Trial design and settings
The Statins In The Elderly (SITE) Study is a 3-year follow-up, open-label comparative multi-centre randomised clinical trial being conducted in two parallel groups in outpatient primary care offices. Our objective is to conduct a randomised controlled trial that is representative of usual care conditions. Our study is being conducted in general practices in France, and the eligibility criteria are limited, apart from regulatory requirements, to age and treatment with statins for the purpose of primary prevention.
All general practitioners working in primary care centres in France can enrol patients in the study. They need only to use informatics devices to complete the electronic case report forms (e-CRF) during patient follow-up.
Eligibility criteria
The study population consists of patients aged 75 and over who have no history of CV disease and who are being treated with a statin and regularly visiting their referent GP in France.
Participants meeting the following criteria are eligible for inclusion: people aged 75 years and older who have been treated with statins for at least 1 year for primary prevention of CV events, and who give informed consent. Patients with hypertension and/or diabetes may be included.
Participants meeting any of the following criteria are excluded: progressive morbid pathology with a life expectancy of 3 months or less, diagnosed with dementia, suffering from known homozygous or double heterozygous familial hypercholesterolemia, or unable to provide informed consent.
Intervention
In the statin-cessation strategy, the statins are stopped and patients are asked to end their treatment on the day of randomisation. In the comparison strategy, patients continue their statin treatment at the usual dosage. Compliance of statins is assessed by the Morisky questionnaire at each visit.
Randomisation
Randomisation is being carried out centrally using the electronic case report form (eCRF) on the day of the inclusion visit after signing the consent form. Participants are considered enrolled in the trial when randomised.
The randomisation is unbalanced, with a 5:4 ratio in favour of the statin-continuation group to take into account a 20% risk that patients allocated to statin continuation group will stop statins after randomization for any reason (side effects, choice of the patients, choice of the doctor).
The allocation sequence was elaborated using computer-generated random numbers, and was not stratified. To reduce predictability of a random sequence, we used blocks of varying size.
A document describing the randomisation procedure is being kept confidential within the Clinical Epidemiology Unit.
Outcome measures
Economic endpoints
The cost-effectiveness of the statin cessation strategy compared to continuing statins will be estimated based on the incremental cost per quality-adjusted life year (QALY) gained at 36 months, from the perspective of the French healthcare system. QALYs will be measured using the Euro-QOL five-dimension, three-level (EQ-5D-3L) questionnaire at inclusion and at each follow-up visit [18]. This questionnaire has been validated in the French population.
Healthcare resources and costs will be extracted from the French Administrative Health Care Database (Système National des Données de Santé, SNDS) by a merging procedure, operated by the French National Health Insurance Fund and using data from the SITE trial. The SNDS gathers all reimbursement data for 98.8% of the French population [19]. The time horizon of the study (36 months) implies a discounting of costs and outcomes. In accordance with the methodological economic guidelines of the French National Health Authority (Haute Autorité de santé, HAS), we will apply a 4% discount rate.
Following cost-effectiveness analysis (CEA), budget impact analysis (BIA) will provide useful information about the sustainability of the statin-cessation strategy if it were applied to the entire French healthcare system. It will consider outpatient costs, hospital costs, medical and non-medical costs induced or avoided by stopping statins, and all budgetary implications related to changes in health management and/or health status improvement. The difference between induced and avoided costs will provide the net benefit to the healthcare system of widespread adoption of the statin-cessation strategy for primary prevention among elderly individuals in France.
Main clinical endpoint
Overall mortality will be the primary clinical endpoint. As the investigators are general practitioners (GPs), they are informed of the death of their patients and will therefore constitute the source of mortality data in this trial.
Secondary endpoints
- QOL score measured by the Quality of Life Scale SF12, which has been validated in geriatric populations
- Incidence of CV events
- Incidence of non-CV events (diabetes and cognitive disorders)
Timeline and recruitment
This study employs the research network of French universities’ primary care departments (PCDs). Each PCD appoints a coordinating physician by geographic area to mobilise investigating physicians (486 expected in the study).
The recruitment of GPs all across France maximises representativeness and limits the impact of potential local practices. The representativeness of the investigators is being documented by collecting data about their sociodemographic profiles, activity, and training.
