Cervical Spine Immobilization After Penetrating Trauma to the Head and Neck Kimberly J. Wilkins, MD; Christian Sommerhalder, MD; Taylor R. Klein, BS; Sebastian D. Schubl MD; Vanessa P. Ho, MD, MPH

Department of Surgery, Jamaica Hospital Medical Center, Jamaica, NY

BACKGROUND C-collars May Result In Adverse Effects: • Skin breakdown • Elevated intracranial pressures • Increased ventilator time • Increased intensive care unit days • Longer hospital stays Prehospital Trauma Life Support Recommendation: “Spinal immobilization may be performed after penetrating injury when a focal neurologic deficit is noted on physical examination although there is little evidence of benefit even in these cases.”

HYPOTHESIS Patients with stab wounds (SW) will have a lower incidence of cervical spinal column injury than patients with gunshot wounds (GSW).

Retrospective analysis of all patients who presented to an urban trauma center between January 2010 - January 2014 with a penetrating injury to the head or neck. Variables Collected:

RESULTS

CONCLUSIONS • Of patients with a GSW to the head and/or neck that survived to be evaluated, 5 (13.8%) had CS fracture. Cervical spine immobilization is appropriate in this population. • Of patients with a SW to the head and/or neck that survived to be evaluated, 1 (0.8%), had a fracture after a stab wound. The patient showed no neurological deficit, had no spinal cord injury, and received no treatment for the fracture. • Further research may allow providers to forego cervical spine immobilization in patients with stab wounds to the head and/or neck.

REFERENCES 1. Como JJ, Diaz JJ, Dunham CM, Chiu WC, Duane TM, Capella JM,

Holevar MR, Khwaja KA, Mayglothling JA, Shapiro MB, et al. Practice management guidelines for identification of cervical spine injuries following trauma: update from the eastern association for the surgery of trauma practice management guidelines committee. J Trauma. 2009 Sep;67(3):651-9.

2. Stuke LE, Pons PT, Guy JS, Chapleau WP, Butler FK, McSwain NE. Prehospital spine immobilization for penetrating trauma--review and recommendations from the Prehospital Trauma Life Support Executive Committee. J Trauma. 2011 Sep;71(3):763-9.

3. Powers J, Daniels D, McGuire C, Hilbish C. The incidence of skin breakdown associated with the use of cervical collars. J Trauma Nurs. 2006;13:198-200.

4. Hunt K, Hallworth S, Smith M. The effects of rigid collar placement on intracranial and cerebral perfusion pressures. Anaesthesia. 2001;56:511-513.

5. Haut ER, Kalish BT, Efron DT, Haider AH, Stevens KA, Kieninger AN, Cornwell EE 3rd, Chang DC. Spine immobilization in penetrating trauma: more harm than good? J Trauma. 2010 Jan;68(1):115-20.

6. Hunt K, Hallworth S, Smith M. The effects of rigid collar placement on intracranial and cerebral perfusion pressures. Anaesthesia. 2001;56:511-513.

Characteristic N=172 % Male Gender 153 89% Age (years, SD) 34.5 ±14.9

Mechanism GSW 48 28% Stab 124 72% Location Head 84 49% Neck 106 62% Head and Neck

19 11%

172 patients had penetrating injury to the head and/or neck: Clinical Outcomes N=172 %

Mortality 24 14% Mortality prior to imaging or clinical evaluation 16 9%

C-spine evaluated and fracture identified 6 4%

Fracture after stab wound (n=120) 1 0.8%

Fracture after GSW (n=36) 5 13.8%

Patient Injuries and Clinical Descriptions

Initi

al G

CS

Initi

al /

Dis

char

ge

Neu

ro

Exam

Trea

tmen

t

Mor

talit

y

1 GSW to face • C1 transverse process fracture, air in spinal canal. • Pseudoaneurysm of the carotid artery • Transferred for neurointerventional radiology

3 Unable

/ Unknown

None No

2

GSWs to face and neck • Right C1 vertebral arch fracture, vertebral artery spasm • Discharged to home with Halo, with no neurologic deficit

15 Normal

/ Normal

Halo No

3 GSW to left face, maxillary area • Displaced fracture to L transverse process of C1 • Expired on HD 0

3 Unable

/ Death

None Yes

4

GSW to zone II of neck • Fractures of C5 transverse process, pedicle, spinous process, and C6 superior facet • Expired after a protracted hospital course

3 Unable

/ Quad

Collar Yes

5

GSWs to neck and legs • Fractures of C2, C3, C4 transverse processes, spinous process of C4 and superior facet of C4 • MRI showed cord contusion at C3/C4 • Discharged to acute rehab

3

Unable /

UE & LE weakness

Collar No

6 Stab Wound to zone II of neck • Teardrop fracture at C6 • MRI showed normal spinal cord

15 Normal

/ Normal

None No

METHODS

• Patient demographics • Date of injury • Wound location • Loss of consciousness • Initial neurologic exam • Imaging Results:

• Vertebral fracture • Spinal cord injury

• Treatment (surgery, halo, collar, none)

• Discharge neurologic exam

• Mortality • Mortality prior to spinal

evaluation

Analysis: Penetrating wounds were stratified into stab wounds vs. gunshot wounds and compared using Pearson’s Chi squared test.

BACKGROUND Percutaneous Endoscopic Gastrostomy (PEG): • PEG tube placement is a safe and effective means to

provide feeding when oral intake is impaired. • Common indications for PEG tube placement include

head/neck trauma, head/neck cancer and neurological diseases.

