Health Economic Evaluation Alongside the OPERAM Trial

Background:

Drug-related morbidity and mortality is an increasing problem in European healthcare systems. Multimorbidity, polypharmacy and old age are important risk factors for drug-related hospital admissions (DRAs). The reported incidence of DRAs in the elderly may be as high as 30% of all acute cases, and about half of DRAs are likely to be preventable. They are mainly related to prescribing problems and non-compliance with drug regimens, and are very costly. Public funds spent on prescription medicines are steadily increasing in many European countries. Under strained budget conditions, healthcare services need to show that they provide value for money, and thus detailed health economic evaluations have become essential for health budget planning. Some studies assessing the impact of medication optimising interventions have assessed the cost-effectiveness of such interventions. A recent Cochrane review looking at interventions to optimise prescribing for the older people in care homes found an unclear effect of interventions on drug costs (Alldred DP 2013). Further assessment of healthcare costs related to inappropriate medication and the cost-effectiveness of pharmacotherapy optimising solutions is essential for a better control of rapidly increasing health care costs.

Design:

European multi-centre, cluster randomised, controlled trial of people aged 70 years or older, with multimorbidity and polypharmacy, being on an ambulatory visit or on a hospital stay in one of the four participating centres in Ireland, Belgium, Switzerland and the Netherlands. A cluster is defined around a treating physician, i.e. the treating physician is randomised and defines the allocation of his patients. Clusters of patients will be randomised to the intervention arm receiving STRIP for optimising therapy or to the control arm undergoing usual clinical care. The patients of physicians who are allocated to the intervention group will undergo a systematic drug review and pharmacotherapy optimisation by a physician and a pharmacist using STRIP, including the STRIPA software. That provides the research team with a recommendation of changes in the patient's medication. Based on STRIPA recommendation and agreement on changes to the patients' pharmacotherapy between the team of the research physician and pharmacist and the prescribing physician, will the patient receive structured counselling about his/her medication; general practitioners will receive a report. Health economic data (covering medical resource use, the provision of informal care and health-related quality of life using the EQ-5D instrument) will be collected from enrolled patients at baseline and during the follow-up phone calls after 2, 6 and 12 months. Patients will be asked to complete a booklet serving as a memory aid. Where relevant, information can also be provided by proxy respondents (e.g. family members, primary care physicians). Additionally, staff time spent on the intervention and comparable activities in the comparator arm will be collected from site staff.

Objectives:

The objectives of the health-economic evaluation are to assess the impact of pharmacotherapy optimisation by the STRIP intervention on economic parameters, health care utilisation and preference-based quality of life.

Statistical considerations:

Over all sites, 80 clusters with a cluster size ranging from 12 to 38 participants will be included. Therefore, 2000 patients, 1000 patients in each arm, will be recruited over 18 months. Health economic analyses will comprise the assessment of cost differences between the trial arms, of differences in quality-adjusted lifetime between the trial arms (expressed as quality-adjusted life years), and of the incremental cost-effectiveness of the intervention in comparison with the comparator, i.e. standard care.

Cost will be calculated by combining the healthcare resource utilisation observed in the trial with national unit costs from sources external to the trial. Health economic analyses involving costs are necessarily country-specific. It is planned to perform these analyses for at least two countries contributing patients to the OPERAM trial, which remain to be specified. This will involve the use of all or a subset of trial data (depending on the degree of heterogeneity observed between participating countries) in combination with a set of national unit costs and a national valuation algorithm for the EQ-5D. Perspectives of cost assessment remain to be decided. Typical perspectives are those of the healthcare system or a nationally relevant third party payer.

All health economic analyses will be subjected to extensive sensitivity analyses to understand the robustness of the results given stochastic uncertainty inherent in the trial data and parameter uncertainty expected for the external unit cost data.

Additionally, performance characteristics of the 5-level version (used in all patients) and the 3-level version (additionally administered in a subsample of 300 patients of the EQ-5D will be compared, e.g. in terms of discriminatory power, responsiveness to change over time and effect size.