Received: 5 February 2024 Revised: 25 April 2024 Accepted: 2 May 2024

DOI: 10.1111/jgs.18977
Journal of the
CLINICAL INVESTIGATION American Geriatrics Society

Impact of inpatient geriatrics consultation on hospital

outcomes in older adults with trauma

Sara Tyree MD 1 | Karen Fischer MPH 2,3 | Daniel Stephens MD 4 |

M. Caroline Burton MD 1 | Sandeep Pagali MD, MPH, AGSF 1,5

1
Division of Hospital Internal Medicine,
Mayo Clinic, Rochester, Minnesota, USA Abstract
2
Department of Medicine Research Hub, Background: Older adults presenting with trauma have worse outcomes than
Mayo Clinic, Rochester, Minnesota, USA younger adults. Starting in 2016, we provided geriatrics consultation (GC) to
3
Department of Quantitative Health older adults admitted to the trauma service. We aimed to analyze the impact
Sciences, Mayo Clinic, Rochester,
Minnesota, USA
of GC on patient outcomes.
4
Department of Trauma, Critical Care, Methods: We performed a retrospective pre–post study and year-matched
and General Surgery, Mayo Clinic, cohort study. We identified patients from the trauma registry at our level
Rochester, Minnesota, USA
1 trauma center. In the pre–post study, we compared patients who received
5
Division of Geriatrics and Gerontology,
GC (2016–2022) with controls (2011–2015). In the cohort study (2016–2022),
Mayo Clinic, Rochester, Minnesota, USA
we compared patients who received GC with controls. We matched for age,
Correspondence race, sex, and injury severity score (ISS) in both studies, as well as admission
Sara Tyree and Sandeep Pagali, Division
of Hospital Internal Medicine, Mayo
year in the cohort study. Outcome variables included mortality (in-hospital,
Clinic, Rochester, MN, USA. 30-day, 90-day), length of stay (LOS), discharge disposition, and hospital read-
Email: [email protected]; pagali. mission rates (30-day, 90-day).
[email protected]
Results: We analyzed 1968 patients in the pre–post study and 2544 patients in
the cohort study. Patients were similar in age, race, and sex. GC patients had a
slightly higher ISS score and a higher rate of ICU stay. Delirium occurrence
was lower among GC patients. GC patients had lower in-hospital mortality
compared to controls (pre–post OR 0.27, p < 0.001; cohort OR 0.31, p < 0.001)
and increased LOS (6 days vs 4 days, p < 0.001; both studies). GC patients in
the cohort study also had lower 30- and 90-day mortality (OR 0.52 and 0.65,
p < 0.01) and were less likely to return home (OR 0.81, p < 0.01); similar
trends, though not statistically significant, were noted in the pre–post study.
Lower readmission rates (statistically non-significant) were noted in the GC
group across both studies.
Conclusions: GC in older adults with trauma has proven benefit with reduced
mortality and a trend toward lower readmission rates but was associated with
increased LOS and higher rates of discharge to skilled facility.

KEYWORDS
acute care, age friendly health system, geriatrics consult, length of stay, models of care,
mortality

J Am Geriatr Soc. 2024;1–9. wileyonlinelibrary.com/journal/jgs © 2024 The American Geriatrics Society. 1

 15325415, 0, Downloaded from https://agsjournals.onlinelibrary.wiley.com/doi/10.1111/jgs.18977 by Readcube (Labtiva Inc.), Wiley Online Library on [25/05/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
2 TYREE ET AL.