• Despite its routine use PEG is associated with

complications and mortality. • Complications include: bleeding, wound infection, tube

dislodgement, necrotizing fasciitis, and buried bumper syndrome.

OBJECTIVE AND HYPOTHESIS The aim of this study was to identify risk factors for the development of post-PEG complications. We hypothesized that patients with low albumin levels, diabetes, thicker abdominal walls and psychomotor agitation would have more complications.

Retrospective analysis of all patients with an order for PEG tube placement during the period of December 2011 - November 2013. Variables Collected:

RESULTS

CONCLUSIONS • 18.7% of patients developed complications after PEG tube placement.

• Age, weight and abdominal wall thickness were identified as risk factors for complications.

• Younger patients (< 51.5 years) who weighed less (< 62.2 kg) and had thinner abdominal walls (< 20.2 mm) were more prone to develop complications.

• Abdominal wall thickness was the only independent risk factor

for complications.

• 14.3% of patients died within 30 days, however; mortality was due to severe underlying injuries and not PEG tube placement.

• There was no association between complications and mortality.

• Preoperative measurement of abdominal wall thickness by pre-procedural imaging can potentially be used to predict post-PEG complications.

METHODS

• Patient demographics • Indication for PEG • Diabetes diagnosis • Albumin levels • Psychomotor agitation • Abdominal wall thickness

(when CT scans were available)

• Anticoagulant use • Length of hospital stay • Complications • Date complication occurred • Mortality

Psychomotor agitation was defined as a need for 1:1 observation or restraints. Analysis: Patients were stratified into No Complications vs. Complications and compared using Fisher’s exact test.

Low abdominal wall thickness predicts percutaneous endoscopic gastrostomy complications. Melissa K. James, Ph. D.; Simon P. Tiu, MD; Richard J. Tom, MD; Vanessa P. Ho, MD, MPH; Taylor R. Klein, BS; Gloria M. Melnic, RN; Sebastian

D. Schubl, MD. Department of Surgery, Jamaica Hospital Medical Center, Jamaica, NY

Variable No Complication,

Total=74 Complication,

Total=17 P value

Age, years 60.7 (SD 17.1) 50.1 (SD 22.7) 0.03*

Male 51 (69.0%) 13 (76.5%) 0.77

Surgical Service 46 (62.2%) 13 (76.5%) 0.40

Trauma Activation 38 (51.3%) 8 (47%) 0.79

Albumin, g/dL 2.8 (SD 0.7) 3.2 (SD 0.9) 0.06

Weight, kg 78.6 (SD 21.6) 63.6 (SD 11.6) 0.01*

BMI, kg/m2 27.1 (SD 6.8) 23.7 (SD 7.6) 0.07

Diabetes 21 (28.4%) 5 (29.4%) 1.00 Abdominal wall thickness, mm

27.6 (SD 8.1) 21.6 (SD 7.6) 0.02*

PrePEG Anticoagulants 36 (48.6%) 6 (35.3%) 0.42

PrePEG Steroids 15 (20.3%) 5 (29.4%) 0.52

Patients on 1:1 observation 8 (10.8%) 4 (23.5%) 0.23

Patients on restraints 30 (40.5%) 11 (64.7%) 0.10

• 91 patients were identified (mean age 58.7, range 19-96, SD 18.6)

• 70.3 % were male, 29.7% were female • 17 patients developed post-PEG complications (18.7%) • The most common complications were surgical site infection,

tube dislodgement and peristomal leakage • The 30-day mortality rate was 14.3%

Table 1. Patient Outcomes

Table 3. Univariate analysis of patients with and without complications

Outcome Mean, Total # SD, %

Complication Rate (n=91) 17 18.7% Patients with 1 complication (n=17) 8 47.1% Patients with > 1 complication (n=17) 9 52.9% Postoperative time to Complication, days

Mean 26.8 Median 16

SD 47.3 Range 0-203

Mortality Rate (n=91) 17 18.7% In-hospital Mortality (n=91) 17 18.7% Death within 30 days (n=91) 13 14.3% Death within 60 days (n=91) 15 16.5% Death within 1 year (n=91) 16 17.6% Postoperative time to Mortality, days 51.8 SD 128.1 Length of Hospital Stay, days Mean 52.2

Median 32 SD 72.3

Range 9-547

Complication Number of patients %

Minor Complications 17/91 18.7

PEG site Infection 10/17 58.8

Peristomal Leakage 7/17 41.2

Tube Dislodgement 5/17 29.4

Tube Dysfunction 2/17 11.8

Minor Bleeding at PEG site 1/17 5.9

Pnuemoperitoneum 1/17 5.9

Central Fever 1/17 5.9

Major Complications 4/91 4.4

Bleeding at PEG site (required surgery) 1/17 5.9

Buried Bumper Syndrome 3/17 17.6

Table 2. PEG related Complications

Background

Ultra Low Dose CT Scanning: A Reliable and Effective Modality with an Improved Patient Safety Profile Sanjit R. Konda, MD; Abraham M. Goch, BS; Philipp Leucht, MD, PhD; Anthony Christiano, BA; Soterios Gyftopoulos, MD; Gideon Yoeli, MD; Kenneth A. Egol, MD

Jamaica Hospital Medical Center, Jamaica, Queens and NYU Hospital for Joint Diseases, New York, New York.

