Rapidly increasing prescribing of proton pump inhibitors in primary care despite interventions: A nationwide observational study

Abstract

Background: Guideline and reimbursement modifications have been introduced to optimize prescribing of antisecretory medication in Danish general practice. Impacts of the interventions have not been evaluated.

Objectives: To analyse developments in prescribing of antisecretory medication in Denmark 2001–2011 and to assess the impacts of interventions on prescribing of antisecretory medication.

Methods: Register-based cohort study covering the entire Danish population of currently 5.5 million inhabitants. Developments in the prescribing of antisecretory medication over time in Denmark between 2001 and 2011 and association with age and gender of users along with the impact of interventions on the prescribing of drug subgroups are analysed.

Results: 96.8% of all antisecretory drugs sold are proton pump inhibitors (PPIs) and 94.4% of the PPIs are prescribed in primary care. Prescribing of PPIs has increased substantially during the past decade. Both number of users and the average individual use have increased. The prescribing of ulcerogenic drugs to the elderly has stagnated in the same time range. Reimbursement modifications and scientific guidelines do not seem to have had a substantial influence on the steadily increasing prescribing of PPIs.

Conclusion: Use of PPIs has increased substantially during the past decade, without a change in indications for use of PPIs in the same time range. Interventions to enhance adherence to guidelines and promote rational use of PPIs do not seem to have had a substantial influence on the overall prescribing rate.

Keywords::

• Reimbursement modifications
• dyspepsia
• pharmacoepidemiology
• primary health care
• proton pump inhibitors

KEY MESSAGE:

• Prescribing of PPIs has increased rapidly from 2001 to 2011 despite interventions to rationalize the prescribing.

• The vast majority of PPIs is prescribed in primary care.

• Reimbursement modification seems to have caused a change in prescribing of subtypes of PPIs but not a change in the overall prescribing of PPIs.

INTRODUCTION

Use of antisecretory medication has increased rapidly in many countries during the past decades without a clear explanation (1,2). It is mainly due to an increasing number of long-term treated patients, and most of them do not have an endoscopically verified indication (3).

PPIs have been thought to be without side effects, but recent knowledge raises concerns about adverse side effects and drug interactions (4). Furthermore, a few weeks of treatment can cause rebound acid hypersecretion and acid-related symptoms in previous asymptomatic individuals when the PPI is withdrawn (5).

Several actions have been taken to optimize the prescribing patterns of PPIs. In 2009, the Danish College of General Practitioners introduced a guideline on management of dyspepsia in primary care. These guidelines are considered to be the reference standard for Danish primary care. The guideline points out that Helicobacter pylori ‘test-and-treat’ strategy is more cost-effective compared to empirical antisecretory treatment. This strategy is also recommended by the American Gastroenterological Association (6).

In 2010, the Danish Health and Medicines Authority modified the reimbursement system, so that rabeprazole, esomeprazole and all histamine-2-receptor- antagonists (H2RAs) were no longer generally reimbursable. The guideline and information about the reimbursement modification were sent to all general practitioners (GPs). The accompanying letter stated that a substantial amount of patients were over treated with PPIs. All patients on H2RAs, rabeprazole and esomeprazole were informed by their pharmacy and encouraged to contact their GP.

Aims of this study were to analyse the development in use of antisecretory medication in the period of 2001–2011 in Denmark and to assess the impact of the interventions on prescribing rates of antisecretory medication.

METHODS

Study design

A register-based study covering the entire Danish population (5.5 million inhabitants) was performed. Approval of the study by a medical ethics committee was not needed.

The Danish public health care system is tax-funded and provides free and equal access to health care for all residents. Approximately 98% are registered with a GP, and 90% of all medications are prescribed in primary care (7). Most PPIs are reimbursable when bought on prescription, and few are available over-the-counter.

Register variables

Information on all sales and redeemed prescriptions nationwide are contained in the Danish Register of Medicinal Product Statistics (1). The register was searched in September 2012 for ATC code A02B (drugs for peptic ulcer and gastro-oesophageal reflux disease); B01AC06 (acetylsalicylic acid); and M01A (anti-inflammatory/ anti-rheumatic products, non-steroids—NSAIDs). The variables turnover, paid reimbursement, volume sold, sector, age, gender and number of users through the years 2001–2011 were used. Volumes are given as defined daily doses (DDDs). The number of users within each age group is presented as relative numbers (n/1000) according to changing demographic composition over time.

Medications other than H2RAs and PPIs within the A02B category were rarely prescribed and, therefore, excluded. Prescriptions for ulcerogenic medication redeemed by persons’ ≥ 65 years included since the elderly are susceptible to medication-associated peptic ulcer (8).

RESULTS

Prescribing of antisecretory drugs in Denmark has increased substantially 2001–2011, primarily for PPIs (Figure 1). PPIs accounted for 96.8% of the total sales within the A02B-group in 2011.

Figure 1. Developments in the use of PPIs in Denmark 2001–2011.

Use of PPIs (DDDs) increased by 243% from 2001 to 2011. In 2011, 108.3 million DDDs of PPI were sold. Of the total use, 94.4% were prescribed in primary care, 2.6% in the hospital sector and 3% were sold over the counter. In 2011, the turnover was €21.4 million and reimbursement paid was €6.9 million.

