Evidence Table

J o u r n a l o f S u r g i c a l R e s e a r c h • m o n t h 2 0 2 1 ( 2 6 4 ) 1 4 9 – 1 5 7

Available online at www.sciencedirect.com

j o u r n a l h o m e p a g e : w w w . e l s e v i e r. c o m / l o c a t e / Y J S R E

Comparison of Geriatric Versus Non-geriatric Trauma Patients With Palliative Care Consultations ✩ , ✩✩

Joanna Wycech, MS,a , c Alexander A Fokin, MDPhD,a , b , ∗

Jeffrey K. Katz, MD,a , b Sari Viitaniemi, RNMSN,a Nicholas Menzione, MD,a

and Ivan Puente, MDFACS,a , b , c , d

a Division of Trauma and Critical Care Services, Delray Medical Center, Delray Beach, Florida b Department of Surgery, Florida Atlantic University, Charles E. Schmidt College of Medicine, Boca Raton, Florida c Division of Trauma and Critical Care Services, Broward Health Medical Center, Fort Lauderdale, Florida d Department of Surgery, Florida International University, Herbert Wertheim College of Medicine, Miami, Florida

a r t i c l e i n f o

Article history:

Received 8 October 2020

Revised 22 January 2021

Accepted 27 February 2021

Keywords:

Geriatric trauma

Palliative care

Palliative care consultations

Do-Not-Resuscitate orders

Mortality

Futile interventions

Propensity matched comparison

a b s t r a c t

Background: Palliative care in trauma patients is still evolving. The goal was to compare char-

acteristics, outcomes, triggers and timing for palliative care consultations (PCC) in geriatric

( ≥65 y.o.) and non-geriatric trauma patients . Materials and methods: Retrospective study included 432 patients from two level 1 trauma

centers who received PCC between December 2012 and January 2019. Non-geriatric ( n = 61) and geriatric ( n = 371) groups were compared for: mechanism of injury (MOI), Injury Sever- ity Score (ISS), Revised Trauma Score (RTS), Glasgow Coma Score (GCS), Do-Not-Resuscitate

(DNR) orders, futile interventions (FI), duration of mechanical ventilation (DMV), ICU admis-

sions, ICU and hospital lengths of stay (ICULOS; HLOS), timing to PCC, and mortality. Fur-

ther propensity matching (PM) analysis compared 59 non-geriatric to 59 Geriatric patients

matched by ISS, GCS, and DNR.

Results: Geriatric patients were older (85.2 versus 49.7), with falls as predominant MOI. Non-

geriatric patients comprised 14.1% of all patients with PCC and were more severely injured

than Geriatrics: with statistically higher ISS (24.1 versus 18.5), lower RTS (5.4 versus 7.0), GCS

(7.1 versus 11.5), with predominant MOI being traffic accidents, all P < 0.01. Non-Geriatrics

had more ICU admissions (96.7% versus 88.1%), longer ICULOS (10.2 versus 4.7 days), DMV

(11.1 versus 4.1 days), less DNR (57.4% versus 73.9%), higher in-hospital mortality (12.5% versus

2.6%), but double the time admission-PCC (11.3 versus 4.3 days) compared to Geriatrics, all

✩ This is an IRB approved study. ✩✩ This manuscript is not under consideration elsewhere, has not been published previously and will not appear online or in print in any media until a decision is made concerning publication in the Journal of Surgical Research.

∗ Corresponding author. Delray Medical Center, Division of Trauma and Critical Care Services 5352 Linton Boulevard Delray Beach, FL 33484.

E-mail address: [email protected] (A.A. Fokin). 0022-4804/© 2021 Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.jss.2021.02.017

150 J o u r n a l o f S u r g i c a l R e s e a r c h • m o n t h 2 0 2 1 ( 2 6 4 ) 1 4 9 – 1 5 7

P < 0.04. In PM comparison, despite same injury severity, Non-geriatrics had triple the time

to PCC, five times the HLOS of geriatrics, and more FI (25.4% versus 3.4%), all P < 0.001.

Conclusions: PCC remains underutilized in non-geriatric trauma patients. Despite higher in-

jury severity, non-geriatrics received more aggressive treatment, and had three times longer

time to PCC, resulting in higher rate of FI than in Geriatrics.

© 2021 Elsevier Inc. All rights reserved.

