INTRODUCTION

MATERIALS AND METHODS
Data source
Study population
Study variables
Outcomes
Statistical analysis

RESULTS
Baseline demographics and hospital characteristics
Table 1.
Variables | Early academic year (July to September) (n=124,934) | later academic year (October to June) (n=356,259) | p-valuea) | ||
---|---|---|---|---|---|
n | % | n | % | ||
Age in years (mean) | 59.3 | 59 | 0.028 | ||
Admission type | 0.016 | ||||
Non-elective | 109,588 | 87.8% | 311,534 | 87.5% | |
Elective | 15,230 | 12.2% | 44,357 | 12.5% | |
Indicator of sex | 0.505 | ||||
Male | 52,213 | 41.8% | 148,491 | 41.7% | |
Female | 72,680 | 58.2% | 207,619 | 58.3% | |
Race | 0.012 | ||||
White | 76,224 | 65.9% | 216,228 | 65.6% | |
African American | 13,008 | 11.2% | 38,133 | 11.6% | |
Hispanic | 17,406 | 15.1% | 49,065 | 14.9% | |
Asian | 4,053 | 3.5% | 11,911 | 3.6% | |
Native American | 739 | 0.6% | 2,049 | 0.6% | |
Others | 4,217 | 3.6% | 12,235 | 3.7% | |
Length of stay in days (mean) | 6.8 | 6.8 | 0.04 | ||
Total hospital charges (mean) | $66,688 | $65,105 | <0.001 | ||
ERCP outcomes | |||||
In-hospital mortality | 1,877 | 1.5% | 5,631 | 1.6% | 0.054 |
Post-ERCP pancreatitis | 1,484 | 1.2% | 3,878 | 1.1% | 0.004 |
Post-ERCP perforation | 259 | 0.2% | 637 | 0.2% | 0.044 |
Post-ERCP hemorrhage | 1,407 | 1.1% | 4,510 | 1.3% | <0.001 |
Post-ERCP sepsis | 11,773 | 9.4% | 31,212 | 8.8% | <0.001 |
Table 2.
Variables | July to September (n=11,773) | Rest of the year (n=31,212) | p-value | ||
---|---|---|---|---|---|
Age in years mean (±SD) | 68 (±16) | 67 (±16) | 0.01 | ||
Sex | 0.082 | ||||
Male | 6,242 | 53.0% | 16,835 | 54.0% | |
Female | 5,532 | 47.0% | 14,368 | 46.0% | |
Admission type | 0.609 | ||||
Non-elective | 10,852 | 92.3% | 28,710 | 92.1% | |
Elective | 906 | 7.7% | 2,447 | 7.9% | |
Admission day | 0.062 | ||||
Weekday (Mon–Fri) | 9,006 | 76.5% | 23,605 | 75.6% | |
Weekend (Sat–Sun) | 2,768 | 23.5% | 7,607 | 24.4% | |
Race | <0.001 | ||||
White | 7,074 | 65.9% | 19,063 | 66.4% | |
African American | 1,133 | 10.6% | 3,354 | 11.7% | |
Hispanic | 1,396 | 13.0% | 3,308 | 11.5% | |
Asian and Pacific Islander | 641 | 6.0% | 1,567 | 5.5% | |
Native American | 69 | 0.6% | 255 | 0.9% | |
Others | 421 | 3.9% | 1,150 | 4.0% | |
Median household income percentile for patient’s zip codea) | <0.001 | ||||
0–25th | 2,826 | 24.5% | 8,076 | 26.5% | |
26–50th | 2,784 | 24.2% | 7,495 | 24.6% | |
51–75th | 2,846 | 24.7% | 7,411 | 24.3% | |
76–100th | 3,057 | 26.6% | 7,507 | 24.6% | |
Primary expected payer | <0.001 | ||||
Medicare | 7,368 | 62.7% | 19,015 | 61.0% | |
Medicaid | 1,074 | 9.1% | 3,260 | 10.5% | |
Private including HMO | 2,577 | 21.9% | 7,038 | 22.6% | |
Self – Pay/no charge/others | 730 | 6.2% | 1,864 | 6.0% | |
Bed size of hospital | 0.027 | ||||
Small | 1,438 | 12.2% | 3,981 | 12.8% | |
Medium | 2,838 | 24.1% | 7,793 | 25.0% | |
Large | 7,497 | 63.7% | 19,438 | 62.3% | |
Region of hospital | <0.001 | ||||
Northeast | 3,059 | 26.0% | 7,846 | 25.1% | |
Midwest | 3,044 | 25.9% | 7,623 | 24.4% | |
South | 3,391 | 28.8% | 9,715 | 31.1% | |
West | 2,279 | 19.4% | 6,029 | 19.3% | |
Outcomes | |||||
Disposition | 0.001 | ||||
Routine | 5,318 | 45.2% | 14,264 | 45.7% | |
Transfer to short-term hospital | 255 | 2.2% | 813 | 2.6% | |
Other transfers (SNF, ICF, other) | 2,876 | 24.4% | 7,207 | 23.1% | |
Home health care | 2,452 | 20.8% | 6,482 | 20.8% | |
Against Medical Advice | 40 | 0.3% | 65 | 0.2% | |
Length of stay (days) mean (±SD) | 12.6 (±15.1) | 12.7 (±15.1) | 0.359 | ||
Total hospital charges mean (±SD) | $133,016 (±197,740) | $130,383 (±200,670) | 0.231 | ||
In-hospital mortality | 823 | 7.0% | 2,340 | 7.5% | 0.072 |
The bed size cutoff points distributed into small, medium, and large. It has been done so that nearly one-third of the hospitals in a given region, location, and teaching status combination would fall within each bed size category.
HMO, health maintenance organization; ICF, intermediate care facility; SD, standard deviation; SNF, skilled nursing facility.
a) Represents a quartile classification of the estimated median household income of residents in the patient’s ZIP Code, derived from ZIP Code-demographic data obtained from Claritas.[14]
All-cause in-hospital mortality and post-ERCP complications
LOS and total hospitalization charges
Patient disposition
Odds and predictors of post-ERCP sepsis at urban teaching hospitals
Table 3.
Predictors of post-ERCP sepsis in procedures performed during the early (July to September) academic months
Table 4.
P<0.05 indicates clinical significance. Multivariate regression model is adjusted for age, sex, race, admission day, type, hospital bed size, region, median household income, length of stay and relevant comorbidities.
CI, confidence interval; ERCP, endoscopic retrograde cholangiopancreatography; LL, lower limit; OR, odds ratio; UL, upper limit.

DISCUSSION
