Trends in same-admission cholecystectomy and endoscopic retrograde cholangiopancreatography for acute gallstone pancreatitis: A nationwide analysis across a decade

BACKGROUND/OBJECTIVES

Gallstones are the leading cause of acute pancreatitis in developed countries. National and international guidelines recommend that a cholecystectomy should be performed during the index hospitalization for acute gallstone pancreatitis. We aimed to delineate the national trends for same-admission cholecystectomy and ERCP for acute gallstone pancreatitis over the last ten years.

METHODS

We used the 2004, 2009 and 2014 National Inpatient Sample database including patients with a principal diagnosis of acute pancreatitis and a secondary diagnosis of choledocholithiasis or cholelithiasis. Exclusion criteria were age <18 years and elective admission. Primary outcome was the trend in incidence rate of same admission cholecystectomy from 2004 to 2014. The secondary outcomes were: 10-year trend in 1) Incidence of gallstone pancreatitis, 2) proportion of gallstone pancreatitis compared to all other etiologies of acute pancreatitis, 3) incidence rate of same-admission ERCP, 4) length of hospital stay, and 5) total hospitalization costs and charges.

RESULTS

The proportion of admissions during which a same-admission cholecystectomy was performed decreased from 48.7% in 2004 to 46.9% in 2009 to 45% in 2014 (trend p < 0.01). During the same time interval, the percentage of admissions during which an ERCP was performed decreased from 25.1% to 18.7% (Trend p < 0.01).

CONCLUSIONS

Adherence to the guidelines for same-admission cholecystectomy for patients admitted with acute gallstone pancreatitis have been declining over the past decade. On the other hand, decline in rate of ERCP in patients with acute gallstone pancreatitis and no signs of cholangitis demonstrates adherence to guidelines in this regard.

Differences in healthcare outcomes between teaching and non teaching hospitals for patients with delirium: a retrospective cohort study

BACKGROUND

The physician workforce at teaching hospitals differs compared to non-teaching hospitals, and data suggest that patient outcomes may also be dissimilar. Delirium is a common, costly disorder among hospitalized patients and approaches to care are not standardized.

OBJECTIVE

This study set out to explore differences in healthcare outcomes between teaching and non-teaching hospitals for patients admitted with delirium.

DESIGN

Retrospective cohort analysis.

SETTING AND PARTICIPANTS

We used the 2014 Nationwide Inpatient Sample database. Adult patients (≥18 years of age) hospitalized in acute-care hospitals in the USA with delirium (defined with ICD-9 code) were studied.

MAIN OUTCOME MEASURES

The primary outcome was in-hospital all-cause mortality. Secondary outcomes were discharge status and several measures of healthcare resource utilization: length of stay, total hospitalization costs and multiple procedures performed.

RESULTS

In 2014, out of 57 460 adult patients admitted to hospitals with delirium, 58.4% were hospitalized at teaching hospitals and the remainder 41.6% at non-teaching hospitals. The in-hospital mortality of delirium patients in teaching hospitals was 1.33% (95% CI 1.08%-1.63%), and 1.26% (95% CI 0.97%-1.63%) in non-teaching hospitals. The mean total hospital costs were $7642 (95% CI 7384-7900) in teaching hospitals, and $6650 (95% CI 6460-6840) in non-teaching hospitals. After adjustment for confounders, total hospitalization costs were statistically significantly different between the hospitals types-with non-teaching providing less expensive care.

CONCLUSIONS

Patients with delirium admitted to non-teaching hospitals had comparable clinical and process outcomes achieved at lower costs. Further research can be conducted to explore the contextual issues and reasons for these differences in healthcare costs.

Racial factors and inpatient outcomes among patients with diabetes hospitalized with foot ulcers and foot infections, 2003-2014

Background

In patients with diabetes, foot amputations among Black patients have been historically higher compared with White patients. Using the National Inpatient Sample database, we sought to determine if disparities in foot amputations and resource utilization have improved over time. We hypothesized there would be improvements and reduced differences in foot amputations between the two races as quality of care and access to healthcare has improved.

