ORIGINAL CONTRIBUTIONS

Incidence of Campylobacter and Salmonella Infections

Following First Prescription for PPI: A Cohort Study
Using Routine Data
Sinead Brophy, PhD1, Kerina H. Jones, PhD2, Muhammad A. Rahman, PhD3, Shang-Ming Zhou, PhD4, Ann John, MD5,
Mark D. Atkinson, PhD6, Nick Francis, PhD7, Ronan A. Lyons, MD8 and Frank Dunstan, PhD9
OBJECTIVES:

To examine the incidence of Campylobacter and Salmonella infection in patients prescribed proton
pump inhibitors (PPIs) compared with controls.

METHODS:

Retrospective cohort study using anonymous general practitioner (GP) data. Anonymised individuallevel records from the Secure Anonymised Information Linkage (SAIL) system between 1990 and
2010 in Wales were selected. Data were available from 1,913,925 individuals including 358,938
prescribed a PPI. The main outcome measures examined included incidence of Campylobacter or
Salmonella infection following a prescription for PPI.

RESULTS:

The rate of Campylobacter and Salmonella infections was already at 3.16.9 times that of non-PPI
patients even before PPI prescription. The PPI group had an increased hazard rate of infection
(after prescription for PPI) of 1.46 for Campylobacter and 1.2 for Salmonella, compared with
baseline. However, the non-PPI patients also had an increased hazard ratio with time. In fact,
the ratio of events in the PPI group compared with the non-PPI group using the prior event rate ratio
was 1.17 (95% CI 0.741.61) for Campylobacter and 1.00 (0.51.5) for Salmonella.

CONCLUSIONS: People who go on to be prescribed PPIs have a greater underlying risk of gastrointestinal (GI)

infection beforehand and they have a higher prevalence of risk factors before PPI prescription.
The rate of diagnosis of infection is increasing with time regardless of PPI use, and there is no
evidence that PPI is associated with an increase in diagnosed GI infection. It is likely that factors
associated with the demographic profile of the patient are the main contributors to increased rate of
GI infection for patients prescribed PPIs.
Am J Gastroenterol advance online publication, 16 April 2013; doi:10.1038/ajg.2013.30

INTRODUCTION
A major role of gastric acid is to protect against bacterial infection (1).
Proton pump inhibitors (PPIs) and histamine receptor antagonists (H2RAs) are among a group of medications used to treat gastric acid-related disorders by reducing gastric acid secretion. PPIs
are considered to be more potent and effective acid suppressers
than H2Ras (2), and as such they are prescribed for conditions
such as gastric or duodenal ulcers, and gastro-oesophogeal reflux
disease, in accordance with the National Institute for Health and
Clinical Excellence guidelines (3,4). H2RAs are available over the
counter in low dosage in many countries, including the United
Kingdom (2).

It follows that PPIs, which act by reducing acid secretion, could

increase the risk of gastrointestinal (GI) infections by raising the
pH environment of the stomach and making it more prone to colonisation by various pathogenic bacteria. Previous case control studies have examined the association between acid suppression and
various GI infections, and they have found an increased risk of GI
infection in patients prescribed PPIs (1,5). For example, PPIs have
been associated with Clostridium difficile (1,59) and the US-FDA
issued a safety announcement on PPI (10). In addition, there have
been conflicting reports for pneumonia (1113) and other infections such as Salmonella and Campylobacter (14). However, the
majority of studies exclude cases where a pre-existing GI infection

Swansea University, Swansea, UK; 2Swansea University, Swansea, UK; 3Swansea University, Swansea, UK; 4Department of StatisticsSwansea University,
Swansea, UK; 5Public Health Wales, Swansea, UK; 6Swansea University, Swansea, UK; 7Cardiff University, Wales, UK; 8Department of Public Health Medicine,
Swansea University, Swansea, UK; 9Department of Medical Statistics, Cardiff University, Wales, UK. Correspondence: Sinead Brophy, PhD, College of Medicine,
Swansea University, Swansea SA2 8PP, UK. E-mail: s.brophy@swansea.ac.uk
Received 18 November 2012; accepted 24 January 2013
2013 by the American College of Gastroenterology