Methods Previously an ultra low dose CT (ULD-CT) protocol was developed to identify articular penetration of the knee joint.3–5

A novel low dose CT protocol was developed by altering multiple parameters and used to prospectively assess fracture patients. Ten orthopaedic fracture surgeons evaluated de-identified images for diagnosis, management, and image quality. The total radiation dose for each imaging type was determined and recorded.

Results • Ultra low dose CT scanning resulted in an estimated effective

dose (ED) 14x lower than that of conventional CT (C-CT) scanning.

• Mean ED for ULD-CT vs. C-CT was 0.03 milliSieverts (mSv) vs. 0.43 mSv (p<.005).

• Sensitivity (Sn), Specificity (Sp), Positive Predictive Value (PPV) and Negative Predictive Value (NPV) of ULD-CT scan to detect fractures was 0.98, 0.89, 0.98, and 0.89 with 2 occult fractures excluded.

• Reliability statistics between ULD-CT and C-CT assessments were comparable reflecting that the diagnostic value of CT technology does not decay with an ULD protocol.

Results

Figure 1. CT scan slices through the intra-articular space. Scan: A, Without air introduced; B, With 0.1 cc of air introduced into the lateral retropatellar space. C, Reduced dose (0.74 mGy) without air introduced; D, Reduced dose (0.74 mGy) with 0.1 cc of air introduced into the lateral retropatellar space.

Figure 2. Ultra low dose CT protocol displaying various adjusted parameters as listed.

Figure 3. Estimated effective radiation dosages for various extremity locations. C-CT= conventional CT scan; ULD-CT= ultra low dose CT scan.

Figure 4. Four ULD-CT images (left of gray bar as follows: top left- elbow, top right- knee, bottom left- ankle, bottom right- acetabulum) vs. corresponding C-CT images (right of gray bar). Nearly indistinguishable resolution with ULD-CT and with a 10x decrease in effective radiation dose.

Conclusions 1. Our ULD-CT protocol appears to represent a significant advancement in the refinement of an ubiquitous diagnostic tool.

2. This ultra low dose protocol appears to provide for high fidelity images in appropriately selected patients, with an improved safety profile over conventional CT. 3. By employing dose reduction strategies as demonstrated here, orthopaedic surgeons can address public concerns over radiation exposure in their medical care.

1.Berrington de González A, Mahesh M, Kim K-P, et al. Projected cancer risks from computed tomographic scans performed in the United States in 2007. Arch Intern Med. 2009;169(22):2071-2077. 2.Brenner DJ, Elliston CD, Hall EJ, Walter EB. Estimated Risks of Radiation-Induced Fatal Cancer from Pediatric CT. AJR Am J Roentgenol. 2001. Feb;176(2):289-96. 3.Brenner DJ, Hall EJ. Computed tomography--an increasing source of radiation exposure. N Engl J Med. 2007. 4.Konda S, Davidovitch R, Egol K. Open knee joint injuries: an evidence-based approach to management. Bull Hosp Joint Dis. 2014. 5.Konda SR, Howard D, Davidovitch RI, Egol KA. The Role of Computed Tomography in the Assessment of Open Periarticular Fractures Associated with Deep Knee Wounds. J Orthop Trauma. 2013;27:509-514.

References

0.009

0.005

0.010

0.248

0.538

Objective To determine whether or not a novel CT imaging protocol could produce similar diagnostic capability as compared to conventional CT scanning while minimizing radiation exposure for a given body part.

• Computerized tomography (CT) was invented in the early 1970s.

• 71.7 million CT scans were performed in 2007.1

• 600,000 head and abdominal CT scans are performed annually for children <15 years old.2

• It is projected that 500 of these individuals will die from cancer

attributable to this irradiation.2

Background

Predicting Mortality and Inpatient Complications: The Score for Trauma Triage in the Geriatric and Middle-Aged (STTGMA) Orthopaedic Trauma Patient

Sanjit R. Konda, MD; Kari J. Broder, BA; Sebastian Schubl, MD; Kenneth A. Egol, MD NYU Hospital for Joint Diseases, New York, New York and Jamaica Hospital Medical Center, Jamaica, Queens.

.

Methods • Inclusion criteria: patients ≥55 years with blunt trauma. • STTGMA variables were collected for 6 months on

orthopaedic and trauma surgery consults. • STTGMA variables: age, pre-existing conditions, injury severity,

neurological status. • Additional variables: anticoagulation status (COAG), albumin

level (ALB), ambulatory status (community, household, non-ambulatory) (AMB), assistive device (walker, cane, wheelchair) (AD).

• Inpatient complications and mortality were monitored. • Hospital cost was calculated to determine if utilizing early

palliative care consults3 resulted in hospital cost savings. • Logistic regressions were used to formulate a JHMC-specific

algorithm and area under receiver operator curve (AUROC) statistics were used to measure specificity and sensitivity of STTGMAJHMC.

Results

Figure 1: Stratification of LE patients by their STTGMAJHMC score. Percentage refers to relative risk of inpatient mortality.

Conclusions • HE-STTGMA and LE-STTGMA can be modified for a

hospital’s patient population to predict inpatient complications and mortality.

• STTGMAJHMC accurately stratifies patients into STTGMA

levels regarding patient mortality. • STTGMA Scale can be used to lower hospital costs by

including Palliative Care earlier in the admission process.

1. Baker SP, O'Neill B, Haddon W, Jr., Long WB. The injury severity score: a method for describing patients with multiple injuries and evaluating emergency care. The Journal of Trauma 1974;14:187-96. 2. Karlekar M, Collier B, Parish A, Olson L, Elasy T. Utilization and determinants of palliative care in the trauma intensive care unit: results of a national survey. Palliative Med 2014;28(8):1062-8. 3. Morrison RS, Penrod JD, Cassel JB, et al. Cost savings associated with US hospital palliative care consultation programs. Arch Intern Med. 2008;168(16):1783-90.

