Weekend and Night Outcomes in a Mature State Trauma System (Text Version)

On September 27, 2010, Brendan Carr made this presentation at the 2010 Annual Conference. Select to access the PowerPoint® presentation (757 KB). 

---

Slide 1

Weekend & Night Outcomes in a Mature State Trauma System

Brendan G. Carr, MD MS

Department of Emergency Medicine
Department of Biostatistics and Epidemiology

University of Pennsylvania School of Medicine

Slide 2

Background

• Outcomes for time-sensitive medical conditions are dependent upon the existence of comprehensive systems of care.
• Variability in outcomes has been demonstrated for a number of time-sensitive conditions including STEMI, cardiac arrest, and ischemic stroke.

Slide 3

The New Jersey STEMI [segment elevation myocardial infarction] "system"

Image: Figure 1. Mortality for Weekend versus Weekday Admissions According to Day of Admission, 1999-2002.

Notes: Myocardial Infarction Data Acquisition System (MIDAS). All first MI admissions in New Jersey from 1987 to 2002 (231,164) were included and grouped in 4-year intervals. 12.9 vs. 12% mortality—effect erased by adjustment for invasive procedures.

Slide 4

Survival From In-Hospital Cardiac Arrest During Nights and Weekends

Image: Figure 2. Survival Rate by Event Location, Illness Category and Monitored Status is shown.

Notes: 507 medical/surgical participating hospitals from January 1, 2000, through February 1, 2007. A total of 58,593 cases of in-hospital cardiac arrest occurred during day/evening hours.

Slide 5

Background

• Trauma Care in the US
  • Verification process
  • Demonstrated survival benefit
  • Explicit criteria required for:
    • Structures (staffing, OR availability)
    • Processes (QI program, prehospital notification)

Image: a book cover titled "Resources For Optimal Care of the Injured Patient 2006" is shown.

Notes: Mackenzie, 2006. NEJM. National Evaluation of the effect of Trauma-Center Care on Mortality.

18 level 1 trauma centers and 51 non-trauma centers
Moderate to severe injury
1,104 patients who died in the hospital
4,087 patients who survived to discharge
Propensity scores to severity adjust

CDC funded

25 percent lower risk of death

Slide 6

Goals of the Investigation

• We sought to determine whether the probability of death or adverse clinical outcomes was higher among injured patients presenting at night or on the weekend.

Slide 7

Hypothesis

• We hypothesized that outcomes after trauma would be similar for patients presenting during nights or on the weekend.

Slide 8

Methods

• Retrospective cohort analysis
• Five years of data (2004-2008)
• Pennsylvania Statewide Trauma Registry
  • 32 accredited trauma centers
  • Admitting diagnosis of injury
  • Age > 18

Slide 9

Methods

• Main Outcomes:
  • In-hospital mortality
• Secondary Outcomes:
  • ICU length of stay
  • Hospital length of stay
  • Delay of more than two hours to laparotomy or craniotomy

Slide 10

Methods

• Exposure:
  • Night presentation
    • 11 pm - 6:59 am
  • Weekend presentation
    • 11 pm Friday - 6:59 am Monday
    • Saturday 12:01 am - Sunday 11:59 pm

Slide 11

Methods—Power

• We calculated the detectable mortality difference given:
  • Known sample size
  • 2-tailed alpha of 0.05
  • Power of 90%
  • Effect size estimate (mortality differences for night and weekend presentation based on pilot data)

Slide 12

Methods—Analysis

• Unadjusted
  • Pearson's chi-square, rank sum, T-test, Logistic regression
• Adjusted
  • Logistic regression
  • Negative binomial regression (LOS)

Slide 13

Methods—Analysis

• Case Mix adjustment
  • Modified Charlson
    • 15 instead of 19 comorbid conditions
• Injury Severity adjustment
  • TRISS (Trauma—Injury Severity Score)
    • Anatomic injury scoring system (ISS)
    • Physiological scoring system (Revised Trauma Score)

Slide 14

Results—Power

• Night as compared to day
  • Powered to detect 0.63% mortality difference
• Weekend as compared to weekday
  • Powered to detect 0.53% mortality difference

Slide 15

Results—Demographics

An image showing the top part of "Table 1. Patient Characteristics" is shown.

	Night (n=20,886)	Day/Evening (n=69,523)
Age (median 25%-75%)	36 (24-52)*	52 (36-74)
Male	14,822 (71.2)	41,358 (59.2)
GCS	15 (14-15)	15 (15-15)
ISS	10 (5-19)*	10 (5-17)
TRISS	0.99 (0.97-0.99)*	0.98 (0.96-0.99)
Race
White	13,488 (69.1)	54,634 (82.3)
Black	5,285 (27.1)	9,638 (14.5)
Other	739 (3.79)*	2,138 (3.2)
Mechanism
Blunt	17,087 (82.0)	64,662 (93.1)
Penetrating	3,741 (18)*	4,809 (6.9)
GSW	752 (3.65)*	960 (1.4)
Charleson Index (median 25%-75%)	0 (0-1)	0 (0-0)
Num. pre-existing conditions (median 25%-75%)	0 (0-4)*	1 (0-4)

Slide 16

Results—Demographics

Image: the top part of "Table 1. Patient Characteristics" is shown.

	Weekend (n=30,505)	Weekday (n=60,169)
Age (median 25%-75%)	45 (20-65)*	51 (33-73)
Male	19,584 (64.2)*	36,807 (61.2)
GCS	15 (14-15)	15 (15-15)
ISS	10 (5-19)*	10 (5-17)
TRISS	0.98 (0.96-0.98)*	0.98 (0.96-0.99)
Race
White	22,639 (78.4)	45,680 (79.7)
Black	5,217 (18.1)	9,788 (17.1)
Other	1,016 (3.5)*	1,871 (3.3)
Mechanism
Blunt	27,315 (89.6)	54,690 (91.0)
Penetrating	3,162 (10.4)*	5,438 (9.0)
GSW	645 (2.1)*	1,086 (1.8)
Charleson Index (median 25%-75%)	0 (0-1)	0 (0-0)
Num. pre-existing conditions (median 25%-75%)	0 (0-4)*	1 (0-4)

Slide 17

Summary Results (Adjusted)

Night

• Mortality -
  • Age < 55 -
  • Severe injury -
  • Blunt -
  • Penetrating -
• Delay to laparotomy +
• Delay to craniotomy -
• ICU LOS ↑
• Hospital LOS ↓

Weekend

• Mortality ↓
  • Age < 55 ↓
  • Severe injury↓
  • Blunt -
  • Penetrating -
• Delay to laparotomy -
• Delay to craniotomy -
• ICU LOS -
• Hospital LOS ↓

Slide 18

Limitations

• Retrospective data
• Single state analysis
• Negative findings raise power concerns
• Inadequate injury severity adjustment
• Inadequate case mix adjustment
• Exclusion of transfer patients

Slide 19

Conclusions

• Patients presenting at night are no more likely to die than patients presenting during the day
• Patients presenting on the weekend are less likely to die than patients presenting on weekdays

Slide 20

Implications

• Explicit staffing and resource requirements for unplanned critical illness protect against the "weekend effect"
• The impact of similar systems based interventions should be tested for other time-sensitive conditions

Slide 21

Acknowledgments

• Co-authors
  • Pat Reilly, MD
  • C. William Schwab, MD
  • Charles C. Branas, PhD
  • Juliet Geiger, RN MSN
  • Douglas J. Wiebe, PhD
• AHRQ K08HS017960
• Pennsylvania Trauma System Foundation

Slide 22

Questions?

Image: a helicopter is shown.