Annals of Internal Medicine' Monika Sarkar, MD Sean Hennessy, PharmD, PhD and YuXiao Yang, MD, MSCE

1
Resident Journal ClubProton-Pump Inhibitor Use
and the Risk for Community-Acquired Pneumonia

• Annals of Internal Medicine. Monika Sarkar, MD
  Sean Hennessy, PharmD, PhD and Yu-Xiao Yang, MD,
  MSCE 16 September 2008 Volume 149 Issue 6
  Pages 391-398
• Nipun Sharma, PGY1

2
Background

• PPI use continues to escalate on both an in and
  out-patient basis
• Two recent studies that have suggested that PPI
  use may increase the risk for CAP A 2004 study
  in JAMA(Laheij RJ et al, 2004) and a 2007 study
  from Arch Internal Medicine(Gulmez SE et al,
  2007) are two studies that were utilized to
  provide the background for this analysis

3
Background- Dutch Cohort

• Nested Case Control Study with a source
  population of 364, 683 persons identified in a
  general practice research database in the
  Netherlands
• A cohort of 19 459 acid-suppressive drug users
  was selected and followed from Jan 1995 to Dec
  2002
• 5551 cases of pneumonia in the study.

4
Background- Dutch Study

• The incidence rates of pneumonia in
  nonacid-suppressive drug users and
  acid-suppressive drug users were 0.6 and 2.45 per
  100 person-years
• The adjusted relative risk for pneumonia among
  persons currently using PPIs compared with those
  who stopped using PPIs was 1.89 (95 confidence
  interval, 1.36-2.62). Current users of
  H2-receptor antagonists had a 1.63-fold increased
  risk of pneumonia (compared with those who
  stopped use).
• our results suggest that acid-suppressive drugs
  such as H2RAs and PPIs are associated with an
  increased risk of community-acquired pneumonia,
  probably because of reduction of gastric acid
  secretion, facilitating oral infections.

5
Background- Danish Cohort

• Population-based case-control study using data
  from the County of Funen, Denmark.
• Cases (n  7642) were defined as all patients
  with a first-discharge diagnosis of
  community-acquired pneumonia from a hospital
  during 2000 through 2004.
• 34 176 controls were used

6
Background- Danish Cohort

• The adjusted odds ratio (OR) associating current
  use of PPIs with community-acquired pneumonia was
  1.5 (95 confidence interval CI, 1.3-1.7).
• No association was found with histamine2-receptor
  antagonists (OR, 1.10 95 CI, 0.8-1.3) or with
  past use of PPIs (OR, 1.2 95 CI, 0.9-1.6).
• Recent initiation of treatment with PPIs (0-7
  days before index date) showed a particularly
  strong association with community-acquired
  pneumonia (OR, 5.0 95 2.1-11.7),
• Subgroup analyses revealed high ORs for users
  younger than 40 years (OR, 2.3 95 CI, 1.3-4.0).
  
• No dose-response effect could be demonstrated.
• Conclusion  The use of PPIs, especially when
  recently begun, is associated with an increased
  risk of community-acquired pneumonia.

7
Purpose of this Study

• A systematic review on management of GERD
  commissioned by the U.S. Agency for Healthcare
  Research and Quality called for further studies
  to elucidate the role of acid suppression on
  development of CAP
• The authors sought to more definitively examine
  the effects of current PPI exposure on
  development of CAP in a large, population-based
  cohort

8
Methods- GPRD

• Nested case control study using the General
  Practice Research Database (GPRD)
• GPRD prospectively collected electronic medical
  records from a sample of general practices
  throughout the U.K.
• It includes demographics, prescription use,
  diagnoses, etc
• Up-to Standard quality marker for GPRD
• Validation studies show that GPRD captures 90-95
  of diagnoses from referral visits and gt90 of
  hospital discharge diagnoses

9
Study Cohort

• Drawn from a cohort of 9 million patients who
  started follow-up from May 1987 to April 2002

