (alone, with spouse/partner, with others, residential care facility, nursing home), marital status (married or with partner), and income (annual $, quartile). Income is reported in NHATS every other year and was obtained for the last available measurement, within 1–2 interviews prior to the index trauma, and inflation-adjusted to 2017 dollars using the imputed income measures provided by NHATs. Health characteristics include probable dementia (self-report of dementia using Alzheimer’s Disease screening tool (AD8), report of clinical diagnosis, or cognitive testing),¹⁷ frailty (composite score of unintentional weight loss, low energy, low physical activity, walk and grip score),¹⁸ depression (score≥3 on Patient Health Questionnaire (PHQ)-2), and social isolation (four or more of: not married or partnered, no visits from family, not attending religious services, no clubs, not talking to family, no friends to talk to).¹⁹ The injury severity score (ISS) is an anatomic measure of injury severity (0–75, higher worse) typically categorized as low 0–9, moderate 10–15, severe 16–75, and ascertained from claims data using validated software.^16,20 Hospital admission characteristics included admission to a trauma center,²¹ length of stay (overall, intensive care unit [ICU]), or any orthopedic procedure during stay (based on ICD9/ICD10 procedure codes, Appendix 1). Healthcare utilization within three months of discharge included any hospital readmission, admission to rehabilitation facility, home healthcare use, or emergency department (ED) visit.

Types of Care, Caregivers and Hours of Care

The amount and types of care patients received was captured during NHATS interviews. Patient report of participation in activities of daily living was characterized as ability to carry out tasks in three categories: mobility (getting outside the home, getting around inside or outside the home, getting out of bed), self-care (eating, showering, toileting, dressing), and household activities (laundry, shopping for groceries or personal items, meal preparation, keeping track of medications). Patients were queried on whether they needed assistance to perform any tasks in the past month, or if they experienced an adverse consequence due to unmet needs (going without because no one was there to assist and/or it was too difficult to do alone).²² Patients identified caregivers, or those who have carried out a household or medical care-related activity with or for the interviewee, and designated whether the caregivers were paid or unpaid (i.e., formal versus informal caregivers), relative or non-relative, and the hours per week of help received. Hours of care received per week were capped at 168 (or 24 hours/day, 7 days/week).

Outcomes

The main outcome was an increase in post-discharge informal caregiving hours per week received following trauma (any increase in hours, ≥8 hours (or one workday)). Outcome was dichotomized as more caregiving hours required after trauma or less/no change in caregiving hours after trauma, with the intent of capturing any increase in caregiving intensity. To assess for change of ≥1 workday per week, the outcome was dichotomized as ≥8 hours/week or less than 8 hours per week/no change.

Statistical Approach

Independent variables and outcomes are described using frequencies, percentages, means/medians, with standard deviation (SD) and interquartile range (IQR) as appropriate. Associations were tested using bivariate linear and logistic regression, with standard errors clustered at the respondent-level to account for repeated measures. Multivariable regression analysis using a logistic model with clustered standard errors was used to test for association between demographic (age, marital status, insurance type, and income pre-trauma; sex, race, education) and clinical characteristics (dementia and frailty pre-trauma, ISS, hospital type) and increase in caregiving hours reported post-trauma. Multiple imputation was used to account for incomplete information. Finally, sensitivity analysis examined the cohort with and without patients living in long-term care facilities to evaluate for any differences in informal care. All statistics used Stata Version 16. P ≤ 0.05 was considered significant.

This study was approved by the Icahn School of Medicine at Mount Sinai Institutional Review Board.

RESULTS:

We identified 430 study participants, who had been admitted to a hospital following traumatic injury and completed pre- and post-trauma interviews. Before trauma, 86.7% of subjects lived at home, and 13.3% lived in a residential care facility. After trauma, 76.5% subjects lived at home and 23.5% lived in a residential care facility or nursing home (p<0.05). Post-trauma, patients were more likely to have dementia, (20.2% vs. 31.6% p<0.01), be frail (46.7% vs. 59.1%, p<0.01), have depression (16.3% vs. 27.4%, p<0.01), and report social isolation (32.8% vs 45.4%, p<0.01)(Table 1).

Table 1.