Introduction

Palliative care, as part of a comprehensive medical plan and in light of serious injury and possible limitation of curative treatment, is an underutilized approach in a trauma setting. Currently, guidelines on the involvement of palliative care in the management of trauma patients are still evolving. The in- tegration of palliative care consultations (PCC) into the care of trauma patients is challenging, as palliative care can be viewed as a treatment failure, with a misconception that it is mutually exclusive with curative treatment.1-3 Such ideas are exacerbated in a trauma setting due to a sudden onset of events and limited time for comprehensive decision mak- ing.1 , 3 , 4 As the population of Geriatric patients is projected to nearly double in the next forty years, and as trauma accidents annually affect over 1 million Geriatric Americans, constitut- ing a third of all trauma admissions, it is crucial to investigate treatment strategies as they compare to the younger trauma population.5 , 6 Geriatric trauma patients carry an increased risk of morbidity and mortality due to pre-existing conditions, decreased metabolic reserve, and attenuated stress response.7

Furthermore, the Do-Not-Resuscitate (DNR) status, which is more prevalent in the Geriatric population, may also affect the management approach.8 , 9

The aim of this study was to compare characteristics, out- comes, triggers and timing for PCC in geriatric ( ≥ 65 y.o.) and non-geriatric (18-64 y.o.) trauma patients. We hypothesized that younger trauma patients will be given a more aggressive treatment approach, and will recover significantly better than the older patients.

Methods

This retrospective cohort study was granted an Institutional Review Board approval and a waiver of informed consent.

Study population

We analyzed 432 patients who were admitted to two level 1 trauma centers between December 2012 and May 2019 and re- ceived a PCC during their hospital stay. A PCC was defined as an end-of-life consultation, and not as a pain management consultation. At both trauma centers the Consultative Open Model was implemented, with an ultimate authority over the initiation of PCC delegated to the trauma surgeon. Exclusion criteria were: age 16 and younger, and death within 24 h of admission to the trauma center.

Metrics

The 432 PCC patients were divided into two groups, based on age: the non-geriatric included 61 patients and the Geriatric included 371 patients. The non-geriatric and geriatric groups were compared in an analysis that included patient character- istics such as: age, race, gender, comorbidities; injury parame- ters such as: mechanism of injury (MOI), injury severity score (ISS), Revised Trauma Score (RTS), Glasgow Coma Score (GCS), traumatic brain injury (TBI) incidence; treatment options such as: packed red blood cells (PRBC) transfused within 24 h of ad- mission, Do-Not-Resuscitate (DNR) orders, futile interventions (FI), mechanical ventilation requirement, duration of mechan- ical ventilation (DMV), intensive care unit (ICU) admissions, ICU length of stay (ICULOS), Hospital LOS (HLOS), discharge disposition, and mortality. In addition, the triggers for PCC that were used at our institution were extracted from patients’ charts and analyzed. Comparison also included various timing aspects of the in-hospital stay, such as: time from admission to PCC, admission to in-hospital DNR, PCC to discharge, and PCC to death. FI were defined as tracheostomy (TCH) or percu- taneous endoscopic gastrostomy (PEG), as these interventions have been described to be the two most frequently performed as FI.10 All variables were identified via ICD-9, ICD-10 codes, and extracted from patients’ medical records.

Propensity score matching

To ensure comparability between the non-geriatric and geri- atric groups, we applied propensity score matching and used ISS, GCS, and DNR order status as covariates, since these variables affect initiation of PCC the most. The propensity matching (PM) was done one-to-one, without replacement, giving priority to exact matches, randomizing the selection for nonexact matches, and with a 0.02 caliper. Propensity score matching resulted in two groups of 59 patients for further comparison (non-geriatric PM group; geriatric PM group). Us- ing the above criteria, the program was unable to provide matches for two non-geriatric patients; therefore they were excluded from this analysis. In PM comparison, we evaluated all the same variables as in the original patient set.

Statistical analysis

Statistical analysis was performed using IBM SPSS Statistics software version 23.0 (IBM, Armonk, NY). The compari- son analyses included group characteristics and bivariate correlations. Chi-squared tests were used for categorical variables and independent samples t tests for variable means, based on adequate sample sizes and normal distribution.

W y c e c h e t a l • 151

Table 1 – Characteristics and outcomes of non-geriatric and geriatric patients with palliative care consultations.