INTRODUCTION
Key points
As the older adult population grows, the number of older
adult trauma patients is also increasing.1 Trauma • Geriatrics consultation for hospitalized older
patients, depending on their injury type and severity, adults with trauma, including a comprehensive
require hospitalization for additional monitoring and geriatrics assessment following the Age
reassessment to intervene promptly. Older adult trauma Friendly Health System's 4Ms framework, is
patients have worse outcomes than younger patients with associated with decreased mortality and hospi-
similar injury severity.2 Factors that contribute to chal- tal readmissions.
lenges in caring for hospitalized older adults include • Opportunities to reduce higher discharge rates
polypharmacy, cognitive impairment that may be unrecog- to skilled nursing facilities and increased hos-
nized, delirium, deconditioning, assessing goals of care, pital length of stay associated with geriatrics
and aligning care plans with patient priorities. A geriatrics consultation need to be explored.
consultation (GC) facilitates a comprehensive geriatrics • Additional studies are needed to evaluate opti-
assessment, which focuses on addressing the above chal- mal criteria for geriatrics consultation, and
lenges as well as supporting acute medical needs. In 2013, value in other hospitalized non-trauma older
the American College of Surgeons Trauma Quality adults.
Improvement Program (TQIP) guidelines emphasized the
importance of GC in older adults to improve patient out- Why does this paper matter?
comes.3 However, subsequent studies evaluating the
This study confirms the impact of geriatrics con-
impact of GC have reported varied results.4–10
sultation by comparing two separate patient
In meta-analyses of hospitalized older adults admitted
groups and validates the value added in terms of
to the hospital for any emergent reason, GC improved
reduced mortality and readmissions. Its findings
mortality and the ability of older adults to return home,
support the current American College of Sur-
but trauma patients comprised a minority of study partici-
geons recommendations about geriatrics consul-
pants.11,12 A protocol combining GC with a panel of diag-
tation for hospitalized older adults with trauma.
nostic tests and protocolized interventions improved
mortality for older adult trauma patients,7,13 as did admit-
ting them to a specialized geriatric trauma unit.14 GC has
been shown to increase adherence to best practice recom- intervention are admitted to the trauma service, while
mendations for older adult trauma patients.15,16 isolated organ injuries (e.g., hip fracture, isolated intra-
The impact of GC alone on outcomes for older adult cranial bleed) are admitted to respective subspecialty ser-
trauma patients is varied across studies. Many studies vices and non-surgical patients with minimal trauma are
have compared outcomes of patients admitted pre- admitted to medical services. The trauma service is
implementation of GC and post-implementation of staffed by general surgeons who have a focused practice
GC. Limitations of this study design include the unmea- in trauma. The geriatrics consultation team is staffed by
sured impact of evolving care protocols, quality projects, board certified geriatricians and a focused group of clini-
and best practices over time. Results reported from these cians with a special interest in geriatrics. Patients admit-
studies have been conflicting on outcome measures incl- ted to non-trauma surgical services can receive a
uding length of the stay (LOS),4,5,8–10,16 mortality,4,5,7–10,13 medicine consultation, staffed by hospitalists and general
incidence of delirium,6,8–10,16 rate of discharge to skilled internists. Patients admitted to medical services receive a
nursing facility (SNF),5,8,16 and readmission.4,5 A meta- surgical consultation if needed, but do not receive a geri-
analysis from 2020 observed that available studies were atrics consultation or another medicine consultation. We
limited and of poor to moderate quality, making meaning- aimed to analyze the impact of GC on patient-related out-
ful conclusions difficult or impossible.17 comes in older adults admitted with trauma.
Following TQIP guidelines, Mayo Clinic Hospital-
Rochester implemented GC for older adult trauma
patients admitted to the trauma service in 2016. The GC METHODS
service follows a consultative practice model, communi-
cating their recommendations with the primary trauma We performed a retrospective pre–post study and year-
service daily. Our consultation criteria include patients matched cohort study at Mayo Clinic Hospitals, Roches-
age ≥75 or age 65–74 with a Frail score ≥3.18–20 Patients ter, an academic level 1 trauma center with 2059 beds.
with polytrauma who require monitoring for surgical We obtained Institutional Review Board approval prior to
 15325415, 0, Downloaded from https://agsjournals.onlinelibrary.wiley.com/doi/10.1111/jgs.18977 by Readcube (Labtiva Inc.), Wiley Online Library on [25/05/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
GERIATRICS CONSULT IN OLDER ADULTS WITH TRAUMA 3

study initiation. We identified participants from the Matching

Rochester Mayo Trauma Registry, a prospectively main-
tained database. Eligible persons were at least 65 years We created a historical matched study population for
old and were hospitalized due to trauma between 2011 pre–post analysis (Figure 1). Our control group consisted
and 2022. We abstracted additional information from the of patients admitted to the trauma service who did not
electronic health record. receive GC (admitted from 2011 to 2015, prior to GC pro-
Baseline demographic data included age, gender, tocol initiation), and our GC group consisted of patients
race, mechanism of injury, injury severity score (ISS), 21 admitted to the trauma service who received GC (admit-
comorbidities (to calculate Charlson Comorbidity ted from 2016 to 2022). We matched participants for age,
Index [CCI]22), admitting service, length of intensive race, sex, and ISS.
care unit (ICU) stay, and delirium diagnosis (ICD To account for the impact of practice evolution on
codes). ISS score ranges from 0 to 75. CCI was calcu- outcomes, we analyzed patients admitted from 2016 to
lated from data from the electronic health record and 2022 separately. In this study, the control group consisted
was not age-adjusted. Outcome variables included mor- of patients who did not receive GC (usually admitted to
tality (in-hospital, 30-day, and 90-day), hospital LOS, services other than the trauma service), and our interven-
discharge destination (home vs skilled nursing facility tion group consisted of patients who received GC (all
[SNF] vs other), and hospital readmission rates (30-day admitted to the trauma service). We matched participants
and 90-day). for age, race, sex, ISS, and year of admission.

