Article type
Year
Abstract
Background: Various gaps may exist in the process from evidence synthesis to its implementation. Evidence synthesis may have been biased and late; guidelines may not have reflected the most up-to-date evidence; clinicians may have made decisions without instructed by relevant evidence.
Objectives: We aimed to examine the gaps between the ideally synthesized evidence, guideline recommendations and real-world clinical practices in the prescription of new generation antidepressants for major depressive disorder (MDD) through the past three decades.
Methods: We used cumulative network meta-analyses (NMAs) to represent ideally synthesized evidence over time. We built a Shiny web app to perform and present NMAs interactively. The analyses were based on a comprehensive dataset of randomized controlled trials of 21 antidepressants in the acute treatment of MDD. The primary outcomes were efficacy (treatment response) and acceptability (all-cause discontinuation), and treatment effects were summarized via odds ratios. We evaluated the confidence in evidence using the CINeMA (Confidence in Network Meta-Analysis) framework for several consecutive NMAs. The Shiny app presents network plots, two-dimensional plots combining efficacy and acceptability for each drug, forest plots, league tables, and funnel plots. We identified and extracted recommendations from several representative practice guidelines (Table 1). We estimated the real-world prescription patterns of antidepressant monotherapy for MDD, using the Medical Expenditure Panel Survey, a nationally representative database in the US. We evaluated the gaps between the results from NMAs, recommendations and prescriptions between 1990 and 2016.
Results: The Shiny app is accessible at https://cinema.ispm.unibe.ch/shinies/GRISELDA/. It indicates dramatic changes of drugs with relative superiority, and potentially exaggerated performance of newly approved drugs. Guidelines are usually updated every 5 to 10 years (Table 1). All proposed specific antidepressant recommendations, and most recommended drugs showed relatively acceptable efficacy and acceptability in the NMA at that time. However, for fluvoxamine and duloxetine, even they had already appeared barely satisfactory in efficacy, acceptability and credibility, yet still being recommended more than 5 years later. In the US, the prescriptions for some newly launched drugs were very large, even without formal recommendations or firm evidence (Figure 1).
Conclusions: Our study revealed the gaps from evidence to real-world practice in the antidepressant treatment of MDD. Considering the initially amplified effects in the evidence, recommendations about new drugs should be made with caution. Since evidence is the cornerstone in the process of evidence-based medicine, it should be kept up to date using rigorous synthesis methods. We also provided an example of how to present and visualize the evidence interactively through a Shiny app, in order to help policy-makers and clinicians comprehend the evidence.
Patient or healthcare consumer involvement: None.
Objectives: We aimed to examine the gaps between the ideally synthesized evidence, guideline recommendations and real-world clinical practices in the prescription of new generation antidepressants for major depressive disorder (MDD) through the past three decades.
Methods: We used cumulative network meta-analyses (NMAs) to represent ideally synthesized evidence over time. We built a Shiny web app to perform and present NMAs interactively. The analyses were based on a comprehensive dataset of randomized controlled trials of 21 antidepressants in the acute treatment of MDD. The primary outcomes were efficacy (treatment response) and acceptability (all-cause discontinuation), and treatment effects were summarized via odds ratios. We evaluated the confidence in evidence using the CINeMA (Confidence in Network Meta-Analysis) framework for several consecutive NMAs. The Shiny app presents network plots, two-dimensional plots combining efficacy and acceptability for each drug, forest plots, league tables, and funnel plots. We identified and extracted recommendations from several representative practice guidelines (Table 1). We estimated the real-world prescription patterns of antidepressant monotherapy for MDD, using the Medical Expenditure Panel Survey, a nationally representative database in the US. We evaluated the gaps between the results from NMAs, recommendations and prescriptions between 1990 and 2016.
Results: The Shiny app is accessible at https://cinema.ispm.unibe.ch/shinies/GRISELDA/. It indicates dramatic changes of drugs with relative superiority, and potentially exaggerated performance of newly approved drugs. Guidelines are usually updated every 5 to 10 years (Table 1). All proposed specific antidepressant recommendations, and most recommended drugs showed relatively acceptable efficacy and acceptability in the NMA at that time. However, for fluvoxamine and duloxetine, even they had already appeared barely satisfactory in efficacy, acceptability and credibility, yet still being recommended more than 5 years later. In the US, the prescriptions for some newly launched drugs were very large, even without formal recommendations or firm evidence (Figure 1).
Conclusions: Our study revealed the gaps from evidence to real-world practice in the antidepressant treatment of MDD. Considering the initially amplified effects in the evidence, recommendations about new drugs should be made with caution. Since evidence is the cornerstone in the process of evidence-based medicine, it should be kept up to date using rigorous synthesis methods. We also provided an example of how to present and visualize the evidence interactively through a Shiny app, in order to help policy-makers and clinicians comprehend the evidence.
Patient or healthcare consumer involvement: None.