The American Journal of GASTROENTEROLOGY

STOMACH

nature publishing group

STOMACH

B Sinead et al.

is present before the prescription of PPI. Therefore, the incidence

of GI infection in a given population before and after the use of PPI
has not been evaluated. This means that it remains unclear if the
increased risk of GI infection is owing to factors in the population
associated with the prescription of PPIs (such as comorbidities,
lifestyle factors, or older age) or to the medications. Therefore, this
study used Campylobacter and Salmonella infections as examples
of significant GI infections, and compared the incidences of these
infections over a 12-month period before and after the prescription of PPIs, thus controlling for individual confounding factors
by using the same patients as their own control. The effect of time
is also examined by comparing these PPI patients with patients
not prescribed a PPI during the duration of the study follow-up
(19902010) and with controls matched for time period. Thus,
there is the exposed group (patient post PPI) and three different
controls; (1) non-exposed self controls (patient pre-PPI), (2) nonPPI patients in the year 1991 and again in 2009, in order to examine incidence owing to time changes, and (3) non-PPI controls
matched on date (day, month, year) with those prescribed a PPI to
control for time changes.

Table 1. Read codes used

PPI

a6b%, a6c%, a6e%, a6f%, a6h%

H2RA

a61.., a611., a612., a613., a614., a615., a616.,

a617. ,a618., a619., a61a., a61A., a61b., a61B.,
a61c., a61C., a61d., a61D., a61e., a61E., a61f.,
a61F., a61g., a61G., a61H., a61I., a61J., a61K.,
a61L., a61M., a61N., a61O., a61P., a61Q., a61R.,
a61s., a61S., a61t., a61T., a61u., a61U., a61v.,
a61w., a61x., a61y., a61z., a62.., a621., a622.,
a623., a624., a625., a626., a627., a628., a629.,
a62A., a62B., a62C., a62D., a62E., a62F., a62G.,
a62H., a62I., a62J., a62K., a62L., a62M., a62N.,
a62O., a62P., a62Q., a62u., a62v., a62v., a62w.,
a62x., a62y., a62z., a68.., a681., a682., a683.,
a684., a69.., a691., a692., a693., a694., a695.,
a696., a697., a698., a6d.., a6d1., a6d2

Campylobacter

A0743

Salmonella

A02., A020., A021., A0220, A022z, A02y., A02z.,

Ayu02, Ayu03

Antibiotic and
antifungals/virals

e%

Arthritis

N%

Immunosupresants

h%

METHODS

Oral steroid

6633.,663a.,663F.,663G.

Routinely collected data

Leukaemia

B64%,B65%,B66%,B67%,B68%,B69%

Lymphoma

B602%

Splenectomy

78401,78403,78410,78404

Aspirin

bu2%, di1%, j11%, blm%

Naproxen

j2c%,di8%

Ibuprofen

di6%, dic%, j28%, j2p%, j2t%, ja%

Bowel infection
(ICD-10)

K58%, R19%

Bowel infection
(OPCS 4)

H23%, H24%, H25%, H62%, Z29%

Campylobacter
(ICD-10)

A045

Salmonella
(ICD-10)