GC protocol Statistical analysis

All trauma patients were evaluated by the trauma service Propensity score matching was conducted using age, race
in the emergency department and a service disposition (categorized into white, other, and unknown), sex, and
recommendation was made by the trauma service in col- ISS for the pre–post study. The same variables were also
laboration with ED. Beginning in 2016, patients admitted used in the cohort study, with the addition of year of
to the trauma service received GC if they met pre- admission. For matching, we utilized the greedy near-
determined consultation criteria. Patients admitted to est neighbor method, with a caliper of 0.25 for the
other services did not receive GC but could receive a pre–post study and 0.1 for the cohort study. Baseline
medicine consultation, typically for a focused pre- characteristics were compared between the two mat-
anesthesia medical evaluation or to answer specific medi- ched groups, as well as all possible cases and controls.
cal questions. Once the GC order was placed, the GC Boxplots of the logit of the propensity score (LPS) were
team typically saw patients within 24 h. The GC team compared between cases and controls and were
consisted of either a board-certified geriatrician or a deemed to have good balance.
member of a focused group of hospitalist physicians We compared demographics and patient outcomes
interested in geriatrics, an advanced practice provider, between the matched cohorts. We performed univariate
a PGY-2 medical resident, and a geriatrics fellow (few analyses using chi-square (binary outcomes) and Kruskal–
weeks/year, based on their learning rotation). The con- Wallis tests (continuous outcomes). We used a multino-
sult was multifaceted, and included performing a com- mial logistic regression model to analyze the discharge
prehensive geriatrics assessment utilizing IHI Age disposition outcome, incorporating propensity matching
Friendly Health System 4M framework from 2017,18 weights. We used a Poisson regression model for the
identifying reversible medical etiologies or need for
additional workup to identify the cause precipitating
trauma, identifying risk for delirium or other hospital-
acquired complications and implementing prevention
strategies, clarifying goals of care and aligning care
plans with patient priorities, and facilitating appropri-
ate disposition at discharge. The 4Ms, mentation,
mobility, medications, and what matters, are evidence-
based elements of high quality care. The GC team con-
tinued to follow the patient until no further input from
GC team was indicated and they signed off or until
hospital discharge. FIGURE 1 Study design.
 15325415, 0, Downloaded from https://agsjournals.onlinelibrary.wiley.com/doi/10.1111/jgs.18977 by Readcube (Labtiva Inc.), Wiley Online Library on [25/05/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
4 TYREE ET AL.

hospital LOS outcome, incorporating propensity matching Cohort study

weights. We used a logistic regression model for in-
hospital mortality, incorporating matching weights. We included total of 2544 patients in the study, 1272 in each
We analyzed 30- and 90-day mortality and readmission group (Table 2). The mean age was 81.9 years (range
outcomes with GC/no GC as unadjusted and then 65–103), 53.9% were female, and 97.6% were white. Patients
adjusted for additional variables (delirium while in were similar in age, sex, race/ethnicity, and year of admis-
hospital, CCI, and ICU stay). Only those who were dis- sion. Though matched for ISS, ISS was statistically higher in
charged from the hospital were included in the post- the GC group with a one-point median difference that corre-
hospital readmission and mortality outcome analyses. lates to unknown clinical significance. CCI was similar
A p-value <0.05 was considered statistically significant. across groups. Fall was the most common type of injury in
We performed all analyses using SAS (SAS Institute both groups, followed by motor vehicle accident injury. The
Inc., version 9.4). proportion of motor vehicle accidents in the GC group was
higher (16% vs 7.1%, p < 0.001). GC patients had higher rates
of ICU stay compared to controls (59.4% vs 39.5%, p < 0.001)
RESULTS and lower rates of delirium (3.5% vs 5.3%, p < 0.001).
Despite the higher ISS and higher rates of ICU stay in
Baseline characteristics and outcomes are shown in the GC group, lower in-hospital mortality (2.3% vs 7.0%,
Table 1 (pre–post) and 2 (cohort), with regression and p < 0.001) was noted. GC was associated with higher hos-
adjusted outcomes in Table 3. The LPS standardized pital LOS (median 6 days vs 4 days, p < 0.001) and a
mean difference after matching was 0.003 for the pre– lower OR for discharge disposition to home compared to
post study and 0.07 for the cohort study. SNF at 0.81 (0.68, 0.95) (Table 3). Using a logistic regres-
sion model, the OR for in-hospital mortality for GC
patients versus controls was 0.31 (95% CI = 0.20–0.47).
Pre–post study Adjusted 30-day mortality (OR 0.52, 95% CI = 0.34–0.78)
and 90-day mortality (OR 0.65, 95% CI = 0.48–0.87) was
We included total of 1968 patients in the study, 984 in also significantly lower for GC patients. There was no sta-
each group (Table 1). The mean age was 80.6 years (range tistically significant difference in 30- and 90-day readmis-
65–101), 48.7% were female, and 97.9% were white. sion rates with a trend toward lower readmission in the
Groups were similar in age, race/ethnicity, sex, and ISS. GC group when adjusted for delirium, CCI and ICU stay.
Fall was the most common type of injury in both groups, Comparing the two studies, both showed increased
followed by motor vehicle accident injury. The propor- hospital LOS and decreased in-hospital mortality for the
tion of motor vehicle accidents in the GC group was GC groups that were clinically and statistically signifi-
lower (16.3% vs 22.5%, p < 0.001). CCI was significantly cant. The cohort study also showed decreased discharge
higher in the GC group (mean 4.6 vs 3.8, p < 0.001). GC to home compared to SNF, decreased 30- and 90-day
patients had higher rates of ICU stay compared to con- mortality, and a trend toward decreased readmission
trols (58.1% vs 26.5%, p < 0.001) and lower rates of delir- rates for the GC group that was not statistically signifi-
ium (2.9% vs 10.5%, p < 0.001). cant; these findings were less pronounced and not statis-
Despite the likely higher medical complexity in the tically significant in the pre–post study.
GC group (based on higher CCI and higher rates of ICU
stay), we noted a lower in-hospital mortality (1.6% vs
5.7%, p < 0.001). GC was also associated with higher DISCUSSION
hospital LOS (median 6 days vs 4 days, p < 0.001) and
an odds ratio (OR) for discharge disposition to home We analyzed two study populations: (1) a pre–post study
compared to SNF of 0.90 (95% confidence interval [CI]: that examined the changes in outcomes of older adult
0.75, 1.09) (Table 3). Using a logistic regression model, trauma patients on the trauma service after the initiation
the OR for in-hospital mortality for GC patients versus of GC, and (2) a year-matched cohort study that com-
controls was significantly lower at 0.27 (95% CI: 0.16– pared older adult trauma patients who received GC from
0.48). Though not statistically significant, a trend 2016 to 2022 with matched controls who did not receive
toward lower 30-day and 90-day mortality rates in GC GC. In both our analyses, we found that GC was associated
group was noted. Adjusted 30- and 90-day mortality with significantly decreased in-hospital mortality and
and readmission rates were similar between groups, increased LOS. In our cohort group, there was also a decr-
although trended in favor of GC. ease in 30- and 90-day mortality and discharge to home.
 15325415, 0, Downloaded from https://agsjournals.onlinelibrary.wiley.com/doi/10.1111/jgs.18977 by Readcube (Labtiva Inc.), Wiley Online Library on [25/05/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
GERIATRICS CONSULT IN OLDER ADULTS WITH TRAUMA 5

