Practice Point: The most current evidence for treating chronic pain with antidepressants favors SNRIs, but influence from pharma may be skewing the data.

EBM Pearl: Umbrella reviews are an emerging form of review: their systematic review of systematic reviews without meta-analysis allows for a bird’s eye view of topics but comes at the expense of less scrutiny and more heterogeneity (which equals more potential bias).

Umbrella reviews are, as the kids say, an “evolving concept.” If you’re not sure what an umbrella review is, you’re not alone — it took a while for the EBM community to sort it out, too. Fourteen years ago, an academic paper described them as similar to what we would nowadays call a “network meta-analysis.”

  • Meta-analysis - combines data from a bunch of similar randomized trials to answer a specific question, for example: “How does amitriptyline compare to placebo in treating neuropathic pain?”
  • Network meta-analysis - compares several different treatment options, such as amitriptyline, nortriptyline, or SSRIs, often using the transitive property such that “if A is better and B and B is better than C, A must also be better than C.”

Our present day understanding of the difference between a network meta-analysis and an umbrella review is that while umbrella reviews are systematic reviews of other systematic reviews, a separate meta-analysis is not necessarily performed. Umbrella reviews may include meta-analysis, network meta-analysis, or just qualitative reviews. They group data by population, intervention, or both, but don’t necessarily re-analyze the data. They do allow us to see overall trends in the focused set of outcomes found by individual systematic reviews. The main downside of an umbrella review is the additional risk of bias that comes with grouping data in potentially heterogeneous ways. And like any review, the validity of umbrella reviews is limited by the quality of the data analyzed. More recently, umbrella reviews are being considered to look at a wider variety of questions in healthcare (not just medication interventions but observational studies and therapies like exercise).

In January, the BMJ published an umbrella review of antidepressants for the treatment of chronic pain. By now, it’s pretty clear that chronic pain and depression are often intertwined. In their introduction, the authors point out that antidepressants are sometimes more frequently prescribed for chronic pain than for depression. The authors followed PRIOR guidelines, using systematic search and reporting methods to focus on higher quality papers, meta-analyses, and network meta-analyses and then evaluate them systematically, including looking for bias. They bundled data from 156 trials involving over 25,000 people with chronic pain from a variety of etiologies treated with various antidepressant medications (tricyclics, SSRIs, SNRIs, even a NDRI noradrenaline dopamine reuptake inhibitor).

Their bottom line is this: there is little high-quality evidence supporting the treatment of chronic pain with antidepressants, and what information there is favors the SNRI class of medication. This umbrella review of research published in the last ten years of treating chronic pain with antidepressants allows us to see a trend we might have missed otherwise. Depending on how you look at the studies, up to three-quarters of recent research on depression and chronic pain has strong ties to industry. Is it little wonder that we have tons of recent data on (relatively new and expensive) SNRI’s compared to (old and generic) TCAs? Most recent data seem to imply that a particular group of drugs is best, but it is also true that most of the funding at this moment in time seems to be directed at this class of drugs. Currency and volume of data are important, but they aren’t the only marker of quality.

For more information, see the topic Management of Chronic Pain in DynaMed.

Reference: BMJ. 2023 Feb 1;380:e072415