Methods and findings

Patients over 18 years old with a principal diagnosis of diabetic foot complications and secondary diagnosis of Diabetes Mellitus were selected. We compared the primary outcome of foot amputations between Black and White patients. Adjusted rates, odds ratios (aOR) and trends of foot amputations among Black and White patients were studied. Healthcare utilization was measured via length of hospital stay (LOS). Of 262,924 patients, 18% were Black. Following adjustment for confounders, major foot amputations decreased among Whites (1.5% in 2003 to 1.1% in 2014) and Blacks (2.1% in 2003 to 0.9% in 2014). On pooled analysis, Black patients had higher adjusted odds of major foot amputations in 2003–2004 [aOR 1.7; (1.16–2.57), p<0.01]. Disparities in major foot amputations disappeared in 2013–2014 [aOR: 0.92 (0.58–1.44), p = 0.70]. Black patients had declining but persistently longer LOS (adjusted mean difference (aMD): 1.1 days (0.52–1.6) p<0.01 in 2003–2004 and 0.46 days (0.18–0.73) p<0.01 in 2013–2014). The main limitation of the study was that the NIS uses ICD-9 and ICD-10 CM codes, and hence prone to incorrect or missing codes.

Conclusions

Major foot amputations declined among Black and White patients hospitalized with Diabetic foot complications between 2003 and 2014. The observed difference for amputations in 2003–2004 was absent by 2013–2014. Future research to determine specific contributors for this reduction in health disparities is needed for ongoing improvements and sustainability.

Clinical outcomes among morbidly obese patients hospitalized with diabetic foot complications

The aim of the study is to investigate effects of morbid obesity on patients hospitalized with diabetic foot ulcers and infections (DFU/Is). This retrospective cohort study of hospitalized adults investigated patients with a principal diagnosis of DFU/Is, and compared outcomes based on weight. The primary outcome was lower limb amputations. Secondary outcomes included in-hospital mortality, morbidity (sepsis and discharge disposition), resource utilization (length of stay [LOS] and total hospitalization charges). Multivariate analyses adjusted for confounders. A total of 31 730 admissions were included; 14% were morbidly obese. Patients with morbid obesity had overall lower rates of amputations (adjusted odds ratio [aOR]: 0.60; 95% confidence interval [CI]: 0.45-0.80). There was no difference in mortality rates between those who were morbidly obese and the non-morbidly obese patients (aOR: 3.89 [95% CI: 0.79-19.30]). Combined data from 2010 to 2014 have found higher odds of sepsis in morbidly obese patients compared with non-morbidly obese patients after adjusting for confounders (aOR: 1.49 [1.09-2.02], P = 0.01). Patients with morbid obesity were less likely to be discharged to home (aOR: 0.82 [95% CI: 0.70-0.95]), had longer LOS (adjusted mean difference [aMD]: 0.47 [95% CI: 0.13-0.81]) and higher hospital charges (aMD: $3205 [95% CI: $3373-$6037]). Despite having lower amputation rates compared to those who are not morbidly obese, patients with morbid obesity hospitalized for DFU/Is succumbed to sepsis more frequently and had less favourable utilization metrics. Specialized care pathways may be needed to enhance the value of care delivered to this high-risk population.

Impact of Insurance Status and Race on Outcomes in Nonvariceal Upper Gastrointestinal Hemorrhage: A Nationwide Analysis

BACKGROUND AND GOALS

We examined the interaction between race, insurance, and important outcomes in nonvariceal upper gastrointestinal hemorrhage (NVUGIH).

STUDY

Adults with NVUGIH were selected from the National Inpatient Sample.

PRIMARY OUTCOME

in-hospital mortality.

SECONDARY OUTCOMES

treatment modalities [esophagogastroduodenoscopy (EGD), early EGD, and endoscopic or radiologic therapy], and resource utilization (length of hospital stay and total hospitalization charges).