A029, A020, A021, A022, A028

The Health Information Research Unit (HIRU) at the College

of Medicine at Swansea University represents part of the Welsh
Assembly Governments commitment to the UK Clinical Research
Collaboration. The main aim of HIRU is to realise the potential of
electronically held routinely collected information to conduct and
support health-related research, and HIRU has set up the Secure
Anonymised Information Linkage (SAIL) databank to achieve this
(15,16). The SAIL databank brings together and links a wide range
of person-based data from multiple sources relevant to health. SAIL
utilises a range of measures to ensure that the data are anonymous
and secure, and that they can be safely utilised for research within
a robust information governance framework (15). The range of
complementary sets of data includes existing routinely collected
data sets such as general practitioner (GP) records, out-patients
clinical data, in-patients episodes, accident and emergency department, pathology, and social services; plus special feeds of data such
as clinical data from rheumatologists. Thus SAIL has been established using disparate data sets, and close to 2 billion records from
multiple health and social care service providers have been loaded
to date, with further growth in progress. Currently 35% (168/484)
of the GP practices in Wales provide data to SAIL, giving 1,913,925
(40%) individuals out of 4,713,970 registered with a GP practice in
Wales over the 20 years follow-up period. The percentage of male
is 49% in both the total population and the SAIL sample, and mean
age is 39.8 in both the population and the sample.

sets to be included in SAIL are subjected to a matching process

using the MACRAL (Matching Algorithm for Consistent Results
in Anonymised Linkage) algorithm to create a unique, encrypted
anonymised linking field for each person (16). The use of an anonymised linking field enables anonymous linkage at the person
level and allows us to follow the patient pathway through the NHS
system both retrospectively and prospectively while preserving
patient privacy.

Data linkage

Ethics

Data linkage of anonymous data sets is made possible by the

allocation of a consistent, unique identifier to each individual.
This is carried out by an NHS Trusted Third Party so that HIRU
does not handle personal identifiable data. In this process, data

The data held by HIRU in the SAIL system are anonymised and
have been obtained with the permission of the relevant Caldicott
Guardian/Data Protection Officer; therefore the National Research
Ethics Service has stated that no ethical review is required.

The American Journal of GASTROENTEROLOGY

H2RAs, histamine receptor antagonists; PPI, proton pump inhibitor.

VOLUME 104 | XXX 2013 www.amjgastro.com

PPI and Camplyobacter/Salmonella Infection

Average age (s.d.)

PPI patients
(n =358,938)

Non-PPIa
(n =1,523,828)

46% (119,313)

49.5% (4107)

58.05 (16.7)

51.04 (19.6)

PPI patients
(n =358,938)

Non-PPIa
(n =1,523,828)

020 years

10/10,222
(0.98 per 1,000)

253/349,755
(0.72 per 1,000)

2140 years

69/74,536
(0.93 per 1,000)

137/474392
(0.29 per 1,000)

4160 years

100/123,424
(0.81 per 1,000)

95/298,093
(0.31 per 1,000)

More than 60

109/167,515
(0.65 per 1,000)

93/321,242
(0.29 per 100)

GI infection by age (Salmonella)

Prescription in previous 12 months

H2RAb

10.3% (36,800)

0.5% (4,181)

Antibiotic

40% (146,377)

6.6% (52,287)

0.12% (440)

0.01% (100)

33.7% (120,820)

3.6% (28,252)

2.6% (9,302)

0.21% (1,642)

Diagnosed with arthritis

(before start date)

23.2% (83,240)

3.4% (26,549)

Bowel surgery
(before start date)

1.15% (4,130)

0.05% (405)

GI infection by season
(Campylobacter)

N=3,447

N=3,533

Spring (MarchMay)

760 (22%)

817 (23%)

Summer
(JuneAugust)

1058 (31%)

1189 (34%)

Autumn
(SeptemberNovember)

909 (26%)

894 (25%)

Winter
(DecemberFebruary)

720 (21%)

633 (18%)

N=128

N=190

Spring (MarchMay)

27 (21%)

46 (24%)

Summer (JuneAugust)

38 (30%)

50 (26%)

Analysis

Autumn
(SeptemberNovember)

38 (30%)

49 (26%)

Winter
(DecemberFebruary)

25 (19%)

45 (24%)

