This protocol has been reported accordingly to the SPIRIT statements requirements. The complete SPIRIT checklist is available as an additional file and in the table at the end of the manuscript (additional file1).
Study design overview
The VESPA (Vortioxetine in the Elderly vs SSRIs: A Pragmatic Assessment) study is a randomized, parallel-group, multicentre, open-label, pragmatic, superiority trial. Over a 12-month recruitment period, psychiatrists from thirteen Italian Psychiatric Services will consecutively enrol in- and outpatients aged 65 or more suffering from an episode of major depression and requiring treatment with an antidepressant. Participants will be randomly allocated to vortioxetine or to one of the SSRIs. Apart from treatment allocation, clinicians and patients will be free of increasing or decreasing the dose according to clinical status and circumstances, as well as of stopping or continuing treatment as clinically indicated. Similarly, the use of concomitant medications during the study will be allowed according to clinical status and circumstances. Routine care outside the trial will continue as usual. During the study, participants will be seen as often as clinically indicated with no extra visits required for the trial. The only requirement will be follow-up visits at one, three, and six months of follow-up (Fig. 1).
As a consequence of these pragmatic characteristics oriented to resemble clinical practice as much as possible, both patients and clinicians will not be blind to pharmacological treatments provided during the trial. Blinding will be applied to outcome assessors and statisticians performing the analyses. The study has been designed according to the principles described in the CONSORT statement (extended version for pragmatic trials) [27] and in agreement with the SPIRIT 2013 statement [28] (see Appendix 1). The study is financially supported by the Italian Medicines Agency (AIFA - Agenzia Italiana del Farmaco) and has already been approved by the Ethics Committee for Clinical Research of Verona and Rovigo (Comitato Etico per la Sperimentazione Clinica delle Province di Verona e Rovigo) (prot. n. 61211 of the 19/09/2018; Protocol version n. 1.5 of the 09/06/2018).
Assessment of pragmatism
To quantify the level of pragmatism of our study, we employed the pragmatic–explanatory continuum indicator summary-2 (PRECIS-2) [29]. This is a validated tool, developed to help investigators make design decisions consistent with the intended purpose of their trial. It explores nine domains (eligibility criteria, recruitment, setting, organisation, flexibility (delivery), flexibility (adherence), follow-up, primary outcome, and primary analysis), for each of which a score from 1 (very explanatory) to 5 (very pragmatic) is provided. The result is graphically summarized in Fig. 2.
Reasons for the scoring are reported in Table 1. A routine use of the PRECIS-2 tool when submitting RCT protocols to funders, research ethics committees and peer-reviewed journals, has been growingly recommended, considering that not all RCTs self-labelled as "pragmatic" or "naturalistic" are actually pragmatic. This process can also help understand the extent to which trial results may be relevant to real-world practice [30].
Table 1. Scoring of PRECIS-2 tool. PRECIS 5-point Likert Scale score: (1) Very Explanatory; (2) Rather Explanatory; (3) Equally Pragmatic/Explanatory; (4) Rather Pragmatic; (5) Very Pragmatic.
Items | Score | Rationale |
Eligibility - to what extent are the participants in the trial similar to those who would receive this intervention if it was part of usual care | 4 | Target population: Elderly with depression. Inclusion criteria are wide. No exclusion criteria will be applied in terms of setting of recruitment, severity of depression, past use of psychotropic drugs, current use of benzodiazepines, number and severity of medical comorbidities, and multiple pharmacotherapy. Diagnosis are based on clinical judgment (guided by DSM-5 criteria), as it is in usual practice. Nevertheless investigator and patient have to agree to discontinue any current antidepressant, second generation antipsychotic, or lithium. |
Recruitment - how much extra effort is made to recruit participants over and above what that would be used in the usual care setting to engage with patients? | 5 | Participants will be recruited without extra efforts. They will be recruited during usual appointments and/or visits. |
Setting - how different is the setting of the trial and the usual care setting? | 4 | The study is multicenter, based in more than 10 psychiatric centers of the National Health System in Italy with a University center. |
Organisation - how different are the resources, provider expertise and the organisation of care delivery in the intervention arm of the trial and those available in usual care? | 4 | We will use usual staff and resources, but some extra resources will be necessary to hire researchers for the study. |
Flexibility (delivery) - how different is the flexibility in how the intervention is delivered and the flexibility likely in usual care? | 5 | The intervention is flexible, similar to usual care. |
Flexibility (adherence) - how different is the flexibility in how participants must adhere to the intervention and the flexibility likely in usual care? | 4 | No extra measures. Participants will be free to assume the intervention or drop it, but drugs will be prescribed and given to the participants during visits. This is different from usual care (patients have a prescription and go to the pharmacy to buy drugs). |
