Identification of Seniors at Risk Scoring in Geriatric Trauma: Exploring Clinical Outcome Correlations.

Introduction: Geriatric trauma patients (GTP) make up an increasing percentage of the overall trauma population. Due to frailty, GTP are at an increased risk of morbidity and readmission. Therefore, it is becoming increasingly important to prognosticate outcomes to assist with resource utilization. We hypothesized that the "Identification of Seniors at Risk" (ISAR) score may correlate with both clinical outcomes and resource utilization for geriatric trauma patients.
Methods: Patients older than 65 years who were admitted to the trauma service were screened using an ISAR scoring algorithm. Outcomes, including 30-day mortality, all-cause morbidity, hospital length of stay (LOS), intensive care unit (ICU) LOS, functional independence measures (FIM) at discharge, and percent discharged to a facility, were analyzed. Both descriptive and data-appropriate parametric and non-parametric statistical approaches were utilized, with significance set at α = 0.05.
Results: One thousand and two hundred seventeen GTP were included in this study. The average age was 81, median injury severity score was 9, and 99% had a blunt trauma mechanism. ISAR scores were generally associated with increasing 30-day mortality (0%, 1.9%, 2.4%, and 2.1% for ISAR 0, ISAR 1-2, ISAR 3-4, and ISAR 5-6, respectively), morbidity (2.6%, 7.6%, 14.7%, and 7.3% for respective categories), longer hospital (3.1, 4.6, 5.1, and 4.3 days, respectively) and ICU stays (0.37, 0.64, 0.81, and 0.67, respectively), lower FIM score at discharge (18.5, 17.1, 15.8, and 14.4, for respective categories), as well as increasing percentage of patients discharged to a facility (29.8%, 58.9%, 72.1%, and 78.8% for respective categories). Conclusions: This exploratory study provides important early insight into potential relationships between ISAR and geriatric trauma outcomes. ISAR screening is a quick and easy-to-use tool that may be useful in GTP triage, level-of-care determination, and disposition planning. Understanding populations at risk, especially those with more intricate discharge needs, is an important step in mitigating those risks and implementing appropriate care plans. Copyright:

INTRODUCTION

It is estimated that adults aged 65 years and greater will constitute more than 19% of the United States population by 2030, compared to approximately 13% in 2010.[1] Concurrent to this trend, it is expected that the overall prevalence of traumatic injuries will increase within the older age group. According to the National Trauma Data Bank 2016 Annual Review, the geriatric (≥65 years) patients comprised 30.8% of the 2015 trauma population.[2] It was also noted that case-fatality rate begins to increase beyond this age threshold.[23] Research shows that the aging process decreases the effective physiologic reserve, and when combined with comorbid conditions, the result is diminished resiliency in the context of traumatic injury.[45] Moreover, the geriatric population faces a disproportionate risk of morbidity and mortality in association with relatively minor injury severity and/or mechanism.[678910] Finally, older adults are likely to require more support and specialty resources after discharge, experience higher readmission rates, and are less likely to return to preinjury baseline function.[11121314]

Cumulatively, these factors make it increasingly more important to accurately quantify frailty and prognosticate outcomes in geriatric trauma patients (GTP).[515] Several clinical measurement scales and scoring systems have been developed for this purpose in the hospitalized geriatric population.[51617] However, beyond the traditional outcome determinants such as the injury severity score (ISS), there is currently no reliable or universally accepted risk assessment system in the GTP. Some investigators have attempted to address this dilemma by examining various frailty indices and other outcome determinants relevant to this important problem.[45]

The “Identification of Seniors at Risk” (ISAR) score was introduced as a simplified frailty screening tool developed for use in emergency departments (EDs).[18] The score consists of six components, each answered simply with a “yes” or “no” designation, where each question is given one point for a “yes” response [Table 1]. An ISAR score of ≥2 has been shown in the general ED population to correlate with an elevated 6-month risk of death, functional decline, the need for discharge to a nursing facility, and elevated readmission rates.[1819]

Table 1

Identification of seniors at risk questionnaire given to patients*

Question asked	Yes	No
1. Did you previously need help on a regular basis?
2. After the injury, have you needed more help than usual?
3. Any previous hospitalizations in the last 6 months?
4. Do you see well?
5. Do you have memory problems?
6. Do you take >3 medications every day?