All records were examined for a diagnosis of the human Campylobacter GI infection or Salmonella GI infection (see Table 1) within
12 months of being prescribed a PPI, or within 12 months of a
randomly selected date for the non-PPI comparison group. Of
the patients prescribed a PPI, 19% were prescribed a PPI between
1990 and 1999 and 81% were prescribed a PPI between 2000 and
2010 (see Figure 1). The number of Campylobacter or Salmonella
GI infections were assessed for (a) the 12-month period before
PPI prescription, (b) the 12-month period post PPI prescription,
(c) patients not prescribed a PPI (12 months before date and 12
months after date). For the non-PPI patients, we examine before
and after the dates of 1991 and before and after 2009. These subsets of the single cohort were chosen to examine the effect of
time on infection rates. Finally, we selected a sample matched for
date with the PPI patients (matched on day, month, year). The
matched samples were used to control for variations in infection
rate with time. STATA version 11 using the st command was used
to examine incidence of infection over a 12-month period in all
groups. Time to event analysis was used to examine infection in
the year before prescription of PPI compared with the year after
the PPI. Cox regression models were used to compare the incidence of infection over each of the two 12-month periods in those
with a PPI prescription and controls, using the cluster command
to produce robust s.e.s. Chi-squared tests were used to examine

Oral steroid
NSAID
Immunosuppressants

GI infection by season
(Salmonella)

GI infection by gender (Campylobacter)

Male

148/119,110
(1.24 per 1,000)

113/412,942
(0.27 per 1,000)

Female

142/140,039
(1.01 per 1,000)

79/421,068
(0.19 per 1,000)

GI infection by age (Campylobacter)

020 years

11/10,222
(1.08 per 1,000)

73/349,755
(0.21 per 1,000)

2140 years

85/74,536
(1.14 per 1,000)

74/474,392
(0.16 per 1,000)

4160 years

146/123,424
(1.18 per 1,000)

51/298,083
(0.17 per 1,000)

More than 60

196/167,515
(1.17 per 1,000)

22/321,242
(0.07 per 1,000)

GI infection by gender (Salmonella)

Male

62/119,313
(0.52 per 1,000)

84/419,703
(0.20 per 1,000)

Female

66/140,248
(0.47 per 1,000)

106/418,904
(0.25 per 1,000)
Table 2 Continued

2013 by the American College of Gastroenterology

H2RAs, histamine receptor antagonists; PPI, proton pump inhibitor.

a
20082009.
b
Histamine Receptor Antagonists (H2RAs).

Approval was obtained from the HIRU Information Governance

Review Panel to use the SAIL system for this research question.
Study population

Anonymised records from 168 general practices out of 484 in

Wales were available within the SAIL databank. These records
were used to select patients who were prescribed a PPI and were
compared with those not prescribed a PPI between the years
1990 and 2010. The first prescription of PPI (see Table 1 and
Table 2) were selected. Data for infection were collected from the
GP records, in-patient and out-patient datasets.

The American Journal of GASTROENTEROLOGY

STOMACH

Table 2. (Continued)

Table 2. Population characteristics

Proportion male

Rate PPI per 1,000,000 Copylobacter

per 100,000,000 Salmonella per 10,000,0000

B Sinead et al.

STOMACH

150,000

PPI
Campylobacter
Salmonella
100,000

50,000

0
1990

1995

2000

2005

2010

Year
Figure 1. Number of people with a precription of PPI (per 1,000,000) or with a diagnosis of Campylobacter or Salmonella (per 100,000,000).

the seasonality of infection between PPI users (before PPI

prescription) and non-users. Prior event rate ratio (PERR) (17)
was used to control for unmeasured confounding. In this method,
neither the exposed nor unexposed patients are treated with the
study drug before the start of the study. This method assumes that
the hazard ratio of the exposed to unexposed for a specific outcome before the start of the study reflects the combined effect of
all confounders (both measured and unmeasured) independent
of any influence of the treatment. To apply the PERR adjustment
method, we divided the unadjusted hazard ratio of date-matched
exposed group versus date-matched unexposed group after PPI
prescription by the unadjusted hazard ratio of exposed versus
unexposed before prescription. PERR-adjusted HR = HRs/HRp
were p = prior events and s = study events.