Variable Non-geriatric PCC ( n = 61) Geriatric PCC ( n = 371) P value Age 49.7 85.2 < 0.001 ∗

Gender (% Male) 73.8% 56.6% 0.012 ∗

Race - - < 0.001 ∗

White 75.4% 95.1% -

Black 13.1% 1.9% -

Hispanic 3.3% 0.5% -

Other 8.2% 2.4% -

Mechanism of injury - - < 0.001 ∗

Fall 37.7% 87.9% -

High energy trauma 42.6% 11.1% -

Penetrating trauma 4.9% 0.0% -

Other 14.8% 1.1% -

Injury severity score 24.1 18.5 0.009 ∗

Glasgow coma score 7.1 11.5 < 0.001 ∗

Revised trauma score 5.4 7.0 < 0.001 ∗

Traumatic brain injury (%) 73.8% 67.7% 0.341

Packed red blood cells transfused ≤24 h (%) 34.4% 11.3% < 0.001 ∗

Futile interventions (%) 24.6% 6.5% < 0.001 ∗

Mechanical ventilation requirement (%) 88.5% 55.8% < 0.001 ∗

Duration of mechanical ventilation [d] 11.1 d 4.1 d < 0.001 ∗

ICU admissions (%) 96.7% 88.1% 0.044 ∗

ICU length of stay [d] 10.2 d 4.7 d < 0.001 ∗

Hospital length of stay [d] 22.2 d 5.8 d < 0.001 ∗

6 mo mortality (%) 78.7% 92.7% < 0.001 ∗

In-hospital mortality (%) 12.5% 2.6% 0.005 ∗

Hospice discharge (%) 66.2% 90.1% < 0.001 ∗

Time admission-PCC [d] 11.3 d 4.3 d < 0.001 ∗

Time Admission-Death [d] 13.7 d 7.7 d 0.088

Time PCC-death [d] 5.2 d 3.4 d 0.084

Time admission to In-hospital DNR [d] 7.6 d 3.6 d 0.003 ∗

PCC = Palliative Care Consultation; DNR = Do-Not-Resuscitate; ICU = Intensive Care Unit. ∗ Denotes a statistically significant difference.

Multiple regression analysis was performed for the mortality outcome. Statistical significance was assumed when the calculated P value was below 0.05.

Results

Characteristics

Out of all patients who received a PCC, non-geriatric patients accounted for 14.1%. Characteristics of the non-geriatric and geriatric groups are presented in Table 1 . Geriatric patients were almost twice the age of non-geriatric patients, they were more often White race, and their injury resulted from falls sig- nificantly more often. Non-geriatric compared to Geriatric pa- tients had significantly higher ISS, lower GCS, lower RTS, and required more PRBC transfusions within the first 24 h of ad- mission. The rate of TBI was not statistically different between the groups. Geriatric patients had a statistically different dis-

tribution of DNR orders, with significantly more preadmis- sion and in-hospital DNR orders signed, which is presented in Fig. 1 . The non-geriatric group had a statistically significant, 2.6 times longer time from admission to PCC, and two times longer time from admission to in-hospital DNR signing than the Geriatric group, despite being more severely injured.

The top three comorbidities in the non-geriatric group in- cluded: hypertension (26.2%), alcoholism (24.6%), and diabetes (16.4%), with 39.3% of that group having no comorbidities. In the Geriatric group only 4.9% had no comorbidity, and the most commonly present comorbidities included: hyperten- sion (72.2%), dementia (23.5%), and preinjury anticoagulation use (23.2%).

Management and outcomes

Non-geriatric patient’s required ICU admission and mechani- cal ventilation more often, stayed on the ventilator, in the ICU,

152 J o u r n a l o f S u r g i c a l R e s e a r c h • m o n t h 2 0 2 1 ( 2 6 4 ) 1 4 9 – 1 5 7

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%

100%

Geriatric Non-Geriatric

12.9% 1.6%

60.1%

55.7%

26.1% 42.6%]

%[ noitubirtsi

D R

N D

Pre-Admission DNR In-Hospital DNR No DNR

p=0.004*

Fig. 1 – Do-Not-Resuscitate order distribution in geriatric and non-geriatric PCC patients.

and in the hospital significantly longer and had FI done four times more often than the Geriatric patients ( Table 1 ).

There were a total of 54 FI performed in 39 patients (9.0%), which included 25 TCH and 29 PEG placements. In the non- geriatric group, a quarter (24.6%) of patients got a FI, while in the Geriatric group that rate was significantly lower, at 6.5%. When it comes to timing of all FI, within the non-geriatric group 18 FI were done before and seven were done after a PCC, while in the Geriatric cohort 22 FI were done before and seven were done after a PCC.