Patients are being recruited through consultations over a period of 36 months. Eligible patients not included in the study are being reported, along with the reason for non-inclusion.
Conduct of research
Patients are approached and screened by GPs during their usual medical visit or for the renewal of their medication. Patients are first pre-included with their referent GP, and then after a time of reflection (that may be few minutes to days or months until the next visit), they signed the consent form with the GP and are randomized between the two strategies. Table 1 outlines the different phases of the study and data collection.
Table 1: Design of the SITE study
|
|
Study period
|
TIMEPOINT
|
|
Pre-inclusion
T -1
|
Inclusion
T 0
|
M3
+/− 1 month
|
M12
+/− 3 months
|
M24
+/− 3
months
|
Close-out
M36
+/- 3
months
|
ENROLMENT
|
|
|
Information/eligibility screen
|
|
ü
|
|
|
|
|
|
Standardised questionnaire 1
|
|
ü
|
|
ü
|
ü
|
ü
|
ü
|
MMSE 2
|
|
ü
|
|
|
|
|
ü
|
Consent signature
|
|
|
ü
|
|
|
|
|
|
|
|
|
|
|
|
|
INTERVENTION
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
ASSESSMENT
|
|
|
SF12 and EQ-5D-3L Questionnaires 3
|
|
|
ü
|
ü
|
ü
|
ü
|
ü
|
Clinical examination 4
|
|
ü
|
ü
|
ü
|
ü
|
ü
|
ü
|
Biological tests 5
|
|
ü(*)
|
|
ü
|
ü
|
ü
|
ü
|
ECG 6
|
|
|
ü
|
|
ü
|
ü
|
ü
|
Clinical event data collection 7
|
|
|
ü
|
ü
|
ü
|
ü
|
ü
|
Compliance 8
|
|
|
ü
|
ü
|
ü
|
ü
|
ü
|
CVRF collection 9
|
|
|
ü
|
ü
|
ü
|
ü
|
ü
|
SE/SAE 10 collection
|
|
|
ü
|
ü
|
ü
|
ü
|
ü
|
- Standardised questionnaire with six questions to search for coronary disease history, stroke, or unnoticed peripheral artery disease.
- Mini-Mental State Examination (French version, GRECO).
- Quality of Life Questionnaire SF12 (self-administered) and EQ-5D 3L (self-administered).
- Clinical examination: examination as recommended by good clinical practices in the field of cardiovascular medicine (blood pressure, CV data, and lung auscultation).
- Laboratory tests: lipid (EAL), blood glucose, HbA1c (if diabetic), electrolytes, creatinine, creatinine clearance, and serum albumin.
(*) In the absence of a balance sheet dated within 12 months before pre-inclusion, laboratory tests are prescribed at pre-inclusion and should be performed until the day of the inclusion visit.
- ECG: electrocardiogram based on recommendations (once every 3 years if permanent arterial hypertension (HTA) and once a year if overt diabetes).
- Reports of significant clinical events occurring between visits
- Compliance assessed by Morisky questionnaire (eight items) at baseline (all patients) and during follow-up (only in the group of patients in whom statin therapy is continued).
- Total CV risk factors: weight, waist circumference, smoking, hypertension, diabetes, previous family history of CV disease among first-degree relatives, HDL and LDL cholesterol.
- SE/SAE: Side effects and serious adverse events, respectively.
Sample size
According to data from the French National Institute for Statistics and Economic Studies (Institut National de la Statistique et des Études Économiques, INSEE) French population aged 75 years or older, we assume that mortality at 3 years in both groups will be 15%.
The SITE’s sample size was computed to achieve 90% power in showing the non-inferiority of statin cessation in comparison to statin continuation, assuming a 2.5% unilateral risk and a non-inferiority margin of 5% of the between-group difference in overall mortality at 3 years after randomisation (nQuery Advisor software, v. 7.0). A 5% non-inferiority margin is fairly common in the literature, but a 3-year follow-up is not. This yields an annual non-inferiority margin of only 1.7%, which we think would be quite acceptable for the scientific community. We do not have to deal with multiple tests, as cost-effectiveness analyses do not require statistical testing [20].
