Download Predictive Factors of Mortality From Nonvariceal Upper

American Journal of Gastroenterology
C 2008 by Am. Coll. of Gastroenterology
Published by Blackwell Publishing
ISSN 0002-9270
doi: 10.1111/j.1572-0241.2008.01865.x
ORIGINAL CONTRIBUTION
Endoscopy
CME
Predictive Factors of Mortality From Nonvariceal Upper
Gastrointestinal Hemorrhage: A Multicenter Study
Riccardo Marmo, M.D., Maurizio Koch, M.D., Livio Cipolletta, M.D., Lucio Capurso, M.D., Angelo Pera, M.D.,
Maria A. Bianco, M.D., Rodolfo Rocca, M.D., Angelo Dezi, M.D., Renato Fasoli, M.D., Sergio Brunati, M.D.,
Ivano Lorenzini, M.D., Ugo Germani, M.D., Giovanni Di Matteo, M.D., Paolo Giorgio, M.D.,
Giorgio Imperiali, M.D., Giorgio Minoli, M.D., Fausto Barberani, M.D., Sandro Boschetto, M.D., Marco
Martorano, M.D., Giovanni Gatto, M.D., Mariano Amuso, M.D., Alfredo Pastorelli, M.D., Elena S. Torre, M.D.,
Omero Triossi, M.D., Andrea Buzzi, M.D., Renzo Cestari, M.D., Domenico Della Casa, M.D., Massimo Proietti,
M.D., Anna Tanzilli, M.D., Giovanni Aragona, M.D., Francesco Giangregorio, M.D., Luciano Allegretta, M.D.,
Salvatore Tronci, M.D., Paolo Michetti, M.D., Paola Romagnoli, M.D., Andrea Nucci, M.D., Francesca Rogai,
M.D., Walter Piubello, M.D., Maria Tebaldi, M.D., Fabrizio Bonfante, M.D., Alessandro Casadei, M.D., Claudio
Cortini, M.D., Giorgio Chiozzini, M.D., Lisa Girardi, M.D., Claudio Leoci, M.D., Giampiero Bagnalasta, M.D.,
Sergio Segato, M.D., Giuseppe Chianese, M.D., Mario Salvagnini, M.D., Gianluca Rotondano, M.D. – For the
Italian Registry on Upper Gastrointestinal Bleeding (Progetto Nazionale Emorragie Digestive – PNED)
OBJECTIVES:
From an Italian Registry of patients with upper gastrointestinal hemorrhage (UGIH), we assessed the
clinical outcomes and explored the roles of clinical, endoscopic, and therapeutic factors on 30-day
mortality in a real life setting.
METHODS:
Prospective analysis of consecutive patients endoscoped for UGIH at 23 community and tertiary
care institutions from 2003 to 2004. Covariates and outcomes were defined a priori and 30-day
follow-up obtained. Logistic regression analysis identified predictors of mortality.
RESULTS:
One thousand and twenty patients were included. A total of 46 patients died for an overall 4.5%
mortality rate. In all, 85% of deaths were associated with one or more major comorbidity. Sixteen of
46 patients (35%) died within the first 24 h of the onset of bleeding. Of these, eight had been
categorized as ASA class 1 or 2 and none of them was operated upon, despite a failure of
endoscopic intention to treatment in four. Regression analysis showed advanced age, presence of
severe comorbidity, low hemoglobin levels at presentation, and worsening health status as the only
independent predictors of 30-day mortality (P < 0.001). The acute use of a PPI exerted a protective
effect (OR 0.23, 95% CI 0.09–0.73). Recurrent bleeding was low (3.2%). Rebleeders accounted for
only 11% of the total patients deceased (OR 3.27, 95% CI 1.5–11.2).
CONCLUSIONS: These results indicate that 30-day mortality for nonvariceal bleeding is low. Deaths occurred
predominantly in elderly patients with severe comorbidities or those with failure of endoscopic
intention to treatment.
(Am J Gastroenterol 2008;103:1639–1647)
INTRODUCTION
Acute upper gastrointestinal hemorrhage (UGIH) remains a
common reason for hospital admission, requiring high levels of assistance and health-care expenditures. The incidence
of acute UGIH and hospital-related 30-day mortality vary
widely, partly because of the changing epidemiology in western countries, in which a number of factors, such as older age,
comorbidites, prescriptions for proton pump inhibitors (PPI),
low-dose aspirin, oral anticoagulants, cyclooxygenase-2 inhibitors, are involved (1–6). Risk assessment is mainly based
on endoscopic recognition of high-risk stigmata at ulcer base
To access a continuing medical education exam for this article, please visit
www.acg.gi.org/journalcme.
(7, 8) and controversy exists in the literature about the predicting value of clinical, biochemical, or therapeutic variables on
the risk of death.
The main aim of this prospective multicenter database
study was to describe the ultimate outcome of patients with
nonvariceal UGIH in a contemporary “real-life” setting. Additional analysis assessed the impact of clinical, endoscopic,
or therapeutic factors on the risk of death in this patient population.
METHODS
The PNED Initiative and Data Collection
A dedicated software including an endoscopic reporting system (Cartella Clinica Endoscopia Digestiva, Bracco, Italy)
1639
1640
Marmo et al.
linked to a project-specific research database was developed. This software was distributed to 23 participating sites
across Italy, establishing a network of centers that received
emergency admissions from which source data were collected
– the Progetto Nazionale Emorragie Digestive (PNED) study
group. Fifteen community hospitals and eight tertiary care
institutions were included. At each site, a “lead” consultant
represented the project locally. At each hospital, a coordinator identified subjects daily in the accident and emergency
department, the wards, the endoscopy unit, the operating theater, and from blood transfusion records and admission data.