Socio-Demographics and Clinical Characteristics of Cohort (N=430)

	Pre-Trauma	Post-Trauma	P-value
Age in years, mean (SD)	83.2 (7.7)	84.2 (7.7)	<0.01
Sex, female	67.7%	67.7%
Race, categorical			1.0
White/Other, Non-Hispanic	83.4%	83.4%
Black, Non-Hispanic	13.1%	13.1%
Hispanic	3.5%	3.5%
Married	32.3%	30.5%	<0.01
Education: ≥high school	74.8%	74.8%
Insurance
Medigap	67.3%	69.7%	0.35
Medicaid	18.3%	20.3%	0.23
Tricare (Veteran’s Insurance)	6.9%	6.5%	0.60
Other	7.5%	3.5%
Living Arrangement			<0.01
Lives alone	33.5%	27.2%
Lives with spouse/partner	32.3%	27.2%
Lives with others (not spouse)	20.9%	22.1%
Residential Care Facility or Nursing Home Resident	13.3%	23.5%
Income in dollars($)*, mean	$44,453.13
Quartile of Income*
Bottom	34.3%
Second	29.3%
Third	22.0%
Top	14.5%
Probable dementia	20.2%	31.6%	<0.01
Depression, PHQ-2≥3	16.3%	27.4%	<0.01
Social Isolation	32.8%	45.4%	<0.01
Frail	46.7%	59.1%	<0.01

^*Income is reported in odd-numbered years only, and is taken from the interview immediately or two-preceding trauma and inflation-adjusted to 2017 $.

One-half of participants were admitted to a trauma hospital, and the most common mechanism of injury was fall (80.8%). Most subjects (78.1%) had low ISS. The median length of stay was 5 days. One-third were readmitted within three months of discharge. More than half were admitted to a skilled nursing facility or received home healthcare services after discharge (Table 2).

Table 2.

Characteristics of Trauma and Healthcare Utilization

Trauma center	51.4%
Injury mechanism
Fall	80.8%
Injury Severity Score (ISS)
ISS Low (0–9)	78.1%
ISS Moderate/High (>10)	21.9%
Length of stay (days), median (IQR)	5.0 (3.0)
Length of intensive care unit (ICU) stay (days), median (IQR)	0.0 (1.0)
Any orthopedic procedure*	26.1%
Healthcare Utilization within 3 months of discharge
Any readmission	33.6%
Skilled nursing facility (SNF) admission	56.1%
Home healthcare use	57.9%
Emergency department (ED) visit	29.5%

IQR: inter-quartile range;

^*list of procedures found in Appendix 1

At the median, participants had two caregivers both before and after trauma. Most participants reported informal or family care but no paid help (76%), and the most common caregiver was a daughter. Post-trauma, participants received more hours of assistance per week (median, pre: 14.8 vs. 8.0, <0.01), the majority of which was unpaid help (Table 3). One-third of patients reported an increase in care received of ≥8 hours per week, and 11% reported an increase in ≥40 hours per week. Following trauma, participants were more likely to report need of assistance with any activity of daily living with increases consistently seen in all activity type (any activity (49.3% vs. 72.4%), self-care (29.8% vs. 53.8%), mobility (29.3% vs. 53.0%), or household (42.1% vs. 62.1%), all p<0.01, Figure 1a). Participants were also more likely to have experienced an adverse consequence due to an unmet need (any activity (22.8% vs. 43.0%), self-care (13.7% vs. 22.3%), mobility (16.1% vs. 28.4%), all p<0.01; or household (4.9% vs. 7.0%, p=0.18), Figure 1b).

Table 3.

Care Needs & Care Received Before & After Trauma as reported by Patient

	Pre-Trauma	Post-Trauma	P-Value
Number of helpers, median	2.0	2.0	<0.01
Types of Informal Caregivers
Has spouse helper	31.1%	28.1%	<0.01
Has daughter helper	46.3%	55.8%	0.03
Has son helper	28.6%	34.5%	0.17
Has other family helper	21.9%	24.7%	0.79
Has other informal helper	22.1%	25.9%	0.57
Hours/week help received (median)	8.0	14.8	<0.01
Types of Caregiving			<0.01
No helpers	9.5%	4.9%
Informal help only	76.3%	75.6%
Has paid help	14.2%	19.5%

Figure 1.

Comparison pre- and post-trauma care needs as reported by patient: 1a) assistance with self-care, mobility, household activities, 1b) adverse consequence (such as going without) due to unmet need in : self care, mobility, household activities.