In-hospital mortality was statistically higher in the Non- Geriatric group, while discharge to hospice was more com- mon (in over 90%) among the Geriatric patients. The overall 6 month mortality rate was statistically higher in the Geri- atric group. The time from PCC to death was not statistically different. A multiple regression analysis for mortality showed that age and DNR order status were independent predictors of mortality, regardless of ISS and GCS. The mortality timing graphs, Figures 2 and 3 , show that the majority of mortalities occurred in the first 10 days of admission. The two graphs fol- low the same tendency in the original Non-Geriatric and Geri- atric comparison, as well as in the propensity matched com- parison.

Commonly used triggers for PCC are presented in Table 2 . The most common trigger for PCC in both groups was “Catas- trophic Injury,” present in 63.9% of non-geriatric and 53.4% of Geriatric patients. The next top two triggers in the non- geriatric group were: “GCS ≤5” (55.7%) and “≥7 d ICU stay with- out improvement” (37.7%). In the geriatric group the next top two triggers were: “≥90 years old” (32.9%) and “family request for palliative care” (19.4%).

Propensity score matching

After propensity matching, the two groups of 59 pa- tients had similarly high ISS, low GCS, and similar dis- tribution of signed DNR orders. The comparison of in- jury characteristics and outcomes of non-geriatric PM

and geriatric PM groups is presented in Table 3 . Results showed that there was no longer statistical difference in gender, blood transfusion requirements, ICU admis- sions, mechanical ventilation requirement, and in-hospital mortality. However, non-geriatric PM patients had 2.6 times longer ICU length of stay, 3 times longer duration of mechan- ical ventilation, 7.5 times higher rate of FI, 3.4 times longer time from admission to PCC, and 3.7 times longer time from admission to in-hospital DNR signing than the Geriatric PM patients, with all P < 0.02. The top two PCC triggers in both groups were the same, with “Catastrophic Injury” remaining the most common (62.7% versus 69.5%, P = 0.437) followed by “GCS ≤5” (54.2% versus 55.9%, P = 0.853).

Discussion

As reported in the literature, the rates of PCC in trauma pa- tients have been steadily increasing over the last few years, from less than 1% of elderly trauma patients receiving PCC in 2006, to 4.3% in 2011.11 The PCC rates at our institutions followed a similar tendency with a gradual increase, between 2012 and 2019 from 1.6% to 6.8%, with an average of 4.3% of annual trauma admissions receiving a PCC.12 It was shown before that PCC patients tend to be older, with higher ISS and mortality rate, as compared to those who did not re- ceive a PCC.6 , 13 However, it is advanced age rather than in- jury or illness severity that appears to be the most consis- tent variable in transitioning to a palliative type of care in a trauma ICU.1 The analysis of 432 trauma patients at our insti- tutions showed that geriatric patients prevail in a PCC cohort, with only 14.1% of PCC patients being non-geriatric. In our study, we aimed to compare the characteristics, differences in the management approach, and resulting outcomes between non-geriatric and geriatric patients who received a PCC. Non- geriatric PCC patients were most often male, who endured high energy trauma, which resulted in significantly higher ISS, RTS, and lower GCS as compared to Geriatric patients. Low en-

W y c e c h e t a l • 153

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

0-5 6-10 11-15 16-20 21-25 26-30 >30

) %(

stneita P

deripx E

Days from Admission

Non-Geriatric Mortality (n=48) Geriatric Mortality (n=344)

Fig. 2 – Mortality timing from admission to the trauma center for non-geriatric and geriatric patients with palliative care consultation.

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

0-5 6-10 11-15 16-20 21-25 26-30 >30

) %(

stneita P

deripx E

Days from Admission

Non-Geriatric Mortality (n=46) Geriatric Mortality (n=58)

Fig. 3 – Mortality timing from admission to the trauma center for non-geriatric and geriatric patients with palliative care Consultation after propensity matching by ISS, GCS and DNR.

ergy trauma, such as falls from standing, was the prevailing mechanism of injury in our Geriatric group, which is in agree- ment with previous studies.14 TBI prevalence in our cohort, was 73.8% in the non-geriatric and 67.7% in the geriatric group, and were almost identical with previously reported data.13

That difference was not statistically significant in the full co- hort comparison, and did not change in our propensity match analysis, further confirming the comparability of our propen- sity matched groups.

