[99] Comparative effectiveness of proton pump inhibitors

[99] Comparative effectiveness of proton pump inhibitors

99Six previous Therapeutics Letters reported information about the benefits and harms of different proton pump inhibitors (PPIs).1-6 These drugs work by irreversibly inhibiting gastric H+K+ ATPase (the proton pump) in the stomach. They inhibit both basal and stimulated acid secretion and are used in a number of clinical settings: gastroesophageal reflux disease (GERD), reflux esophagitis, peptic ulcer disease (PUD), and symptoms associated with stomach acid such as heartburn and acid indigestion.

Are there any important differences between the different drugs in the class?

We have conducted two systematic reviews to compare the efficacy and safety of different PPIs: one in patients with symptomatic GERD and one in patients with PUD. A search up to March 2014 was done for all relevant randomized controlled trials (RCTs). RCTs were included if they compared two or more PPIs for at least 4 weeks in duration. We found 38 unique RCTs, in patients with GERD and 25 RCTs for patients with PUD. These 63 studies were retrieved, read and critiqued in detail. The critique included assessing each trial for features that may introduce bias in the trial findings.

Comparative trials of PPIs in adult patients with symptomatic GERD

Based on 26 RCTs in 23,789 patients, esomeprazole was not significantly different from other PPIs for most outcome measures: time to first resolution of symptoms; mortality; serious adverse events; withdrawal due to adverse events; and patients with at least one adverse event.7 Quality of life scores were not reported.

Based on 13 RCTs in 7,532 patients, lansoprazole was not significantly different from other PPIs for most outcome measures: total relief of symptoms; relief of retrosternal pain; relief of dysphagia; time to first resolution of symptoms; endoscopic healing of esophagitis; recurrence or relapse of symptoms; mortality; serious adverse events; withdrawal due to adverse events; and patients with at least one adverse event.7 Quality of life scores were not reported.

Comparative trials of PPIs in adult patients with symptomatic PUD

Based on 6 RCTs in 1753 patients, esomeprazole was not significantly different from other PPIs for most outcome measures: relief of heartburn; relief of epigastric pain; endoscopic healing of ulcer; mortality; serious adverse events; withdrawal due to adverse events; patients with at least one adverse event; or any specific adverse event. Total symptomatic relief, time to first resolution of symptoms, recurrence or relapse of symptoms, and quality of life scores were not reported. H. pylori eradication at 6 to 8 weeks did not differ between esomeprazole and omeprazole.7

Based on 19 randomized trials in 3,649 patients, lansoprazole was not significantly different from other PPIs for most outcome measures: H. pylori eradication; mortality; serious adverse events; withdrawal due to adverse events; and patients with at least one adverse event. Total symptomatic relief, time to first resolution of symptoms, recurrence or relapse of symptoms and quality of life scores were not reported.7

Overall Risk of Bias assessment

RCTs remain our best source evidence, but unfortunately can be biased.8 Selection bias occurs if randomization and allocation of the patient to treatment are compromised. Performance and detection bias can result if the patients, investigators and outcome assessors are not blinded to the treatment group. Reporting bias occurs if the outcomes reported are those that show the desired results and if trials with positive results are published and those with negative results are not.9

The 63 RCTs in these reviews were judged to have a high risk of selection, performance, detection and reporting bias. Thus the few significant differences in outcome measures found were thought to be most likely due to bias. In addition differences reported were small and were judged to not be clinically important.7

Applicability of evidence

The studies were mostly short-term (8-12 weeks) and excluded patients with any complications or co-morbidities so are mainly applicable to a relatively healthy population treated for short durations.7

What proportion of patients treated with PPIs will experience benefit of therapy?

The majority of patients (60 to 85%) with GERD experienced relief of heartburn at 4-8 weeks. In general symptom response rates were lower in patients with PUD compared to GERD. H. pylori eradication at 6 to 8 weeks was seen in 76 to 85% patients with PUD. Endoscopic healing at 4 to 8 weeks was seen in 66 to 82% patients with GERD or PUD. There is a wide variation in median time to first resolution of symptoms, ranging from 1 to 9 days in patients with GERD.7

Comparative overall safety of PPI

Harms were underreported in these short-term RCTs that directly compared different PPIs. Longer duration, head-to-head comparative RCTs specifically designed to monitor adverse effects have not been conducted. Based on observational studies, PPIs have been associated with an increased risk of the following adverse events: enteric infections (e.g. C difficile), spontaneous bacterial peritonitis, hospital and community acquired pneumonia, fractures, hypomagnesemia, acute interstitial nephritis, iron deficiency and vitamin B12 deficiency.10,11 There have also been concerns that long-term treatment with PPI may lead to development of gastric polyps, gastric cancer, carcinoids and colorectal cancer.1

Are there differences in cost of PPIs?