After the guideline and the reimbursement modification were introduced, prescribing of the no longer reimbursable drug subtypes decreased substantially, while use of other PPIs increased (Figure 2).

Figure 2. Use of antisecretory medication in Denmark around the time of reimbursement modification (year 2010).

In primary care, 96.5% of the total use was prescribed to adults (≥ 20 years). The average personal use of PPIs (mean DDDs per user per year) increased by 45% from 2001 to 2011. In 2011, patients were averagely prescribed 218 DDDs each (Figure 1).

From 2001 to 2011, the relative percentage of female users increased from 55.4% to 57.1%. Most PPIs are used by persons < 65 years measured in absolute terms, but when measured in relative terms (number of users per 1000 individuals within the age group) use of PPI is more common among the elderly (data not shown). The elderly also had the highest increase in the number of users and volume sold. Number of persons ≥ 65 years using ulcerogenic drugs increased in the first years after 2001 and has stagnated since 2008 (data not shown).

DISCUSSION

Main findings

This study captured all redeemed prescriptions of antisecretory medication in Denmark 2001–2011 and showed substantially increasing redeeming of PPIs prescribed in primary care. Both, number of users and volume sold increased. No similar increase was detectable in the use of ulcerogenic drugs among the elderly.

After the guideline and the reimbursement modification were introduced, prescribing of the no longer reimbursable PPIs decreased while use of other PPIs increased. The increasing prescribing of PPIs in total seems unaffected by the interventions.

Strengths and limitations

A major strength of the study is that it is register-based, enabling us to include the entire Danish population. The register is comprehensive for all redeemed prescriptions, thus including almost all PPIs sold, as the over- the-counter use is low.

Limitations are the lacking information about indication for the drug and that the information provided by the register is not personally identifiable, which makes it impossible to characterize the users as incident or prevalent users and to identify associations between use of PPIs and other medications. In this study design, we are unable to determine what considerations the GPs have for prescribing PPIs and thereby analyse what reasoning might precede the prescribing.

Possible explanations for increased prescribing of PPIs

One explanation for increased prescribing of PPIs could be that the incidence of the indications has changed over time. Peptic ulcer disease—requiring short-term treatment—is decreasing, while gastro-oesophageal reflux disease and its complications—requiring long-term treatment—are increasing in the Western world. These trends may be associated with reduced H. pylori prevalence and the obesity epidemic. Incidence of oesophagitis is especially increasing among the elderly, probably due to endoscopy selection criteria.

Another reason could be an increased need for ulcer prophylaxis. We observed that prescribing of ulcerogenic medication to the elderly has stagnated the past several years. However, the increase in prescribing of PPIs to the elderly could be due to increasing awareness of prescribing ulcer prophylaxis to patients at risk, but we cannot know that as the indication is not registered.

The effect of reimbursement modification

Introduction of a guideline did not seem to have influenced the increasing use of PPIs significantly. We expected that promoting test-and-treat strategy superior to empirical antisecretory treatment and information to the GPs about overtreatment with PPIs would cause a slower increase in prescribing of PPIs. However, as previously demonstrated disseminating evidence to health care professionals did not influence prescribing (9). Reimbursement modification appears to have more potential as other studies found that the reimbursement restrictions can lead to substantial changes in drug utilization (10). We found that the use of no longer zreimbursable subtypes of antisecretory drugs (especially esomeprazole) decreased. However, prescribing of pantoprazole increased almost inversely and the most frequently prescribed PPIs, lansoprazole and omeprazole, increased by the same rate as in the previous years. Therefore, the reimbursement modification caused changes in prescribing of PPI subtypes, but did not seem to have had an important influence on the steadily- increasing total prescribing of PPIs.

Implications: A proper diagnosis and on-demand treatment

The potential side effects and drug interactions emphasize the need for critical evaluation of the indication for continuous antisecretory medication in each patient. The literature shows that the duration of antisecretory treatment is only discussed between patients and GPs in 10% of the cases (11). A Dutch study showed that many of the patients treated with antisecretory drugs without proper indication are not motivated for discontinuation (12). Therefore, moderation and accuracy are essential aspects to keep in mind if initiating antisecretory medication empirically. Dyspepsia as a symptom is fluctuating, and a clinical diagnosis of its cause is unreliable. The substantial placebo response in uninvestigated dyspepsia might convince both patient and GP that further treatment is needed, if initial symptoms are relieved by a few weeks of treatment. Consequently, a subgroup of empirically-treated patients may redeem repeated prescriptions because of a positive response to therapy, although their initial symptoms are not acid related because of the placebo response or rebound acid hypersecretion induced by the empirical treatment.

Prevention of long-term treatment without proper indication might be the most effective way to reserve PPIs to only patients with a real need for antisecretory treatment. With patients not being motivated for discontinuation, a realistic goal might be to reduce the individual use of PPIs among long-term treated patients without a clear indication. This might be obtained by encouraging on-demand treatment instead of continuous treatment (12).

Conclusion

Prescribing of PPIs in primary care in Denmark has increased substantially from 2001 to 2011. Both number of users and the average individual use have increased. Neither reimbursement modification nor information about overtreatment nor introduction of guidelines seems to have had a substantial influence on the overall prescribing rate. The reimbursement modification caused a change in prescribing of PPI subtypes causing economic savings but not a decrease in dosage or persons on PPIs.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.