RESULTS
Sample characteristics

In this study, 18.7% of all patients were prescribed a PPI. Patients

prescribed a PPI were older with a greater frequency of females
compared with controls (Table 2). The PPI patients were more
likely to have other factors associated with GI symptoms such
as antibiotic use, oral steroid use, NSAID prescription, immunosuppressant medication, bowel surgery, or a diagnosis of arthritis.
Incidence of infection

Examining the effect of time: The rates of Campylobacter and

Salmonella infections were higher after a prescription for PPI
compared with rates in the same individual for the period before
prescription (Table 3). In fact, the risk of Campylobacter infection is 1.46 times higher after a PPI prescription compared with
before prescription. However, this increased rate is also seen in
the non-PPI patients. In fact, patients who are prescribed a PPI
have a higher risk of infection compared with controls even before
their prescription. Patients prescribed a PPI were already at 6.91
(95% CI: 5.169.26) times the risk of Campylobacter infection
compared with people not prescribed an acid suppression drug
The American Journal of GASTROENTEROLOGY

(see Table 3. Patients matched for dateBefore). The absolute

risk of Campylobacter infection increases by 0.55 per 1,000 person years of follow-up after a PPI prescription (Table 3) and by
0.03 per 1,000 person years of follow-up for Salmonella infection
(Table 4). However, this increase could be owing to changes in
diagnosis with time, as the same increase in trend is seen in the
non-PPI controls (Tables 3, 4 and Figure 1). For example, in the
control group the increase was 0.07 extra diagnosed with Campylobacter per 1,000 person years of follow-up between years 2008
and 2009 and 2009 and 2010 (Table 3 and Figure 1).
Examining the effect of unmeasured confounders: The PERR for
Campylobacter was 1.17 (95% CI 0.741.61) based on an unadjusted prior hazard ratio for the PPI/non-PPI date-matched sample of 6.54 (1.184/0.181) and an unadjusted post hazard ratio of
7.68 (1.73/0.225).
The PERR for Salmonella was 1.00 (95% CI: 0.51.5) based on an
unadjusted prior hazard ratio for the PPI/non-PPI date-matched
sample of 2.8 (0.11/0.03) and an unadjusted post hazard ratio of
2.78 (0.14/0.05).
Sensitivity analysis

Patients prescribed a PPI were more likely to have previously been prescribed a H2RA. In these patients, this may have
increased their prior rate of infection. We examined the prior
rate in those matches where H2RA had not previously been
prescribed (n = 321200 pairs out of a previous 358938 pairs).For
Campylobacter, this gave a prior unadjusted HR of 6.29 (1.07/0.17)
and a post unadjusted ratio of 7.86 (1.69/0.22) with a PERR of
1.22 (95% CI: 0.791.71).
The controls have fewer risk factors (antibiotic prescription,
NSAID use, etc.) compared with PPI users. Therefore, we examined those pairs where the control does have a risk factor (e.g.,
prescription for oral steroids, NSAIDS, antibiotics, diagnosis of
diabetes, or bowel surgery). The prior unadjusted HR was 1.65
(1.07/0.65) and post unadjusted ratio was 1.85 (1.74/0.94) with a
PERR of 1.12 (95% CI: 0.591.6).
VOLUME 104 | XXX 2013 www.amjgastro.com

PPI and Camplyobacter/Salmonella Infection

PPI patients
Campylobacter infection
Person years follow-up
Incidence per 1,000 per
year (95% CI)

Non-exposed (before
PPI prescription)

Exposed (post PPI

prescription)

425

622

358,892

358,651

1.18 (1.0761.3)

1.73 (1.61.8)

1.46 (1.291.65)