Follow-up - how different is the intensity of measurement and follow-up of participants in the trial and the likely follow-up in usual care? | 4 | The primary outcome will be assessed after 1, 3 and 6 months, as it is usually done in everyday practice. Six months represent a clinically sound time frame for assessing the overall tolerability of medications, including both acute, short-term and medium-long-term effects. Nevertheless, visits could be longer than usual to assess all the scales and long-term effects and adverse events could occur after 6 months. |
Primary outcome - to what extent is the trial's primary outcome relevant to participants? | 5 | Primary outcome is relevant to participants and policy makers. |
Primary analysis - to what extent are all data included in the analysis of the primary outcome? | 5 | The Intention to Treat (ITT) population will consist of all randomized patients. This ITT population will be used for the analysis of both primary and secondary outcomes. Missing values in rating scales will be imputed using the Last Observation Carried forward (LOCF) approach. |
Participants
The following inclusion criteria will be applied:
- The participant is 65 years old or above;
- The participant is willing to participate by signing an informed consent form;
- The participant is suffering from an episode of major depression, based on clinical judgment (guided by DSM-5 criteria);
- Treatment with an antidepressant is appropriate, based on clinical judgment;
- There is agreement between investigator and participant to discontinue any of the following concomitant drugs: antidepressant, second generation antipsychotic, or lithium. All other concomitant medications are allowed;
- Uncertainty about which trial treatment would be best for the participant.
Participants will be excluded in case of:
- Dementia, of any type and stage, as formally diagnosed by a specialist (geriatrician, neurologist, or others);
- Diagnosis of schizophrenia or bipolar disorder;
- Clinical conditions or treatments that contraindicate the use of oral vortioxetine or SSRIs, according to clinical/medical judgment (for example conditions or treatments that increase risk of bleeding, seizures, serotoninergic syndrome, hyponatraemia, etc.).
All medications will be prescribed according to routine clinical practice, in compliance with the Summary of Product Characteristics (SPC) registered in the AIFA databank (https://farmaci.agenziafarmaco.gov.it/bancadatifarmaci/home).
No exclusion criteria will be applied in terms of setting of recruitment, severity of depression, past use of psychotropic drugs, current use of benzodiazepines (as long as SPC indications are respected), number and severity of medical comorbidities, and multiple pharmacotherapy. Such criteria will select participants similar to those who require antidepressant treatment under usual care, including patients with multiple medical comorbidities. The recruitment will be pragmatic, as participants will be selected among people attending inpatient and outpatient community services. There will not be overt recruitment effort. Also, allowing different recruitment settings, having multiple sites of recruitment, and selecting patients similar to those who are treated in every day clinical practice, will increase the generalizability of trial results. To control for a potential risk of excessive heterogeneity between centres, the randomization will be stratified by centre. According to these features, the PRECIS-2 “setting” domain has been evaluated as pragmatic.
Interventions
Patients will be randomized to either vortioxetine or one of the SSRIs. Doctors will be free to choose which SSRIs is more appropriate among those marketed in Italy and commonly used in clinical practice in the elderly (sertraline, citalopram, escitalopram, paroxetine, fluoxetine, fluvoxamine). A flexible dosing schedule, within the licensed dose range and in line with the summary of product characteristics (SPC), will be suggested (Table 2) in order to resemble clinical practice as much as possible.
Table 2. Treatments and dosing schedule
Medication | Licensed dose range in the elderly | Notes from the registered Summary of Product Characteristics |
vortioxetine | 5–20 mg/day | The minimum effective dose of 5 mg vortioxetine once daily should always be used as an initial dose for participants aged ≥ 65 years. Caution should be exerted when prescribing to elderly participants at doses above 10 mg vortioxetine once daily. |
sertraline | 50–200 mg/day | Caution is required in the elderly, because these patients may be at greater risk of hyponatraemia. |
paroxetine | 20–40 mg/day | In the elderly, increased plasma concentrations of paroxetine have been reported, however within the range observed in younger subjects. The treatment should start at the same doses used in adults. |
citalopram | 10–20 mg/day | In the elderly, half of the dose range prescribed in adults is required. |
escitalopram | 5–10 mg/day | In the elderly, half of the dose range prescribed in adults is required. |
fluoxetine | 20–60 mg/day | Caution is required when the dose is increased in the elderly, and generally the daily dose should not be above 40 mg/day. The maximum recommended dose is 60 mg/day. |
fluvoxamine | 100–300 mg/day | In elderly participants, titration should be slower and the dosage should always be established with caution. |
Formulation choice (tablets versus drops) will be made by clinicians and participants following every day practice, and no measures will be implemented to optimise treatment adherence.