RESULTS

Mortality was similar for Medicare and private insurance [adjusted odds ratios (aOR): 1.15 95% confidence interval (CI) 0.90 to 1.47), P=0.24], but higher for under/uninsured patients [aOR: 1.84 (CI: 1.42 to 2.40), P<0.01]. Compared with Medicare, patients with private insurance had more EGDs [aOR: 1.35 (CI: 1.23 to 1.48), P<0.01], early EGDs [aOR: 1.29 (CI: 1.21 to 1.38), P<0.01], and endoscopic [aOR: 1.19 (CI: 1.11 to 1.27), P<0.01], or radiologic therapy [aOR:1.35 (CI: 1.06 to 1.71), P=0.01]. Patients who were under/uninsured had less EGDs [aOR: 0.84 (CI: 0.76 to 0.91), P<0.01] or endoscopic therapy [aOR: 0.74 (CI: 0.68 to 0.81), P<0.01], but similar odds of early EGD [aOR: 0.95 (CI: 0.88 to 1.02), P=0.13] or radiologic therapy [aOR: 1.01 (CI: 0.75 to 1.37), P=0.75]. Compared with whites, blacks had lower [aOR: 0.73 (CI: 0.58 to 0.93), P=0.01] and Native Americans higher mortality [aOR: 2.60 (CI: 1.57 to 4.13), P<0.01]. Blacks were less likely [aOR: 0.86 (CI: 0.79 to 0.94), P<0.01] and Asians more likely [aOR: 1.24 (CI: 1.05 to 1.47), P=0.01] to have EGDs. Both blacks and Hispanics had lower, whereas Asians had higher early EGD rates. Patients with private insurance had lower total charges [adjusted mean difference: -$2761 (CI: -$4617 to -$906), P<0.01].

CONCLUSIONS

Insurance and race have independent effects on NVUGIH mortality, therapeutic modalities used, and resource utilization. Black and under/uninsured patients have the worst outcomes.

Trends in aggressive inpatient care at the end-of-life for stage IV lung cancer patients

26

Background: Metastatic lung cancer patients are treated with palliative intent. Aggressive inpatient care at the end-of-life is seen as a marker of poor quality care. However, national trends and factors related to aggressive inpatient care at the end-of-life for these patients have not previously been evaluated. Methods: A total of 3,030,866 hospitalizations of metastatic lung cancer patients were identified in the National Inpatient Sample database between 1998 and 2014. Longitudinal analysis was conducted to determine trends in aggressive inpatient care at the end-of-life and multivariate logistic regression was performed to determine associations with age, race, region, and hospital characteristics. Results: From 1998 to 2014, in-hospital mortality for metastatic lung cancer patients decreased from 17% to 11%. However, the proportion admitted to the intensive care unit during the terminal hospitalization increased from 12.9% to 26.4%. Reflecting this aggressive end-of-life care, mean total charges for a terminal hospitalization increased from $29,386 to $72,469, adjusted for inflation. Among patients who died in the inpatient setting, the intensive care unit stay translated into higher total costs (+$18,461 CI: $17,460 to $19,463). Promisingly, palliative care encounters for terminal hospitalizations increased during this period from 8.7% to 53.0% and were correlated with a decrease in aggressive care at the end-of-life. A multivariable model showed variation by patient and hospital characteristics in aggressive inpatient care utilization. Conclusions: Among patients with metastatic lung cancer there has been a substantial increase in intensive care unit use during terminal hospitalizations, resulting in high cost for the health care system.

National trends in admissions for potentially preventable conditions among patients with metastatic solid tumors, 2004-2014

1

Background: The Centers for Medicare and Medicaid Services (CMS) has identified 10 conditions for hospitalization among patients receiving chemotherapy that are potentially preventable through appropriately managed outpatient care. CMS plans to measure hospitals’ performance based on frequency of admission for: anemia, dehydration, diarrhea, emesis, nausea, neutropenia, fever, pain, pneumonia, and sepsis. Our objective was to measure hospital utilization patterns for these conditions. Methods: Nationally representative data from the 2004 and 2014 National Inpatient Sample were analyzed. Adults with stage IV solid tumors admitted with a principal diagnosis of one of the ten conditions were identified using ICD-9 codes provided by CMS. The primary outcome was number of admissions. Secondary outcomes were total hospitalization costs, length of stay (LOS), and in-hospital mortality rate. Proportions and count data were compared using chi-square and binomial test, respectively. Results: Between 2004 and 2014, potentially preventable hospitalizations increased from 64,053 to 85,740 (p < 0.01). The most frequent reason for admission was pneumonia (32%) in 2004 and sepsis (41%) in 2014. Table 1 provides the results for the year 2014 by condition. Hospitalizations for sepsis had the longest mean LOS (p < 0.01), highest mean total costs (p < 0.01) and most frequently resulted in death (p < 0.01). Conclusions: The Department of Health and Human Services states, “Improving patients’ quality of life by keeping patients out of the hospital is a main goal of cancer care.” Despite supportive care advances, the increased frequency of hospitalization demonstrates a need for continued symptom management innovation. Infectious symptoms should be a focus of these technologies given their high prevalence, mortality and resource utilization. [Table: see text]

One Step Forward, Two Steps Back: Trends in Aggressive Inpatient Care at the End of Life for Patients With Stage IV Lung Cancer

BACKGROUND

Patients with metastatic lung cancer are treated with palliative intent. Aggressive care at the end of life is a marker of poor-quality care. National trends and factors related to aggressive inpatient care at the end of life for these patients have not been evaluated.