Specially trained research assistants collected data directly
from the patient’s point of care into case report forms. The
coordinator was then responsible for checking and returning a completed form for each patient correctly identified.
All data were denominalized and downloaded into a central
database on a monthly basis. They were reviewed at a single national location for internal logic of patient flow and
biological plausibility. All data queries were resolved within
30 days following original data entry. To help assure the internal validity of the registry, there was an independent data
validation of a random subset of all information collected;
quality of data was validated on a quarterly basis by randomly
comparing 5% of all records to the source data recorded in
the hospital charts. Personnel from a clinical research organization trained all research staff at a common startup meeting
and at each initial on-site visit prior to the first patient entry.
On that occasion, all endoscopists participating in the registry
were invited to review a wide set of video images of different
bleeding stigmata in order to estimate inter-rater variability
and find as much agreement as possible on the diagnosis of
stigmata.
Patient Population
Patients were considered for the study if they had clinical
evidence of overt UGIH on admission or a history of hematemesis/coffee ground vomiting, melena, hematochezia,
or a combination of any of the above within 24 h preceding the admission, or clinical evidence of acute UGIH while
hospitalized for any other reason (in-hospital bleeding), independently of their age. UGIH was confirmed only if either hematemesis, melena, or dark, tarry materials on rectal
examination was documented and witnessed by nursing or
medical staff. Patients were entered in the registry only if
an upper GI endoscopy was performed. In case of bleeding
from esophago-gastric varices, data were initially computed
as patients having an UGIH, but were then excluded from the
prospective database.
An audit of all patients presenting over a fixed time period,
at each participating institution, was performed to rule out
any selection bias in the way in which the study population
was enrolled. Patients initially assessed at another hospital for
the bleeding episode and subsequently transferred to one of
the participating centers were excluded from the analysis.
Study Variables
The following independent variables were included in the
electronic form: demographics (age, sex, site, and date of
endoscopy); historical data (presenting signs or symptoms,
any significant comorbidity, the patient’s physical status on
presentation using the American Society of Anesthesiologists [ASA] classification (9), relevant past medical history,
any concomitant intake of medications in the 7 days preceding the bleeding episode, time elapsed from the onset of
bleeding); physical examination findings and laboratory data
(hemodynamic data, rectal exam, nasogastric tube use, complete blood count and coagulation parameters); and resuscitative measures. The endoscopic report included identification of the bleeding lesion, stigmata of recent hemorrhage
at ulcer base according to the Forrest classification (active
bleeding, visible vessel, or adherent clots were classified as
“high-risk,” whereas flat spot and clean based ulcer as “lowrisk”) (8), detailed description of any method of endoscopic
hemostasis delivered, number of therapeutic attempts, timing, and negative outcomes of endotherapy. The presence of
fresh blood into the stomach in such amount to hamper endoscopic diagnosis of the source of bleeding was annotated
and underscored as a “bleeding but source not identified at
EGD.” The impossibility to conclude an already started endoscopic procedure with the intention to treat the source of
hemorrhage was defined as “failed intention to endoscopic
treatment.” Both surgical and angiographic therapies were
recorded as well as any pharmacologic therapy administered
for hemorrhage (types, amount, and route of administration).
The outcomes evaluated were the frequency of death, recurrent bleeding, and need for surgery. Such outcomes were
monitored from the admission to the hospital or the onset
of bleeding for in-hospital patients up to 30 days after the
endoscopic examination. Both investigators and nurses all
worked with the same operational definitions of outcome. A
priori definitions for all outcomes were adopted according to
established definitions (10). Thirty-day mortality was the primary investigated outcome; a “bleeding-related” death was
defined as any death occurring within 30 days of the indexbleeding episode. Continued or persistent bleeding was defined as (a) failure to control arterial bleeding endoscopically;
(b) presence of bloody nasogastric aspirate; (c) shock with a
pulse greater than 100 beats/min, a systolic blood pressure
<100 mmHg, or both; and/or (d) the need for substantial
replacement of blood or fluid volume (need of more than 4
units of blood in 6 h). Persistent hemorrhage was usually considered an indication for emergency surgery or percutaneous
embolization. Nonetheless, the decision as to when to resort
to surgery in the individual case was left to the discretion of
the local teams. Recurrent bleeding was defined by recurrent
vomiting of fresh blood, melena, or both with either shock or a
decrease in hemoglobin concentration of at least 2 g/L following initial successful treatment and after initial stabilization of
24 h. Rebleeding had to be confirmed by a second endoscopy
(visible bleeding or stigmata of recent hemorrhage). Patient
transfer to and from other institutions and any readmission
was also noted. All data were recorded for the full duration of
the initial medical encounter. To ensure the completeness of
follow-up information, the study nurses called all patients or
their families at 30 days. Furthermore, after PNED had been
Thirty-Day Mortality for Nonvariceal GI Bleeding in Italy
completed, administrative databases were consulted and all
charts of included patients were reviewed for a full 30 days
following admission or onset of bleeding while in hospital.
Data Analysis
Descriptive data were generated for independent and dependent variables. All categorical data were expressed as proportions, whereas all continuous data were expressed as mean
± standard deviations. Where appropriate, alternate descriptive statistics such as quartile ranges for categorical data and
medians with ranges for continuous data are reported. All
calculations were performed using the SPSS statistical package version 8.0 (MJ Norusis, Chicago, IL). Variance analysis
was used for comparisons of continuous data using discriminant analysis to select the cutoff values that best separated
those deceased from those nondeceased. Such cutoff values
were used to dichotomize continuous variables. Differences
in proportions were analyzed using the χ 2 test and odds ratio
with 95% confidence intervals when appropriate. Backwards
step-wise logistic regression analysis was performed to determine independent predictors of 30-day mortality. A P value
<5% was considered statistically significant. To measure the
goodness of fit (i.e., strength of association in a nonlinear
regression model), both the Cox and Snell’s R-Square and
Nagelkerke’s R-Square were used.