The wide range of costs of PPIs in British Columbia is shown in the Table.

Table: PPI Cost Comparisons

Canadian $
Esomeprazole Nexium 20
Lansoprazole Prevacid 15
Lansoprazole Prevacid FasTab 15




Pantoprazole sodium Pantoloc 20
Pantoprazole magnesium Tecta 40 0.20
Rabeprazole Pariet 10
Dexlansoprazole Dexilant 30


  • There is currently no convincing RCT evidence that one PPI is preferable to another for the management of GERD or PUD related symptoms or for endoscopically confirmed healing of esophagitis.
  • The risk of bias assessment of the 63 included RCTs indicated a high risk of selection, reporting, performance and detection bias.
  • There are no long-term, head-to-head comparative RCTs specifically designed to monitor adverse effects of PPIs.
  • Observational studies suggest that long-term use of PPIs is associated with a number of serious adverse effects.
  • The cost of different PPIs vary by over tenfold.


The draft of this Therapeutics Letter was submitted for review to 65 experts and primary care physicians in order to correct any inaccuracies and to ensure that the information is concise and relevant to clinicians.
The Therapeutics Initiative is funded by the BC Ministry of Health through a grant to the University of BC. The Therapeutics Initiative provides evidence-based advice about drug therapy, and is not responsible for formulating or adjudicating provincial drug policies .


  1. Therapeutics Initiative. Treatment of Gastoesophageal Reflux Disease (GERD). Therapeutics Letter. 1994 (December); 3:1-2.
  2. Therapeutics Initiative. Review and Update. Therapeutics Letter. 1995 (September); 9:1-2.
  3. Therapeutics Initiative. New Drugs. Therapeutics Letter. 1996 (March-April); 13:1-2.
  4. Therapeutics Initiative. Review and Update 1996. Therapeutics Letter. 1996 (November-December); 16:1-2.
  5. Therapeutics Initiative. New Drugs IV. Therapeutics Letter. 1998 (September-October); 26:1-2.
  6. Therapeutics Initiative. Do Single Stereoisomer Drugs Provide Value? Therapeutics Letter. 2002 (June-September); 45:1-2.
  7. Therapeutics Initiative. A systematic review of the comparative effectiveness of proton pump inhibitors for the treatment of adult patients with gastroesophageal reflux disease or peptic ulcer disease. 2016; British Columbia Ministry of Health, Pharmaceutical Services Division. http://www2.gov.bc.ca/assets/gov/health/health-drugcoverage/pharmacare/derp-ppi.pdf (Accessed June 1, 2016)
  8. Savovic J, Jones HE, Altman DG et al. Influence of reported study design characteristics on intervention effect estimates from randomized, controlled trials. Ann Intern Med. 2012;157(6):429-38. DOI:10.7326/0003-4819-157-6-201209180-00537
  9. Turner EH, Matthews AM, Linardatos E, et al. Selective publication of antidepressant trials and its influence on apparent efficacy. N. Engl. J. Med. 2008;358(3):252-60. DOI:10.1056/NEJMsa065779
  10. Provincial Academic Detailing. Proton Pump Inhibitors. http://www2.gov.bc.ca/gov/content/health/practitioner-professionalresources/pad-service/proton-pump-inhibitors (accessed March 2016)
  11. Reimer C. Safety of long-term PPI therapy. Baillieres Best Pract Res Clin Gastroenterol. 2013;27(3):443–54. DOI:10.1016/j.bpg.2013.06.001


  • Hans Baer
    Posted at 15:25h, 29 June Reply

    Thanks for this excellent and important review. I just want to point out that maybe there are some differences between PPIs in terms of conveniences to patients, or to individual patients. For example, one of the PPIs is claimed to be effective when taken shortly before a meal because it survives stomach acid (Na-bicarbonate content?) while others (all??) have to be taken well ahead of a meal (1 h) to assure that the bioavailability is not compromised. I also have a one-patient observation that showed that one of the PPIs caused an intolerance symptom while others did not (I am sorry, I cannot remember which one the troublesome was, maybe pantoprazol, and I cannot look up my notes right now because of traveling).