Each question of the ISAR questionnaire is worth up to 1 point, with the lowest score being 0 and the highest score being 6. Based on the cumulative score, patients were placed into 1 of 4 ISAR categories which were lowest (ISAR 0), mild (ISAR 1-2), moderate (ISAR 3-4), and highest (ISAR 5-6). ISAR: Identification of seniors at risk. *Modified for brevity and clarity, without context loss

This retrospective study explores the potential use of ISAR as a prognostic tool in the geriatric trauma population. We hypothesized that increasing ISAR scores in our institutional GTP subset would correlate with 30-day mortality, all-cause morbidity, discharge destination, and patient functional outcomes.

METHODS

The current study is a single-site, Institutional Review Board (IRB) exempt, retrospective medical record review conducted at a regional Level 1 Trauma Center. Geriatric (age ≥65 years) patients underwent ISAR screening as a standard part of the tertiary trauma evaluation. In accordance with established protocol, patients and/or their caregivers responded to the ISAR questionnaire [Table 1] administered by either an advanced practitioner or physician. This typically occurred on hospital day 2. Patients with an ISAR ≥2 received a geriatrics consult. The geriatrics team assisted with dementia management, determining additional outpatient resource requirements and facilitating their acquisition, helping with discharge facility planning, leading and/or participating in family discussions, as well as the performance of both inpatient and outpatient medication reviews while making necessary medication changes as indicated. For the purposes of this study, inclusion criteria specifically consisted of all trauma patients aged >65 years who completed the ISAR questionnaire in full. Trauma patients aged >65 years who did not complete the ISAR questionnaire were excluded from the study.

After obtaining IRB exemption, patient records were abstracted between August 2013 and December 2017. In addition to the ISAR scores, collected variables included patient demographics (age and gender); mechanism of injury; Injury Severity Score (ISS); Glasgow Coma Scale (GCS); 30-day mortality; all-cause morbidity; hospital length of stay (HLOS); intensive care unit LOS (ICULOS); Functional Independence Measures (FIM) on discharge; and discharge disposition (home versus facility). A complete list of comorbidities captured in the Trauma Registry is shown in Table 2. All-cause morbidity was captured by the Pennsylvania Trauma Systems Foundation and includes myocardial infarction, deep-vein thrombosis, pulmonary embolism, pneumonia, acute kidney injury, urinary tract infection, pressure ulcers, and unplanned admission to the ICU. FIM is an 18-component scale of physical, social, and psychological function used to assess disability.[20] Specific FIM factors assessed in this study include components such as eating, grooming, toilet transfers, ambulation, and memory.

Table 2

List of comorbidities captured in the trauma registry

Categories	Categories (continued)
Cardiovascular	Psychological/mental disability
CAD	Mental/personality disorder
CHF	ADD
MI	ADHD
HTN	Intellectual disability
PAD	Autism spectrum
Prehospital cardiac arrest with	Substance use disorder
CPR	Alcohol use disorder
CVA
Angina pectoris	Vaping and e-cigarette use
	Current smoker
Gastrointestinal	Cancer/hepatobiliary
Peptic ulcer disease	Documented history of cirrhosis
Gastric or esophageal varices	Undergoing current therapy
Bariatric surgery	Concurrent existence of metastasis
Endocrine/metabolic	Autoimmune
Diabetes mellitus	Arthritis
Obesity	SLE
Thyroid disease
Hematologic	Pulmonary/renal
Bleeding disorder	COPD
Chronic aspirin use	Serum creatinine >2 mg% (on admission)
Anticoagulant therapy	Dialysis
Immunosuppression	Neurologic
HIV/AIDS	Spinal cord injury
Routine steroid therapy	Cerebral palsy
Transplants (major organ transplants only)	Dementia
Active chemotherapy	Previous history of head trauma
Health status	Pediatric/genetic
Advanced directive limited care	Prematurity
Functionally dependent health status	Congenital disorder
	OI
Obstetric/gynecologic
Pregnancy

CAD: Coronary artery disease, CHF: Congestive heart failure, MI: Myocardial infarction, HTN: Hypertension, PAD: Peripheral arterial disease, CVA: Cerebrovascular accident, ADD: Attention deficit disorder, ADHD: Attention deficit hyperactivity disorder, SLE: Systematic lupus erythematosus, COPD: Chronic obstructive pulmonary disease, OI: Osteogenesis imperfect, CPR: Cardiopulmonary resuscitation

Similar to the work by Asomaning and Loftus,[21] ISAR was introduced at our institution as a clinical tool intended to better identify injured patients who would most benefit from the resources provided by our geriatric medicine service. Previous studies evaluating ISAR in the emergency department population utilized ISAR scores of > 2 as the primary cut-off point between patients being deemed “low risk” versus “elevated risk”.[1821] To gain further insight into more granular ISAR characteristics, we stratified patients into four groups based on the following ISAR score ranges: lowest (ISAR 0), mild (ISAR 1–2), moderate (ISAR 3–4), and highest (ISAR 5–6). The intent was to determine, with enhanced granularity levels, the degree of correlation between increasing ISAR scores and pre-determined study outcomes.