Ethics
The registry was approved by the Institutional Review Board
of all participating centers. In addition, all eligible patients
were asked to sign a written informed consent.
RESULTS
Study Population
A total of 1,263 cases were identified and entered in the PNED
registry between March 2003 and March 2004 (with data
queries collected until April 2004). Thirty-nine patients who
were in transfers from other hospitals were excluded. Also
excluded were 205 patients whose source of bleeding was
esophago-gastric varices. Therefore, 1,020 patients affected
with nonvariceal hemorrhage were analyzed. The mean enrolment per site was 53 ± 48 patients. The 450 cases (44%
of all PNED population) enrolled by the top five recruiting centers represented 80% of all patients with UGIH seen
in those institutions during the study period. Patient characteristics are outlined in Table 1. A recent history of drugs’
consumption was recorded in a high proportion of upper GI
bleeders, mostly nonsteroidal anti-inflammatory drugs (34%)
and low-dose aspirin (17.5%). One or more comorbidities
were recorded at the time of presentation in 60.6% of cases
(618/1,020), with 25% (255/1,020) having more than two
comorbidities. The median number of comorbid conditions
per patient was 1.0 (IQR 0–2.0), mainly cardiovascular diseases affecting almost half of the patients. Mean length of
stay was 7.2 ± 6.3 days (median 4.0, IQR 2.0–9.0). Full 30-
1641
Table 1. Characteristics of the Study Population
Total Population (N = 1,020)
Mean
(Median, IQR)
Male
62%
Age (yr)
68 (70, 53–81)
Number of comorbidities
1.0 (1.0, 0–2)
ASA score
1–2
42%
3
45%
4
13%
In-hospital bleeding
14%
Symptoms on presentation
Melena
78%
Hematemesis
28%
Hematochezia
9%
Timing of endoscopy (h)
8 (10, 2–14)
Medications at presentation
NSAIDs
34%
Aspirin
17.5%
Anticoagulants
8.2%
Ca++ -antagonists
7.5%
Steroids
6.7%
Nitrates
6.3%
Coxib
3.3%
Initial hemodynamic instability
27%
Laboratory results
Initial mean hemoglobin (g/dL) 9.6 (9.8, 6.9–11.4)
Hematocrit
31 (30, 24–35)
INR
1.4 (1.2, 1.0–1.3)
Source of bleeding at endoscopy
Duodenal ulcer
37.8%
Gastric ulcer
28.3%
Gastroduodenal erosions
9.1%
Neoplasia
5.7%
Mallory-Weiss tears
4.5%
Esophagitis
3.6%
Dieulafoy’s lesion
1.4%
Other vascular source
2.4%
No lesion identified
7.2%
SRH at ulcer base (N = 675)
Clean base
33.2%
Flat spot
22.0%
Adherent clot
11.7%
Visible vessel
17.7%
Active bleeding
15.4%
95%
CI/SD
58.8–64.8
±16
±1.3
39.2–44.5
42.5–47.3
11.8–15.7
12–16
76.0–81.1
25.0–30.0
7.9–11.5
±10
31.1–36.9
15.3–20.0
6.7–10.0
6.0–9.2
5.5–7.8
4.9–7.9
2.5–4.0
24.7–30.2
±2.4 SD
±7 SD
±1.3 SD
34.9–40.9
25.6–31.2
7.4–10.9
4–0–6.9
3.8–5.5
3.0–5.0
0.9–2.4
1.5–3.4
5.7–8.9
29.7–36.8
18.9–25.2
9.4–14.3
15.0–20.8
12.8–18.3
IQR = interquartile range; SD = standard deviation; CI = confidence interval;
NSAIDs = nonsteroidal anti-inflammatory drugs; SRH = stigmata of recent
hemorrhage.
Mean daily dose of aspirin 100 ± 25 mg.
ASA score refers to the American Society of Anesthesiologists classification of a
patient’s severity and acuity of disease index.
day follow-up information was available for 1,008 (98.8%)
patients.
Endoscopic Data
Overall, 1,844 endoscopic procedures were performed with
a median of one procedure per patient (range 1–5). Upper
GI endoscopy was first performed in 78% of patients within
24 h of presentation to the emergency room or of the onset of bleeding in patients already hospitalized. Peptic ulcers
1642
Marmo et al.
Table 2. Endoscopic and/or Medical Treatment and Outcomes in
Patients With Nonvariceal Bleeding
No endoscopic treatment
Endoscopic treatment
Injection
Thermal coagulation
Hemoclips
Combined injection + thermal
Medical therapy
PPI
H2 RA
PPI and H2 RA
Vasoactive drugs (somatostatin
or octreotide)
Outcome
Clinical rebleeding
Need for emergency surgery
30-day-mortality
No. of
Patients
Percent
(95% CI)
633
387
168
113
74
32
62.0 (58.2–65.3)
38.0 (35.7–41.8)
43.4 (39.8–47.3)
29.2 (25.7–33.7)
19.1 (15.3–22.5)
8.3 (6.8–10.3)
586
194
100
96
57.5 (54.4–60.5)
19.0 (16.2–20.8)
9.8 (8.1–11.7)
9.4 (7.7–11.3)
33
13
46
3.2 (2.2–4.5)
1.3 (0.7–2.2)
4.5 (3.3–6.0)
PPI = proton pump inhibitors; H2 RA = histamine-2 receptor antagonists.
were by far the most common cause of bleeding, accounting for 66% of the cases (gastric 28%). The distribution of diagnoses was similar for inpatients and emergency
admissions except that Mallory-Weiss lesions were more
common in those admitted urgently. No lesion could be identified as source of the hemorrhage in 7.2% of the cases
(Table 1). No patient received erythromycin or other prokinetic agents to attempt to clean out the gastric blood before
endoscopy.