TABLE 1 Characteristics and outcomes of the pre–post comparison groups.

Geriatrics consult
No consult (n = 984) (n = 984) P-value
Patient characteristics
Age, mean (SD) 80.2 (8.57) 80.9 (8.13) 0.113a
Median (range) 80 (65, 100) 81 (65, 101)
Female, n (%) 479 (48.7%) 479 (48.7) 1.000b
Race/ethnicity, n (%) 0.470b
White 967 (98.3%) 960 (97.6%)
Black or African American 3 (0.3%) 3 (0.3%)
Hawaiian Native or Pacific Islander 2 (0.2%) 0 (0.0%)
Hispanic 3 (0.3%) 4 (0.4%)
Asian 2 (0.2%) 7 (0.7%)
Unknown 7 (0.7%) 10 (1.0%)
Injury severity score 0.292a
Mean (SD) 11.2 (7.59) 11.2 (6.96)
Median (range) 9.0 (1.0, 51.0) 10.0 (1.0, 50.0)
CCI, mean (SD) 3.8 (3.22) 4.6 (3.69) <0.001a
Mechanism of injury, n (%) <0.001b
MVA 212 (21.5%) 160 (16.3%)
Fall 666 (67.7%) 760 (77.2%)
Other 106 (10.8%) 64 (6.5%)
ICU stay, yes, n (%) 261 (26.5%) 572 (58.1%) <0.001b
Delirium, yes, n (%) 103 (10.5%) 29 (2.9%) <0.001b
Outcomes
In-hospital mortality, n (%) 56 (5.7%) 16 (1.6%) <0.001b
30-Day mortality, n (%) 33 (3.6%) 33 (3.4%) 0.862b
90-Day mortality, n (%) 65 (7.0%) 73 (7.5%) 0.653b
Hospital length of stay <0.001a
Mean (SD) 6.1 (6.71) 7.4 (7.25)
Median (range) 4 (1, 97) 6 (1, 82)
Discharge disposition, n (%) <0.001b
SNF 514 (55.4%) 535 (55.3%)
Home 375 (40.4%) 353 (36.5%)
c
Other 39 (4.2%) 80 (8.3%)
30-Day readmission, n (%) 145 (15.6%) 146 (15.1%) 0.743b
90-Day readmission, n (%) 215 (23.2%) 243 (25.1%) 0.325b

Abbreviations: CCI, Charlson Comorbidity Index; GC, geriatric consult; ICU, intensive care unit; MVA, motor vehicle accident; SNF, skilled nursing facility.
a
Kruskal–Wallis p-value.
b
Chi-square p-value.
c
Includes hospice, psychiatry, home health, and other categories.