Key outcome variables included 30-day mortality, all-cause morbidity, HLOS, ICULOS, FIM on discharge, and discharge to facility. Descriptive statistics included measures of central tendencies and percentages/proportions reflective of general patient sample characteristics. For categorical data, Chi-squared or Fisher's exact testing were utilized, as appropriate. Since ISAR subgroups were heterogeneous in terms of patient sample characteristics (e.g., age, gender, GCS and ISS), statistical adjustments for these parameters were performed using analysis of covariance (ANCOVA) when calculating estimates of ISAR's effects on key outcome variables, as applicable. Due to a small corresponding sample size, 30-day mortality analyses were performed using Fisher's exact test. Finally, secondary analyses of mortality, HLOS, and ICULOS were performed, excluding patients with active advanced directive/do-not-resuscitate (AD/DNR) status. Data were reported as frequencies, means with standard deviation (standard error when specified), or medians with interquartile range (IQR), as appropriate. Statistical significance was set at α = 0.05.

RESULTS

A total of 1217 patients met geriatric trauma inclusion criteria and underwent ISAR screening during the study period. This was 28.3% of the total number of 4303 geriatric trauma patients whose records were available for screening. In terms of demographics, the median patient age was 81 (IQR 74–88) years, 58% were female, 99% experienced blunt injury mechanism, with the median ISS of 9 (IQR 4–10). The majority of patients had an ISAR score of 1–2 (n = 567, 46.6%) and 3–4 (n = 464, 38.1%), while 92 (7.6%) patients had an ISAR of 0 and 94 (7.7%) patients had an ISAR of 5–6 [Table 3].

Table 3

Basic characteristics of the study sample stratified by identification of seniors at risk scores

	Age†	Gender (% male)†	ISS*	GCS†
ISAR 0 (n=92)	72.6±9.46	52.7	9.15±6.44	14.8±0.95
ISAR 1-2 (n=567)	78.1±11.9	44.8	8.47±5.79	14.8±1.00
ISAR 3-4 (n=464)	83.1±8.18	38.6	8.63±5.75	14.5±1.17
ISAR 5-6 (n=94)	84.4±10.9	32.3	7.27±5.95	14.2±1.16

*P<0.20, †P<0.05. ISAR: Identification of seniors at risk, ISS: Injury severity score, GCS: Glasgow Coma Scale

Due to inherent differences in subgroup characteristics [Table 3], statistical corrections were performed, as applicable (see methods for detailed statistical procedure). We noted an association between ISAR scores and 30-day mortality. More specifically, 30-day mortality was 0% for ISAR 0, 1.9% for ISAR 1–2, 2.4% for ISAR 3–4, and 2.1% for ISAR 5–6 [Table 4]. After excluding patients with AD/DNR (n = 67), mortality was 0% for ISAR 0, 1.5% for ISAR 1–2, 1.6% for ISAR 3–4, and 1.2% for ISAR 5–6. Higher ISAR scores were also associated with greater all-cause morbidity [Table 4, P < 0.001]. Here, all-cause morbidity was 14.7% for ISAR 3–4 compared to 2.6% for ISAR 0.

Table 4

Outcomes in relation to the various identification of seniors at risk groups

Outcome category	ISAR score (n=1217)
	ISAR 0 (n=92)	ISAR 1-2 (n=567)	ISAR 3-4 (n=464)	ISAR 5-6 (n=94)
30-days mortality rate (%)	0.0	1.9	2.4	2.1
All-cause morbidity (%)	2.6	7.6	14.7	7.3
Hospital LOS (days)	3.1±0.59	4.6±0.20	5.1±0.21	4.3±0.45
ICU LOS (days)	0.37±0.28	0.64±0.10	0.81±0.13	0.67±0.24
Discharge FIM score	18.5±0.29	17.1±0.08	15.8±0.14	14.4±0.38
Rate of discharge to facility (%)*	29.8	58.9	72.1	78.8

