Gastroprotection in Low-Dose Aspirin Users for Primary and Secondary Prevention of ACS: Results of a Cost-Effectiveness Analysis Including Compliance

Abstract

Purpose

Low-dose aspirin (ASA) increases the risk of upper gastrointestinal (GI) complications. Proton pump inhibitors (PPIs) reduce these upper GI side effects, yet patient compliance to PPIs is low. We determined the cost-effectiveness of gastroprotective strategies in low-dose ASA users considering ASA and PPI compliance.

Methods

Using a Markov model we compared four strategies: no medication, ASA monotherapy, ASA+PPI co-therapy and a fixed combination of ASA and PPI for primary and secondary prevention of ACS. The risk of acute coronary syndrome (ACS), upper GI bleeding and dyspepsia was modeled as a function of compliance and the relative risk of developing these events while using medication. Costs, quality adjusted life years and number of ACS events were evaluated, applying a variable risk of upper GI bleeding. Probabilistic sensitivity analyses were performed.

Results

For our base case patients using ASA for primary prevention of ACS no medication was superior to ASA monotherapy. PPI co-therapy was cost-effective (incremental cost-effectiveness ratio [ICER] €10,314) compared to no medication. In secondary prevention, PPI co-therapy was cost-effective (ICER €563) while the fixed combination yielded an ICER < €20,000 only in a population with elevated risk for upper GI bleeding or moderate PPI compliance. PPI co-therapy had the highest probability to be cost-effective in all scenarios. PPI use lowered the overall number of ACS.

Conclusions

Considering compliance, PPI co-therapy is likely to be cost-effective in patients taking low dose ASA for primary and secondary prevention of ACS, given low PPI prices. In secondary prevention, a fixed combination seems cost-effective in patients with elevated risk for upper GI bleeding or in those with moderate PPI compliance. Both strategies reduced the number of ACS compared to ASA monotherapy.

Technical Appendix

Input Parameters

We performed a structured search using PubMed and Embase databases limiting our results to English language and using combinations of relevant entry terms (aspirin, proton pump inhibitor, gastrointestinal, acute coronary syndrome, prevention, compliance, adherence, incidence, risk, relative risk, cost-effectiveness). Where available, we used meta-analyses or systematic reviews reporting intention-to-treat summary estimates. In order to derive annual transition probabilities, we multiplied placebo risks (on the development of ACS, upper GI bleeding and dyspepsia) by the relative risks of aspirin and, if necessary, PPI. In case placebo risks were unknown, we divided the risk with aspirin monotherapy by the relative risk of aspirin. Utility values of the combined health states for which no data was available (e.g. Post ACS+dyspepsia) were derived by multiplying the separate utilities of the involved health states (Table 4).

Model Assumptions

• 1) A patient who develops dyspepsia visited his/her primary care provider and received a 4 week trial of PPI therapy (omeprazole 20 mg daily). Patients previously treated with PPIs were assumed to be given a dose of 40 mg/day. Should this be ineffective (approximately 45 % of patients), the patient is referred to a gastroenterologist. The patient receives diagnostic endoscopy including a H.pylori test. H.Pylori eradication therapy is given if appropriate, and eradication is confirmed by a breath test. Patients receive another 8 weeks of PPI therapy and are assumed to visit their primary care provider a total of three times per year.

• 2) All patients with persistent dyspepsia receive PPI therapy. Patients who were allocated to no medication or aspirin monotherapy receive 20 mg PPI daily during the complete cycle, whereas patients who were allocated aspirin+PPI or a single tablet formulation receive additional PPI (40 mg omeprazole daily in total). All patients are assumed to visit their primary care provider annually.

• 3) Patients who develop an upper GI bleeding are admitted to the hospital after reporting to the emergency department. Sixty percent of patients need a blood transfusion and all receive endoscopic therapy, intravenous PPI, H.pylori testing and H.pylori eradication therapy plus breath test confirmation if necessary. A second therapeutic endoscopy is performed in case of therapy failure, followed by percutaneous embolization if therapeutic endoscopy remains unsuccessful. A second look endoscopy is performed in patients with an ulcus ventriculi. The average duration of hospitalization is 10 days. At discharge all patients receive PPI therapy for the remainder of the time horizon: 20 mg omeprazole in case the patient was allocated to no medication or aspirin monotherapy and 40 mg omeprazole in case the patients was allocated aspirin+PPI or a single tablet formulation. Patients allocated to the single tablet formulation continue their assigned medication and are prescribed an additional 20 mg PPI (instead of changing to 40 mg PPI concomitant to low-dose aspirin). In case of primary prevention of ACS, low-dose aspirin therapy is interrupted for 1 year. Patients are assumed to visit the outpatient clinic once in the following year. During the first year, 6.7 % of patients experiences a rebleeding.

• 4) Patients experiencing an ACS report to the emergency department where an ECG is made and cardiac marker levels (including troponin) are determined. We assumed that coronary angiography is performed in 90 % of patients, whereas 70 % of patients receive an additional percutaneous intervention and 5 % of patients require coronary artery bypass grafting surgery. In hospital, all patients receive low-dose aspirin and β-blockers, and some patients receive clopidogrel (60 %), ACE-inhibitors (55 %), nitroglycerin (70 %) and heparin (35 %). The average duration of hospitalization is 5 days. At discharge, all patients receive low-dose aspirin. In addition, patients receive β-blockers, statins and ACE-inhibitors for the remainder of the time horizon, whereas 80 % also receive clopidogrel for 1 year. During the first year rehospitalization is necessary in 30 % of patients. Patients are assumed to visit the outpatient clinic four times during the first year and once annually thereafter

Table 4 Derivation of health state utilities

Table 5 Health care costs

Table 6 Distributions of the probabilistic sensitivity ananlyses

Fig. 7

(Risk of death all causes) by age

Fig. 8

Probability distributions of compliance. N.B. The compliance to the single tablet formulation before complications have occurred, was assumed to equal the compliance to aspirin before complications have occurred. The compliance to the single tablet formulation after ACS or GI complications, was assumed to equal the compliance to aspirin after an ACS

Fig. 9

Cost-effectiveness of PPI co-therapy for a range of PPI prices

Fig. 10

Tornado diagram of one way sensitivity analyses comparing the fixed combination with ASA+PPI co-therapy for average risk patients using aspirin for primary prevention

Fig. 11

Tornado diagram of one way sensitivity analyses comparing the fixed combination with ASA+PPI co-therapy for average risk patients using aspirin for secondary prevention