Almost all (96.7%) of our non-geriatric patients required an ICU admission, and their average ICU stay was twice as long as

the average stay of geriatric patients. Our non-geriatric group also required mechanical ventilation statistically more often than the geriatric patients, which together with the ICU ad- missions can be attributed to their significantly higher ISS. The reported estimations are that up to 20.0% of ICU admis- sions meet criteria for PC consultations.15

The mean hospital length of stay in the non-geriatric PCC group was statistically longer at 22.2 d compared to 5.8 d in the geriatric group. The same, statistically significant difference can be seen in the propensity matched cohort, where non- geriatric PM had a mean hospital length of stay of 22.7 d, and

154 J o u r n a l o f S u r g i c a l R e s e a r c h • m o n t h 2 0 2 1 ( 2 6 4 ) 1 4 9 – 1 5 7

Table 2 – Commonly Used Triggers for PCC in Non-Geriatric and Geriatric Groups.

Triggers Non-Geriatric PCC ( n = 61) Geriatric PCC ( n = 371) P value Catastrophic injury 63.9% 53.4% 0.124

≥ 90 y-o 0.0% 32.9% < 0.001 ∗

Glasgow coma score ≤5 55.7% 17.3% < 0.001 ∗

ICU stay ≥7 days without improvement 37.7% 15.6% < 0.001 ∗

Patient or family request for palliative care 13.1% 19.4% 0.241

Feeding tube 31.1% 14.6% 0.001 ∗

> 21 days in hospital 27.9% 3.8% < 0.001 ∗

Presence of advanced directives and complexity of decision making 23.0% 17.5% 0.309

Counseling need regarding prognosis and goals of care 21.3% 16.2% 0.321

Depression / Uncontrolled psychosocial issues 14.8% 6.7% 0.031 ∗

Significant weight loss / Cachexia 14.8% 4.6% 0.002 ∗

End stage dementia with inability to communicate 13.1% 13.2% 0.984

Requirement for transfer to ventilator facility 13.1% 7.3% 0.122

> 2 admissions from long term care facility / frequent ER visits for the same diagnosis in 3 months

11.5% 8.1% 0.381

Septicemia 9.8% 6.5% 0.341

Metastatic cancer 9.8% 5.4% 0.176

Patient / family questions about hospice qualifications 6.6% 10.0% 0.399

Multi-system organ failure 6.6% 2.2% 0.053

End stage HIV not responding to antiviral treatment 6.6% 0.0% < 0.001 ∗

Two or more resuscitations during the hospital stay 4.9% 2.2% 0.204

Pain management in patient with life threatening disease 4.9% 1.6% 0.094

Heart failure on maximum therapy and symptomatic 3.3% 17.3% 0.005 ∗

Dyspnea uncontrolled by standard therapy 3.3% 2.4% 0.695

Stroke 1.6% 8.1% 0.071

COPD with chronic dyspnea at rest 1.6% 5.1% 0.230

Long-term care resident with Do Not Resuscitate orders 1.6% 5.1% 0.230

PCC = Palliative Care Consultation; ICU = Intensive Care Unit; ER = Emergency Room; HIV = Human Immunodeficiency Virus; COPD = Chronic Obstructive Pulmonary Disease.

∗ Denotes a statistically significant difference.

Geriatric PM patients’ of 4.3 d, having similar in-hospital mor- tality in both groups. In a trauma registry analysis of 124,000 trauma patients the reported median hospital length of stay was 4 d, with a mean of 6 d.16

PCC timing

Previous studies reported that the median time from admis- sion to PCC can vary between 2.9 and 4.5 d, but that most of the consultations are done in the last 24 h of patient’s life.6 , 17

In our non-geriatric group the average time from admission to PCC was 11.3 d and from admission to death was 13.7 d; and in the Geriatric group were 4.3 d and 7.7 d, respectively, showing that Geriatric patients got a PCC faster ( P < 0.001), but did not die earlier ( P = 0.09). This coupled with our finding that the ISS was significantly higher in the non-geriatric group than in the Geriatric group shows that age is a strong factor in the decision to order a PCC. It was further corroborated in our propensity matched comparison, where despite being similarly injured, the Non-Geriatric PM patients again had statistically longer time to PCC, 11.4 versus 3.4 d ( P < 0.001).

Models of Palliative care and triggers

The perceived benefits of using a PCC in severely injured trauma patients have been noted as a reduction of non- beneficial resources and establishment of end of life goals.6 , 18

It has also been recommended that palliative care for both the patients and the families after trauma can be useful to assist in decision making and pain management.18 There are currently two main models of PCC integration into patients’ care. In the “consultative model” a PCC is done in patients with a high risk of a poor outcome by consultation initiated by the attending surgeon, while in the “integrated model” pal- liative principles and interventions are embedded into daily ICU practice in all patients facing critical illness.3 , 19

At our institutions a “consultative model” for PCC is used, which incorporates the most commonly utilized trigger method.20 However, the final decision for a PCC is based on the attending trauma surgeon’s opinion.