METHODS

Patients with stage IV lung cancer and a terminal hospitalization were identified in the National Inpatient Sample database between 1998 and 2014. Longitudinal analysis was conducted to determine trends in aggressive inpatient care at the end of life and multivariate logistic regression was performed to determine associations with patient and hospital characteristics.

RESULTS

A total of 412,946 patients met the inclusion criteria. From 1998 to 2014, the proportion of patients admitted to the intensive care unit (ICU) during the terminal hospitalization increased from 13.3% to 27.9% (P < .001). The ICU stay translated into a higher mean total cost of care (+$18,461; 95% CI, $17,460 to $19,463). Promisingly, palliative care encounters for terminal hospitalizations also increased during this period from 8.7% to 53.0% (P < .01) and were correlated with a decrease in aggressive care at the end of life. However, this did not offset the trend in increased ICU use; mean total costs for a terminal hospitalization increased from $14,000 to $19,500, adjusted for inflation. A multivariable model demonstrates variation by patient and hospital characteristics in aggressive care use.

CONCLUSIONS

Among patients with metastatic lung cancer there has been a substantial increase in ICU use during terminal hospitalizations, resulting in high cost for the health care system.

Impact of primary payer status on outcomes among patients with burn injury: A nationwide analysis

OBJECTIVE

To study the relationship between insurance provider and important outcomes among patients with burn injury.

METHODS

Adults with burn injury were selected from the National Inpatient Sample. The primary outcome was inpatient mortality. Secondary outcomes were morbidity (septic shock and prolonged mechanical ventilation (PMV)), treatment metrics (time to surgery and parenteral or enteral nutrition (P/E-nutrition)) and resource utilization (length of stay (LOS) and total hospitalization costs and charges). Confounders were adjusted for using multivariate regression analysis.

RESULTS

Insurance did not affect in-hospital mortality rate. Compared with private insurance, Medicaid was associated with higher septic shock rate (aOR: 2.14 (1.04-4.39), longer LOS (adjusted mean difference (aMD): 2.79 (0.50-5.08) days) and higher costs (aMD: $16,161 ($4789-$27,534) while uninsured patients has shorter LOS (aMD: -2.57 (-4.59--0.55) days), lower charges (aMD: $-37,792 $(-65,550-$-10,034) and costs (aMD: $-8563 ($15,581-$-1544)). Insurance did not affect PMV rates or time to surgery or P/E-nutrition.

CONCLUSIONS

Primary payer does not affect in-hospital mortality or treatment metrics among patients admitted for burn injury. However, compared with private insurance, Medicaid was associated with both higher morbidity and resource utilization, whereas uninsured patients had lower resource utilization.

Thirty-Day Readmission Among Patients With Non-Variceal Upper Gastrointestinal Hemorrhage and Effects on Outcomes

BACKGROUND & AIMS

We aimed to determine the rate of hospital readmission within 30 days of non-variceal upper gastrointestinal hemorrhage and its impact on mortality, morbidity, and health care use in the United States.

METHODS

We performed a retrospective study using the Agency for Healthcare Research and Quality's Healthcare Cost and Utilization Project Nationwide Readmission Database for the year 2014 (data on 14.9 million hospital stays at 2048 hospitals in 22 states). We collected data on hospital readmissions of 203,220 adults who were hospitalized for urgent non-variceal upper gastrointestinal hemorrhage and discharged. The primary outcome was rate of all-cause readmission within 30 days of discharge. Secondary outcomes were reasons for readmission, readmission mortality rate, morbidity (shock and prolonged mechanical ventilation) and resource use (length of stay and total hospitalization costs and charges). Independent risk factors for readmission were identified using Cox regression analysis.