On endoscopy, actively bleeding ulcers were found in
15.4% of patients, whereas visible vessels and adherent clots
at ulcer base were diagnosed in 17.7% and 11.7% of cases,
respectively. Low-risk endoscopic stigmata were detected in
55.2% of ulcer patients (22% pigmented material, 33.2%
clean base). Overall, endoscopic treatment was delivered in
387 patients, and all patients with “high-risk stigmata” received endoscopic therapy. Injection therapy alone with diluted epinephrine was performed in 43.4%, isolated thermal
probe coagulation in 29.2%, hemoclips in 19.1%, and combined endotherapy with injection followed by thermal coagulation in 8.3% (Table 2).
Pharmacotherapy
The vast majority of the bleeding patients (67.5%) were given
PPIs, either intravenously or orally, during hospitalization or
emergency room visit. Among patients with high-risk lesions
at endoscopy, the mean daily doses of IV and oral PPI were
212 ± 80 mg and 80 ± 40 mg, respectively. Overall, 19%
of UGI bleeders were treated with IV ranitidine. Ten percent
of patients received both PPI and H2 -receptor antagonists
and 9.4% received vasoactive drugs such as somatostatin or
octreotide. Ninety-seven percent of patients undergoing endoscopic hemostasis also received a PPI.
Helicobater pylori Management
Testing for H. pylori was performed during hospital stay in
62% of patients (all by means of histologic evaluation, as per
protocol). Among patients who were tested, H. pylori infection was detected in 44.3% (95% CI 41.3–47.4) of patients
overall. Eradication therapy was initiated during hospitalization in less than 10% of cases.
Outcomes
Recurrent bleeding was recorded in 33 patients (3.2%, [2.2–
4.5]), whereas emergency surgery was deemed necessary in
only 13 cases (1.3%, [0.7–2.2]). Overall, 46 patients died
with a 30-day mortality rate of 4.5% [3.3–6.0], with no significant differences for the inpatient group compared with
emergency admissions (Table 2). The median time to death
was 4 days (95% CI 2–6). In the absence of any comorbidity, mortality was 0.7%. It increased to 8.4% if one severe
comorbidity was recorded, and to 23.1% if two or more severe comorbidities were present (P < 0.01). The mean age
of the patients who died was 76.6 ± 14.0 yr. In all, 85% of
deaths (39 out of 46) were associated with one or more severe
comorbidities. Sixteen patients (34.8%) died within the first
24 h of the onset of bleeding, with 10 patients deceased within
8 h of the endoscopic examination. Eight patients had been
categorized as severely unstable or moribund (ASA class 3
or 4), whereas eight patients as healthy or moderately healthy
(ASA class 1 or 2) and none of them was operated upon. In
four of them, there had been a failure of endoscopic intention
to treatment. No patient was treated radiologically by means
of percutaneous embolization. For the remaining 30 patients,
death occurred within 7 days in 50% and within 12 days
in 75% of cases. Subgroup analysis according to the health
status showed that 3.4% (33/934) of patients categorized as
ASA 1 or 2 died versus 25.5% (13/38) of patients categorized
as ASA 3 or 4 (P < 0.0001, OR 9.68 [4.7–19.8]). During the
first 24 h, 8/960 (0.8%) ASA 1 or 2 patients died versus
9/42 (17.6%) of ASA 3 or 4 patients (P < 0.0001) [OR 29.4
(9.1–96.6)].
Advanced age, hemodynamic instability, hemoglobin concentration, and size of the ulcer(s) were all significantly associated with the risk of death at univariate analysis (Table 3).
Diastolic pressure <70 mmHg (OR 2.46), concomitant use of
nitrates (OR 1.82), systolic pressure <115 mmHg (OR 1.78),
hematemesis (OR 1.66), and active bleeding at endoscopy
Table 3. Predictors of 30-Day Mortality at Univariate Analysis
(Continuous Variables)
Predictors
Mean age (± SD) in years
Diastolic pressure (mmHg)
Heart rate (bpm)
Hemoglobin (g/dL)
Size of duodenal ulcer (mm)
Size of gastric ulcer (mm)
bpm = beats per minute.
Death
(N = 46)
76.6 ± 14.0
66.3 ± 15.2
96.7 ± 18.7
8.2 ± 2.8
12.7 ± 7.3
17.9 ±12.8
No Death
(N = 974)
P
Value
67.5 ± 16.0 <0.001
72.5 ± 11.6 0.001
89.1 ± 14.4 0.001
9.2 ± 2.5
0.001
9.4 ± 5.4
0.04
11.3 ± 6.7
0.003
Thirty-Day Mortality for Nonvariceal GI Bleeding in Italy
1643
Table 4. Predictors of 30-Day Mortality at Univariate Analysis (Categorical Data)
Predictors
Clinical
In-hospital bleeding
Hematemesis
Renal failure
Neoplasia
Liver cirrhosis
Endoscopic
Blood in gastric cavity, source not identified
Failed intention to endoscopic treatment
Gastric cancer
Recurrent bleeding
Need for surgery
No. of Deaths/No. of
Patients With Predictor
No. of Deaths/No. of
Patients Without Predictor
P
Value
Odds Ratio
(95% CI)
10/142 (7.0%)
20/282 (7.1%)
8/71 (11.3%)
14/81 (17.3%)
7/69 (10.1%)
36/878 (4.1%)
26/738 (3.5%)
38/949 (4.1%)
32/939 (3.4%)
39/951 (4.1%)
0.126
0.018
0.011
0.0001
0.03
1.77 (0.8–3.6)
2.0 (1.1–3.8)
3.0 (1.4–6.8)
5.9 (3.0–11.6)
2.6 (1.1–6.1)
5/13 (38.5%)
6/8 (75%)
7/48 (14.6%)
5/33 (15.2%)
2 /13 (15.4%)
41/1,007 (4.1%)
40/1,012 (4%)
39/972 (4.0%)
41/987 (4.2%)
44/1,007 (4.4%)
<0.001
<0.001
0.004
0.003
0.057
14.7 (4.6–46.9)
72.9 (14.2–372)
4.0 (1.7–9.7)
4.12 (1.5–11.2)
3.9 (0.85–18.5)
EGD = esophago-gastro-duodenoscopy.
