Digestive and Liver Disease 35 (2003) 157–164

www.elsevier.com / locate / dld

Alimentary Tract

Dyspepsia and Helicobacter pylori infection: a prospective multicentre

observational study
F. Perri a , *, V. Festa a , E. Grossi b , N. Garbagna b , G. Leandro c , A. Andriulli a , on behalf of the
‘‘NUD-LOOK’’ Study Group 1
a
Division of Gastroenterology, ‘‘ Casa Sollievo della Sofferenza’’ Hospital, I.R.C.C.S., San Giovanni Rotondo 71013, Italy
b
BRACCO SpA, Ethical Medical Department, Milan, Italy
c
‘‘ De Bellis’’ Hospital, I.R.C.C.S., Castellana Grotte, Italy
Received 18 June 2002; accepted 29 October 2002

Abstract

Objectives. Dyspepsia still represents an unsolved clinical enigma.

Aim. The aims of this study were to determine whether symptoms and Helicobacter pylori infection are predictors of organic disease in
uninvestigated dyspepsia, and if H. pylori eradication improves symptoms in functional dyspepsia.
Methods. An observational study was performed on outpatients with uninvestigated dyspepsia. Symptoms were scored and H. pylori
status determined. Patients with functional dyspepsia and H. pylori infection were randomly given either a standard eradicating treatment
or a 1-month course of empirical treatment. The latter was also given to functional dyspeptic patients without infection. Symptoms were
re-assessed in functional dyspeptic patients at 2- and 6-month follow-up visits. Patients receiving eradicating treatment were re-tested for
H. pylori at the 2 month visit.
Results. A total of 860 patients were studied and 605 (70.3%) were affected by functional dyspepsia. H. pylori infection was diagnosed
in 71.8% of patients with organic dyspepsia and in 65.0% with functional dyspepsia ( p50.053). Male sex, anaemia, smoking habit, age
over 45 years, and severe epigastric pain, but not H. pylori infection, were independent predictors of organic disease. Symptoms
significantly improved in most functional dyspeptic patients regardless of their H. pylori status and type of treatment.
Conclusion. H. pylori infection is not a strong predictor of organic disease in uninvestigated dyspepsia. H. pylori eradication is not
essential to improve symptoms in functional dyspepsia.
 2003 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Science Ireland Ltd. All rights reserved.

Keywords: Dyspepsia; Helicobacter pylori; Observational study

1. Introduction
Dyspepsia is a frequently occurring syndrome in the
general population [1,2]. The majority of patients with
dyspepsia do not have any definite anatomical or bio-
chemical explanation for their symptoms and are classified
as affected by functional dyspepsia [3]. These patients
often undergo unnecessary endoscopies, increasing both
the workload and waiting lists of Gastrointestinal Units.
For this reason, exhaustive research for the best predictors
of organic dyspepsia has been performed with the follow-
*Corresponding author. Tel.: 139-0882-410568; fax: 139-0882-
411879.
E-mail address: perrisgr@tin.it (F. Perri).
1
‘‘NUD-LOOK’’ Study Group participants are listed in , Appendix A.
ing risk factors identified: age over 45 years, male sex,
anaemia, abdominal mass, weight loss, family history of
gastric cancer or peptic ulcer, chronic use of NSAIDs,
dysphagia, cigarette smoking, jaundice, and vomiting [4].
Recently, H. pylori infection has been added to the list of
risk factors, and non-invasive screening tests have been
proposed to decide which patients should undergo endo-
scopy [5]. This strategy, however, does not seem to be
cost-effective and misses most H. pylori-negative patients
with esophagitis for whom endoscopy could be beneficial
[6].
The pathogenesis of functional dyspepsia is still obscure
[7]. Therefore, a clinical classification of functional
dyspeptic patients on the basis of their symptoms has been
proposed to possibly elucidate the underlying etiologic
factors and offer physicians empirical guidelines for