RESULTS

The 30-day rate of readmission was 13%. Only 18% of readmissions were due to recurrent non-variceal upper gastrointestinal bleeding. The rate of death among patients readmitted to the hospital (4.7%) was higher than that for index admissions (1.9%) (P < .01). A higher proportion of readmitted patients had morbidities requiring prolonged mechanical ventilation (1.5%) compared with index admissions (0.8%) (P < .01). A total of 133,368 hospital days was associated with readmission, and the total health care in-hospital economic burden was $30.3 million (in costs) and $108 million (in charges). Independent predictors of readmission were Medicaid insurance, higher Charlson comorbidity score, lower income, residence in a metropolitan area, hemorrhagic shock, and longer stays in the hospital. Older age, private or no insurance, upper endoscopy, and prolonged mechanical ventilation were associated with lower odds for readmission.

CONCLUSIONS

In a retrospective study of patients hospitalized for non-variceal upper gastrointestinal hemorrhage, 13% are readmitted to the hospital within 30 days of discharge. Readmission is associated with higher mortality, morbidity, and resource use. Most readmissions are not for recurrent gastrointestinal bleeding.

Assessment of the July effect in post-endoscopic retrograde cholangiopancreatography pancreatitis: Nationwide Inpatient Sample

AIM

To assess incidence of post-endoscopic retrograde cholangiopancreatography (post-ERCP) pancreatitis in the early (July/August/September) vs the late (April/May/June) academic year and evaluate in-hospital mortality, length of stay (LOS), and total hospitalization charge between these time periods.

METHODS

This was a retrospective cohort study using the 2012 Nationwide Inpatient Sample (NIS). Patients with International Classification of Diseases, 9^th Revision, Clinical Modification (ICD-9 CM) procedure codes for ERCP were included. Patients were excluded from the study if they had an ICD-9 CM code for a principal diagnosis of acute pancreatitis, if the ERCP was performed before or on the day of admission or if they were admitted to non-teaching hospitals. Post-ERCP pancreatitis was defined as an ICD-9 CM code for a secondary diagnosis of acute pancreatitis in patients who received an ERCP as delineated above. ERCPs performed during the months of July, August and September was compared to those performed in April, May and June in academic hospitals. ERCPs performed at academic hospitals during the early vs late year were compared. Primary outcome was incidence of post-ERCP pancreatitis. Secondary outcomes included in-hospital mortality, length LOS, and total hospitalization charge. Proportions were compared using fisher’s exact test and continuous variables using student t-test. Multivariable regression was performed.

RESULTS

From the 36480032 hospitalizations in 2012 in the United States, 6248 were included in the study (3065 in July/August/September and 3183 in April/May/June) in the 2012 academic year. Compared with patients admitted in July/August/September, patients admitted in April/May/June had no statistical difference in all variables including mean age, percent female, Charleston comorbidity index, race, median income, and hospital characteristics including region, bed size, and location. Incidence of post-ERCP pancreatitis in early vs late academic year were not statistically significant (OR = 1.03, 95%CI: 0.71-1.51, P = 0.415). Similarly, the adjusted odds ratio of mortality, LOS, and total hospitalization charge in early compared to late academic year were not statistically significant.

CONCLUSION

Incidence of post-ERCP pancreatitis does not differ at academic institutions depending on the time of year. Similarly, mortality, LOS, and total hospital charge do not demonstrate the existence of a temporal effect, suggesting that trainee level of experience does not impact clinical outcomes in patients undergoing ERCP.

Trends in aggressive care at the end-of-life for stage IV lung cancer patients

6614

Background: Prior studies have demonstrated that high-intensity end-of-life care improves neither survival nor quality of life for cancer patients. The National Quality Forum endorses dying from cancer in an acute care setting, ICU admission in the last 30 days of life, and chemotherapy in the last 14 days of life as markers of poor quality care. Methods: Discharge data from the National Inpatient Sample database was analyzed for 3,030,866 acute care hospitalizations of metastatic lung cancer patients between 1998 and 2014. Longitudinal analysis was conducted to determine trends in aggressive care at the end-of-life and multivariate logistic regression was performed to determine associations with age, race, region, hospital characteristics, and aggressive care. Results: In-hospital mortality for metastatic lung cancer patients decreased from 17% to 11%. Among terminal hospitalizations, utilization of radiation therapy and chemotherapy decreased from 4.6% to 3.0% and from 4.8% to 3.0%, respectively. However, the proportion admitted to the ICU increased from 13.3% to 27.9% and invasive procedures increased from 1.2% to 2.0%. Reflecting this aggressive end-of-life care, mean total charges for a terminal hospitalization rose from $29,386 to $72,469, adjusted for inflation. Among patients who died in the inpatient setting, the ICU stay translated into higher total costs (+$16,962, CI: $15,859 to $18,064) compared to patients who avoided the ICU. Promisingly, palliative care encounters for terminal hospitalizations increased during this period from 8.7% to 53.0% and was correlated with a decrease in inpatient chemotherapy (OR = 0.56, CI: 0.47 to 0.68), radiotherapy (OR = 0.77, CI: 0.65 to 0.92), and ICU admissions (OR = 0.48, CI: 0.45 to 0.53) but had only a modest impact on terminal hospitalization cost (-$2,992, CI: -$3,710 to -$2,275). Multivariable analysis showed variation by patient and hospital characteristics in aggressive care utilization. Conclusions: Among patients with metastatic lung cancer there has been a substantial increase in ICU use during terminal hospitalizations, resulting in high cost for the health care system. Inpatient palliative care has the potential to reduce aggressive end-of-life interventions.