While some of our findings (decreased mortality, no dif- Reviewing the possible variations across the study
ference in readmission rates) are consistent with previous groups, the groups in our year-matched cohort study had
studies examining the impact of GC on older adult some significant differences, inherent to the fact that
trauma patients,4,5,7,8,10,13,16 others (increased LOS and patients with different injuries were typically admitted to
SNF admission) are not.5,8,10,16 different specialties, which is a practice-based parameter.
 15325415, 0, Downloaded from https://agsjournals.onlinelibrary.wiley.com/doi/10.1111/jgs.18977 by Readcube (Labtiva Inc.), Wiley Online Library on [25/05/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
6 TYREE ET AL.

TABLE 2 Characteristics and outcomes of the cohort comparison groups.

No consult Geriatrics consult

(n = 1272) (n = 1272) P-value
Patient characteristics
Age, mean (SD) 81.9 (8.52) 81.5 (8.05) 0.185a
Median (range) 83 (65, 103) 82 (65, 101)
Female, n (%) 685 (53.9%) 685 (53.9%) 1.000b
Race/ethnicity, n (%) 0.877b
White 1242 (97.6%) 1240 (97.5%)
Native American 5 (0.4%) 1 (0.1%)
Black or African American 2 (0.2%) 4 (0.3%)
Hawaiian Native or Pacific Islander 1 (0.1%) 0 (0.0%)
Hispanic 5 (0.4%) 6 (0.5%)
Asian 5 (0.4%) 8 (0.6%)
Unknown 12 (1.0%) 13 (1.1%)
Injury severity score <0.001a
Mean (SD) 11.5 (7.76) 11.8 (7.26)
Median (range) 9 (1, 48) 10 (1, 50)
CCI, mean (SD) 4.4 (3.58) 4.6 (3.65) 0.123a
Mechanism of injury, n (%) <0.001b
MVA 90 (7.1%) 204 (16.0%)
Fall 1104 (86.8%) 990 (77.8%)
Other 78 (6.1%) 78 (6.1%)
ICU stay, yes, n (%) 490 (38.5%) 756 (59.4%) <0.001b
Delirium, yes, n (%) 52 (4.1%) 42 (3.3%) 0.293b
Outcomes
In-hospital mortality, n (%) 89 (7.0%) 29 (2.3%) <0.001b
30-Day mortality, n (%) 63 (5.3%) 44 (3.5%) 0.032b
90-Day mortality, n (%) 116 (9.8%) 98 (7.9%) 0.095b
Hospital length of stay <0.001a
Mean (SD) 6.2 (7.11) 7.5 (7.04)
Median (range) 4 (1, 90) 6 (1, 82)
Discharge disposition, n (%) <0.001b
SNF 620 (52.4%) 711 (57.2%)
Home 463 (39.1%) 427 (34.4%)
Other 100 (8.4%) 105 (8.4%)
30-Day readmission, n (%) 196 (16.6%) 190 (15.3%) 0.388b
90-Day readmission, n (%) 297 (25.1%) 316 (25.4%) 0.858b

Abbreviations: CCI, Charlson Comorbidity Index; GC, geriatric consult; ICU, intensive care unit; MVA, motor vehicle accident; SNF, skilled nursing facility.
a
Kruskal–Wallis p-value.
b
Chi-square p-value.

GC patients, admitted to the trauma service under the intracranial hemorrhage. The confounding introduced by
care of general surgeons, were more likely to have poly- this difference is difficult to quantify and may not have
trauma, rib fractures, and hematomas. Controls, who been overcome by matching. In the pre–post study, CCI
were more often admitted to orthopedics and neurosur- was significantly different between groups (higher in
gery, were more likely to have isolated hip fractures and GC). In the year-matched cohort study, ISS was
 15325415, 0, Downloaded from https://agsjournals.onlinelibrary.wiley.com/doi/10.1111/jgs.18977 by Readcube (Labtiva Inc.), Wiley Online Library on [25/05/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
GERIATRICS CONSULT IN OLDER ADULTS WITH TRAUMA 7

TABLE 3 Adjusted outcomes: GC versus control; both studies.

Pre–post study Cohort study

Outcome Odds ratio (95% CI) P-value Odds ratio (95% CI) P-value
a
Hospital mortality 0.27 (0.16, 0.48) <0.001 0.31 (0.20, 0.47) <0.001
30-Day mortalityb 0.85 (0.49, 1.49) 0.569 0.52 (0.34, 0.78) 0.002
b
90-Day mortality 0.83 (0.56, 1.23) 0.373 0.65 (0.48, 0.87) 0.004
c,d
Discharge to home 0.90 (0.75, 1.09) 0.298 0.81 (0.68, 0.95) 0.012
c–e
Discharge to other 1.96 (1.32, 2.94) 0.001 0.92 (0.68, 1.23) 0.557
30-Day readmissionb 0.86 (0.65, 1.14) 0.292 0.82 (0.65, 1.03) 0.085
b
90-Day readmission 0.97 (0.77, 1.22) 0.799 0.93 (0.76, 1.12) 0.428