Hazard ratio
(exposed vs unexposed) a
Non-PPI patients
(selected years)
Campylobacter infection
Person years follow-up
Incidence per 1,000 per
year (95% CI)

Incidence per 1,000 per

year (95% CI)

54

57

577,620

577,604

0.09 (0.070.12)

0.1 (0.080.13)

Person years follow-up

Incidence per 1,000 per
year (95% CI)

126

199

997,476

994,729

0.13 (0.1 to 0.15)

0.20 (0.170.23)

1.58 (1.261.97)
PPI

Incidence per 1,000 per

year (95% CI)

39

50

359,139

359,158

0.11 (0.080.15)

0.139 (0.10.18)

1.2 (0.841.9)

Salmonella infection
Person years follow-up
Incidence per 1,000 per
year (95% CI)

Years 19901991

Years 19911992

43

41

577,640

577,631

0.074 (0.050.1)

0.071 (0.050.09)

0.95 (0.621.5)
Years 20082009

Salmonella infection
Person years follow-up
Incidence per 1,000 per
year (95% CI)

Non-PPI

Years 20092010

42

46

782,652

782,667

0.05 (0.040.07)

0.06 (0.040.07)

Hazard ratio (before/after)

425

65

358,938

359,120.9

1.184 (1.071.3)

0.181 (0.140.23)

6.91 (5.169.26)

1.04 (0.681.59)

Patients matched for date

PPI

Non-PPI

Salmonella infection

39

14

Person years follow-up

Incidence per 1,000 per
year (95% CI)

359,139

359,150

0.11 (0.080.15)

0.039 (0.020.065)

3.1 (1.75.7)

Hazard ratio
(PPI/non-PPI)b

After
Campylobacter infection

Exposed (post PPI

prescription)

Before

Hazard ratio
(PPI/non-PPI)b

Person years follow-up

Incidence per 1,000 per

year (95% CI)

Years 20092010

Before
Campylobacter infection

Person years follow-up

Non-exposed (before
PPI prescription)

Hazard ratio
(before/after)a

1.061 (0.731.53)

Hazard ratio (before/after)a

Patients matched for date

Salmonella infection

Non PPI patients

Years 19911992

Years 20082009

Person years follow-up

PPI patients

Hazard ratio (exposed vs.

unexposed)a

Years 19901991

Hazard ratio (before/after)a

Campylobacter infection

Table 4. Incidence rate for Salmonella infection

After
622

81

358,651

359,147

1.73 (1.61.88)

0.225 (0.1810.28)

Hazard ratio
(PPI/non-PPI)b

9.50 (7.412.2)

CI, confidence interval; PPI, proton pump inhibitor.

a
Clustered on individual ID.
b
Adjusted for age and gender and clustered on matched pair ID.

Salmonella infection
Person years
follow-up
Incidence per 1,000 per
year (95% CI)
Hazard ratio
(PPI/non-PPI)b

50

18

359,158

359,176.9

0.139 (0.10.18)

0.05 (0.030.08)

3.1 (1.825.3)

CI, confidence interval; PPI, proton pump inhibitor.

a
Clustered on individual ID.
b
Adjusted for age and gender and clustered on matched pair ID.

DISCUSSION
This study does not find evidence that PPI is associated with
increased rates of GI infection. Patients prescribed a PPI had a
higher rate of Salmonella and Campylobacter infection before
receiving their PPI prescription compared with those who did
not receive a PPI prescription during the study period. Both those
prescribed a PPI and those who were not prescribed a PPI had
an increase in the rate of Salmonella and Campylobacter infection
with time. Thus, increases in GI infection after a PPI prescription
2013 by the American College of Gastroenterology

are likely to be owing to a general increase in infection diagnosis

rather than a result of the PPI prescription.
This is the first known study to examine the incidence of
Campylobacter and Salmonella GI infection before and after first
prescription for PPI. This method means that the PPI patients act
as their own exposed/non-exposed controls and thus confounding
owing to patient characteristics is not an issue. The large sample
numbers available in routine data means rare conditions such as
The American Journal of GASTROENTEROLOGY

STOMACH

Table 3. Incidence rate for Campylobacter infection

STOMACH

B Sinead et al.