The analysis of triggers for PCC among our patients showed the most common triggers to include: the brain condition and cognitive ability (“GCS ≤5” and “Dementia”); the lack of

W y c e c h e t a l • 155

Table 3 – Characteristics and outcomes of non-geriatric and geriatric patients with palliative care consultations after propensity matching by ISS, GCS and DNR.

Variable Non-geriatric PCC ( n = 59) Geriatric PCC ( n = 59) P value Injury severity score 23.3 22.1 0.663

Glasgow coma score 7.2 7.1 0.896

DNR status 57.6% 61.0% 0.931

Pre-admission DNR 1.7% 1.7% -

In-hospital DNR 55.9% 59.3% -

No DNR 42.4% 39.0% -

Age 50.5 81.8 < 0.001 ∗

Gender (% Male) 74.6% 59.3% 0.078

Race - - 0.059

White 76.3% 91.5% -

Black 11.9% 1.7% -

Hispanic 3.4% 0.0% -

Other 8.5% 6.8% -

Mechanism of injury - - 0.001 ∗

Falls 39.0% 74.6% -

High energy trauma 40.7% 20.3% -

Penetrating trauma 5.1% 0.0% -

Other 15.3% 5.1% -

Revised trauma score 5.5 5.6 0.760

Traumatic brain injury (%) 72.9% 69.5% 0.684

Packed red blood cells transfused ≤24 h (%) 32.2% 18.6% 0.091 Futile interventions (%) 25.4% 3.4% 0.001 ∗

Mechanical ventilation requirement (%) 88.1% 79.7% 0.210

Duration of mechanical ventilation [d] 11.3 days 3.8 days < 0.001 ∗

ICU admissions (%) 96.6% 93.2% 0.402

ICU length of stay [d] 10.3 days 4.0 days < 0.001 ∗

Hospital length of stay [d] 22.7 days 4.3 days < 0.001 ∗

6 mo mortality (%) 78.0% 98.3% 0.001 ∗

In-hospital mortality (%) 8.5% 6.8% 0.729

Hospice disposition (%) 71.2% 89.8% 0.011 ∗

Time admission-PCC [d] 11.4 days 3.4 days < 0.001 ∗

Time admission-death [d] 10.0 days 6.2 days 0.023 ∗

Time PCC-death [d] 5.3 days 2.8 days 0.029 ∗

Time admission to in-hospital DNR [d] 9.9 days 2.7 days 0.014 ∗

PCC = Palliative Care Consultation; DNR = Do-Not-Resuscitate; ICU = Intensive Care Unit. ∗ Denotes a statistically significant difference;

improvement in function; and the presence of advanced di- rectives.

Do-Not-Resuscitate orders and FI in PCC patients

The presence of DNR orders, especially in geriatric patients with multiple comorbidities can be a factor affecting manage- ment and outcomes in severely injured patients. Preadmission and in-hospital DNR orders may significantly influence the treatment strategy, facilitate shift from curative to supportive care, trigger the use of palliative care services, and affect dis- charge disposition and ultimately mortality.14 The previously reported prevalence of DNR orders in hospitalized trauma pa- tients was between 0.0% and 16.5%.8 , 21-24 In our comparison

of non-geriatric and geriatric PCC groups, we found that non- geriatric patients had a significantly lower rate of signed DNR orders, with 42.6% of patients with no DNR prior or throughout their hospital stay, as compared to 26.1% in the geriatric group. In-hospital DNR was signed by 55.7% of non-geriatric patients, on average 7.6 d after admission, while in the geriatric group it was signed by 60.9% of patients, on average 3.6 d after admis- sion. In the propensity match comparison, when the presence of a DNR order was controlled, we still saw in-hospital DNR or- ders being signed almost three times later from admission in non-geriatric PM at 9.9 d compared to 2.7 d in Geriatric PM .

In our study 9.0% of PCC patients received a FI, with a sig- nificantly higher rate of non-geriatric than geriatric patients getting a FI during their stay. The significant difference in FI

156 J o u r n a l o f S u r g i c a l R e s e a r c h • m o n t h 2 0 2 1 ( 2 6 4 ) 1 4 9 – 1 5 7

rates was also confirmed in the propensity matched cohorts, after controlling for injury severity and DNR order status. In both age groups there were at least two times more FI done before PCC than after PCC .