We also assume that 20% of the individuals randomised to the statin-continuation group may spontaneously stop their treatment during follow-up. Therefore, to ensure 90% statistical power in under-treatment analysis, it was decided to include 20% more individuals in the statin-continuation group. In total, 2,430 individuals will be enrolled in the SITE study, including 1,080 in the statin-cessation group and 1,350 in the statin-continuation group.
Data management
Electronic case report forms (eCRFs) are the primary data-collection instrument for the study. Medical information for individual participants obtained as a result of this study are considered confidential, and disclosure to third parties is prohibited. eCRFs are being labelled with a unique trial number.
Consent forms sent to sponsors may contain patient identifiers for the purpose of monitoring, as described in the trial risk assessment. Such information will be kept in secure, locked storage.
Monitoring is mainly performed through e-CRF and phoning. However, on –site monitoring is performed after the 3 months-follow-up and after the last follow-up in each of the recruitment sites.
Statistical methodology
In the non-inferiority analysis, the bilateral 95% confidence interval (CI) for the between-group difference in the 3-year mortality rate will be estimated. Considering <5% change as clinically non-significant, we will consider non-inferiority demonstrated only if the 95% CI upper bound of the 3-year mortality difference between the two groups is below 5%. This analysis will primarily be performed according to an under-treatment principle. Formally, patients who do not receive the trial treatment as planned in the protocol will be excluded from the analysis. An intention-to-treat analysis will also be performed with the aim of maintaining the initial effect of randomisation. The missing = failure strategy (failure being defined as death) will be used in the non-inferiority analysis.
Secondary clinical analyses will be conducted according to the intention-to-treat principle: all randomised patients will be analysed in the group in which they were initially randomised, and all of their data will be used regardless of eventual treatment changes during the study. As economic analysis is not based on statistical tests, no adjustment of the α-level is needed to deal with multiple comparisons. In all analyses, p < 0.05 will be taken to indicate statistical significance.
The economic analysis will also be conducted according to the intention-to-treat principle. First, the point estimate of the incremental cost/utility ratio and its bootstrap 95% confidence interval will be calculated. We will then estimate the incremental net monetary benefit (INMB) adjusted for potential confounding factors (adherence to statins at baseline, treatment length, and particular statin dosage) using multiple linear regression models. Acceptability curves will then be plotted. The value of information will also be examined according to the principles proposed by Claxton [20].
Safety and adverse event reporting
The adverse events expected during this research are mainly those related to the age of the study population, such as CV events, diabetes and/or its complications, cognitive disorders and dementia, falls, cancers, deaths, and adverse effects of concomitant therapies. Given that one group of patients will continue to use statins, adverse reactions listed in the summaries of product characteristics are also expected.
The sponsor unit charged with security and vigilance must be informed immediately of any serious adverse events (SAEs), regardless of whether expected or unexpected, via the eCRF. This ensures a continuous monitoring of safety of included patients throughout the trial. Moreover, annual safety evaluations will be conducted by the security and vigilance unit of the sponsor, with particular focus on CV morbidity and all-cause mortality in both groups. The Data Safety Monitoring Board (DSMB) of the study will meet when 50% of the patients will reach 12, and 24 months of follow-up. We did not anticipate criteria by which the trial may be stopped, but advices of DSMB could be solicited in case of unexpected safety issues or new facts. In addition, all CV events will be validated by an external adjudication committee, which includes cardiologist, neurologist, internal and geriatric medicine specialists.
The sponsor’s security and vigilance unit must communicate serious unexpected adverse reactions (SUSARs) and new facts in the research in a timely manner in accordance with current regulations:
- to the French National Agency for the Safety of Medicines and Health Products, (Agence Nationale de Sécurité du Médicament et des Produits de Santé, ANSM).
- to the research ethics committee (Comité de Protection des Personnes, CPP). The committee will ensure, if necessary, that the research participants were informed about the side effects and confirm their consent.
On the date of first inclusion, the security and vigilance unit prepares an annual safety report including:
- a list of serious side effects of the research,
- a concise and critical analysis of the safety of participants in the research.
This report is sent to the ANSM and the CPP within 60 days of the date of first inclusion.
Trial registration
This research has been registered with clinicaltrials.gov under the number NCT02547883.
Amendment
To extend inclusions, the protocol was amended in March 2019 to also include patients from hospital departments.