According to the PRECIS-2 “flexibility-delivery” and “flexibility-adherence” domains, treatment delivery has been rated as pragmatic, although a full score of 5 could not be reached as we were formally required to follow the EU pharmacovigilance regulation [31, 32].
Outcome measures
The number of participants withdrawing from allocated treatment due to adverse events at the end of the study (6 months) will represent the primary outcome. This measure may be considered a pragmatic proxy of tolerability [33] as it occurs when adverse events actually reach an unbearable burden, as perceived by patients and/or relatives and/or carers and/or clinicians. Antidepressant treatment will be considered withdrawn due to adverse effects when the drug is stopped for more than two consecutive weeks following the occurrence of any adverse event, based on clinical judgment and/or as reported by participants. Participants will be additionally evaluated after also one and three months from randomization, collecting relevant clinical information and assessing scales, as showed in table 3. Side effects responsible for treatment withdrawal, and their severity, will be recorded the follow-up form and an ad hoc form for Severe Adverse Events (SAE).
Secondary outcomes will include:
- acceptability: withdrawals from allocated treatment due to any cause (this outcome measure will include withdrawals for side-effects plus withdrawals for any other issues);
- overall mortality;
- any episode of deliberate self-harm;
- suicide mortality;
- adverse events, measured as the mean change in scores at the Antidepressant Side-Effect Checklist (ASEC) [34] at each time point. ASEC is a validated rating scale measuring the occurrence and severity of 21 antidepressant adverse events;
- response to treatment, defined as a reduction of at least 50% of the baseline score of the Montgomery–Åsberg Depression Rating Scale (MADRS) [35] at each time point. MADRS is a validated, ten-item questionnaire for assessing the severity of depression;
- efficacy, measured as mean change scores at MADRS at each time point;
- quality of life, measured as mean change scores of the self-administered scale EQ-5D,[36], at each time point. EQ-5D explores five areas, including mobility, self-care, usual activities, pain/discomfort, anxiety/depression, and assesses the overall subjective perception of health with an analogic scale;
- cognitive performance, measured as mean change scores of the Short Blessed Scale (SBT) [37],at each time point. SBT is a validated, six-item weighted instrument, originally designed to identify dementia, which assesses orientation, registration, and attention.
Rating scales to assess the secondary outcomes will be administered by blind assessors at one, two and three months after randomization. In addition, the Charlson Age-Comorbidity Index (CACI) [38] will be employed. This is a validated rating scale used to evaluate the degree of medical comorbidity, and to predict the 10-year survival in participants with multiple comorbidities. All study tools and phases are shown in Table 3.
Table 3. Study phases and tools
Procedures and tools | T0 Enrolment phase (duration: 12 months) | T1 (1 month) | T2 (3 months) | T3 (6 months) |
Review of criteria for inclusion in the study | X | | | |
Informed consent document signed | X | | | |
Randomization (allocation to treatment and number assigned) | X | | | |
Recruitment Form | X | | | |
ASEC | | X | X | X |
MADRS | X | X | X | X |
EQ-5D | X | X | X | X |
CACI | X | X | X | X |
SBT | X | X | X | X |
Follow-up form | | X | X | X |
Severe Adverse Event (SAE) Form | | ⇓ ⇓ any time ◊ ◊ |
Safety
The VESPA study will operatively employ the definitions endorsed by the EC Directive 2001/20/EC,[39]. As soon as a severe adverse event occurs, an ad hoc form for Severe Adverse Events (SAE) will be filled in and forwarded to the coordinating centre (University of Verona), in accordance with the EU regulation about pharmacovigilance in clinical research [31]. If, for any reasons, the disadvantages of participation will appear to be significantly greater than foreseen, the Principal Investigator of the site will inform trial participants and the bodies providing ethical oversight.
Considering that the study medications are already in the Italian market, and considering that they will be prescribed for licensed indications without altering clinical practice, the VESPA study has not appointed an ad hoc data safety and monitoring committee.
Randomization
Participants will be randomly assigned to vortioxetine or SSRIs with an allocation ratio of 1:1. A centralized web-based randomization procedure will be employed to guarantee the concealment of allocation. The trial biostatistician will prepare the sequence of treatments randomly permuted in blocks of constant size. The site investigators will not know the block size. Allocation will be stratified by recruiting centre. By using the web-based application RedCap [40], investigators will be able to screen participants for inclusion, administer instruments maintaining the blindness to treatment allocation, and randomize them.