1590-8658 / 03 / $30  2003 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Science Ireland Ltd. All rights reserved.
doi:10.1016 / S1590-8658(03)00023-9
158 F. Perri et al. / Digestive and Liver Disease 35 (2003) 157–164
treatment [3]. The existence of patients suffering mainly
from fasting epigastric pain relieved by food and antacids,
and others complaining primarily of discomfort exacer-
bated by food and relieved by prokinetic drugs, recently
led a panel of experts to suggest the existence of at least
two main subgroups of functional dyspeptic patients,
namely those with ulcer-like dyspepsia and those with
dysmotility-like dyspepsia [3]. Ulcer-like dyspepsia shares
similarities with ulcer disease and could represent part of
its spectrum [8]. Several studies have shown that H. pylori
infection is present in more than 90% of patients with
gastric and duodenal ulcer [9–11]. Eradication of the
infection results in permanent ulcer healing and no ulcer
relapse [12,13]. Due to symptom similarities with ulcer
disease, a pathogenic role of H. pylori infection has been
proposed in ulcer-like dyspepsia [14,15]. Until now,
however, clinical studies on the possible association be-
tween H. pylori infection and specific dyspeptic symptoms
have produced conflicting results [16,17]. Similarly, it is
still a matter of debate whether eradication of H. pylori
infection may improve dyspeptic symptoms in all func-
tional dyspeptic patients [18,19] or, at least, in a specific
patient subgroup [20].
The aims of this prospective observational study were to
evaluate the proportion of patients referred to endoscopy
for uninvestigated dyspepsia who are affected by func-
tional dyspepsia, and to determine whether H. pylori status
is a risk factor for organic dyspepsia. Secondary aims were
to evaluate whether H. pylori infection is associated with a
specific symptom pattern and to study the influence of H.
pylori eradication on the outcome of symptoms in func-
tional dyspepsia.
2. Materials and methods
2.1. Study population
This study was conducted in Italy between 2000 and
2001. A total of 42 Italian gastrointestinal centres partici-
pated in the study. Each centre enrolled at least 20
outpatients with ‘‘uninvestigated dyspepsia’’ consecutively
referred for upper gastrointestinal endoscopy. Exclusion
criteria were any previous H. pylori eradication therapy,
treatment with antibiotics, H 2 -receptor antagonists, bis-
muth or proton pump inhibitors during the previous month,
concomitant occurrence of clinical symptoms due to either
an underlying biliary or pancreatic disease or irritable
bowel syndrome (IBS) and gastroesophageal reflux disease
(GERD). IBS was diagnosed on the basis of the presence
of at least three Manning’s criteria [21]: pain relieved by
defecation, looser stool at pain onset, more frequent stools
at pain onset, visible abdominal distension, feeling of
incomplete evacuation, mucus per rectum, and abdominal
bloating. GERD was defined by the presence of at least
one typical symptom (heartburn or acid regurgitation)
severe enough to affect patients’ activities.
All patients underwent careful history taking and phys-
ical examination, as well as upper gastrointestinal endo-
scopy. Alcohol use, consumption of antibiotics, NSAIDs
and anti-H 2 secretory drugs or proton pump inhibitors
within the last 4 weeks were recorded. Risk factors for
organic diseases were also investigated: anaemia, recent
weight loss, abdominal mass, family history of gastric
cancer and peptic ulcer disease, chronic use of NSAIDs,
smoking habit (.10 cigarettes a day), jaundice and
vomiting. Informed signed consent was obtained by all
participants before entering into the study.
2.2. Symptom questionnaire and patient subgroups
Each patient filled out a previously validated ques-
tionnaire [22]. Two well-categorized dyspeptic symptoms
were assessed: epigastric pain and epigastric discomfort.
Epigastric discomfort was defined as a subjective, unpleas-
ant sensation or feeling that is not interpreted as pain
according to the patient and which, if fully assessed, can
include any of the following symptoms: early satiety,
fullness, bloating and nausea. Each symptom was graded 0
to 3 according to its frequency (0, no symptoms; 1, less
than one episode a week; 2, more than one episode a week;
3, every day) and severity (1, mild (not influencing
patients’ activities); 2, relevant (diverting from but not
urging modifications of usual activities); and 3, severe
(interfering with ordinary activities of patients)). A global
score for epigastric pain was obtained by multiplying the
intensity score by the frequency score. For epigastric
discomfort, we used the highest score among those ob-
tained by multiplying the intensity scores by the frequency
scores of each of the aforementioned symptoms (i.e., early
satiety, fullness, bloating and nausea). Patients with un-
explained dyspepsia were included only if symptoms lasted
for at least 3 months with a global symptom score of at
least 2 or greater. They were grouped into three main
subgroups [3] according to the symptom score:
1. patients with ‘‘prevalent epigastric pain’’ (ulcer-like
dyspepsia) when the score for epigastric pain was
higher than that for epigastric discomfort;
2. patients with ‘‘prevalent epigastric discomfort’’
(dysmotility-like dyspepsia) when the score for epi-
gastric discomfort was higher than that for epigastric
pain;