Mortality

Previous studies on trauma patients reported mortality rang- ing from 11.2% to 31.3%, with 7.2% of trauma patients dying within 1 y of trauma admission.13 , 25 ,26 It was also reported that patients with GCS < 9 had the highest risk of death within 1 y at 44%. 26 The mortality rates in both of our PCC groups were far higher, at 78.7% in the non-geriatric and 92.7% in the geriatric groups, and these numbers included hospice mor- tality. The multiple regression analysis for mortality showed that age and DNR order status were independent predictors of mortality, regardless of ISS and GCS. In our study, 68.9% of patients in the geriatric PCC group were discharged to hospice, and 92.7% expired in hospice. With high post discharge mor- tality rates in severely injured trauma patients, it is justified to consider implementing palliative care as a crucial part of patients’ treatment strategy.

The mortality timing graphs, in Figures 2 and 3 , present time from admission to a level 1 trauma center to death. The majority of deaths in both the non-geriatric and geriatric groups occurred within the first 10 d of admission, which can be attributed to high injury severity. Due to the high mortality rate in the first 10 d of admission and knowing that PCC oc- curred at 4.3 d in the geriatric, but was delayed on average to 11.3 d in the non-geriatric patients, we recommend that non- geriatric severely injured patients get a PCC earlier in their course of stay, to mirror the approach to the geriatric popu- lation. The high rate of hospice discharge in our PCC patients shows that our two institutions have a sensible timing of PCC, compared to Rivet et al ., who reported that their palliative care was implemented too late, as half of their population died in the ICU/step down without being discharged to end-life care or hospice.17

Practical recommendations

The significant discrepancy of about 2 wk in hospital length of stay between the analyzed non-geriatric and geriatric pa- tients, together with the significantly longer stay in the ICU, significantly higher rate of FI, three times longer time from admission to PCC, and double the time from admission to DNR, shows that management of non-geriatric trauma pa- tients tends to be more aggressive and persistent, despite the high mortality in that group of 78.7%. Although the discharge disposition is different for non-geriatric than for geriatric pa- tients, a discharge to a skilled nursing facility or home with service with a poor prognosis of a full recovery does not war- rant the non-geriatric patient’s delay in receiving PCC and en- suring that their goals of care discussion takes place. A timely PCC is warranted for all severely injured trauma patients and should not be reserved only for the geriatric population. The recommended timely PCC based on a full evaluation of the patient is crucial in streamlining the goals of care, limits to aggressive management, hospital discharge options, and out- look on the financial burden.

Conclusions

Palliative care remains underutilized in non-geriatric trauma patients. Non-geriatric patients had a significantly longer stay in the intensive care unit and in the hospital, were four times more likely to receive a futile intervention, and waited for a palliative care consultation and a DNR order three times as long as Geriatric patients, showing a more aggressive man- agement approach to their care, despite being more severely injured. Even after controlling for injury severity and DNR or- der distribution, the more aggressive approach in treatment of younger patients resulted in significantly more futile inter- ventions, longer hospital stay, and less hospice discharge dis- position.

Author contributions

Joanna Wycech MS-was responsible for data acquisition from Trauma Database and statistical data analysis, drafting of manuscript and final approval; Alexander A Fokin MD, PhD – was the primary author who designed the study, analyzed obtained data and drew conclusions, and responsible for the writing of the manuscript; Jeffrey K. Katz MD – was a contrib- utor to the design of the study, aided with data acquisition, drafting of manuscript and final approval; Sari Viitaniemi RN, MSN, CCRN- was a contributor in collection of the data, de- sign of the study, final editions to the manuscript and pallia- tive care specialized opinions required for a revision; Nicholas Menzione MD – was a contributor to the design of the study, aided with data acquisition, drafting of manuscript and final approval; Ivan Puente MD, FACS- was the Principal Investiga- tor for the study, contributor to the design of the study, pro- vided input for and finalized the manuscript.

Acknowledgment

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Declaration of Competing Interest

None. The authors report no proprietary or commercial inter- est in any product mentioned or concept discussed in this ar- ticle.

references

1. Mosenthal AC , Murphy PA . Trauma care and palliative care: time to integrate the two. J Am Coll Surg . 2003;199:509–516 .

2. Baggs JG , Norton SA , Schmitt MH , et al. Intensive care unit cultures and end-of-life decision making. J Crit Care . 2007;22:159–168 .

3. Mosenthal AC , Weissman DE , Curtis JR , et al. Integrating palliative care in the surgical and trauma intensive care unit: a report from the Improving Palliative Care in the Intensive Care Unit (IPAL-ICU) Project Advisory Board and the Center to Advance Palliative Care. Crit Care Med . 2012;40:1199–1206 .