The following variables were all NOT associated with the risk of death: previous nonvariceal hemorrhage, history of peptic ulcer disease, previous gastric resection, comorbid
illnesses such as cardiovascular disease or diabetes, medications (clopidogrel, low-dose aspirin, COX-2 inhibitors, other NSAIDs, anticoagulants, heparin, steroids or antidepressants),
syncope or melena on presentation, naso-gastric tube findings of coffee-ground material or bright red blood mixed with clots; type of endoscopic lesions such as Mallory-Weiss
tears, Dieulafoy’s ulcers, angiodysplasias, esophagitis, gastro-duodenal erosions; site of either gastric or duodenal ulcer; major endoscopic stigmata of bleeding; type of endoscopic
therapy delivered; timing of endoscopy; size of the hospital, community or university affiliation, geographic location (by region).
(OR 1.30) were all clinically relevant OR but did not reach
statistical significance and were therefore not inserted in the
final multivariate model.
Similar prognostic impact had the presence of “severe”
comorbid illnesses such as renal failure, liver cirrhosis and
neoplasia, or the presence of blood in the gastric cavity in such
amount to hamper both diagnostic and therapeutic endoscopy
(Table 4). A total of 345 patients (33.8%) received one or
more transfusions with packet red cells. Thirty-day mortality was 7.5% in patients with transfusion requirements (OR
2.67, 95% CI 1.46–4.85). Thirty-day mortality was 15.2%
in patients with recurrent bleeding and recurrent bleeders
accounted for 10.9% of the total patients deceased (OR 4.12,
95% CI 1.5–11.2). Resort to emergency surgery was rare;
only 13 patients (1.3%) were operated and 2 of them died.
Thirty-day mortality was therefore 15.4% in patients needing
surgery who accounted for 4.3% of the total deaths (OR 3.97,
95% CI 0.85–18.5) (Table 4).
Adjuvant medical therapy with antisecretory drugs is likely
to impact on the risk of death after endoscopic treatment.
Thirty-day mortality rate was significantly higher in patients
treated with H2 -receptor antagonists than in those treated with
PPIs (15% vs 4%, P < 0.0001) (OR 4.28 [95% CI 1.36–
11.34]). Conversely, the acute use of a PPI showed a protective effect (OR 0.23, 95% CI 0.09–0.73), with a relative risk
reduction of 76.6% (95% CI 34.5–90.5%) and an absolute
risk reduction of 11.5% (95% CI 8.6–14.3).
In the logistic regression analysis, independent predictors
of 30-day mortality included advanced age (over 80 yr), low
hemoglobin levels at presentation (<7 g/dL), presence of severe comorbidity (renal failure, neoplasia, liver cirrhosis),
and worsening health status (ASA class 1 or 2 vs ASA class
3 or 4) (Table 5). The Nagelkerke R2 of our final model was
0.79. The interaction between the two major prognostic factors, i.e., advanced age and severe comorbidity, on the risk of
30-day mortality is outlined in Figure 1.
DISCUSSION
Acute UGIH remains a common medical problem that
has significant associated morbidity, 30-day mortality, and
health-care resource use. The PNED data indicate that ulcers
are by far the most common cause of nonvariceal UGIH, accounting for 66% of all diagnoses. More than a half of the
patients in the present study were taking at least one ulcerogenic drug, mainly NSAIDs, whereas the use of low dose
aspirin was much lower than reported in North America (11,
12) and this should be taken into account in terms of generalizability of results.
An important feature of the PNED study is that it represents
a consortium of practice sites that use a structured endoscopy
reporting system to collect information in a centralized endoscopic database. Such data are useful because they reflect
“real world” endoscopic practice from a wide range of practice settings and minimize patient selection or potential referral bias. Of the 23 participating centers, in fact, only eight
Table 5. Independent Predictors of 30-Day Mortality (Logistic Regression Analysis)
Predictors
Rebleeding
ASA class 3
Time to admission∗
Hemoglobin <7 g/dL
Age >80 yr
ASA class 4
Renal failure
Advance neoplasia†
Liver cirrhosis
Failure of endoscopic
intention to treatment
∗
†
Odds
Ratio
95% CI
P
Value
3.27
3.32
4.35
4.70
4.70
7.24
5.29
9.81
11.01
11.35
−1.1–7.7
0.56–6.1
−0.1–8.8
2.3–7.0
2.3–7.0
4.0–10.4
2.8–7.8
7.3–12.3
7.5–14.5
−1.4–24.1
0.1166
0.0206
0.0544
0.0004
0.0005
0.0009
0.0004
0.0000
0.0002
0.0616
Time elapsed between patient’s perception of malaise and seeking medical attention;
Karnofsky index ≤70.