3. patients with ‘‘unclassifiable dyspepsia’’, when there
was no difference in the scores for epigastric pain and
discomfort.
2.3. H. pylori status
H. pylori status was determined according to the inves-
 F. Perri et al. / Digestive and Liver Disease 35 (2003) 157–164
tigators’ preference on the basis of at least one of the
following tests: histology, serology, and rapid urease test
(RUT). All patients performed the urea breath test (UBT),
which was centralized in a single laboratory based in
Milan (Centro Diagnostico Italiano). Patients were consid-
ered infected when the UBT and at least one of the
previously mentioned tests were positive. After endoscopy,
patients were sorted into two main groups: patients with
organic dyspepsia (presenting with endoscopic lesions) and
patients with functional dyspepsia (absence of lesions
detected by endoscopy). A diffuse mucosal redness with-
out erosive changes of oesophageal, gastric or duodenal
mucosa was considered in accordance with functional
dyspepsia.
2.4. Treatment and follow-up
Patients with organic dyspepsia were treated according
to the endoscopic findings and dropped out of the study.
Patients with functional dyspepsia were sorted into two
groups: one with H. pylori infection and the other without
infection. Before starting the study, each centre decided
whether H. pylori-infected patients would have been
treated or not with an eradicating regimen. In each centre,
the investigator chose the eradicating treatment on the
basis of his / her personal experience. Using this approach,
different clusters of patients (each belonging to a single
centre policy) were created and combined into three main
treatment groups: infected patients who were given an
eradicating regimen, infected patients who were treated
with symptomatic therapy for 1 month, and non-infected
patients who received either antisecretory or prokinetic
drugs for 1 month. At 2 and 6 months after the initial visit,
all patients were re-evaluated by means of the same
clinical questionnaire, which was again filled out by the
patient. The subgroup of H. pylori-infected patients, who
were given antibiotics, underwent a centralized repeat
UBT to check eradication at the 2-month follow-up visit.
No antibiotics or antisecretory drugs were allowed before
the UBT assessment. At the 6-month visit, a global
assessment of symptoms was performed by clinical inves-
tigators, according to the following categories: 1, worsen-
ing of symptoms; 2, no improvement; 3, significant
improvement; 4, disappearance of symptoms.
2.5. Statistical analysis
A post-hoc analysis was preliminarily performed to
check for any significant difference in the results among
the centres. Clustered sampling was controlled by means
of multivariate analysis, with an a priori decision to
eradicate or not the infected patients taken as the outcome
variable. Univariate analysis was performed for age, sex,
alcohol consumption, smoking habits, and symptom scores
by sorting patients according to their endoscopic findings
and H. pylori status. Symptom scores were expressed as
159
means (61 SD). Student’s t-test was used for age and
symptom scores. Student’s t-test for paired data was used
for evaluating whether symptom scores significantly
changed during the follow-up. Pearson’s x 2 test was used
for sex, alcohol and smoking habits. Both univariate and
multivariate analysis was carried out to search for any
significant association between ‘‘alarm symptoms’’ (such
as anaemia, recent weight loss, abdominal mass, family
history of gastric cancer and peptic ulcer disease, chronic
use of NSAIDs, smoking, jaundice and vomiting), age,
severity of symptoms and H. pylori infection, and organic
disease. A cut-off value of $45 years was chosen for age.
A cut-off value of $4 was chosen for the global score
(obtained by multiplying the intensity score by the fre-
quency scores) for both epigastric pain and discomfort.
Odds ratios (OR) with 95% confidence intervals (CI) were
computed by means of x 2 analysis only for the indepen-
dent variables which entered into the model. Statistical
evaluation was performed by means of the statistical
software package SPSS / PC1 (SPSS Inc., Chicago, IL,
USA).
3. Results
3.1. Study population
A total of 860 patients (411 males; mean age 46.2614.8
years, range 16–83 years) were enrolled into the study:
730 (84.8%) were referred by general practitioners and
105 (12.2%) by gastroenterologists. A total of 25 (3%)
patients underwent endoscopy on their personal request. A
total of 269 (31.2%) patients were smokers and 131
(15.4%) were ex-smokers. A total of 722 (84%) patients
were non-drinkers while 138 (16%) drank more than 15 g
alcohol a day.
After endoscopy was performed, 605 (70.3%) patients
were diagnosed as being affected by functional dyspepsia:
203 (33.5%) were classified as suffering from ulcer-like
dyspepsia, 209 (34.5%) from dysmotility-like dyspepsia
and the remaining 193 (32%) from undetermined dyspep-
sia. The remaining 255 (29.7%) patients were affected by
organic disease. Final diagnoses are shown in Table 1
Table 1
Endoscopic findings in patients with organic dyspepsia
n %
Esophagitis 63 24.7
Gastric ulcer 17 6.7
Erosive gastritis 47 18.4
Atrophic gastritis 6 2.4
Gastric polyps 3 1.2
Gastric neoplasia 5 2.0
Duodenal ulcer 94 36.8
Duodenal scar 10 3.9
Erosive duodenitis 57 22.4
160 F. Perri et al. / Digestive and Liver Disease 35 (2003) 157–164