GI infections can be followed up using a cohort design. However,

there are limitations in working with routine data. Cases of infection may be missed as this method can only record the cases who
presented to their GP and had a stool sample sent to microbiology
for testing and diagnosis. Thus, we can only report on incidence of
diagnosed Campylobacter or Salmonella infection and cannot estimate true incidence including undiagnosed infections. However,
this limitation applies to all studies in this area. We chose major
significant infections (Campylobacter and Salmonella) in order to
minimise this ascertainment problem. However, people prescribed
a PPI are more likely to be diagnosed as they are more likely to be
attendees at GP practice. Thus, non-PPI control may have a higher
rate, closer to PPI patients, than we present here.
Patients prescribed a PPI may have been more likely to buy overthe-counter medicines before the prescription of the PPI, this is
seen in the higher proportion also prescribed a H2RA. However,
in Wales where prescriptions are free, more people seek a prescription rather than pay for the same medication over the counter.
Therefore, we feel this bias may be small and the sensitivity analysis
removing patients with prior prescription for H2RA did not demonstrate a real change in the findings.
The accuracy of diagnosis will have improved with time, and
testing methods used changes with time, moving to more standardised and genomic (e.g., PCR) testing. This may explain higher
diagnosis rates among all people examined with time.
Previous studies have used a case control design matching on
those with GE infection compared with those without (18,1), and
shown an association with acid suppression and increased risk of
enteric infection. Recommendations (1) have been for prospective
studies to examine if the association is causal. This study performs a
cohort study and shows, in terms of a temporal relationship between
PPI and infection, this association is unlikely to be causal. Previous
case control studies may have suffered from a healthy control bias,
as the sensitivity analysis using non-healthy controls (i.e., those
with prescriptions for different medications (NSAIDS, antibiotics,
and steroids)) showed similar rates to those taking PPI medication.
In summary, there is an association between taking a PPI and GI
infection with Salmonella or Campylobacter. However, this risk is
largely owing to difference between those who are prescribed PPIs
and those who are not, rather the PPI prescription. PPI use is associated with an increased rate of 0.6 per 1,000 person years of follow-up for Campylobacter and 0.18 per 1,000 years of follow-up for
Salmonella, and this increase is comparable to that seen in controls
and can be explained by changing time trend rates in diagnosis.
CONFLICT OF INTEREST

Guarantor of the article: Sinead Brophy, PhD.

Specific author contributions: All authors were involved in the
study design; data were analysed by S.B., M.A.R., and S.M.Z.; and
the manuscript was drafted by S.B., K.H.J., M.A.R., and M.D.A.
Input and amendments were made by A.J., N.F., R.A.L., and F.D.
Financial support: This study is funded by a grant from the
National Institute Social Care and Health Research (NISCHR)
within the Welsh Assembly.
Potential competing interests: None.
The American Journal of GASTROENTEROLOGY

Data access: This study makes use of anonymised data held in the
Secure Anonymised Information Linkage (SAIL) system, which
is part of the national e-health records research infrastructure for
Wales. We would like to acknowledge all the data providers who
make anonymised data available for research.

Study Highlights
WHAT IS CURRENT KNOWLEDGE

3PPIs are among the most commonly prescribed drug groups

worldwide.
3PPIs have been reported to increase the rates of GI
infections such as Salmonella and Campylobacter.

WHAT IS NEW HERE

3Using a cohort study of 358,938 people prescribed PPI, we

did not find evidence that the increased risk is attributable
to the PPI.

3It is more likely that association with infection is owing to

inherent factors within the patient that are associated with

PPI prescription.

3Work is needed to examine why people prescribed a PPI are

more prone to infection even before their PPI prescription.