Abbreviations: GC, Geriatric consult; SNF, skilled nursing facility.

a
Using a logistic regression model with propensity matching weights.
b
Adjusted for delirium while in hospital, Charlson Comorbidity Index, and intensive care unit stay.
c
SNF is the reference category.
d
Using a multinomial logistic regression model with propensity matching weights.
e
Includes hospice, psychiatry, home health, and other categories.

significantly higher in the GC group, although it is observations by others.4,7,10,13 While both the pre–post
unclear if this difference translates to clinical signifi- and cohort studies showed significantly decreased in-
cance. There were higher rates of delirium in the control hospital mortality, the lack of statistical difference in 30-
groups and it is unclear if this was due to the lack of GC; and 90-day mortality for the pre–post study leads us to
certainly our delirium rates are lower than expected.19 question the mechanism whereby GC improved 30- and
In both studies, GC patients had higher rates of ICU 90-day mortality in the cohort study. There is also the
stay, which could be due to higher medical acuity associ- potential for selection bias, as patients had to survive
ated with polytrauma. In the cohort study, GC patients the trauma and resuscitation to receive a GC. In the pre–
were also less likely to discharge to home (and more post group, we hypothesize that mortality could have
likely to discharge to SNF), which could be another indi- been influenced by evolving and improving management
cator of higher medical comorbidity and injury severity practices and protocols over time. In the cohort group,
resulting in subsequent need for SNF; this is contrast to we wonder if the injury differences associated with
other studies that found GC was associated with higher admitting service could have contributed to mortality dif-
likelihood of discharge home.5,8,16 It is possible that the ferences (for instance, the control group had more intra-
increased LOS for GC patients, which is in contrast to cranial hemorrhage).
previous studies,4,5,8–10,16 was mediated by various factors The 2013 TQIP guidelines highlighted the importance
associated with discharge to SNF, as arranging discharge of GC for older adults who met certain criteria.3 While
to SNF can contribute to extension of hospital LOS by there was little direct evidence to support GC for older
several days. Appropriate identification of need for dis- adult trauma patients before the guidelines, further evi-
charge to SNF could have also contributed to lower read- dence has accumulated, and the results are mixed. Out-
mission rates and 30- and 90-day mortality following comes evidence may be more reliably in favor of GC for
hospital discharge. In patients hospitalized with heart patients with hip fractures, but it is unclear if these
failure, there is a U-shaped relationship between LOS results can be generalized to all older adult trauma
and readmission rates, with short and long LOS being patients. As a Cochrane review from 2020 noted, GC is
associated with increased readmissions for cardiovascular an important clinical tool, but high-quality studies are
reasons.20 LOS has been correlated with increased hospi- lacking in how it may be best applied to patients requir-
tal mortality,23 but this relationship was not borne out in ing planned or emergent surgery.24 At the very least, no
our study. This may suggest that the relatively longer evidence suggests that GC causes harm. Important out-
LOS in our GC group was beneficial. comes that may be positively impacted by GC, such as
Given that GC patients had higher medical acuity or alignment of care plans with patient priorities, have not
comorbidity than controls, as noted by higher CCI in the been assessed by studies to date.
pre–post study, higher ISS scores in the cohort study, and We strongly believe there is a role for geriatrics
higher ICU stay in both studies, their decreased mortality involvement in the care of older adults with trauma even
rates are even more impressive, consistent with with our consultative practice model. Evidence also
 15325415, 0, Downloaded from https://agsjournals.onlinelibrary.wiley.com/doi/10.1111/jgs.18977 by Readcube (Labtiva Inc.), Wiley Online Library on [25/05/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
8 TYREE ET AL.