1644
Marmo et al.
Figure 1. The interaction between advanced age and severe comorbidity on the risk of 30-day mortality.
were tertiary institutions, unlike the Canadian RUGBE study
in which tertiary institutions accounted for two-thirds of the
participating centers and 62% of patients were enrolled at six
centers (13).
All ulcer-bleeding patients with high-risk stigmata had an
endoscopic treatment. About 43% of patients who were given
endoscopic treatment received epinephrine injection therapy,
probably a less than optimal endoscopic therapy according to
consensus recommendations (14), although in line with other
recent European surveys on this matter (15, 16). Thermal
probe coagulation or mechanical securing of hemostasis by
means of hemoclips were employed in one-third and one-fifth
of the cases, respectively, whereas combination therapy was
seldom applied.
Outcomes in the PNED registry were comparable to previously published large databases, except the recurrent bleeding
rate of 3.2% which is much lower than previously reported (1,
5, 7, 13, 15, 16). Such difference may be possibly explained
with the strict definition of “clinically relevant” recurrent
bleeding (confirmation endoscopy required), which may in
part have reduced the number of patients labeled as such; the
relatively high percentage (44%) of nonulcer bleeding; the
relatively low proportion of patients with actively bleeding
lesions (only 15.4% vs a 27% rate in the RUGBE registry)
(13); the fact that all patients with high-risk stigmata were
treated endoscopically and the widespread use of PPIs after endoscopic therapy. Also, the mean dosages of adjuvant
PPI therapy were sensibly higher than those reported in the
Canadian study where most patients may not have received
their PPI in the appropriate manner (58% had a mean daily
dose ≤100 mg). This may have played a role in preventing
rebleeding, particularly in patients with adherent clots (17).
Although it is possible that some patients may have rebled and
gone to another hospital at that time, such an event is unlikely
in the manner in which patients were cared for in the participating institutions. The usual timing of a recurrent bleeding
and the mean duration of hospital stay of patients included
in PNED make us quite confident that the number of missed
negative outcomes is very low. Nonetheless, the inability to
identify prognosticators of rebleeding due to its low prevalence should be acknowledged as an important limitation of
the present study.
Mortality from nonvariceal UGIH is usually in the range
of 5–10% in most studies that emphasize prompt resuscitation, early endoscopic diagnosis, and treatment followed
by dedicated multidisciplinary postoperative care (13, 15).
The 4.5% 30-day mortality rate recorded in the PNED study
is very similar to the 5.4% death rate reported in the wellconducted RUGBE study, which ran similar analysis of prognostication of death (13). Nonetheless, one or more severe
or life-threatening comorbidity was present in 85% of the
patients who died, a figure comparing favorably with other
reports (13, 15).
Risk stratification in patients with UGIH is essential for
optimal management of bleeders, both for triage of those at
high risk to inpatient care and for identification of patients at
low risk of adverse outcome who can be safely managed as
outpatients (14, 18, 19). The Rockall score, prospectively and
externally validated as effective in different patient populations, was selected for use by the investigators involved in the
PNED initiative. However, as pointed out in a recent review
on the topic by Das and Wong (20), at present there is not an
Thirty-Day Mortality for Nonvariceal GI Bleeding in Italy
ideal risk score. For acute bleeders, urgent EGD should not
be an essential part of the initial risk score; however, EGD
should be an important component of the risk score at a subsequent point in patient care. Hence, the need to categorize
the patient’s risk profile with the aid of accurate and userfriendly clinical predictors that are typically available during
initial patient triage.
Our study identified a number of prognostic clinical factors
independently associated with the risk of 30-day mortality after an acute UGI bleed. The most relevant predictors were advanced age, liver cirrhosis, chronic renal failure and advanced
neoplasia, and the low hemoglobin level at presentation. In the
presence of any of these prognosticators, emergency physicians as well as managing gastroenterologists should pursue
a high level of care to prevent death, because about 30% of
deaths occur within the first 24 h of admission to hospital,
and part of this mortality within the first 8 h. Interestingly, the
same RUGBE investigators recently reported that a modified
clinical risk score including hemoglobin level, hemodynamic
instability, presence of liver disease, or cardiac failure was
significant in multivariate analysis for rebleeding and mortality (21).
The impossibility, for whatever reason, to perform endoscopic therapy should be considered as a strong predictor of
negative outcome, which was recorded in 6 out of 8 ASA 1
or 2 patients who died within the first 24 h. The absence of
statistical significance is clearly due to a high standard error
related to the small number of patients in which endoscopic
therapy could not be performed. Patients who fail endoscopic
therapy are best served with immediate surgery or considered
for percutaneous embolization. A surprising and somewhat
awkward finding was that half of those who died within the
first 24 h were not operated upon despite being at good surgical risk nor did they go for an angiogram instead, and in
four of them there was a failure at endoscopic control of
hemorrhage. Whether surgeons were reluctant to bring such
persistent bleeders to the operation theater or gastroenterologists unwisely insisted on attempts at hemostasis needs to
be verified, and clear-cut management protocols should be
shared in this setting.
The main strength of the PNED registry is that it focused
on the most relevant outcome measure, i.e., death. Previous
studies have noted that a large proportion of deaths occur
in those with severe comorbidity in whom even minor hemorrhage is poorly tolerated, independently of any effective
endoscopic intervention (5, 12, 14–16, 22). Furthermore, we
identified in detail that advanced age, low hemoglobin levels
at presentation, chronic renal failure, neoplasia, and liver cirrhosis were all powerful independent predictors of death. It
is difficult to establish what is and what is not a preventable
death; however, the identification of potential risk factors may
help us select those elements that need to be closely monitored in an attempt to further reduce 30-day mortality. The
fact that most deaths occur in elderly patients with severe
comorbidities should prompt us to switch our focus to improve management of this selected high-risk subgroup, to be
1645
included into specifically designed trials. Furthermore, major
preventive steps should be taken in order to reduce the risk
of NSAIDs’ related hemorrhage, particularly in the highest
risk population.