REFERENCES
1. Leonard J, Moayyedi JK. Systematic review of the risk of enteric infection
in patients taking acid suppression. Am J Gastroenterol 2007;102:204756,
quiz 2057.
2. Huang JQ, Hunt RH. Pharmacological and pharmacodynamic essentials of
H(2)-receptor antagonists and proton pump inhibitors for the practising
physician. Best Pract Res Clin Gastroenterol 2001;15:35570.
3. National Institute for Clinical Excellence. Guidance on the use of proton
pump inhibitors for the treatment of dyspepsia, 2000.
4. National Institute of Clinical Excellence. Managment of dyspepsia in adults
in primary care, 2005.
5. Stevens V, Dumyati G, Brown J et al. Differential risk of Clostridium
difficile infection with proton pump inhibitor use by level of antibiotic
exposure. Pharmacoepidemiol Drug Saf 2011;20:103542.
6. Deshpande A, Pant C, Pasupuleti V et al. Association between proton pump
inhibitor therapy and Clostridium difficile infection in a meta-analysis.
Clin Gastroenterol Hepatol 2012;10:22533.
7. Howell MD, Novack V, Grgurich P et al. Iatrogenic gastric acid suppression
and the risk of nosocomial Clostridium difficile infection. Arch Intern Med
2010;170:78490.
8. Linsky A, Gupta K, Lawler EV et al. Proton pump inhibitors and risk
for recurrent Clostridium difficile infection. Arch Intern Med
2010;170:7728.
9. Turco R, Martinelli M, Miele E et al. Proton pump inhibitors as a risk factor
for paediatric Clostridium difficile infection. Aliment Pharmacol Ther
2009;31:7549.
10. FDA. Warning on Sereral Proton Pump Inhibitors. http://www.medicalreference.net/2012/02/breaking-news-fda-issues-warning-on.html,
2012.
11. Dublin S, Walker RL, Jackson ML et al. Use of proton pump inhibitors and
H2 blockers and risk of pneumonia in older adults: a population-based
case-control study. Pharmacoepidemiol Drug Saf 2010;19:792802.
12. Johnstone J, Nerenberg K, Loeb M. Meta-analysis: proton pump inhibitor
use and the risk of community-acquired pneumonia. Aliment Pharmacol
Ther 2010;31:116577.
13. Ran L, Khatibi NH, Qin X et al. Proton pump inhibitor prophylaxis
increases the risk of nosocomial pneumonia in patients with an
intracerebral hemorrhagic stroke. Acta Neurochir Suppl 2011;111:4359.
14. Bavishi C, Dupont HL. Systematic review: the use of proton pump
inhibitors and increased susceptibility to enteric infection. Aliment
Pharmacol Ther 2011;34:126981.

VOLUME 104 | XXX 2013 www.amjgastro.com

PPI and Camplyobacter/Salmonella Infection

2013 by the American College of Gastroenterology

18. Garcia Rodriguez L, Ruigomez A, Panes J. Use of acid-suppressing drugs and

the risk of bacterial gastroenteritis. Clin Gastroenterol Hepatol 2007;5:141823.

This work is licensed under the Creative Commons

Attribution-NonCommercial-Share Alike 3.0
Unported License. To view a copy of this license, visit http://
creativecommons.org/licenses/by-nc-sa/3.0/

The American Journal of GASTROENTEROLOGY

STOMACH

15. Ford DV, Jones KH, Verplancke JP et al. The SAIL Databank: building a
national architecture for e-health research and evaluation. BMC Health
Serv Res 2009;9:157.
16. Lyons RA, Jones KH, John G et al. The SAIL Databank: linking multiple
health and social care datasets. BMC Med Inform Decis Mak 2009;9:3.
17. Tannen RL, Weiner MG, Xie D. Use of primary care electronic medical record
database in drug efficacy research on cardiovascular outcomes: comparison of
database and randomised controlled trial findings. BMJ 2009;338:b81.