* denotes p<0.05

Multivariable logistic regression did not identify any demographic or health variables consistently associated with any increase in caregiving hours needed after trauma, including ISS (Table 4). Although female sex was independently associated with an increase in any caregiving hours, female sex was not associated with caregiving hours when evaluated at ≥8 hours/week, or in a sensitivity analysis excluding long-term care residents. There were no other statistically significant changes in the model when long term care residents were excluded. When the threshold for increased care was ≥8 hours/week, frailty was a predictor of increased needs.

Table 4.

Logistic Regression of the Association between Demographic and Clinical Characteristics of Cohort with an Increase in Informal Caregiving Hours Received Following Trauma, stratified by any increase (compared to no change/decrease hours) and increase of >8 hours/week (or one work day, versus no change/less than 8 hours/week)

	Any increase in caregiving hours / week	Increase of ≥8 hours / week
ISS≥9	0.84 (0.48–1.48)	1.01 (0.55–1.85)
Age	1.01 (0.98–1.04)	1.02 (0.99–1.05)
Sex, female	1.62 (1.12 – 2.56)*	0.91 (0.55–1.52)
Race
White	[ref]	[ref]
Black	0.95 (0.49–1.82)	1.39 (0.70–2.76)
Hispanic	0.58 (0.18–1.86)	1.187(0.34–4.06)
Top quartile of income pre-trauma	1.02 (0.56–1.86)	1.26 (0.67–2.38)
Married	1.22 (0.75–1.97)	1.52 (0.91–2.54)
Education: ≥high school	0.92 (0.56–1.55)	0.85 (0.49–1.46)
Medicaid	0.82 (0.45–1.51)	1.17 (0.61–2.21)
Frail	1.00 (0.65–1.56)	1.84 (1.15–2.97) *
Trauma hospital	1.38 (0.92–2.07)	1.25 (0.81–1.93)
Probable dementia	0.77 (0.45–1.30)	0.96 (0.55–1.67)

^*in a sensitivity analysis excluding long-term care residents, female sex was no a longer statistically significant predictor (OR: 1.59 (0.96–2.62))

DISCUSSION

We found that older adults who fall and are hospitalized already have high care needs pre-trauma, including assistance with routine activities and one-quarter reporting unmet needs, indicating pre-existing disability within this cohort. Following trauma, older adults are more likely to be frail and less likely to live at home, and rates of depression and social isolation increase. More than 20% required new assistance, and 40% reported unmet needs, despite receiving increased caregiving hours per week compared to pre-trauma. No individual patient characteristics predicted any increase in caregiving hours after trauma, although frailty was associated with an increase in informal care of ≥8 hours/week. These findings demonstrate the prevalence of needing assistance among older adults at risk for trauma and suggest that high caregiving needs are common among older trauma patients, regardless of injury severity.

Our findings are consistent with previous research of ground-level falls among the elderly, showing high rates of post-discharge healthcare utilization and change in mental health post-fall.^{23 24,25} Post trauma, we found that over half of our cohort utilized rehabilitation or home healthcare services, one-third were evaluated in the ED or readmitted, and rates of depression and social isolation increased. Additionally, traumatic falls are recognized as an indicator of declining function: patients with a poor pre-trauma performance status have worse long-term functional outcomes, and higher rates of severe pain and depression, than patients who were more independent pre-trauma.²⁶ In our cohort, one in five patients had unmet needs, or going without a routine activity such as self-care without assistance, suggesting poor performance status. It is particularly concerning that over 20% have new needs, and almost 40% have unmet needs in the months following trauma; previous research has shown that regardless of pre-existing disability, only 50% of adults who fall will regain their baseline functional status in the year following injury, suggesting that some new caregiving needs will be permanent.¹³