Data management
At baseline, before randomization, and after one, three and six months, a number of socio-demographic and clinical information will be collected, along with the administration of the above-mentioned validated rating scales (MADRS, EQ-5D, CACI, SBT, ASEC). All data on other medications will be registered at every visit. The ASEC scale will be administered only during follow-up.
All study data will be collected with RedCap and digitally stored by the Istituto di Ricerche Farmacologiche Mario Negri IRCCS, a not-for-profit biomedical research organization based in Milan (Italy), where also the statistical analysis will be performed. RedCap will allow an immediate data validation at the moment of data collection. Moreover, a set of electronic and manual edit checks will be performed. The local coordinator of each recruiting centre will store and safely preserve hard copy documents (signed informed consent and self-administered questionnaires) for at least 7 years after the end of the study, according to the Italian law. At the end of the study the full dataset will be made available upon motivated request as a spreadsheet file in an online repository (e.g. Dryad Digital Repository). This is in line with FAIR principles [41], aimed at enhancing the accessibility and reutilization of novel research data.
The accuracy and completeness of data collection will be monitored by site visits. At least one visit for each recruiting centre is planned. Furthermore, auditing will be also carried out remotely, as the data manager of the study will be able to regularly check the trial dataset through the web application RedCap.
Power analysis
Considering the differential rate of withdrawals due to adverse events between SSRIs and vortioxetine on the basis of a meta-analysis of antidepressants for older people [9] and of three clinical trials of vortioxetine in older patients with depression [21, 26, 42] we expect the vortioxetine group to show a clinically significant advantage by reducing this rate from about 17% [9] to about 5% [21, 26, 42]. A sample size of 276 participants (138 in each group) achieves 90% power to detect a difference of 12% between the two withdrawal proportions in favour of vortioxetine. The test statistics will be the two-sided Z test with pooled variance. The significance level of the test is targeted at 5%. On the basis of the above-mentioned studies, we can assume that about 23% of the participants could be lost within 6 months (the mean of the total dropout rates of vortioxetine and SSRI studies in the elderly). Therefore 358 participants (179 in each group) will be enrolled in order to obtain at least 276 evaluable subjects. The sample size calculation was performed according to the methodology described by Pocock [43].
Statistical Analysis
According to the pragmatic principle of intention-to-treat (ITT), efforts will be made to follow each participant until the end of the study. The ITT population will consist of all randomized participants, and will be used for the analysis of both primary and secondary outcomes. The absolute risk of the primary outcome will be calculated on the ITT population. Subjects with missing primary outcome data will be allocated to the worst outcome. When possible, in addition to the primary analysis, appropriate statistical methods will adjust for the potential confounding effect of prognostic factors (sex, age, living condition, severity of comorbid medical conditions, previous psychiatric history, MADRS score at baseline). Missing rating scales scores will be imputed using the Last Observation Carried forward (LOCF) approach: ratings will be carried forward from the last available assessment to the 6-month follow-up assessment. As a secondary analysis, missing scores will be imputed following a multiple imputation approach [44].
In order to check the results of the ITT approach, though for confirmatory purposes only, the primary outcome will also be analysed using a per-protocol (PP) approach. According to the PP approach, analysis will be restricted to subjects with primary outcome assessment available at six-months. Subjects withdrawing for reasons not related to adverse effects will be excluded from the analysis.
The proportion of participants withdrawing from the study due to adverse events within 6 months of follow-up will be compared between the two groups of treatment using a logistic regression with centre (random variable) as a covariate. A multivariable analysis (secondary analysis) will be performed through a Poisson regression model with a robust error variance, given that this procedure allows to estimate relative risks directly [45].
For dichotomous secondary outcomes, the proportion of participants withdrawing from the study due to adverse events within 6 months will be compared between the two groups of treatment using a logistic regression with centre (random variable) as a covariate. When possible, a multivariable analysis will be performed through a Poisson regression model with a robust error variance. For continuous secondary outcomes, the 6-month estimate will be compared between the two groups of treatment with an analysis of covariance with baseline value as an additional covariate, or with Mann-Whitney test on changes, according to the variable’s distribution. Same outcomes will be studied using linear mixed models taking into account all assessments to evaluate the rate of change with shorter repeated evaluations and no need of missing imputation. Further, score changes in subscales will be evaluated in order to detect possible specific treatment-related side-effects.
A Cox proportional hazard model will be used to explore time to treatment withdrawal due to adverse events (secondary analysis). The proportional hazard assumption of the effects will be tested.
Adverse events will be tabulated. Nominal value for statistical significance will be set at 0.05, two-tailed. A specific Statistical Analysis Protocol will be produced and made publicly available before the inclusion of the last participant. All analyses will be performed using STATA [46], release 15 or higher.