HARBISON v. LITTLE

Reset A A Font size: Print

United States Court of Appeals,Sixth Circuit.

Edward Jerome HARBISON, Plaintiff-Appellee, v. George LITTLE, in his official capacity as Tennessee's Commissioner of Correction, et al., Defendants-Appellants.

No. 07-6225.

Decided: July 02, 2009

Before:  SILER, CLAY, and COOK, Circuit Judges. ARGUED:  Joseph Frederick Whalen, Office of the Tennessee Attorney General, Nashville, Tennessee, for Appellants.  Stephen M. Kissinger, Federal Defender Services of Eastern Tennessee, Inc., Knoxville, Tennessee, for Appellee.   ON BRIEF:  Joseph Frederick Whalen, Office of the Tennessee Attorney General, Nashville, Tennessee, for Appellants.  Stephen M. Kissinger, Dana Hansen Chavis, Federal Defender Services of Eastern Tennessee, Inc., Knoxville, Tennessee, for Appellee.

OPINION

Edward Jerome Harbison is a Tennessee prisoner under death sentence who has exhausted all appeals and was denied a writ of habeas corpus.   In 2006, Harbison filed a complaint under 42 U.S.C. § 1983, challenging Tennessee's lethal injection protocol.   The district court granted judgment in favor of Harbison, holding that the protocol violated the Eighth Amendment.   The state defendants (State) appealed, relying on the Supreme Court's decision in Baze v. Rees, --- U.S. ----, 128 S.Ct. 1520, 170 L.Ed.2d 420 (2008), which was decided after the district court decision in this case.  Baze upheld Kentucky's lethal injection protocol and held that a substantially similar protocol would not violate the Eighth Amendment.   Finding Tennessee's protocol substantially similar, we vacate the district court's judgment and remand for further proceedings.

I. Background

In 1984, Harbison was convicted of first-degree murder, second-degree burglary, and grand larceny, and sentenced to death.   On direct appeal, the Tennessee Supreme Court affirmed his convictions and sentence.  State v. Harbison, 704 S.W.2d 314 (Tenn.1986).   The state courts also denied Harbison's claims for post-conviction relief.   See Harbison v. State, No. E2004-00885-CCA-R28-PD, 2005 WL 1521910 (Tenn.Crim.App. June 27, 2005) (unpublished opinion);  Harbison v. State, No. 03C01-9204-CR-00125, 1996 WL 266114 (Tenn.Crim.App. May 20, 1996) (unpublished opinion).

In 1997, Harbison filed a 28 U.S.C. § 2254 petition for a writ of habeas corpus in federal district court.   In 2001, the district court dismissed Harbison's petition as meritless, and we affirmed.   See Harbison v. Bell, 408 F.3d 823 (6th Cir.2005).   Harbison pursued additional relief in federal court, but his efforts were not successful.   See Harbison v. Bell, 503 F.3d 566 (6th Cir.2007), cert. denied, --- U.S. ----, 128 S.Ct. 1479, 170 L.Ed.2d 302 (2008), cert. granted, --- U.S. ----, 128 S.Ct. 2959, 171 L.Ed.2d 884 (2008), and rev'd on other grounds, --- U.S. ----, 129 S.Ct. 1481, 173 L.Ed.2d 347 (2009).

In 2006, Harbison filed this 42 U.S.C. § 1983 action, challenging the lethal injection protocol used by Tennessee.   After a bench trial, the district court granted judgment in favor of Harbison, holding that Tennessee's lethal injection protocol violated the Eighth Amendment because the State knowingly disregarded the protocol's substantial risk of inflicting unnecessary pain.  Harbison v. Little, 511 F.Supp.2d 872 (M.D.Tenn.2007).   The State appeals from this decision.

Harbison argues that the lethal injection protocol utilized by the State violates his Eighth Amendment rights because it involves the unnecessary and wanton infliction of pain.   In 1998, Tennessee adopted lethal injection as an option for the execution of prisoners sentenced to death, and it implemented a “three-drug” protocol for carrying out lethal injection.   See Workman v. Bredesen, 486 F.3d 896, 901 (6th Cir.2007), cert. denied, 550 U.S. 930, 127 S.Ct. 2160, 167 L.Ed.2d 887 (2007).   The three drugs utilized are sodium thiopental, pancuronium bromide, and potassium chloride.  Id. at 902.   The sodium thiopental anesthetizes the prisoner, the pancuronium bromide paralyzes the prisoner's muscles, and finally the potassium chloride induces cardiac arrest.  Id.