(some patients were affected by different diseases simul- Table 2

Global scores for epigastric pain in the study population at the 0-, 2-, and
taneously). The prevalence of gastric cancer in our patients
6-month visits
with uninvestigated dyspepsia was 0.6%.
The post-hoc analysis showed no significant difference Epigastric pain Time (months)
in the results among the centres. Importantly, the propor- 0 2 6
tion of patients with either organic dyspepsia or H. pylori OD 2.962.3 a

infection was similar among all the centres. Moreover, no FD 2.062.0 a

significant difference was observed between the two FD Hp1 2.162.1 b
groups of infected patients who were given or not eradicat- FD Hp2 2.062.0 b
ing treatment on the basis of clustered sampling. FD Hp1ER 1.862.0 a,c 0.461.2 a,d 0.461.3 c,d
FD Hp1NER 2.562.2 a,c 0.961.6 a,d 0.360.7 c,d
FD Hp1empirical Rx 2.262.1 a,c 0.761.2 a,d 0.661.3 c,d
3.2. H. pylori infection FD Hp2empirical Rx 2.062.0 a,c 0.460.8 a,b 0.260.6 c,b
OD, organic dyspepsia; FD, functional dyspepsia; ER, eradicated patients;
H. pylori infection was detected in 183 of 255 (71.8%) NER, non-eradicated patients after treatment; Rx, treatment.
a
patients with organic dyspepsia and in 393 of 605 (65.0%) p,0.0001.
b
p5n.s.
with functional dyspepsia ( p50.053; OR51.37; 95% d
p50.05.
CI50.996–1.89) (Fig. 1). The prevalence of H. pylori c
p,0.0001.
infection was significantly higher in patients with duodenal
ulcer or scar (90.4%) than in patients with other organic Table 3
diseases (62.3%) ( p,0.0001). On the contrary, no signifi- Global scores for epigastric discomfort in the study population at the 0-,
cant difference in the prevalence of H. pylori infection was 2-, and 6-month visits
observed among functional dyspeptic patients sorted by the Epigastric Time (months)
predominant symptoms into ulcer-like, dysmotility-like discomfort
0 2 6
and unclassifiable dyspepsia (Fig. 1). a
OD 1.461.9
FD 1.961.8 a

3.3. Symptoms FD Hp1 1.961.7 b

FD Hp2 1.861.9 b

For symptoms (pain and discomfort), the global score FD Hp1ER 2.362.0 a,c 0.861.4 a,d 0.460.9 c,d
FD Hp1NER 2.062.1 a,c 0.560.9 a,b 0.360.5 c,b
obtained by multiplying the intensity by the frequency
FD Hp1empirical Rx 1.861.9 a,c 0.861.3 a,d 0.561.0 c,d
score was considered. Patients with organic dyspepsia had FD Hp2empirical Rx 1.861.7 a,c 0.560.9 a,d 0.360.7 c,d
a significantly higher global score for epigastric pain at
OD, organic dyspepsia; FD, functional dyspepsia; ER, eradicated patients;
entry into the study than functional dyspeptic patients NER, non-eradicated patients after treatment; Rx, treatment.
(2.8662.27 vs. 2.0362.07; p,0.0001) (Table 2). In a
p,0.0001.
contrast, patients with functional dyspepsia had a sig- b
p5n.s.
c
nificantly higher global score for epigastric discomfort than p,0.0001.
d
patients with organic dyspepsia (1.8561.82 vs. 1.3761.86; p50.05.
p,0.0001) (Table 3). No significant difference in the
global score for epigastric pain or discomfort was found 3.4. Risk factors for organic dyspepsia
between functional dyspeptic patients with H. pylori and
those without the infection (Tables 2 and 3). A significant difference in the prevalence of male sex,

Fig. 1. Dyspeptic patients sorted by endoscopic findings (OD, organic dyspepsia; FD, functional dyspepsia) or pattern of symptoms (ULD, ulcer-like
dyspepsia; DLD, dysmotility-like dyspepsia; UD, unclassifiable dyspepsia) and H. pylori status.
 F. Perri et al. / Digestive and Liver Disease 35 (2003) 157–164 161