    • jauca
      Posted at 00:32h, 13 July Reply

      Thank you for your comment. Many claims are made about various PPIs (as well as any other drugs) but few of these claims are supported by any evidence. We made it clear that our conclusions refer to what we know about these drugs based on the best available evidence at this time.

    • Tom Perry, MD, FRCPC (clinical pharmacologist) UBC TI
      Posted at 18:58h, 12 October Reply

      Hello Hans Baer,

      I’m sorry I just noticed these comments tonight. Please see my comment posted in response to Kathy’s comment below.

      Tom Perry, MD, FRCPC (clinical pharmacologist, UBC TI)

  • Paul
    Posted at 18:51h, 08 July Reply

    I’m left scratching my head at the pricing table.

    Is it really true that the lower dose omeprazole, and the lower dose pantoprazole, are both twice (or more) the cost of the corresponding higher dose pills?

    If so, I find it really odd pricing; and if correct, here’s a good argument for choosing a scored tablet, if available, and having the patient split it, resulting in really cheap dosing.

    • jauca
      Posted at 00:27h, 13 July Reply

      It is strange, but true: the lower dose of omeprazole and pantoprazole are twice or more the cost of the higher dose. Since there is no conceivable scientific reason for this, it is probably a marketing decision intended to maximize profits.

    • Kathy
      Posted at 09:24h, 30 July Reply

      All the ones that come in tabs are enteric coated, they do not survive the acidic environment of the stomach. You could open the capsules and guess half of the granules, but splitting the tablets would waste all their money!

      • Tom Perry, MD, FRCPC (clinical pharmacologist) UBC TI
        Posted at 18:56h, 12 October Reply

        So far as we can tell, the argument that PPIs are “destroyed” by stomach acid if their coating is disrupted is complete nonsense. The industry has used this successfully to prevent pill-splitting for years. This has increased sales by hundreds of millions of dollars or perhaps billions of dollars, compared to what prudent and cost-conscious consumers would have spent if well informed.


        When Astra-Zeneca decided to market the pure enantiomer esomeprazole (one-half of omeprazole, which is a racemic mixture of S-omeprazole and R-omeprazole), it gave aqueous solutions of esomeprazole to volunteers to demonstrate absorption. They used a 250 mL water chase to ensure the oral dose was absorbed promptly.

        Think: if the drug in aqueous solution survives passage through the stomach, then why wouldn’t a fragmented pill (or capsule), if chased with a reasonable amount of water.

        Consider also this: PPIs routinely increase pH by a factor of 2-2.4 (from stomach pH of about 2 to stomach pH or about 4-4.4). This is 2-2.4 log units to base 10, or approximately 100-200 fold reduction of hydrogen ion concentration ( 100-fold (pH rises by > 2)?

        If there is anyone out there who knows differently, I would sure love to see the evidence that pill-splitting “ruins” the pills and would lead people to “waste their money”.

  • Anonymous patient
    Posted at 23:16h, 14 March Reply

    I was provided with a two week supply of Dexilant as a trial to see if it would have a calming effect on my persistent hacking cough from gerd. Within three days the constant cough had diminished in severity and frequency and after a week, it was occurring less than three times in 24 hours. This is the first drug to manage my symptoms and I wonder why. No other rx has been able to achieve this dramatic result. Further the damage from my constant cough over several years is scarring in my throat and causing difficulty sleeping- both issues for long term damages if not treated. So BCBS does not cover this drug at all. Frustrated by insufficient research. Frustrated by bouts of dry coughing that leave me exhausted. Need a cure for gerd that is better than surgery outcomes of 50% cure. Tired of sounding like a fog horn. Nothing else has worked until this drug which Is not covered by insurers. 😡

    • Alan Cassels @ TI
      Posted at 15:33h, 19 March Reply

      We cannot comment on why Dexlansoprazole (Dexilant)is not covered by private insurers.

Post A Reply to Tom Perry, MD, FRCPC (clinical pharmacologist) UBC TI Cancel Reply