References

Objectives • To create a new model to predict mortality and inpatient

complications for middle-aged and geriatric trauma patients in the Emergency Department setting.

• To customize the STTGMA formula for the JHMC patient

population. • To optimize patient triage to the ICU or floor and to reduce

costs.

• To initiate a Palliative Care consult earlier for high-risk patients.

• Each trauma patient is assigned an Injury Severity Score (ISS) upon discharge from the hospital.

• However, on admission, there is no ISS to assist in the

treatment and care of middle-aged and geriatric trauma patients.1

• Triage of each trauma patient varies depending on the

mechanism of injury.

• There is no ISS that separates patients by low-energy (eg. falling from standing) vs. high-energy (eg. motor vehicle accident, falling from height, pedestrian struck) mechanism of injury.2

• We developed a novel injury severity score specific to

geriatric trauma: Score for Trauma Triage in the Geriatric and Middle-Aged (STTGMA) orthopaedic trauma patient.

Figure 2: Stratification of HE patients by their STTGMAJHMC score. Percentage refers to relative risk of inpatient mortality.

Figure 3: Ability of the HE-STTGMA to predict inpatient complications using AUROC. * - denotes statistically significant improvement in ability of HE-STTGMA-JHMC to predict complication compared to the original HE-STTGMA.

Figure 3: The Ability of the Original STTGMA vs. Modified STTGMAJHMC to Predict In-Hospital Complications

Figure 1: Patient Distribution and Relative Risk of Mortality in Low-Energy Patients using STTGMAJHMC Scale

Figure 2: Patient Distribution and Relative Risk of Mortality in High-Energy Patients using STTGMAJHMC Scale

• 94 high energy (HE) patients, 6 deaths (7.4%) • 154 low energy (LE) patients, 6 deaths (3.9%) • HE-STTGMAJHMC=HE-STTGMA+AD+AMB+ALB • LE-STTGMAJHMC=LE-STTGMA+AD+AMB+COAG • HE-STTGMA and HE-STTGMAJHMC showed ability to predict complications (Figure 3) • STTGMAJHMC stratified patients by mortality (Figure 1,2) • Estimated reduction in hospital costs over 1 year using

STTGMAJHMC + palliative care consult on admission • LE= $205,984 • HE=$192,416 • Total=$398,400

• Every 2 hours in NYC, 1 pedestrian or cyclist is either killed or seriously injured.

• In Queens, 22,760 pedestrians/cyclists were struck by motor vehicles in 2013. 384 pedestrians/cyclists sustained fatal injuries.

• Distracted driving accounts for some pedestrian

casualties, however there is no established mechanism that explains traffic injuries and fatalities that occur while drivers are not distracted.

• Pedestrians/cyclists are always perceived as having the “right-of-way”. This perception often encourages unsafe practices which include:

Acknowledgements Methods

Results Introduction

According to NYC Department of Transportation: Morning Rush Hours are 5AM to 10 AM Evening Rush Hours are 3PM to 8PM

Figure 3. Motor Vehicle Driver Data

Figure 2. Dangerous times on the roadways Table 1. Patient Characteristics

Table 2. Patient Outcomes

• A higher number of pedestrians were struck by motor vehicles on Mondays.

• The highest number of both pedestrians and cyclists were struck during evening rush hours (3 – 8PM).

• More pedestrians were struck during the fall and winter seasons.

• Most cyclists were struck during the

summer and fall seasons.

• 336 of the drivers were male, 164 were female and 569 were unknown. • 124 (12%) of the accidents were ‘hit and run’ accidents. • 44 drivers ran the red light.

• 280 drivers were making a turn at the time of the accident.

• We would like to thank the ED Registration for their contribution to the project.

• This project is funded by the New York State Governor’s Traffic Safety Committee.

• This project is ongoing and data collection will continue

until September, 2015.

Pedestrians Total= 898

Cyclists Total = 171

Male 443(49.3%) 161 (94.2%)

Female 455 (50.7%) 10 (5.8%)

Age (years) Mean: 44 Range: 18 - 92

Mean: 35 Range: 18 - 88

Data was prospectively obtained from all pedestrians and cyclists 18 years and older who were struck by motor vehicles and presented to JHMC for treatment during the period of September 16th, 2013 – April 30th, 2015. Medical records were reviewed to collect demographic data, injuries, outcomes, and incident locations. Pedestrian and cyclist practices were obtained via interviews with patients, emergency medical services personnel, and from police reports.

Objective The objective of this study is to determine risk factors that contribute to accidents involving pedestrians and cyclists. By collecting prospective data on each patient who present to JHMC after being hit by a motor vehicle we hope to match the circumstances and location with medical outcome for each incident. By using the data we collect we will be able to design better targeted interventions to reduce the number of pedestrians and cyclists struck by motor vehicles in Queens.

Outcomes Pedestrians Cyclists

Number of admitted patients 192 (21.4%) 30 (17.5%)

- Admitted to SICU 60 (31.3%) 8 (26.7%) - Admitted to Surgical Floor 132 (68.7%) 22 (73.3%)

Number of patients requiring surgery 107 (11.9%) 9 (5.3%)

- Number of surgeries performed 210 12

Length of hospital stay (days)

Average: 2 Range: 1 - 100

Average: 1 Range: 1- 44

Number of patients requiring Rehabilitation

45 (5.0%) 2 (1.2%)

Mortality 19 (2.1%) 3 (1.7%)

Figure 1. Patient Injury Severity • 144 pedestrians and 33 cyclists had a loss of consciousness.

