WP08 - Randomized Controlled Trial (RCT) of SENATOR vs. ‘Standard Care’


  • Conduct training of research staff and confirm preparedness of the 6 international research sites.
  • Perform clinical trial according to GCP and highest ethical standards.
  • Enroll over a 3 year period 1800 acute hospitalized patients, aged >65 years with at least 3 active co-morbid conditions.
  • Participant allocation to either SENATOR-guided optimization of drug and non-drug therapy as compared with standard therapy.
  • Follow enrolled subjects during index hospitalization until discharge or for initial hospital month (whichever is appropriate) and for up to 4 months post randomization and ascertain all pre-specified adverse drug reactions and all other suspected moderate/severe ADRs. 
  • Perform blinded adjudication of all potential endpoints by Endpoints Committee.
  • Complete scientific reports, manuscript preparation and dissemination.

Workpackage Description

A computer-assisted, sequential randomization (initially to ‘control’ arm (Phase I) , subsequently to ‘Active’ arm (Phase II)), adaptive (using minimization algorithm to achieve frequency matching of key risk factors between groups), trial of SENATOR-guided therapy (intervention) vs. standard clinical practice (control) of older multimorbidity patients (i.e. those with ≥ 3 chronic diseases) who are hospitalized with acute illness in one of six European university medical centres will be performed. Subjects will be followed daily during hospitalization (or for the first 30 days, whichever applies) for occurrence of any one of 10 pre-specified ADRs or any non-trivial, non pre-specified event believed by the primary researcher to be a possible ADR. Patients will be followed 3.5 months (+/- 2 weeks) post-discharge by telephone. The Primary Outcome will be the proportion of subjects with at least one moderate/severe ADR judged by the blinded Potential Endpoint Adjudication Committee as being probably/definitely an ADR. Secondary outcomes will include (i) mean number of ADRs per subject, (ii) number of pre-specified ADRs per subject (iii) Medication Appropriateness Index score, and (iv) use of appropriate non-drug therapy in subjects with one of core conditions included in the literature review in WP02. The primary analysis will include all enrolled subjects and will use a logistic regression model adjusting for demographic and important predictors of baseline ADR risk (as identified in WP01). To reject the null hypothesis, there must be both a significant difference between the groups in primary outcome and at least a trend (p<0.10) toward a similar reduction in the pre-specified endpoints. The per-protocol population will consist of all randomized subjects in whom at least 60% of the SENATOR-generated drug recommendations were at least initially applied. The trial recruitment will reflect the underlying profile of the source populations with regard to age, gender, ethnic and socio-economic status. In the data analysis, we will examine for effect modification through the use of stratified analysis with regard to these variables.

Task 1: Train research staff.

Develop training manual (which will provide all the documentations and education components necessary for site initiation), hold investigator meetings and workshops, conduct training in GCP and hold pre-initiation site visits.

Task 2: Develop Investigator/physician web portal, data entry and randomization facility in conjunction with ClinInfo (WP10).

The study will maintain a dedicated web-portal and provide a regular newsletter to physicians participating in the active arm of study in order to reinforce the study’s purpose and progress, discussing topical issues and detailing some of the common ADRs, and non-drug therapy opportunities identified by SENATOR.

Task 3: Potential subject identification.

Potential study subjects will be screened by the primary researcher under supervision from the local clinician co-PI and their permission sought for initiating the study consent process.

Task 4: Intervention allocation and application.

To randomize 1800 participants to SENATOR-guided or routine therapy. Generate SENATOR recommendations as appropriate, provide and discuss with primary attending team, and record physician response to SENATOR-generated recommendations.

Task 5: Follow study population.

Actively follow 1800 participants on a daily basis until index hospital discharge (or for up to 30 hospital days) and ascertain all prescribed medications, application of non-drug therapy for conditions identified in WP02 and all pre-specified ADRs and any additional suspected non-trivial ADRs.

Task 6: Post-discharge follow-up at 3.5 (+/- 2 weeks) months post-randomization.

Task 7: Data clarification form and data query reconciliation.

WP Leader

Prof. Dr. Joseph Eustace (UCC)

Participating Partners

University College Cork (UCC)
Ghent University Hospital (UGENT)
Servicio Madrileño de Salud (SERMAS)
Instituto Nazionale di Riposo e Cura per Anziani, Ancona (INRCA)
University of East Anglia (UEA)
Landspitali University Hospital Reykjavik (LUH)
University of Iceland (UoI)
Grampian Health Board (NHSG)