suggests benefits with more comprehensive approaches. GC with those who do not. Social determinants of health
While the term GC highlights the clinician team, imple- and additional outcome variable should be explored. A
mentation of the recommendations related to the 4Ms mixed methods approach may also be useful to explore
(mentation, mobility, medications, and what matters) factors that could play a role in the effectiveness of
requires a multidisciplinary approach.18,25 It is not acci- GC. Additionally, a diagnosis-based outcomes approach
dental that the 4Ms are designed to contribute to an age could be insightful, as different injury types may have dif-
friendly health system. Metanalyses have found that older ferent outcome trajectories.
adults do better on hospital wards that are specifically
designed to care for older patients.11 In a study cited by
TQIP guidelines, older adult trauma patients had CONCLUSION
improved outcomes when admitted to a geriatric trauma
unit that utilized a multidisciplinary approach.14 Studies Geriatrics consultation has added value in older adults
of older adult trauma patients with orthopedic injuries with trauma and resulted in reduced mortality with
have found improvements with patients who are cared trends toward lower readmission rates. GC was also asso-
for by a combined orthopedics-geriatrics team model.26,27 ciated with higher length of stay and decreased discharge
There are some limitations in our study. Our study to home compared to SNF. Future prospective studies
represents data from a single tertiary care center, but we can further analyze the impact across diagnoses and help
hope the large sample size preserves aspects of its gener- refine consultation criteria to optimize resources.
alizability. The study population was predominantly
white, although we did not exclude any patients from this AUTHOR CONTRIBUTIONS
study due to their ethnicity, and we were unable to iden- Study concept and design: ST, KF, DS, MCB,
tify additional social determinants of health. Also, some SP. Acquisition of data: DS, SP. Statistical analysis:
patients admitted to non-trauma service teams were seen KF. Interpretation of the data: ST, SP. Drafting of the man-
by a non-geriatric hospital medicine team for co- uscript: ST, SP. Critical revision of the manuscript: ST, KF,
management, which may have minimized differences in DS, MCB, SP. Supervision: SP.
patient outcomes between groups given many hospitalists
might also follow appropriate geriatrics principles in their A C KN O WL ED G EME N T S
practices. There could still be underlying confounders for The Department of Medicine Research Hub at Mayo
why a person was chosen for GC that were not adjusted Clinic, Rochester, Minnesota provided support for statisti-
for by the matching process. Other limitations equally cal analysis.
impacting both control and GC groups include: difficulty
obtaining baseline functional data on patients prior to C O N F L I C T O F I N T E R E S T S T A TE M E N T
admission which would confound the disposition desti- The authors declare no conflicts of interest.
nation; lack of post-discharge metrics if the patient estab-
lished follow up outside our organization, or was SP ONS O R 'S RO LE
readmitted elsewhere; lower than expected rates of delir- None.
ium diagnosis likely due to under-diagnosis.19 Addition-
ally, this was a retrospective study relying on data F IN A NCI A L D IS C LO SU RE
abstracted from the trauma registry and EHR. We per- The Department of Medicine Research Hub at Mayo
formed review on a limited subset of patient medical Clinic, Rochester, Minnesota, provided statistical support.
records to verify accuracy but were unable to directly
review every chart. ORCID
Despite the above limitations, our study has several Sandeep Pagali https://orcid.org/0000-0002-0838-1026
strengths including its large sample size, a robust data-
base maintained at a tertiary care level 1 trauma center RE FER EN CES
with diverse types of injury, and propensity matching 1. Jiang L, Zheng Z, Zhang M. The incidence of geriatric trauma
that minimizes variability across comparison groups. is increasing and comparison of different scoring tools for the
Most significantly, we had the ability to compare groups prediction of in-hospital mortality in geriatric trauma patients.
World J Emerg Surg. 2020;15(1):59. doi:10.1186/s13017-020-
both during the same period and in a pre–post analysis,
00340-1
which is a unique feature in this study compared to pre- 2. Hashmi A, Ibrahim-Zada I, Rhee P, et al. Predictors of
vious studies. mortality in geriatric trauma patients: a systematic review and
We believe there is a clear need for prospective stud- meta-analysis. J Trauma Acute Care Surg. 2014;76(3):894-901.
ies comparing older adult trauma patients who receive doi:10.1097/TA.0b013e3182ab0763
 15325415, 0, Downloaded from https://agsjournals.onlinelibrary.wiley.com/doi/10.1111/jgs.18977 by Readcube (Labtiva Inc.), Wiley Online Library on [25/05/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
GERIATRICS CONSULT IN OLDER ADULTS WITH TRAUMA 9