W y c e c h e t a l • 157

4. Karlekar M , Collier B , Parish A , et al. Utilization and determinants of palliative care in the trauma intensive care unit: results of a national survey. Palliat Med . 2014;28:1062–1068 .

5. Zafar SN . Trauma care does not discriminate: the association of race and health insurance with mortality following traumatic injury. J Trauma Acute Care Surg . 2015;79:515–516 .

6. Kupensky D , Hileman BM , Emerick ES , Chance EA . Palliative medicine consultation reduces length of stay, improves symptom management, and clarifies advance directives in the geriatric trauma population. J Trauma Nurs . 2015;22:261–265 .

7. Olufajo OA , Tulebaev S , Javedan H , et al. Integrating Geriatric Consults into Routine Care of Older Trauma Patients: One-Year Experience of a Level I Trauma Center. J Am Coll Surg . 2016;222:1029–1035 .

8. Fokin AA , Wycech J , Katz JK , et al. Palliative care consultations in trauma patients and role of do-not-resuscitate orders: propensity-matched study. Am J Hosp Palliat Care . 2020;37:1068–1075 .

9. Wooster M , Stassi A , Hill J , Kurtz J , Bonta M , Spalding MC . End-of-life decision-making for patients with geriatric trauma cared for in a trauma intensive care unit. Am J Hosp Palliat Care . 2018;35:1063–1068 .

10. Hoffman MR . Tracheostomies and PEGs: when are they really indicated? Surg Clin North Am . 2019;99:955–965 .

11. Lilley EJ , Lee KC , Scott JW , et al. The impact of inpatient palliative care on end-of-life care among older trauma patients who die after hospital discharge. J Trauma Acute Care Surg . 2018;85:992–998 .

12. Wycech J , Fokin AA , Katz JK , et al. Reduction in futile interventions in trauma patients following a palliative care consultation. J Palliat Med . 2020;23 .

13. Toevs CC , Philp FH . The benefit of palliative medicine in trauma. J Am Coll Surg . 2014;219:S49 .

14. Jawa RS , Shapiro MJ , McCormack JE , Huang EC , Rutigliano DN , Vosswinkel JA . Preadmission Do Not Resuscitate advanced

directive is associated with adverse outcomes following acute trauma injury. Am J Surg . 2015;210:814–821 .

15. Hua MS , Li G , Blinderman CD , et al. Estimates of the need for palliative care consultation across united states intensive care units using a trigger-based model. Am J Respir Crit Care Med . 2014;189:428–436 .

16. Davidson GH , Hamlat CA , Rivara FP , Koepsell TD , Jurkovich GJ , Arbabi S . Long-term survival of adult trauma patients. JAMA . 2011;305:1001–1007 .

17. Rivet EB , Ferrada P , Albrecht T , et al. Characteristics of palliative care consultation at an academic level one trauma center. Am J Surg . 2017;214:657–660 .

18. Owens D . The palliative care in trauma. Crit Care Nurs Q . 2012;35:223–227 .

19. O’Connell K , Maier R . Palliative care in the trauma ICU. Curr Opin Crit Care . 2016;22:584–590 .

20. Nelson JE , Bassett R , Boss RD , et al. Models for structuring a clinical initiative to enhance palliative care in the intensive care unit: a report from the IPAL-ICU Project (Improving Palliative Care in the ICU). Crit Care Med . 2010;38:1765–1772 .

21. Marco CA , Michael S , Bleyer J , Post A . Do-not-resuscitate orders among trauma patients. Am J Emerg Med . 2015;33:1770–1772 .

22. Salottolo K , Offner PJ , Orlando A , et al. The epidemiology of do-not-resuscitate orders in patients with trauma: a community level one trauma center observational experience. Scand J Trauma Resusc Emerg Med . 2015;23:29 .

23. Wade CE , del Junco DJ , Fox EE , et al. Do-not-resuscitate orders in trauma patients may bias mortality-based effect estimates: an evaluation using the PROMMTT study. J Trauma Acute Care Surg . 2013;75(Suppl 1):S89–S96 .

24. Sise MJ , Sise CB , Thorndike JF , et al. Withdrawal of care: a 10-year perspective at a Level I trauma center. J Trauma Acute Care Surg . 2012;72:1186–1193 .

25. Michetti CP , Fakhry SM , Brasel K , et al. Trauma ICU prevalence project: the diversity of surgical critical care. Trauma Surg Acute Care Open . 2019;4 .