*Discharge to facilities includes nursing homes, rehabilitation, or acute care facilities. Means are reported with SE. n: Number of patients in the sample, SE: Standard error, ISAR: Identification of Seniors at Risk, LOS: Length of stay, ICU: Intensive care unit, FIM: Functional independence measures

Increased HLOS was seen with greater ISAR score ranges [Table 4, P < 0.001]. The highest average HLOS was noted in the ISAR 3–4 groups (5.1 days) compared to 3.1 days for ISAR 0. When patients with AD/DNR are excluded, the intergroup difference in LOS continues to be significant (3.33 ± 0.42 days for ISAR 0; 4.31 ± 0.187 for ISAR 1–2; 5.19 ± 0.20 for ISAR 3–4; and 4.16 ± 0.32 for ISAR 5–6, P < 0.01). ICULOS for ISAR scores ≥1 was significantly longer when compared to the ISAR 0 group [Table 4, P < 0.001]. However, after excluding patients with AD/DNR, intergroup differences in ICULOS became nonsignificant (0.35 ± 0.14 days for ISAR 0; 0.62 ± 0.13 days for ISAR 1–2; 0.75 ± 0.09 days for ISAR 3–4; and 0.37 ± 0.12 days for ISAR 5–6, P = 0.43).

Increasing ISAR was associated with lower FIM scores at discharge [Table 4, P < 0.001]. More specifically, patients with an ISAR of 0 had a mean discharge FIM of 18.5 ± 1.45, compared to a mean FIM of 14.4 ± 4.39 for the ISAR 5–6 group.

Overall, 85.3% of the population came from home, 7.2% presented from a skilled nursing facility, and 7.5% were from assisted living. Regarding pre-injury location given ISAR grouping, 96.4% of ISAR 0, 94.3% of ISAR 1–2, 82.2% of ISAR 3–4, and 61.6% of ISAR 5–6 came from home. With higher ISAR ranges, there also was an increase in the percentage of patients discharged to an intermediate destination, such as nursing home, rehabilitation, or long-term acute care facility [Table 4, P < 0.001]. For this outcome variable, 29.8% of ISAR 0 patients were discharged to a facility, while 78.8% of those with ISAR 5–6 required postdischarge placement.

DISCUSSION

Human longevity continues to increase, with a progressively greater percentage of the U.S. population leading longer and more productive lives.[522] At the same time, this important trend is associated with increasing frequency of chronic comorbidities, polypharmacy, and frailty.[523] The geriatric population is projected to continue to require an increasing percentage of trauma system resources. In addition, a traumatic event can initiate a cascade of declining functional status and an exacerbation of underlying medical conditions, all of which may lead to a spiral of increased hospital readmissions, a longer HLOS, health-care-related complications, escalating health-care expenditures, decreased return to original baseline, and significantly lower quality of life.[324252627] Therefore, it is becoming more important for clinicians who manage GTP to develop methods for prognosticating overall outcomes, complications, as well as identifying patients who are at high risk for readmission. When this is performed correctly and routinely, more optimized, patient-specific support can be employed early and efficiently. At present, available support resources include geriatric medicine and social work consults, as well as physicial therapy, occupational therapy and rehabilitation planning for prompt discharge when the patient is ready.

Previous studies have focused on attempts to prognosticate outcomes in the injured geriatric population.[28293031] This is, at least in part, because frailty has become an important patient characteristic, increasingly associated with higher morbidity and mortality, prolonged hospitalization, along with a greater need for posthospitalization support.[32333435] The recognition of these associations led to the observation that even a simple frailty screening assessment may be beneficial in elective surgical patients, facilitating improved morbidity and mortality through appropriate early intervention(s).[36] ISAR is also being investigated in a similar fashion to determine its ability to prognosticate, manage resources, and hopefully improve outcomes.[37] In 2017, Gronewold et al. used ISAR screening with or without a Comprehensive Geriatric Assessment (CGA).[38] Three hundred twenty-seven patients were screened and found to be ISAR “positive” (i.e., defined as scores 2 or higher). The authors found that the addition of CGA to ISAR was useful in prognosticating an increased LOS, increased number of falls, and an increasing requirement for nursing and physiotherapy support services. However, in multivariate analysis, ISAR was not predictive of longer hospitalization, whereas activities of daily living, signs of depression, younger age, male sex, and number of admission diagnoses were predictive of prolonged hospitalization. Subsequently, Scharf expanded ISAR/CGA screening to include hospitalized internal medicine patients and found that it may be useful in identifying patients requiring greater overall resources.[39] Our study is an attempt at using the ISAR screening tool in GTP while additionally stratifying ISAR scores into more granular ranges. The ISAR screening questionnaire can be easily completed by hospital staff. Similar to Asomaning and Loftus,[21] the introduction of ISAR at our institution was initiated by a multidisciplinary trauma team and more specifically championed by advanced practice providers and rotating residents, with the intent of improving the quality of injury care for geriatric patients. In this study, a slightly modified ISAR questionnaire [Table 1] was administered by residents, physician assistants, or nursing practitioners, but it could also have been facilitated by other staff, including nursing. The brevity and simplicity of binary responses make it quick and easy to complete. Prior studies focused primarily on patients with ISAR scores >2. The current study subdivided ISAR scores into four groups: ISAR 0, 1–2, 3–4, and 5–6. As demonstrated by our data, increasing ISAR scores were found to be associated with greater 30-day mortality, all-cause morbidity, longer HLOS and ICULOS, lower FIM at discharge, and a greater proportion of patients discharged to a facility. The strongest correlations to ISAR scores were noted for FIM and facility discharge.