Table 4 ing 207 (52.6%) were treated with other empirical

Risk factors at univariate and multivariate analyses therapies (antisecretory or prokinetic drugs for 1 month).
FD OD p The 212 uninfected patients were also treated empirically
(%) (%) (for 1 month).
Univariate analysis At 2 months, 538 of 605 patients returned to the follow-
Age $45 years 44.7 53.7 ,0.05 up visit with a mean drop-out rate of 11% (Fig. 2). The
Male sex 44.2 56.2 ,0.001 proportion of patients who dropped out of the study was
Anaemia 2 5 ,0.02
Abdominal mass 0 0.8 ,0.03 similar in the groups sorted according to their H. pylori
Weight loss 3 4.3 n.s. status and treatment. Infected patients who received anti-
Family history of gastric cancer 5.3 3.5 n.s. biotics showed a mean eradication rate (CI) of 77.3%
Family history of peptic ulcer 19.3 23.1 n.s. (71–84) (on PP analysis) and 66.1% (59–73) (on ITT
Use of NSAID 6 5.9 n.s. analysis).
Dysphagia 41.7 58.1 n.s.
Cigarette smoking 15.1 33.3 ,0.0001 At 6 months, 472 patients returned to the final visit with
Jaundice 0.2 0.8 n.s. a drop-out rate of 12% (Fig. 2).
Vomiting 1.8 3.9 n.s. The global scores for epigastric pain and discomfort
H. pylori infection 64.9 71.5 n.s. were assessed at entry into the study and at the 2- and
Epigastric pain $4 30.1 47.7 ,0.0001 6-month visits. To evaluate symptom outcome after treat-
Discomfort $4 26.5 20.5 n.s.
ment, four groups of functional dyspeptic patients were
OR 95% CI considered (Fig. 2): infected patients who were uninfected
after treatment; infected patients who were still infected
Multivariate analysis
Anaemia 1.77 1.2022.62 after therapy; infected patients who had not been treated,
Male sex 1.40 1.1421.72 and uninfected patients (Tables 2 and 3). In all functional
Epigastric pain $4 1.26 1.1521.41 dyspeptic patients with H. pylori infection, the global
Cigarette smoking 1.22 1.1121.33 symptom scores for either pain or discomfort significantly
Age ($45 years) 1.08 1.0121.18
decreased at the 2- and 6-month follow-up, regardless of
the eradicating treatment or the success of therapy (Tables
2 and 3). In uninfected functional dyspeptic patients, a
anaemia, abdominal mass, smoking habits, age more than significant improvement of pain and discomfort was also
45 years, and severe epigastric pain was found between observed. In all patients the clinical improvement was
patients with organic and those with functional dyspepsia more impressive within 2 months after the endoscopic
at univariate analysis (Table 4). At multivariate analysis, procedure than in the following 4 months of follow-up
all mentioned factors except for abdominal mass were visits. For all patients with functional dyspepsia, the
independently associated with an increased risk for organic clinical judgement at the 6-month visit was the following:
disease (Table 4). 145 (30.8%) patients were symptom-free; 217 (45.9%)
improved; 98 (20.8%) had no symptom change; and only
3.5. Treatment and symptom outcome in patients with
functional dyspepsia

After the first clinical and endoscopic evaluation, the

605 patients with functional dyspepsia were treated differ-
ently according to the pre-determined policy of single
centres. Among the 393 patients with H. pylori infection,
186 (47.4%) were given an eradicating regimen (mainly a
1-week PPI-based triple therapy) (Table 5) and the remain-