• Most patients had a GCS of 12-15. • Most patients had minor injuries.

• 41 pedestrians had critical injuries, 13 died within 24 hours of injury.

• 5 cyclists had critical injuries, 2 died within 24 hours of injury.

Cell phone and headphones use Inebriation due to alcohol consumption Pedestrians walking along roadsides Pedestrians crossing midblock Pedestrians crossing against the signal Cyclists riding against the flow of traffic Cyclists riding in between lanes of traffic Cyclists riding too closely to parked vehicles Cyclists riding without protective gear (helmets, elbow pads, knee pads, and wrist guards) Cyclists failing to abide by traffic laws and signals

Characterization of pedestrians and cyclists struck in the Jamaica Hospital catchment area Melissa K. James, Ph.D.; Mauricia C. Victor, MS, MPH; Taylor R. Klein, BS; Jonathan Robitsek, Ph.D.; Sebastian D. Schubl, MD Department of Surgery, Jamaica Hospital Medical Center, Jamaica, New York

(A) (B)

(C) (D)

(E) (F)

(G) (H)

Risky Motorcyclist Behavior Correlates with Small Motorcycle Engine Size R. Jonathan Robitsek, Richard Tom, Simon P. Tiu, Taylor R. Klein, Marianne J. Mylan, Sebastian D. Schubl, Vanessa P. Ho Department of Surgery, Jamaica Hospital Medical Center, Jamaica, NY

HYPOTHESIS Riders of MCs with larger engines are more likely to exhibit risky behavior than riders of MCs with smaller engines.

BACKGROUND • Motorcycles have the greatest crash costs per

person-mile of all vehicles in the US (Miller, 1993)

• In 2010, motorcycles made up 3% of all registered vehicles in the US, and accounted for 0.6% of all vehicle miles traveled (NHTSA, 2010)

• From 2001-2010, US motorcyclist fatalities increased by 41% (NHTSA, 2010)

• Per vehicle mile traveled, motorcyclists are ~30 times more likely to perish and 5 times more likely to be injured than drivers or passengers in caged vehicles (NHTSA, 2010).

• Across all classes of motorcycles, there has been a

trend towards an increase in the average size of engines in fatal motorcycle crashes, suggesting that larger engine motorcycles are inherently riskier to ride than those with smaller engines.

• “You are what you drive”: Motorcycle size and engine type may reflect the riding style and attitude of the rider

• Rider style and attitudes towards risk taking may be an important causative factor contributing to the higher injury and fatality statistics observed for larger engine size motorcycles, rather than the large engine of the motorcycle itself (Houston, 2011)

METHODS • Retrospective analysis of adult inpatients after MC

accident from 4/2002 to 3/2007 • Demographics, helmet fastening, MC licensure, and

MC insurance were collected from charts and/or prospective patient interviews.

• ES were categorized as: small (<500cc), medium (500-850cc), large (>850cc), or unknown.

• Risky behavior : A failure to fasten or wear a helmet,

or operating a motorcycle without a license or without insurance.

• Univariate analysis was performed to determine

factors associated with risky behavior (chi-square and t test); multivariable logistic regression was performed.

Behavior By Engine Size Behavior By Age

Ove

rall

Ris

ky

Beh

avio

r H

elm

et F

aste

ning

Li

cens

ure

Insu

ranc

e

• 190 motorcyclists admitted between 4/2002-3/2007

• 115 (61%) had complete safety profile, collected

during admission • 108 Male (94%) • Mean age 28 ± 8.1 years • 59 (51%) showed safe behavior • 56 (49%) showed risky behavior

• Risky riders were significantly younger than safe

riders (B; p < 0.001), and more likely to be riding small engine motorcycles (A; p < 0.001).

• Age (OR = 0.87, p < 0.001) and small engine size (OR = 10.75, p = 0.03) were independently predictive of risky behavior.

• Medium (OR = 4.7; p = 0.02) and large (OR = 5.6;

p = .02) engine MC riders were more likely to fasten their helmet than riders of small engine MCs (C). Riders with unfastened helmets were younger than fastened riders (D; p = 0.04).

• Medium (OR = 16.8; p < 0.001) and large (OR = 33.6; p < 0.001) engine MC riders were more likely to be properly licensed than riders of small engine MCs (E). Riders without proper licensure were younger than licensed riders (F; p < 0.001).

• Medium (OR = 24.05; p = 0.003) and large (OR = 58.5; p < 0.001) engine MC riders were more likely to be properly insured than riders of small engine MCs (G). Riders without proper licensure were younger than licensed riders (H; p < 0.001).

• Length of stay was 56% longer for risky riders than safe riders (OR = 1.59; p < 0.001).

RESULTS

DISCUSSION • Within our sample:

• Small motorcycle engine size was significantly and independently associated with risky behavior • Younger age was significantly and independently associated with risky behavior • Hospital length of stay was significantly longer for risky riders as compared to safe riders.