There are several limitations to the present registry. The
adopted study design is not an experimental one, and therefore not as rigorous as that of a randomized controlled trial.
The demonstrated associations should be seen as suggestive,
but require prospective independent validation, which is actually under way. One specific limitation may be the robustness
of our final model, given that multiple covariates for only 46
events were analyzed. Some important clinical variables did
not reach statistical significance at univariate analysis despite
having clinically relevant odds ratio values. A beta-error cannot be excluded and with larger number of events also other
variables (e.g., recurrent bleeding) may likely provide additional information.
We attempted to address other possible shortcomings such
as completeness of follow-up, ascertainment of outcomes,
possible patient selection bias and inadequate adjustment for
confounders by establishing conservative and a priori definitions for all study variables including outcomes, by training all research staff in a standardized fashion, by enforcing
strict data verification and validation protocols and by ensuring complete 30-day follow-up. Although defining all 30-day
mortalities to be “bleeding-related” could be considered a
weakness of the study, any death within this time frame can
be logically related to the bleeding episode, acting as either
determining or precipitating event for the negative outcome.
In conclusion, death after an acute nonvariceal UGIH occurs mostly among elderly patients with severe comorbidity
or those with failure of endoscopic intention to treatment.
These factors should be taken into account in a struggle toward a further reduction of overall 30-day mortality. Future
studies should reconsider early resort to surgery, especially
for young and surgically fit patients, to reduce the risk of death
during the first 24 h of the bleeding episode. Also, the use
of preventive strategies to reduce the bleeding risk especially
in patients with advanced age, neoplasia, renal failure, and
liver cirrhosis deserves appropriate investigation. Improving
the ultimate outcome of patients with nonvariceal UGIH will
take an integrated approach by a team focused on treating the
patient and not just the source of bleeding.
ACKNOWLEDGMENTS
The PNED initiative was a collaborative effort supported by
the Italian Society for Digestive Endoscopy (SIED) and the
Italian Association of Hospital Gastroenterologists (AIGO).
We acknowledge the great deal of work performed by medical
and nursing staff in each of the participating units.
APPENDIX
The PNED investigators’ group includes Riccardo Marmo,
Ospedale Curto, Polla; Livio Cipolletta, Gianluca Rotondano,
1646
Marmo et al.
and Maria A. Bianco, Ospedale Maresca, Torre del Greco;
Lucio Capurso, Maurizio Koch and Angelo Dezi, ACO San
Filippo Neri, Rome; Angelo Pera and Rodolfo Rocca, Ospedale Mauriziano Umberto I, Torino; Fausto Barberani and
Sandro Boschetto, Ospedale San Camillo De Lellis, Rieti;
Alfredo Pastorelli and Elena Sanz Torre, Ospedale Bel Colle,
Viterbo; Sergio Brunati and Renato Fasoli, Ospedale Cantù,
Abbiategrasso; Ivano Lorenzini and Ugo Germani, AO Umberto I, Ancona; Giorgio Minoli and Giorgio Imperiali, Ospedale Valduce, Como; Giovanni Gatto and Mariano Amuso,
AO Villa Sofia, Palermo; Massimo Proietti and Anna Tanzilli,
Ospedale Del Prete, Pontecorvo; Walter Piubello, Maria
Tebaldi and Fabrizio Bonfante, Ospedale di Desenzano, Desenzano del Garda; Renzo Cestari and Domenico Della Casa,
AO Ospedali Civili, University of Brescia, Brescia; Paolo
Michetti and Paola Romagnoli, Ospedali Galliera, Genova;
Omero Triossi and Andrea Buzzi, Ospedale Santa Maria delle
Croci, Ravenna; Alessandro Casadei and Claudio Cortini,
AO Morgagni, Forlı̀; Giorgio Chiozzini and Lisa Girardi,
Ospedale Umberto I, Mestre; Luciano Allegretta and Salvatore Tronci, Ospedale Santa Caterina Novella, Galatina; Giovanni Aragona and Francesco Giangregorio, Ospedale Civile,
Piacenza; Sergio Segato and Giuseppe Chianese, AO Ospedale Circolo e Fondazione Macchi, Varese; Andrea Nucci
and Francesca Rogai, AO Ospedale Careggi, Firenze; Giampiero Bagnalasta and Claudio Leoci, Ospedale Civile,
Manerbio; Giovanni Di Matteo and Paolo Giorgio, IRCCS
Ospedale De Bellis, Castellana Grotte; Marco Martorano, Ospedale dell’Immacolata, Sapri; Mario Salvagnini, Ospedale
San Bortolo, Vicenza.
STUDY HIGHLIGHTS
What Is Current Knowledge
r
Mortality for nonvariceal upper GI bleeding varies
widely and has been reported from 5% to 14%. Predictors of recurrent bleeding and mortality have been
identified in previous registries, showing endoscopic
intervention and adjuvant proton pump inhibitor (PPI)
therapy to exert a protective effect.
What Is New Here
r
In this Italian multicenter registry, 30-day mortality was
low (4.5%). Mortality occurred in 30% of cases within
the first 24 h of the index bleeding. Recurrent bleeding was infrequent and had a lower impact on the risk
of mortality. Multivariate logistic regression identified
advanced age, presence of severe comorbidities, and
low hemoglobin levels at presentation as major clinical predictors of poor outcome. All these parameters
can be used pre-endoscopy to identify a subset of highrisk patients in which a high level of care is warranted
to prevent negative outcomes.