Table 5
Eradicating treatments prescribed to 186 FD patients with H. pylori
infection
Regimen n %
One week OME1CLA1AMO 101 54.2
One week OME1CLA1TIN 48 25.8
One week OME1CLA1MTZ 25 13.4
Two weeks RBC-CLA 8 4.2
Other therapies 4 2.3
OME, omeprazole; CLA, clarithromycin; AMO, amoxicillin; TIN,
tinidazole; MTZ, metronidazole; RBC, ranitidine bismuth citrate. Fig. 2. FD patients at the 0-, 2-, and 6-month visit (Tx, treatment).
162 F. Perri et al. / Digestive and Liver Disease 35 (2003) 157–164
Table 6
Clinical judgment on dyspeptic symptoms in patients with functional dyspepsia at the 6-month visit. Percentages within parentheses
Patient Symptom outcome
Symptom-free
FD Hp1ER 30
FD Hp1NER 12
FD Hp1empirical Rx 41
FD Hp2empirical Rx 62
Total 145 (30.8)
ER, eradicated patients; NER, non-eradicated patients after treatment; Rx, treatment.
12 (2.5%) became worse (Table 6). No significant differ-
ence in symptom outcome was observed between the four
subgroups of patients with functional dyspepsia (Table 6).
4. Discussion
Observational studies are believed to overestimate treat-
ment results when compared to randomized, controlled
trials (RCTs) on the same topic. This belief is mainly
based on systematic comparisons between RCTs and
observational studies performed before 1970. More recent-
ly, however, it has been shown that there is no significant
difference in the magnitude of the effects of treatment
when the results of observational studies are compared
with those obtained from RCTs [23,24]. Therefore, the
fundamental criticism that, in observational studies unre-
cognized confounding factors may distort the results, is
likely to be invalid.
Observational studies are generally less prone to hetero-
geneity in the results than RCTs, since each observational
study is more likely to include a broad representation of
the population at risk. For this reason, observational
studies are generally preferred to RCTs in identifying risk
factors or prognostic indicators. When we consider one of
the questions to which the present study was addressed,
that is whether H. pylori eradication is beneficial in
functional dyspeptic patients, the answer we found from
the two largest RCTs published thus far was paradoxically
opposite [18,19]. Therefore, we decided to perform an
observational study adopting some principles of the design
of RCTs. In particular, we established some criteria for
determining the patient’s eligibility, the base-line features
(symptoms and H. pylori status), and the outcome mea-
sures. More importantly, concerning the eradicating treat-
ment, H. pylori-infected patients with functional dyspepsia
were a priori assigned to one of two groups according to a
cluster randomization design. For all these reasons, we
believe that the results obtained in this large, observational
study are accurate and reliable.
This study shows that about two-thirds of Italian patients
with uninvestigated dyspepsia are affected by functional
dyspepsia. Both the workload and the waiting lists of
endoscopic centres would be reduced by identifying the
Improved Unchanged Worse
70 12 1
17 2 0
61 60 5
69 24 6
217 (45.9) 98 (20.8) 12 (2.5)
predictors of organic dyspepsia. These factors would
enable clinicians to select patients who should be referred
to prompt endoscopy. In our study, anaemia, smoking,
male sex and age over 45 years, but not H. pylori
infection, were found to be independently associated with
organic disease as well as the type of predominant
symptom (pain) and its severity. However, the clinical
application of this finding is questionable since if we had
decided to perform endoscopy only in patients with at least
one of these risk factors, more than 95% of our patients
would have required endoscopy. Intriguingly, H. pylori
infection was not a predictor of organic disease. Although
the prevalence of infection was slightly higher in patients
with organic than functional dyspepsia, this difference was
not statistically significant. This finding, however, should
be interpreted cautiously for two reasons. First, the power
of the statistical test (0.44) was below the desired power of
0.80, such that a negative finding cannot be considered
absolutely true. Second, this finding is related in some way
to the proportion of organic diseases not associated with H.
pylori infection. The decision to perform endoscopy only
in patients infected by H. pylori should be considered on
the basis of the prevalence of H. pylori-negative ‘‘organic
diseases’’ in patients with uninvestigated dyspepsia.
Another interesting finding is that, at least in patients
with functional dyspepsia, neither a particular symptom
nor symptom severity are significantly associated with the
presence of infection, suggesting that H. pylori plays a
minimal role, if any, in causing symptoms. This hypothesis
is strengthened by the finding that, in our population, H.
pylori eradication was not essential to improve dyspeptic