3. The American College of Surgeons. ACS TQIP Geriatric 16. Lenartowicz M, Parkovnick M, McFarlan A, et al. An evaluation
Trauma Management Guidelines [On-line]. Accessed January of a proactive geriatric trauma consultation service. Ann Surg.
4, 2024. https://www.facs.org/media/314or1oq/geriatric_ 2012;256(6):1098-1101. doi:10.1097/SLA.0b013e318270f27a
guidelines.pdf 17. Eagles D, Godwin B, Cheng W, et al. A systematic review and
4. Olufajo OA, Tulebaev S, Javedan H, et al. Integrating geriatric meta-analysis evaluating geriatric consultation on older trauma
consults into routine care of older trauma patients: one-year patients. J Trauma Acute Care Surg. 2020;88(3):446-453. doi:10.
experience of a level I trauma center. J Am Coll Surg. 2016; 1097/TA.0000000000002571
222(6):1029-1035. doi:10.1016/j.jamcollsurg.2015.12.058 18. Institute for Healthcare Improvement. Age-Friendly Health Sys-
5. Dugan JP, Burns KM, Baldawi M, Heidt DG. Impact of geriat- tems [On-line]. Accessed January 16, 2024. https://ihi.org/
ric consultations on clinical outcomes of elderly trauma initiatives/age-friendly-health-systems
patients: a retrospective analysis. Am J Surg. 2017;214(6):1048- 19. Marcantonio ER. Delirium in hospitalized older adults. N Engl
1052. doi:10.1016/j.amjsurg.2017.09.001 J Med. 2017;377(15):1456-1466. doi:10.1056/NEJMcp1605501
6. Southerland LT, Gure TR, Ruter DI, Li MM, Evans DC. Early 20. Sud M, Yu B, Wijeysundera HC, et al. Associations between
geriatric consultation increases adherence to TQIP geriatric short or long length of stay and 30-day readmission and mortal-
trauma management guidelines. J Surg Res. 2017;216:56-64. ity in hospitalized patients with heart failure. JACC Heart Fail.
doi:10.1016/j.jss.2017.03.023 2017;5(8):578-588. doi:10.1016/j.jchf.2017.03.012
7. Bradburn EH, Gross BW, Jammula S, Adams WH, Miller JA, 21. Baker SP, O'Neill B. The injury severity score: an update. J Trauma.
Rogers F. Improved outcomes in elderly trauma patients with 1976;16(11):882-885. doi:10.1097/00005373-197611000-00006
the implementation of two innovative geriatric-specific 22. Charlson M, Szatrowski TP, Peterson J, Gold J. Validation of a
protocols—final report. J Trauma Acute Care Surg. 2018;84(2): combined comorbidity index. J Clin Epidemiol. 1994;47(11):
301-307. doi:10.1097/TA.0000000000001752 1245-1251. doi:10.1016/0895-4356(94)90129-5
8. Sharma J, Parulekar M, Stewart P, et al. Geriatric consultation 23. Lingsma HF, Bottle A, Middleton S, Kievit J, Steyerberg EW,
reduces high-risk medication usage at discharge in elderly trauma Marang-van de Mheen PJ. Evaluation of hospital outcomes: the
patients. Cureus. 2018;10(11):e3649. doi:10.7759/cureus.3649 relation between length-of-stay, readmission, and mortality in
9. Roberts ARA, Falank CR, Ontengco JB, Carter EL, Hallen SAM. a large international administrative database. BMC Health Serv
Collaborative care when older adults fall: the benefits of geriatric Res. 2018;18(1):116. doi:10.1186/s12913-018-2916-1
consultation for trauma patients aged 75 years and older. J Am 24. Negm A, Bhandari M. Cochrane in CORR(R): comprehensive
Geriatr Soc. 2022;70(4):1284-1286. doi:10.1111/jgs.17637 geriatric assessment for older people admitted to a surgical ser-
10. Ibitoye SE, Braude P, Carter B, et al. Geriatric assessment is vice. Clin Orthop Relat Res. 2020;478(7):1409-1415. doi:10.1097/
associated with reduced mortality at 1 year for older adults CORR.0000000000001326
admitted to a major trauma center: a prospective observational 25. Mate KS, Berman A, Laderman M, Kabcenell A, Fulmer T.
study. Ann Surg. 2023;277(2):343-349. doi:10.1097/SLA. Creating age-friendly health systems—a vision for better care
0000000000005092 of older adults. Healthc (Amst). 2018;6(1):4-6. doi:10.1016/j.
11. Ellis G, Whitehead MA, Robinson D, O'Neill D, Langhorne P. hjdsi.2017.05.005
Comprehensive geriatric assessment for older adults admitted 26. Grund S, Roos M, Duchene W, Schuler M. Treatment in a cen-
to hospital: meta-analysis of randomised controlled trials. BMJ. ter for geriatric traumatology. Dtsch Arztebl Int. 2015;112(7):
2011;343:d6553. doi:10.1136/bmj.d6553 113-119. doi:10.3238/arztebl.2015.0113
12. Deschodt M, Flamaing J, Haentjens P, Boonen S, Milisen K. 27. Baroni M, Serra R, Boccardi V, et al. The orthogeriatric coman-
Impact of geriatric consultation teams on clinical outcome in agement improves clinical outcomes of hip fracture in older
acute hospitals: a systematic review and meta-analysis. BMC adults. Osteoporos Int. 2019;30(4):907-916. doi:10.1007/s00198-
Med. 2013;11:48. doi:10.1186/1741-7015-11-48 019-04858-2
13. Bradburn E, Rogers FB, Krasne M, et al. High-risk geriatric proto-
col: improving mortality in the elderly. J Trauma Acute Care Surg.
2012;73(2):435-440. doi:10.1097/TA.0b013e31825c7cf4
14. Mangram AJ, Mitchell CD, Shifflette VK, et al. Geriatric How to cite this article: Tyree S, Fischer K,
trauma service: a one-year experience. J Trauma Acute Care Stephens D, Burton MC, Pagali S. Impact of
Surg. 2012;72(1):119-122. doi:10.1097/TA.0b013e318241f0ba inpatient geriatrics consultation on hospital
15. Fallon WF Jr, Rader E, Zyzanski S, et al. Geriatric outcomes
outcomes in older adults with trauma. J Am
are improved by a geriatric trauma consultation service.
J Trauma. 2006;61(5):1040-1046. doi:10.1097/01.ta.0000238652.
Geriatr Soc. 2024;1‐9. doi:10.1111/jgs.18977
48008.59