Reprint requests and correspondence: Riccardo Marmo, M.D.,
Division of Gastroenterology, Hospital “L.Curto,” Polla, Via Sottobraida 32, 84037 Sant’Arsenio, Italy.
Received May 25, 2007; accepted January 10, 2008.
REFERENCES
1. Longstreth GF. Epidemiology of hospitalization for acute
upper gastrointestinal hemorrhage: A population-based
study. Am J Gastroenterol 1995;90:206–10.
2. Yavorski RT, Wong RK, Maydonovitch C, et al. Analysis
of 3294 cases of upper gastrointestinal bleeding in military medical facilities. Am J Gastroenterol 1995;90:568–
73.
3. Rockall TA, Logan RF, Devlin HB, et al. Incidence of and
30-day mortality from acute upper gastrointestinal haemorrhage in the United Kingdom. BMJ 1995;311:222–6.
4. Higham J, Kang JY, Majeed A. Recent trends in admissions
and 30-day mortality due to peptic ulcer in England: Increasing frequency of haemorrhage among older subjects.
Gut 2002;50:460–4.
5. Vreeburg EM, Snel P, de Bruijne JW, et al. Acute upper gastrointestinal bleeding in the Amsterdam area: Incidence, diagnosis, and clinical outcome. Am J Gastroenterol
1997;92:236–43.
6. Esrailian E, Gralnek IM. Nonvariceal upper gastrointestinal bleeding: Epidemiology and diagnosis. Gastroenterol
Clin North Am 2005;34:589–605.
7. Rockall TA, Logan RF, Devlin HB, et al. Risk assessment after acute upper gastrointestinal haemorrhage. Gut
1996;38:316–21.
8. Laine L, Peterson WL. Bleeding peptic ulcer. N Engl J
Med 1994;331:717–27.
9. Owens W, Felts J, Spitznagel E. ASA physical status classifications: A study of consistency of ratings. Anesthesiology
1978;49:239–43.
10. Cooper GS, Chak A, Lloyd LE, et al. The accuracy of
diagnosis and procedural codes for patients with upper GI
haemorrhage. Gastrointest Endosc 2000;51:423–6.
11. Lewis JD, Bilker WB, Brensinger C, et al. Hospitalization and 30-day mortality rates from peptic ulcer disease
and GI bleeding in the 1990s: relationship to sales of nonsteroidal anti-inflammatory drugs and acid suppression
medications. Am J Gastroenterol 2002;97:2540–9.
12. Peura DA, Lanza FL, Gostout CJ, et al. The American College of Gastroenterology Bleeding Registry: Preliminary
findings. Am J Gastroenterol 1997;92:924–8.
13. Barkun A, Sabbah S, Enns R, et al. Endoscopic hemostasis and proton pump inhibition are associated with improved outcomes in a real-life setting. Am J Gastroenterol
2004;99:1238–46.
14. Barkun A, Bardou M, Marshall JK, and the Nonvariceal
Upper GI Bleeding Consensus Conference Group. Consensus recommendations for managing patients with nonvariceal upper gastrointestinal bleeding. Ann Intern Med
2003;139:843–57.
15. van Leerdam ME, Vreeburg EM, Rauws EA, et al. Acute
upper GI bleeding: Did anything change? Time trend analysis of incidence and outcome of acute upper GI bleeding between 1993/1994 and 2000. Am J Gastroenterol
2003;98:1494–9.
16. Ramsoekh D, Leerdam van ME, Rauws EA, et al. Outcome
of peptic ulcer bleeding, nonsteroidal anti-inflammatory
drug use, and Helicobacter pylori infection. Clin Gastroenterol Hepatol 2005;3:859–64.
Thirty-Day Mortality for Nonvariceal GI Bleeding in Italy
17. Sung JJ, Chan FK, Lau JY, et al. The effect of endoscopic
therapy in patients receiving omeprazole for bleeding ulcers with nonbleeding visible vessels or adherent clots.
A randomized comparison. Ann Intern Med 2003;139:
237–43.
18. Cebollero-Santamaria F, Smith J, Gioe S, et al. Selective
outpatient management of upper gastrointestinal bleeding in the elderly. Am J Gastroenterol 1999;94:1242–
7.
19. Cipolletta L, Bianco MA, Rotondano G, et al. Outpatient
management for low-risk nonvariceal upper gastrointestinal bleeding: A randomized controlled trial. Gastrointest
Endosc 2002;55:1–5.
20. Das A, Wong RC. Prediction of outcome of acute GI hemorrhage: A review of risk scores and predictive models.
Gastrointest Endosc 2004;60:85–93.
21. Romagnuolo J, Barkun AN, Enns R, et al. Simple clinical predictors may obviate urgent endoscopy in selected
1647
patients with nonvariceal upper gastrointestinal tract bleeding. Arch Intern Med 2007;167:265–70.
22. Marmo R, Rotondano G, Piscopo R, et al. Dual therapy
versus monotherapy in the endoscopic treatment of highrisk bleeeding ulcers: A meta-analysis of controlled trials.
Am J Gastroenterol 2007;102:279–89.
CONFLICT OF INTEREST
Guarantor of the article: Riccardo Marmo, M.D.
Specific author contributions: Each author contributed in
data entrance from his/her own center. Drs. Marmo and Koch
developed the study protocol, Dr. Marmo analyzed data, and
Dr Rotondano wrote the draft of the manuscript.
Financial support: None.
Potential competing interests: None.