H. Pylori as a predictor of marginal ulceration: A nationwide analysis

OBJECTIVE

Helicobacter pylori has been implicated as a risk factor for development of marginal ulceration following gastric bypass, although studies have been small and yielded conflicting results. This study sought to determine the relationship between H. pylori infection and development of marginal ulceration following bariatric surgery in a nationwide analysis.

METHODS

This was a retrospective cohort study using the 2012 Nationwide Inpatient Sample (NIS) database. Discharges with ICD-9-CM code indicating marginal ulceration and a secondary ICD-9-CM code for bariatric surgery were included. Primary outcome was incidence of marginal ulceration. A stepwise forward selection model was used to build the multivariate logistic regression model based on known risk factors. A P value of 0.05 was considered significant.

RESULTS

There were 253,765 patients who met inclusion criteria. Prevalence of marginal ulceration was 3.90%. Of those patients found to have marginal ulceration, 31.20% of patients were H. pylori-positive. Final multivariate regression analysis revealed that H. pylori was the strongest independent predictor of marginal ulceration.

CONCLUSIONS

H. pylori is an independent predictor of marginal ulceration using a large national database. Preoperative testing for and eradication of H. pylori prior to bariatric surgery may be an important preventive measure to reduce the incidence of ulcer development.

Trends in Adult-to-Adult Living Donor Liver Transplant Organ Donation: The Johns Hopkins Experience

Adult-to-adult living donor liver transplantation is an increasingly important option for 17000 patients awaiting liver transplantation in the United States. However, adult-to-adult living donor liver transplantation volumes peaked in 2001 (N=518), and have gradually fallen in 2002 (N=362), 2003 (N=321), and 2004 (N=323). Recent concerns about donor safety and ethical considerations have made careful analysis of donor availability and selection criteria critically important. We conducted a retrospective review of our active liver transplant recipient registry (N=251) and compared it to our living donor registry (N=231), which included all potential living donors before the selection process. Fifteen percent of recipients accounted for the majority (53%) of donor evaluations, whereas 42% of recipients did not have even a single donor evaluation. Recipient diagnosis appears to have a significant impact on donor availability, with donors rarely evaluated for patients with alcoholic cirrhosis. Careful and stringent selection criteria rule out 67% of potential donors.

The in-hospital mortality rate for upper GI hemorrhage has decreased over 2 decades in the United States: a nationwide analysis

BACKGROUND

Despite major advances in upper GI hemorrhage (UGIH) treatment, UGIH mortality has been reported as unchanged for the past 50 years.

OBJECTIVE

To measure the UGIH in-hospital mortality rate and other important outcome trends from 1989 to 2009.

DESIGN

A longitudinal study of UGIH hospitalizations by using the Nationwide Inpatient Sample.

SETTING

Acute-care hospitals.

PATIENTS

All patients admitted for UGIH. Patients who bled after admission were excluded.

MAIN OUTCOME MEASUREMENTS

UGIH in-hospital mortality rate, incidence, in-hospital endoscopy and endoscopic therapy rates, length of hospital stay, and total in-hospital charges.

RESULTS

The non-variceal UGIH mortality rate decreased from 4.5% in 1989 to 2.1% in 2009. The non-variceal UGIH incidence declined from 108 to 78 cases/100,000 persons in 1994 and 2009, respectively. In-hospital upper endoscopy and endoscopic therapy rates increased from 70% and 10% in 1989 to 85% and 27% in 2009, respectively. The early endoscopy rate increased from 36% in 1989 to 54% in 2009. The median length of hospital stay decreased from 4.5 days in 1989 to 2.8 days in 2009. Median total hospitalization charges increased from $9249 in 1989 to $20,370 in 2009. At the national level, the UGIH direct in-hospital economic burden increased from $3.3 billion in 1989 to $7.6 billion in 2009. Similar trends were found for variceal UGIH.