• Riders of motorcycles with larger engines may attempt to mitigate risk by practicing safe behaviors • Limitations include small sample size and potential selection bias as some patients were not or could not be interviewed • Injury prevention providers should emphasize safe riding practices amongst young motorcyclists and those with all sizes of engines • Future research should study whether community and in-hospital interventions can modify rider behavior

REFERENCES (1) Miller, T. (1993). Costs and functional consequences of United States roadway crashes. Accident analysis and Prevention, 593:607. (2) National Highway Traffic Safety Administration. (2010). Traffic Safety Facts: Motorcycles. DOT HS 811 639. Washington, D.C.: US Department of Transportation. (3) Houston, D.J. (2011). Motorcyclists. In B.E. Porter (Ed.) Handbook of Traffic Psychology (pp. 375-387). Academic Press.

References

Discussion

Methods

Development of a Palliative Care Trigger Tool for Trauma Patients at Jamaica Hospital Angelo P. Canedo DO PGYI, R. Jonathan Robitsek PhD, Sebastian D. Schubl MD, Alan R. Roth DO, Angelo R. Canedo PhD,

Introduction Background

Trauma accounts for 41 million annual ED visits, 2.3 million annual hospitalizations, and 30% of all life years lost in the US [1]. Traumatic Injury incurs over $406 billion in combined health care costs and lost productivity [1]. In 2010, The Center to Advance Palliative Care Consensus Panel addressed the selection of consultation criteria, proposing a checklist approach, to be used in the Hospital Setting for the purpose of prospectively capturing those patients at high risk for unmet palliative care needs [2]. The panel recommended that each institution devise a systematic approach to identify patients with high risk of unmet Palliative Care needs [2]. The consensus panel also advised that screening should occur at the time of admission and continue throughout the hospital course to ensure the provision in a timely manner. Consultation Checklists and Trigger Tools for Palliative Care case finding have been designed, implemented and assessed within the confines/context of the Intensive Care Unit. Proactive case finding in the MICU through the use of Trigger Criteria has been shown to result in decreased length of stay in ICU without changing mortality rates or discharge disposition [3-5]. We hope to develop a novel set of Trigger criteria for identifying Trauma patients warranting Palliative Care consult on admission to the Hospital from the Emergency Department within the first 24 hours. Ultimately we aim to develop a trigger tool composed of largely quantitative criteria or elements easily extracted from the Electronic Medical Record that can be automated within the EMR system to screen all Trauma patients, regardless of age or trauma stage/setting, at the time of admission to the Hospital. At the present moment no such tool exists. This novel trigger tool would aid in capturing trauma patients with unmet Palliative Care needs, while circumventing the need for additional staff devoted to manually screening all patients admitted each day to the hospital. This work may potentially serve as the basis of future studies evaluating the impact of implementing these criteria via established outcome measures.

.

Prognosis within the setting of traumatic Injury is often uncertain with limited aid from quantitative surgical severity/mortality and morbidity prediction systems [6, 7]. There are inherent difficulties and conflict in establishing goals of care [6, 7]. These difficulties are only exacerbated by the frequent presence of a surrogate decision maker for the patient [6, 7]. These factors contribute to the increased likelihood of unmet Palliative Care needs in the trauma patient. Significant barriers to the provision of palliative care services within the hospital setting exist: including workforce shortages, program resource constraints and timeliness of referral for consultation [2]. Much emphasis has been placed on the role of Palliative Care within the setting of the ICU. There are a myriad of justifications for this focus on ICU populations including: ideally capturing all critically ill patients, the imperative to cure being strongest in this setting which tends to detract from attention to end-of-life care and the fact that up to 20% of patients will die, while as little as 5 % of patients may actively participate in decision making, regarding their care [8]. Limiting the use of Palliative Care Consultation services until admission to the TICU where patients may have a length of stay as short as 2-3 days prior to death, may contribute to a decreased likelihood of meeting the Palliative Care needs of patients. Proactive case finding in the MICU through the use of Trigger Criteria has been shown to result in decreased length of stay in ICU without changing mortality rates or discharge disposition [3-5]. Bradley et al. [9] developed trigger criteria for the SICU through a consensus panel of Palliative Care specialists. The implementation of these criteria, require a dedicated individual to gather the information daily [9]. Several criteria are less quantitative in nature. Retrospective analysis pre and post-implementation of these trigger criteria did not result in an increased number of Palliative Care consults for SICU patients [10]. There was also no change in the time from trigger identification to receiving a Palliative Care consult between the pre and post-intervention (trigger criteria) groups [10]. There was no change in the number of SICU stays or in-hospital mortality between groups despite the fact that meeting a single trigger conferred at least a 50% risk of in-hospital mortality [10]. Bryczkowski and colleagues [11] recently presented the results of a retrospective study examining potential triggers in 92 Geriatric (age >55 yrs.) ICU patients. Outcomes defining Palliative Care needs included death, Glasgow Outcome Score and discharge to facility [11]. Data were analyzed using a non-parametric Kruskal-Wallis ANOVA [11]. The authors suggest that the following criteria may be useful for triggers of Palliative consultation in trauma patients > 55 years old: Age > 75, Glascow Coma Scale < 13, Injury Severity Score > 18, Palliative Performance Scale on admission < 80 and any requirement for transfusion [11].