symptoms. In fact, at the 6-month follow-up visit, three-
quarters of patients with functional dyspepsia were im-
proved regardless of their H. pylori status and the adopted
treatment. Treatment itself (but not successful H. pylori
eradication), spontaneous fluctuation of dyspeptic symp-
toms in time, and patient reassurance by a ‘‘negative’’
endoscopy could be crucial in improving symptoms.
Potential drawbacks of the present study are the rela-
tively short (6 months) follow-up period and the lack of a
control group without active treatment.
When we consider the largest published controlled
clinical trials on functional dyspepsia [18,19,25], the
proportion of patients who dropped out of the studies
 F. Perri et al. / Digestive and Liver Disease 35 (2003) 157–164
increases the longer the follow-up period. In our study, this
proportion approaches 25% at 6 months and would have
been even higher by prolonging the follow-up to 1 year.
Extrapolation of clinical status at 1 year, based on the
symptoms previously recorded at the last follow-up visit,
has been suggested by some authors to avoid the problem
of drop out patients [25]. However, this approach seems to
be methodologically incorrect since patients might have
been lost to follow-up because they sought new medical
advice for their worsening symptoms [26].
In general, observational studies do not enrol control
patients to whom a placebo is given. From a methodo-
logical point of view, this represents one of the main
differences between RCTs, in which both physicians and
patients are blinded to the treatment, and prospective
observational studies, in which a supposedly active treat-
ment is randomly given to patients. Before starting this
study, each participating centre determined its own thera-
peutic option (i.e., to treat or not to treat H. pylori
infection) and this makes the results of this study reliable.
A potential advantage of observational studies is that
patients do not fear being treated with placebo, and this is
extraordinarily important in a disease like functional
dyspepsia, in which anxiety is always over-expressed.
In conclusion, this prospective observational multi-cen-
tre study found that about two-thirds of Italian patients
with uninvestigated dyspepsia suffer from functional
dyspepsia, and H. pylori infection is associated neither
with an increased risk of organic disease nor with specific
symptoms. An improvement of dyspeptic symptoms is
observed at 6 months in most functional dyspeptic patients,
regardless of their H. pylori status and the type of
treatment.
Conflict of interest statement
None declared.
Appendix A. The ‘‘NUD-LOOK’’ Study Group
S. Lupo, Ospedale ‘‘S. Maria della Misericordia’’,
Albenga (SV); N. Casu, B. Chessa, Ospedale Civile,
Alghero (SS); G. Pomarico, Ospedale Civile, Andria (BA);
G. Assenza, Ospedale Civile ‘‘E. Muscatello’’, Augusta
(SR); D. Iannuziello, Clinica ‘‘S. Rita’’, Bari; G. Fuggi,
Ospedale ‘‘Fatebenefratelli’’, Benevento; P. Castagnoli,
Ospedale Civile, Civitavecchia (RM); G. Dante, Ospedale
Civile, Cairo Montenotte (SV); S. Bellentani, Ospedale
‘‘Ramazzini’’, Carpi (MO); L. Vitale, Ospedale Maggiore,
Chieri (TO); U. Bigazzi, Ospedale Civile, Citta` di Castello
(PG); G. Giudici, Ospedale ‘‘S. Biagio’’, Clusone (BG);
G. Minoli, Ospedale ‘‘Valduce’’, Como; T. Luisi, Presidio
Ospedaliero, Conversano (BA); R. Marin, Ospedale Civile
163
Dolo (VE); V. Di Ciommo, Ospedale ‘‘Bambin Gesu’’, `
Fiumicino (RM); F. Montanaro, Ospedale Civile ‘‘S.
Giovanni di Dio’’, Frattamaggiore (NA); B. Gaetano,
Ospedale Civile, Giugliano (NA); S. Rosati, Ospedale
Civile, Iesi (AN); A. Maurano, Ospedale Civile ‘‘Curteri’’,
Mercato San Severino (SA); F. Pietropaolo, Ospedale
Civile, Maddaloni (CE); V. Dambruoso, Ospedale Civile,
Montefiascone (VT); L. Tomarelli, R. Curto, Ospedale
Civile ‘‘SS. Benvenuto e Rocco’’, Osimo (AN); C. Di
Franco, Clinica Noto, Palermo; L. Barresi, Ospedale
‘‘Buccheri La Ferla-Fatebenefratelli’’, Palermo; L. Marzio,
R. Grossi, Casa di Cura ‘‘L. Pierangeli’’, Pescara; G.
Sbolli, Ospedale Civile, Piacenza; R. Marmo, Ospedale
Civile, Polla (SA); E. Suriani, Ospedale Civile degli
Infermi, Rivoli (TO); L. Mazzucca, Ospedale Civile,
Rogliano (CS); L. D’Anna, Policlinico Militare ‘‘Celio’’,
Roma; A. Grassi, Istituto Regina Elena, Roma; S. Monaco,
Ambulatorio Endoscopia Digestiva, San Giovanni La
Punta (CT); G. Scarpulla, Presidio Ospedaliero ‘‘M.
Raimondi’’, San Cataldo (CL); R. Clemente, M.
Quitadamo, Ospedale ‘‘Casa Sollievo della Sofferenza’’,
San Giovanni Rotondo (FG); A. Carrato, Clinica Lourdes,
San Sebastiano al Vesuvio (NA); S. Ghione, D. Cassine,
Ospedale Civile ‘‘SS. Annunziata’’, Savigliano (CN); A.
Longhini, P. Borelli, Ospedale Civile, Sondrio; C. Cor-
telezzi, Ospedale Multizonale di Circolo, Varese; A.
Spadaccini, Ospedale Civile, Vasto (CH); F. Biandrate,
Ospedale Civile, Vigevano (PV); N. Belluardo, Ospedale
