Skip to main content

Table 3 Examples of degree of concordance between 2009 prediction and updated conclusion a

From: Assessment of a method to detect signals for updating systematic reviews

Examples Predictions and conclusions
Example 1  
Original conclusion (from CER on analgesics for osteoarthritis) No clear differences between various nonaspirin, nonselective NSAIDs or partially selective NSAIDs with regard to efficacy for pain relief or improvement
2009 surveillance assessment [14] Conclusion still valid
Conclusion from 2011 CER update [23] No clear difference in efficacy for pain relief, or withdrawals due to lack of efficacy
Concordance Good
Example 2  
Original conclusion (from CER on analgesics for osteoarthritis) Etoricoxib is associated with fewer gastrointestinal adverse events than nonselective NSAIDs
2009 surveillance assessment [14] Possibly out-of-date
Conclusion from 2011 CER update [23] No comparable conclusion, as etoricoxib was not included because it did not gain FDA approval for sale in the United States
Concordance Good
Example 3  
Original conclusion (from CER on second-generation antidepressants) Overall discontinuation rates did not differ significantly between SSRIs as a class and bupropion, mirtazapine, nefazodone, trazodone and venlafaxine. In the case of venlafaxine compared with SSRIs, higher discontinuation rates due to adverse events appeared to be balanced by lower discontinuation rates due to lack of efficacy.
2009 surveillance assessment [16] Conclusion is possibly out-of-date, and this portion may need updating based on new analysis showing lower dropout rate with escitalopram.
Conclusion from 2011 CER update [24] Meta-analyses of numerous efficacy trials indicate that overall discontinuation rates are similar. Duloxetine and venlafaxine have a higher rate of discontinuations due to adverse events than SSRIs as a class. Venlafaxine has a lower rate of discontinuations due to lack of efficacy than SSRIs as a class.
Concordance Fair: Escitalopram data did not end up in the conclusions
Example 4  
Original conclusion from CER on second-generation antidepressants Three head-to-head RCTs suggest that no substantial differences exist between fluoxetine and sertraline, fluvoxamine and sertraline, and trazodone and venlafaxine regarding relapse. Twenty-one placebo-controlled trials support the general efficacy and effectiveness of most second-generation antidepressants for preventing relapse or recurrence. No evidence exists for duloxetine.
2009 surveillance assessment [16] Conclusion is possibly out-of-date, and this portion of the CER may need updating to include evidence for duloxetine.
Conclusion from 2011 CER update [24] On the basis of results of six efficacy trials and one naturalistic study, no significant differences exist between escitalopram and desvenlafaxine, escitalopram and paroxetine, fluoxetine and sertraline, fluoxetine and venlafaxine, fluvoxamine and sertraline, and trazodone and venlafaxine for preventing relapse or recurrence.
Concordance Fair: No duloxetine evidence ended up being included with regard to this key question
Example 5  
Original conclusion (from CER on management of GERD) Medical therapy with PPIs and surgery (fundoplication) appeared to be similarly effective for improving symptoms and decreasing esophageal acid exposure.
2009 surveillance assessment [18] Conclusion is still valid, and this portion of the CER does not need updating.
Conclusion from 2011 CER update [25] The 2005 CER concluded that medical therapy with PPIs and antireflux surgery were similarly effective in improving GERD-related symptoms and decreasing esophageal acid exposure, although some surgical patients required ongoing medical therapy postprocedure. With the addition of long-term follow-up data (7 to 12 years) from two previously reviewed studies and results from two new RCTs, our updated review found that patients who underwent antireflux surgery experienced a greater improvement in heartburn and regurgitation at follow-up than did patients who received medical treatment alone.
Concordance Poor: Update indicates symptoms are better with surgery
  1. aCER, comparative effectiveness review; FDA, US Food and Drug Administration; GERD, gastroesophageal reflux disease; NSAID, nonsteroidal anti-inflammatory drug; PPI, proton pump inhibitor; RCT, randomized controlled trial; SSRI, selective serotonin reuptake inhibitor.