To isolate the quantitative patient variables involved in triggering a palliative care consult, a database was created via retrospective review of all trauma activation and trauma consult patients at JHMC between 05/01/2014 and 07/31/2014 from the JHMC Trauma Registry. Variables collected include patient demographics, GCS on admission, disposition post trauma bay, advanced directives on admission, presence or absence of a Palliative Care consult during the admission, vital signs and select laboratory values. All pre-existing conditions will be collected to calculate the Charlson Comorbidity Index (Deyo’s modification). The HPI of each patient will also be reviewed to identify the following conditions: pelvic/Hip fractures, metastatic fractures, TBI, neurodegenerative disease, trauma mechanism and functional status pre-trauma. The Shock Index will be calculated for each patient, which is an effective measure of hemodynamic instability. We will utilize the categorization suggested by Zarzaur et al [15], which defines four groups of worsening shock index. Mean Arterial Pressure will also be calculated for each patient to provide an indicator of tissue perfusion. All statistical analyses are being conducted using R version 3.1.2. A fully saturated logistic regression model will be employed using all of the variables described previously that are not suspect of collinearity to identify the variables with the greatest odds of being associated with a palliative care consult being ordered. In order to assess our proposed model, the data will also be modeled using a much smaller set of predictors. All of the relevant data collected from the EMR will be also be entered into the APACHE II scoring system, an ICU scoring system that is effective in predicting mortality. While we are not trying to predict mortality per se, many of the variables that we are collecting overlap with those included in the APACHE II. As such, a second logistic regression will be performed using the APACHE II score, the Charlson Comorbidity index, intubation status, disposition, presence of advanced directive, and the 6 variables obtained from the HPI as independent predictors. The predicted probabilities from the two logistic regression models will be utilized in an ROC curve analysis, with the dependent variable (presence/absence of a palliative care consult) as the classification variable. Area under the curve (AUC) will be compared between the two models, with the higher AUC taken as reflecting better prediction. The Hosmer-Lemeshow goodness of fit test will also be used to compare models. Lastly, the AIC (Akaike’s Information Criterion) will be compared between the two models. A priori, we predict that our proposed model will have better predictive power than the model that consists of only the APACHE II score + CCI; thus, the Type-I error rate will be lower following the Hosmer-Lemeshow simulations for our model than the alternative model, as will the AIC. In order to assess how accurate our model is in predicting who needs a palliative care consult, based on quantitative medical data alone, data from the 3 month time period described above, where the status of a palliative care consult is known will serve as the “training” data set, but only 75% of the responses will be utilized. The remaining 25% of the dataset will serve as the “test” data set. The test dataset does not contain the variable that indicates the presence/absence of a palliative care consult. That is what we will try and predict, using the model developed above using the training data. Predicted probabilities will be computed using the model fits developed with the training data set, and a confusion matrix created. The confusion matrix allows visualization of the “actual” state of whether a patient received a palliative care consult vs. what the model predicted. The confusion matrix also allows the calculation of multiple parameters used to judge model fit including accuracy, sensitivity, false positive rate, misclassification rate, specificity and precision. We will also calculate Cohen’s Kappa, which measures how well the model performs as compared to how well it would perform by chance alone. An ROC curve will be generated, which will allow calculation of the AUC for the model as to how well it predicts outcome. Depending on the performance of the logistic regression model, we may apply one, or more of other modeling techniques, including: Random forest decision trees, partial linear discriminant analysis, neural networks, and support vector machines. Any other modeling techniques employed will be compared to the logistic regression model described above, and compared using statistics from the confusion matrices. All modeling will be guided using Kuhn and Johnson [16], which was developed in the Global Non-Clinical Statistics Unit at Pfizer, and is one of the preeminent modern references on these methods.

We hope to develop a novel set of Trigger criteria for identifying Trauma patients warranting Palliative Care consult on admission to the Hospital from the Emergency Department within the first 24 hours. Ultimately, we aim to develop a trigger tool composed of largely quantitative criteria or elements easily extracted from the Electronic Medical Record that can be automated within the EMR system to screen all Trauma patients, regardless of age or trauma stage/setting, at the time of admission to the Hospital. This automated tool, composed of criteria extracted from the primary literature known to contribute to increased mortality/morbidity risk and complicated hospital course, will then provide patient lists each day to the Palliative Care consultation service for further consideration of the need for Palliative Consultation. At the present moment no such tool exists. This novel trigger tool would aid in capturing trauma patients with unmet Palliative Care needs, while circumventing the need for additional staff devoted to manually screening all patients admitted each day to the hospital. This work may potentially serve as the basis of future studies evaluating the impact of implementing these criteria via established outcome measures including but not limited to: Hospital Length of Stay, ICU Length of Stay, time to palliative care consultation, rate of entry into DNR, time to DNR entry, withdrawal of life-sustaining care, patient/family satisfaction, discharge disposition (i.e. Hospice) as well as other potential measures of life-altering morbidity/functional limitation and mortality rates.