Our findings build upon current research of the long-term outcomes of geriatric trauma patients and highlight the need to characterize informal care of injured older adults, support informal caregivers, and advocate for greater access to high-quality paid home care. In our cohort, most everyday needs were addressed by family members who provided an average of 30 hours of unpaid care each week. This is five times the care reported via formal paid care. This is likely due in part to current home healthcare reimbursement: although Medicare will cover medically necessary services, criteria can be restrictive, such as requirement of being “homebound”, and care designated as “non-medical” (e.g. self-care or household tasks) are not covered.^27–29 The Centers for Medicare & Medicaid define assistance with non-medical activities of daily living as personal or “custodial” care, and such coverage is at the discretion of individual supplementary plans (Medigap, Medicaid, private/employer-sponsored).³⁰ Access to custodial and personal care services is also dependent on politics and geography: as Medicaid is the primary payer for long-term personal care or custodial services, paid home healthcare use increased among low-income, middle-aged adults who lived in states who expanded Medicaid under the Affordable Care Act versus non-expansion states.³¹ However, the system for obtaining such services is complex, and the quality of such services is variable: an analysis of older adults enrolled in Medicare and Medicaid who had home and community-based services showed comparable rates of hospitalization to older adults living in long-term care facilities.³² Ultimately, when caregiving resources are inadequate or unreachable, the burden of care is shouldered by informal caregivers.³³ Unfortunately, while home health coverage is limited, disability is not; the majority of older adults report needing new assistance with at least one ADL six months after trauma.^34,35 In addition to everyday tasks, informal caregivers often absorb additional medical fees and care-related costs, such as medical equipment/supplies and home modifications, while at the same time they have decreased productivity and earning potential in the workforce due to new responsibilities.^36–38 This amounts to both the physical and financial burden of care being transferred to family and friends, and yet the direct and indirect costs are overlooked within our healthcare system, and family and sick leave policies remain extremely limited.^39,40 Current events, such as the ongoing COVID-19 pandemic, have exacerbated these issues, creating new caregiver roles and making caregiver responsibilities more difficult to carry out at the same time that many were reporting increased financial hardships and financial insecurity due to loss of employment.⁴¹

These findings underscore the need for transitions of care to account for the long-term needs of older adult trauma patients and can inform pre-discharge communication with informal caregivers. The complex process of transitioning care from inpatient to outpatient settings relies heavily on written discharge documentation and less is known about patient and family expectations around recovery and care needs. Current best practice guidelines for geriatric trauma recommend providing the patient and caregivers with written documentation of the discharge plan and follow-up, but not strategies for optimal communication about a complex post-discharge care plan.⁴² Qualitative studies have demonstrated that patient perception of the discharge planning process does not accurately reflect understanding of or ability to carry out a discharge plan: although the vast majority of patients can verbalize their reason for hospital admission, 20–40% of patients reported that they were not asked if they had adequate support at home to carry out the discharge plan.^43,44 However, inclusion of informal caregivers in the discharge planning process has specific value to health systems: a meta-analysis by Rodakowski et al. showed that integration of caregivers in discharge planning was associated with 25% fewer readmissions at 90- and 180-days post-discharge. Documentation of family caregivers in the medical record and provision of education for caregiver duties has been formalized with the Caregiver Advise, Record and Enable (CARE) act, which has become law in over 40 states, but implementation remains inconsistent.⁴⁵ Although caregiver experience and caregiver burden have been explored for those caring for older adults who have chronic conditions such as dementia and cancer^10–12, research focused on the experience of post-trauma caregivers is lacking and essential to understand the impact of care responsibilities on the wellbeing of informal caregivers and patients alike.

Limitations

There are several limitations to note. First, the sample of patients interviewed may not be generalizable to all older patients experiencing traumatic injury. The median ISS in our cohort was lower than that of other analyses of injured older adults who required hospitalization^2,34,46 which may be associated with their community-dwelling status at time of follow-up. In utilizing the NHATS data set, patients had to be alive within a year of injury and have completed interviews in consecutive years, therefore patients who died before the second interview were excluded. On average, interviews were conducted approximately 5 months before and 5 months after trauma, but it is possible that patient participation in routine activities may have varied depending on time from injury to interview. Additionally, these interviews assess caregiving as perceived by the patient, and not the caregiver, and additional research is needed to understand the caregiver perspective and experience. There is a companion study to NHATs, the National Study of Caregiving (NSOC), which recruits family and unpaid caregivers to report on caregiver demographics, help provided, duration, intensity and impact of caregiving on caregivers; however, as NSOC is a sample of the NHATs cohort, we were unable to identify a sufficient number of caregivers of patients who met our inclusion criteria for analysis. To further elucidate the experience of caregivers, an ongoing study at our institution is conducting interviews regarding the impact of the roles and responsibilities of informal caregivers of older adult trauma patients. Finally, we were underpowered to perform a multivariate analysis of the most extreme change in care needs (≥40 hours/week), and this population warrants further investigation given high reported needs.

Conclusion

Older adults who experience trauma often have high care needs prior to trauma and increased care needs after trauma, including 20% with new care needs and 43% with unmet needs. However, higher injury severity does not independently predict increased caregiving hours after trauma. These findings help quantify the types and amount of care assumed by informal caregivers after trauma and should be used to set expectations for patients and families about recovery and facilitate care transitions after hospitalization.