concurring.
Although I fully agree with the Court’s order, I write separately to address a number of issues raised by the dissent.
The dissent argues that the State’s eighteen unsuccessful efforts to run an intravenous line into Romell Broom’s veins over the course of two hours demonstrates the sensibility of Ohio’s execution protocol because the State ultimately halted the execution attempt. The argument continues that, if the same problems arise during Reynolds’ execution, this execution similarly would be stopped. Thus, the likelihood that he would ultimately prevail in his § 1983 claim is too slim to warrant a stay of execution. This argument misses the mark.
Preliminarily, it overlooks the possibility that Broom has already suffered an Eighth Amendment violation by being subjected to this failed execution attempt. Even if Reynolds’ execution similarly was halted partway through, at that point he already may have suffered constitutional harm under Ohio’s protocol. Although the details of the failed Broom execution have not yet fully emerged, the initial reports suggest that the execution attempt could provide uniquely relevant evidence in support of the proposition that there is a “demonstrated risk of severe pain” under the revised Ohio protocol. See Baze v. Rees, 553 U.S. 35, 128 S.Ct. 1520, 1537, 170 L.Ed.2d 420 (2008).
The failed Broom execution raises concerns about the risks of maladministration under the Ohio protocol, and its intravenous siting provisions in particular. Ohio’s protocol allows for “as much time as is necessary to establish two [intravenous] sites” and the changes specifically grant the execution team members discretion in deciding if and when to abandon such efforts if problems arise. Reynolds raised concerns about the discretion granted to the execution team by the May 2009 revisions prior to the failed Broom execution and has argued that the risks presented by maladministration are part of his “core complaints.” That, in the interim, the State’s procedures further were called into question increases both the likelihood that Reynolds could ultimately succeed on the merits of his § 1983 claim and the likeli*958hood that, if no stay were to be granted, he would be harmed irreparably.
The dissent also emphasizes that Ohio revised its protocol precisely to alleviate Eighth Amendment concerns. While I have no doubt that Ohio did not revise its execution protocol in order to make executions more cruel or unusual, the State’s intent is not at issue. The question is whether the changes to the protocol amount to a factual predicate sufficient to revive Reynolds’ Eighth Amendment challenge under Cooey II. Even given the two previous instances when Ohio ran into difficulties administering its lethal injection protocol, the halting of Broom’s lethal injection operation prior to its completion was unprecedented. This event only strengthens Reynolds’ argument that the May 2009 changes sufficiently raised the risks of maladministration to revive his claims.
Finally, the dissent argues that “the Constitution allows the people to make policy mistakes, ... and correct them for themselves over time, and we should let that process run its course.... ” However, in considering a motion for a stay of execution, we must balance both the likelihood that the prisoner will prevail on the merits and the likelihood that, if no stay is granted, irreparable harm will occur. Indeed, the State has agreed not to attempt another execution of Broom until the district court can reconsider the matter. In this context, where allowing the process to run its course could result in the severest of consequences, it is more prudent to allow the district court to take these new circumstances into consideration.