Further research into the geriatric trauma population is needed to better determine various nuances of the relationship between ISAR and clinical outcomes. In addition, future studies comparing ISAR directly to other dedicated frailty screening tools could be useful. Validation studies featuring a much larger number of subjects screened with ISAR will be important in order to clarify some of the observed preliminary trends with greater granularity. A key aspect of future investigations should be the determination of more precise, clinically relevant ISAR cutoff values in the context of care planning, early intervention, and optimal resource allocation. The somewhat inconsistent behavior of certain outcomes examined in this study, especially in the highest ISAR range, may have several potential causative factors. First, we propose that there may be some degree of ISAR reporting variability, especially at the higher end of the spectrum, secondary to patients being unable to provide sufficiently detailed history due to a number of potential confounders (e.g., dementia, complexity of comorbid conditions, acute injury state). Second, when patients themselves were unable to provide ISAR-related information, their family members were asked to furnish the information on the patient's behalf. This may have inherently introduced some bias into our existing reporting. As a result of the above-proposed phenomena, patients with higher actual ISAR scores may have been misclassified as having lower reported scores, thus leading to our stated observations. Third, the proportion of complications captured in the Trauma Registry among those with higher ISAR scores may be underreported for a variety of reasons (e.g., early in-hospital mortality, de-escalation of care/transition to hospice, and patient-specific factors such as dementia). Finally, it is important to note that ISAR was originally intended to be a binary “all-or-none” tool for screening of at-risk geriatric patients.[21] Thus, its predictive utility may not be maintained across all levels of the scale.

Despite the above limitations, our study generated sufficient preliminary evidence regarding the correlation between increasing ISAR scores and clinical outcomes to warrant further research in this new area of investigation. The current study suggests that it may be beneficial to mobilize appropriate resources earlier in the care of the GTP to optimize clinical outcomes and to potentially reduce various adverse occurrences (e.g., mortality, morbidity, longer lengths of stay). It would be interesting to see if ISAR-based interventions, such as earlier high intensity physical and occupational therapy, geriatrics consultation, enhanced home nursing visits, and the provision of home-based physical therapy support in the postdischarge period, when applied in patient-specific fashion, have the potential to positively impact outcomes. Indeed, studies have shown that geriatric consults in the setting of trauma may be beneficial when managing the patient's comorbidities, coordinating care, reducing complications, and improving functional recovery.[404142] In addition, geriatric experts can assist in end-of-life and goals-of-care discussions and planning, especially in the context of rapid functional deterioration and the presence of advance directives.[43]

CONCLUSION

This is a large-sample study examining the relationship between ISAR score and clinical outcomes in geriatric trauma patients. Our data suggest a possibility of an association between ISAR scores and all-cause morbidity, 30-day mortality, HLOS, discharge to a facility, as well as decreased functional independence. Given the paucity of published clinical data in this area, the reader must be cautioned regarding the exploratory nature of this study. Further research, including independent validation of our findings and employing larger patient samples, will be necessary to better define and categorize the above findings.

Research quality and ethics statement

This clinical investigation received an exemption from their organization's IRB. The corresponding IRB exemption number is SLIR-2017-77. The authors followed applicable EQUATOR Network (http:// www.equator-network.org/) guidelines during the conduct of this research project.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.