10
Subjects

• Case patients- Eligible patients diagnosed with
  their first episode of CAP within the follow-up
  period. Date Index Date
• Control patients- up to 10 selected using
  incidence density sampling, matched by age, sex
• 10 controls per case to increase statistical
  power and analysis efficiency
• All controls selected from the same general
  practice site as the case and all having started
  follow-up within one month of the case

11
Exposure

• Current PPI therapy (prescription that would last
  to within 30 days of the index date)
• Daily dose also examined
• Duration of use before index date (lt30 days,
  30-180 days, gt180 days) taken into account

12
Statistical Analysis

• Conditional logistic regression used to calculate
  ORs and 95 CIs. A 2-tailed P value less than
  0.05 was considered statistically significant
• Potential Confounders- factors with known
  association with PPI therapy, pneumonia risk, etc
• In this study current smoking, total gp visits
  in the past yr, hospitalizations in the past yr,
  COPD, asthma, MI, CHF, CRF, cirrhosis, DM,
  stroke, prior hx CAP, hx CA, dementia, alcoholism
• Drug exposures- H2RAs, anxiolytics,
  antidepressants, antipsychotics, abx,
  anti-parkinson drugs, barbiturates, anxiolytics,
  opiates, corticosteroids

13
(No Transcript)
14
Results

• 80,066 Case patients of which 20,192 required
  hospitalization
• 799,881 Control patients
• OR for CAP assoc. with PPI use within 30 days of
  index date 2.05. After adjustment for
  additional covariates, PPI use was no longer
  associated with increased risk of CAP
• With Case patients restricted to hospitalized
  patients, Adjusted OR was 1.01
• Adjusted OR for H2RAs was 0.99

15
(No Transcript)
16
(No Transcript)
17
Case-Control with expanding matching criteria
18
Conclusions

• Current PPI use overall not associated with an
  increased risk for CAP, but markedly increased
  risk associated with PPI use started within a
  few days of index date
• Residual confounding may be responsible for much
  of the association previously seen. In this
  study, much of the initial association decreased
  as additional variables were adjusted for. The
  most significant of these were gp and hospital
  visits in the year prior to index date. The
  Danish study included none of these while the
  Dutch study adjusted only for gp visits. Opiates
  was the third most influential variable and
  neither of these other studies accounted for it.
  Seperate case-control gave similar results.

19
Conclusions

• Inverse temporal relationship- contrary to
  expected.
• Acid suppressive effects of PPIs takes at least 7
  days to reach maximum effect
• Same trend and a more pronounced increase in risk
  seen among H2RA users. H2RAs have weaker acid
  suppressive effects that PPIs
• Of note, modest protective effect associated with
  current PPI use for more than 6 months. No
  explanation known for this finding

20
Conclusions

• Protopathic bias- unlikely to be a major
  contributor as empirical PPI use is for subacute
  cough.
• PPIs may be associated through NSAID use for
  early CAP symptoms.
• Another possibility is dyspepsia from
  antibiotics used to treat CAP (antibiotics
  started within 7 days of index date not adjusted
  for). This was first suggested by the 2007,
  however, in that group there was a higher PPI-CAP
  association in patients who had not used abx
  prior to the index date.

21
Study Limitations

• PPI prescription does not PPI use
• Accuracy of CAP diagnosis- Read and Oxford
  Classification codes used. GPRD does not contain
  enough radiographic data to corroborate diagnoses
  of CAP. If it can be assumed that
  hospitalization for CAP is more specific, the
  study showed a near identical analysis in this
  case.
• Validity of GPRD diagnoses
• Adjustment of variables occasionally leading to
  data with high p-values
• Observational nature of study prevents causal
  relationships from being established
• Indication for PPI use not explored

22
Study Strengths

• Large cohort with 10 controls per case patient
• Selection of cohorts by incidence density
  sampling
• Thorough analysis of confounding factors
• In depth analysis of specific factors (temporal,
  dose-dependent)

23
Future Studies

• Randomized, controlled clinical trial
• Ethical considerations
• Would need a large number of subjects which would
  be more difficult to obtain if there is a
  restriction based on PPI indication.
  Additionally, it would be difficult to find a
  sufficient number of subjects if a significant
  number of matching variables are used