Civile Quartiere Celle, Vittoria (RG).
List of abbreviations
DLD, dysmotility-like dyspepsia; ER, eradicated
patients; FD, functional dyspepsia; GERD, gas-
troesophageal reflux disease; IBS, irritable bowel
syndrome; NER, non-eradicated patients after treat-
ment; OD, organic dyspepsia; RUT, rapid urease
test; Rx, treatment; UBT, urea breath test; UD,
unclassifiable dyspepsia; ULD, ulcer-like dyspepsia.
References
[1] Adami HO, Agenas I, Gustavsson S, Loof L, Nyberg A, Nyren O, et
al. The clinical diagnosis of ‘‘gastritis’’. Aspects of demographic
epidemiologic and health care consumption based on a nation wide
sample survey. Scand J Gastroenterol 1984;19:216–9.
[2] Heatley RV, Rathbone BJ. Dyspepsia: a dilemma for doctors. Lancet
1987;ii:779–82.
[3] Talley NJ, Colin-Jones D, Koch KJ, Koch M, Nyren ´ O, Stanghellini
V. Functional dyspepsia: a classification with guidelines for diag-
nosis and management. Gastroenterol Int 1991;4:145–60.
´ L, Talley N. Challenges in managing dyspepsia in general
[4] Agreus
practice. Br Med J 1997;315:1284–8.
[5] Sobala GM, Crabtree JE, Pentith JA, Rathbone BJ, Shallcross TM,
164 F. Perri et al. / Digestive and Liver Disease 35 (2003) 157–164
Wyatt JI, et al. Screening dyspepsia by serology to Helicobacter
pylori. Lancet 1991;338:94–6.
[6] Hession PT, Malagelada J. Review article: the initial management of
uninvestigated dyspepsia in younger patients: the value of symptom-
guided strategies should be reconsidered. Aliment Pharmacol Ther
2000;14:379–88.
[7] Funch-Jensen P, Talley NJ, Drossman DA. Review in depth:
functional dyspepsia. Eur J Gastroenterol Hepatol 1992;4:595–625.
[8] O’Morain C, Buckley M. Helicobacter pylori and dyspepsia. Scand
J Gastroenterol 1996;31(Suppl. 214):28–30.
[9] Rauws EAJ, Langenberg W, Houthoff HJ, Zanen HC, Tytgat GN.
Campylobacter pyloridis-associated chronic active antral gastritis. A
prospective study of its prevalence and the effect of antibacterial and
antiulcer treatment. Gastroenterology 1988;94:33–40.
[10] Sipponen P, Seppala K, Aarynen M, Helske T, Kettunen P. Chronic
gastritis and gastroduodenal ulcer: a case-control study on risk of
coexisting duodenal or gastric ulcer in patients with gastritis. Gut
1989;30:922–9.
[11] Tytgat GNJ, Rauws EAJ. Campylobacter pylori and its role in
peptic ulcer disease. Gastroenterol Clin North Am 1990;19:183–96.
[12] Marshall BJ, Goodwin CS, Warren JR, Murray R, Blincow ED,
Blackbourn SJ, et al. Prospective double-blind trial of duodenal
ulcer relapse after eradication of Campylobacter pylori. Lancet
1988;ii:1437–42.
[13] Rauws EAJ, Tytgat GNJ. Cure of duodenal ulcer associated with
eradication of Helicobacter pylori. Lancet 1990;i:1233–5.
[14] McColl K. Role of gastric acid in the aetiology of dyspeptic disease
and dyspepsia. Baillieres Clin Gastroenterol 1998;12:489–502.
[15] Perri F, Clemente R, Festa V, Annese V, Quitadamo M, Rutgeerts P,
et al. Patterns of symptoms in functional dyspepsia: role of
Helicobacter pylori infection and delayed gastric emptying. Am J
Gastroenterol 1998;93:2082–8.
[16] Lambert JR. The role of Helicobacter pylori in non-ulcer dyspepsia:
a debate—for. Gastroenterol Clin North Am 1993;22:141–51.
[17] Talley NJ. The role of Helicobacter pylori in non-ulcer dyspepsia: a
debate—against. Gastroenterol Clin North Am 1993;22:153–67.
[18] Blum A, Talley N, O’Morain C, van Zanten SV, Labenz J, Stolte M,
et al. Lack of effect of treating Helicobacter pylori infection in
patients with nonulcer dyspepsia. New Engl J Med 1998;339:1875–
81.
[19] McKoll K, Murray L, El-Omar E, Dickson A, El-Nujumi A, Wirz A,
et al. Symptomatic benefit from eradicating Helicobacter pylori
infection in patients with nonulcer dyspepsia. New Engl J Med
1998;339:1869–74.
[20] Asante M, Mendall M, Northfield T. Which Helicobacter pylori-
positive dyspeptics are likely to respond symptomatically to empiri-
cal H. pylori eradication? Eur J Gastroenterol Hepatol 1998;10:265–
8.
[21] Manning A, Thompson W, Heaton K, Morris AF. Toward positive
diagnosis of the irritable bowel. Br Med J 1978;2:653–4.
[22] Tucci A, Corinaldesi R, Stanghellini V, Tosetti C, Di Febo G,
Paparo GF, et al. Helicobacter pylori infection and gastric function
in patients with chronic idiopathic dyspepsia. Gastroenterology
1992;103:768–74.
[23] Concato J, Shah N, Horwitz RI. Randomized, controlled trials,
observational studies, and the hierarchy of research designs. New
Engl J Med 2000;342:1887–92.
[24] Benson K, Hartz AJ. A comparison of observational studies and
randomized, controlled trials. New Engl J Med 2000;342:1878–86.
[25] Talley N, Vakil N, Ballard E, Fennerty MB. Absence of benefit of
eradicating Helicobacter pylori in patients with nonulcer dyspepsia.
New Engl J Med 1999;341:1106–11.
[26] Perri F, Andriulli AM. Absence of benefit of eradicating Helicobac-
ter pylori in patients with nonulcer dyspepsia (letter). New Engl J
Med 2000;342:589–90.