What is the evidence. See evidence matrix in Epistemonikos later

We found three systematic reviews [1],[2],[3] including seven randomized controlled studies comparing iloprost vs placebo [4],[5],[6],[7],[8],[9],[10].

One of the studies did not report usable data for our analysis, so it is not considered in this table or in the summary of findings [5].  

All studies included patients with systemic sclerosis exclusively, except one that also included patients with severe primary Raynaud’s [6], but presented data in a way that allowed to separate the information of the systemic sclerosis group.

Patients included in the studies presented a mean of 18 to 28 attacks per week [4],[7],[9],[10]. Two studies included patients with more than 12 episodes per week [6], but did not report average baseline frequency, and one study did not report any data on this regard [8].

What types of interventions were included.

What types of outcomes were measured.

 To whom this evidence does and does not apply

• Studies have included patients with severe Raynaud’s phenomenon secondary to systemic sclerosis. If the intervention has any benefit, this would be of less magnitude in less severe patients. 

• None of the studies evaluated patients with Raynaud’s phenomenon secundary to other diseases, and only one study included a few patients with Primary Raunaud’s, so evidence should be applied cautiously to these groups.

• Outcomes selected for this summary are those for which there is more information, and have been used in systematic reviews and guidelines. However, some research has shown the more important outcomes for patients might be others, insufficiently researched up to know [11],[12]. 

• It is not possible to provide a risk/benefit ratio, since there is uncertainty about the latter.
• The use of iloprost in these patients is not supported by evidence and it is associated to adverse effects and costs. However, in non-resource-constrained scenarios  some patients might be inclined to use an unproven treatment. It is particularly important to inform the patient about the certainty of the evidence.

• This summary considers the outcomes more commonly used in guidelines. However, these migh not be the key outcomes for decision-making, so patients and clinicians will face substantial uncertainty.

• Iloprost has high cost, and even though there is variability about the administration period in the studies, it is usually used for long periods in clinical practice.
• Considering uncertainty about benefits, specially on the more important outcomes, it is not possible to provide an adequate cost/benefit analysis. However, given the scarce magnitude of benefits, if any, and the high cost, this balance is probably not favorable to using iloprost in these patients.

 Differences between this summary and other sources

• This summary differs substantially from individual systematic reviews identified. The Cochrane review [2] is the more complete one, but did not include one study that evaluated more patients than all of the other studies combined [7]. Regarding the other reviews, one of them did not conduct meta-analysis of all of the studies, presenting  the results of the Cochrane review and the newer study, but does not provide a pool estimate [1]. The third review is only focused on studies measuring digital ulcer healing outcomes [3].
• Our summary differs from the main guideline identified, which strongly recommends the intervention [13]. 

• Taking into account the current certainty of the evidence, the likelihood of new studies changing what we already know about the effects of iloprost on secundary Raynaud’s phenomenon is high. 
• We identified at least four registered randomized studies evaluating this question that have not been published yet nor provided results in any other format up to now [14],[15],[16],[17]. So, a new systematic review might provide discordant information from the body of evidence evaluated in this summary.