LIMITATIONS

Retrospective data, administrative database.

CONCLUSION

In-hospital mortality from UGIH has been decreasing over the past 2 decades, with a concomitant increase in rate of endoscopy and endoscopic therapy. However, despite decreasing length of stay, the total economic burden of UGIH is increasing.

Impact of day of admission on mortality and other outcomes in upper GI hemorrhage: a nationwide analysis

BACKGROUND

Studies have reached varying conclusions regarding the association between day of admission and outcomes in patients with upper GI hemorrhage (UGIH).

OBJECTIVES

To evaluate whether important outcomes in UGIH, including in-hospital mortality, differ between patients admitted on weekends versus weekdays.

DESIGN AND SETTING

Retrospective cohort study by using the 2009 Nationwide Inpatient Sample.

PATIENTS

Patients were included if they were adults with a principal diagnosis of acute UGIH. Patients admitted between midnight Friday and midnight Sunday were classified as weekend admissions.

MAIN OUTCOME MEASUREMENTS

In-hospital mortality, in-hospital endoscopy, endoscopic therapy, length of stay, and total hospitalization charges.

RESULTS

The study included 199,008 patients with nonvariceal UGIH and 3251 patients with variceal UGIH. Compared with patients admitted on weekdays, patients with nonvariceal UGIH admitted on weekends had similar adjusted in-hospital mortality rates (odds ratio [OR] 1.11; 95% confidence interval [CI], 0.93-1.30), endoscopic therapy rates (OR 0.98; 95% CI, 0.92-1.04), and length of stay (P = .09), but had lower early endoscopy rates (within 24 hours)(OR 0.64; 95% CI, 0.60-0.67), lower in-hospital endoscopy rates (OR 0.84; 95% CI, 0.78-0.91), and higher hospitalization charges (mean increase, $1558; P = .01). Patients with variceal UGIH admitted on weekends and weekdays did not differ in any of these outcomes.

LIMITATIONS

Retrospective data, administrative database.

CONCLUSIONS

Compared with patients admitted on weekdays, patients with nonvariceal UGIH admitted on weekends had similar mortality rates and lengths of stay, but lower endoscopy rates and higher hospitalization charges. Patients with variceal GI hemorrhage had similar outcomes regardless of day of admission.

The AIMS65 score compared with the Glasgow-Blatchford score in predicting outcomes in upper GI bleeding

INTRODUCTION

We previously derived and validated the AIMS65 score, a mortality prognostic scale for upper GI bleeding (UGIB).

OBJECTIVE

To validate the AIMS65 score in a different patient population and compare it with the Glasgow-Blatchford risk score (GBRS).

DESIGN

Retrospective cohort study.

PATIENTS

Adults with a primary diagnosis of UGIB.

PRIMARY OUTCOME

inpatient mortality.

SECONDARY OUTCOMES

composite clinical endpoint of inpatient mortality, rebleeding, and endoscopic, radiologic or surgical intervention; blood transfusion; intensive care unit admission; rebleeding; length of stay; timing of endoscopy. The area under the receiver-operating characteristic curve (AUROC) was calculated for each score.

RESULTS

Of the 278 study patients, 6.5% died and 35% experienced the composite clinical endpoint. The AIMS65 score was superior in predicting inpatient mortality (AUROC, 0.93 vs 0.68; P < .001), whereas the GBRS was superior in predicting blood transfusions (AUROC, 0.85 vs 0.65; P < .01) The 2 scores were similar in predicting the composite clinical endpoint (AUROC, 0.62 vs 0.68; P = .13) as well as the secondary outcomes. A GBRS of 10 and 12 or more maximized the sum of the sensitivity and specificity for inpatient mortality and rebleeding, respectively. The cutoff was 2 or more for the AIMS65 score for both outcomes.

LIMITATIONS

Retrospective, single-center study.

CONCLUSION

The AIMS65 score is superior to the GBRS in predicting inpatient mortality from UGIB, whereas the GBRS is superior for predicting blood transfusion. Both scores are similar in predicting the composite clinical endpoint and other outcomes in clinical care and resource use.