1. http://www.cdc.gov/injury/overview/leading_cod.html 2. Weissman, D.E. and D.E. Meier, Identifying patients in need of a palliative care assessment in the hospital setting: a consensus report from the Center to Advance Palliative Care. J Palliat Med, 2011. 14(1): p. 17-23. 3. Campbell, M.L. and J.A. Guzman, Impact of a proactive approach to improve end-of-life care in a medical ICU. Chest, 2003. 123(1): p. 266-71. 4. Campbell, M.L. and J.A. Guzman, A proactive approach to improve end-of-life care in a medical intensive care unit for patients with terminal dementia. Crit Care Med, 2004. 32(9): p. 1839-43. 5. Norton, S.A., et al., Proactive palliative care in the medical intensive care unit: effects on length of stay for selected high-risk patients. Crit Care Med, 2007. 35(6): p. 1530-5. 6. Mosenthal, A.C. and P.A. Murphy, Trauma care and palliative care: time to integrate the two? J Am Coll Surg, 2003. 197(3): p. 509-16. 7. Mosenthal, A.C., et al., Changing the culture around end-of-life care in the trauma intensive care unit. J Trauma, 2008. 64(6): p. 1587-93. 8. Mosenthal, A.C. and P.A. Murphy, Interdisciplinary model for palliative care in the trauma and surgical intensive care unit: Robert Wood Johnson Foundation Demonstration Project for Improving Palliative Care in the Intensive Care Unit. Crit Care Med, 2006. 34(11 Suppl): p. S399-403. 9. Bradley, C.T. and K.J. Brasel, Developing guidelines that identify patients who would benefit from palliative care services in the surgical intensive care unit. Crit Care Med, 2009. 37(3): p. 946-50. 10. Bradley, C., J. Weaver, and K. Brasel, Addressing access to palliative care services in the surgical intensive care unit. Surgery, 2010. 147(6): p. 871-7. 11. Bryczkowski, S.B.e.a., Missed Opportunities for Palliative Care: Identifying the Palliative Care Needs in the Elderly Trauma Population. Journal of the American College of Surgeons, 2014. 219(3): p. S54. 15. Zarzaur BL, Croce M, Fischer PE, Magnotti LJ, Fabian TC. New vitals after injury: shock index for the young and age x shock index for the old. J Surg Res 2008, 147:229–236. 16. Kuhn M and Johnson K. Applied Predictive Modeling. Springer-Verlag. 2013. 17. Harrell, F. E. Regression modeling strategies: With applications to linear models, logistic regression, and survival analysis. Springer-Verlag, New York. 2001.

Occupational Therapy Leadership in Designing and Implementing Injury Prevention Programs in Level I Trauma Centers

Occupational Therapy Leadership in Designing and Implementing Injury Prevention Programs in Level I Trauma Centers

Patricia A. Gentile, DPS, OTR/L

Mark Dekki, MPA Jamaica Hospital Medical Center, Jamaica, NY, USA

In the US, traumatic injuries are a major public health problem across the lifespan and a leading cause of death in individuals 1 to 44 years of age. More than 2.8 million people are hospitalized and 31.7 million people treated in emergency rooms as a result of violence and injuries each year.

The American College of Surgeons, the organization responsible for verifying Trauma Centers across the country, mandates that Trauma Centers be actively involved with injury prevention efforts.

Occupational therapists working with injured patients have always provided these individuals with information and practical tools to make healthy choices to prevent injuries from reoccurring.

Occupational therapists working in Trauma Centers now need to take it to the next level by assuming leadership roles in injury prevention programming. This leadership includes using concepts of occupational therapy to design, implement and evaluate broad scope prevention initiatives. It also requires occupational therapists to increase their involvement and presence in advocacy and health policy related to trauma injury prevention. Increased involvement in leadership activities of this sort will also support the profession to move forward consistent with the AOTA Centennial Vision.

INTRODUCTION

TYPICAL PROGRAMS

JHMC LEVEL I TRAUMA CENTER SNAPSHOT

REFERENCES

Centers for Disease Control and Prevention, National Center for Injury Prevention and Control. Web–based Injury Statistics Query and Reporting System (WISQARS) [online]. Accessed February 10, 2015.

American College of Surgeons (2014), Resources for optimal care of the injured patient 2014 (online). Retrieved from http://www.facs.org/trauma/verification/resources-preview/index.html

McDonald E.M, MacKenzie E.J, & Teitelbaum S.D, et al. (2007) Injury prevention activities in U.S. trauma centers: Are we doing enough? Injury, 38, 538-547.

Cohen, L. & Swift, S. (1999) The spectrum of prevention: developing a comprehensive approach to injury prevention Injury Prevention, 5:203–207

Hildenbrand, W.C. and Lamb, A.J. (2013). Health policy perspectives: Occupational therapy in prevention and wellness: Retaining relevance in a new health care world. American Journal of Occupational Therapy, 67,266-271.

OT APPROACH AND INFLUENCING FACTORS

• Helmet Safety • Violence Prevention Initiatives • Fall Prevention • Distracted Driving Education • Car-Seat Education • Street Crossing/ Pedestrian Safety

OT LEADERSHIP COMPETENCIES

Contact Information: pgentile@jhmc.org

Person

Occupation Environment

• Utilizes Center’s Trauma Registry for Data Driven Problem Identification

• Develops Collaborative, Community-Based, Strategies • Designs and Implements PEC Population Based

Interventions • Employs Careful Evaluation and Measurement • Demonstrates Post-Implementation Persistence • Identifies Funding Opportunities • Engages in Health Policy & Activism

Population Need

Financial Support

Public Policy

Media Attention

Local/National Events

Influencing Factors PEO Approach

*Injury Severity Score (ISS): • Used to define major trauma • Correlates with mortality, morbidity and

hospitalization time. • Major/polytrauma is defined as ISS > 15

• Injury develops through a process. • Identifying factors underlying injury can be aided by

using conceptual models. • Conceptual models can guide the development of

interventions. • Injury prevention involves several disciplines. • Multi-disciplinary interventions are often more

successful

KEY POINTS IN INJURY PREVENTION

4%

41%

31%

24% 0 -18

19-35

36-55

56 +

Age

37%

26%

29%

4% 4%

MVA

Falls

Assault

Suicide

Other

Mechanism

78%

21%

1%

Blunt

Penetrating

Other

Types of Trauma 9%

47% 22%

22% 0

1 to 8

9 to 15

>15

Injury Severity Score (ISS)*