In 2007, the governor issued an executive order directing the Department of Correction to review Tennessee's lethal injection protocol due to concerns raised about the written procedures.   State of Tennessee, Executive Order by the Governor No. 43 (Feb. 1, 2007), available at http://​state.​tn. ​us/​sos/​pub/​execorders/​exec-​orders-​bred​43.​pdf. The Commissioner of the Department of Correction, George Little, formed a committee which conducted research, sought expert opinions, and studied the approaches of other jurisdictions.  Workman, 486 F.3d at 902.   Following this review, Tennessee issued an amended protocol, including an updated execution manual.  Id. The amended protocol set forth each step of the execution process in detail, as well as the qualifications, selection process, and training requirements for the execution team members.  Id. Although it considered other lethal injection alternatives, Tennessee decided to retain the three-drug protocol that it had been using for its lethal injection procedure.  Id. at 902-03.

That same year, we reviewed the amended protocol in the context of a temporary restraining order to suspend an execution in Workman.   We concluded that the inmate did not demonstrate a likelihood of success on the merits of his Eighth Amendment claim.  Id. at 905-06.   We noted that the amended three-drug protocol was designed to avoid the needless infliction of pain, rather than cause it.  Id. at 907.   Further, we found that the State's efforts in amending its protocol demonstrated an intent to not just meet the requirements of the Eighth Amendment, but to exceed them.  Id. at 907.

Later in 2007, the district court in this case concluded that the amended protocol retained an inherent risk of the sodium thiopental's being improperly administered and therefore Harbison would not be unconscious when the second and third drugs are administered.  Harbison, 511 F.Supp.2d at 884.   The court cited several reasons for reaching this conclusion.   First, the amended protocol did not provide a test for determining whether the inmate was conscious before administering the second drug, pancuronium bromide.  Id. at 884-86.   Second, the State did not carefully select and adequately train the individuals performing the execution.  Id. at 886-91.   Third, the protocol did not provide for tactile monitoring of the IV lines during the administration of the drugs.  Id. at 891-92.   The court also noted that the State protocol review committee had recommended several safeguards as part of its review process, including the adoption of a “one-drug” protocol, but that the Department of Correction did not adopt these recommendations when issuing the amended protocol.  Id. at 895.   In light of these factors, the court concluded that the State knowingly disregarded an excessive risk of causing pain to the inmate when it issued the amended protocol.  Id. at 903.

The district court distinguished our prior decision upholding the protocol in Workman because there was no evidentiary hearing before the Workman district court.   It found the assumptions we relied on in Workman to have been proven false by the testimony at the bench trial in this case.  Id. The district court criticized this court's reliance on the committee report, finding that after a four-day evidentiary hearing, “despite the hard work of the Protocol Committee, none of the recommendations that were the fruit of its hard work were accepted by Corrections Commissioner Little nor integrated into the new protocol.”  Id. at 899.   It also found two of the assumptions of Workman to be proven erroneous by the evidentiary hearing in Harbison's case.   First, it found that the risk of the sodium thiopental's failing to render the inmate unconscious, resulting in the inmate's feeling the effects of the two subsequent drugs, was more significant than we described given the lack of a check for consciousness before administration of the two final drugs and remote placement of the doctor in another room.  Id. at 900.   It also found important that we assumed that the committee, after all of its research, recommended the three-drug protocol.  Id. In actuality, the committee recommended the one-drug protocol, but Commissioner Little, who did not participate in the committee meetings or consult with medical experts, rejected the one-drug protocol, and recommended an amended three-drug protocol.  Id.

 Subsequent to the district court's decision, the Supreme Court addressed Kentucky's three-drug lethal injection protocol in Baze v. Rees, --- U.S. ----, 128 S.Ct. 1520, 170 L.Ed.2d 420 (2008).   The Court issued several opinions in that case, including Chief Justice Roberts's plurality opinion (writing for two other justices), one concurring opinion, four other opinions concurring in the judgment, and one dissenting opinion.   Under these circumstances, Chief Justice Roberts's plurality opinion is controlling.   See Emmett v. Johnson, 532 F.3d 291, 298 n. 4 (4th Cir.2008);  see also Walker v. Epps, 287 Fed. App'x 371, 375 (5th Cir.2008) (relying on plurality opinion for controlling legal standard).   A prisoner cannot successfully challenge a method of execution merely by showing that the method may result in pain, either by accident or as an inescapable consequence of death, or that a slightly safer alternative is available.  Baze, 128 S.Ct. at 1531.   In order for a lethal injection protocol to violate the Eighth Amendment, the inmate must show it “creates a demonstrated risk of severe pain.   He must show that the risk is substantial when compared to the known and available alternatives.”  Id. at 1537.   With respect to the disposition of future challenges to state protocols, the plurality opinion stated:  “A State with a lethal injection protocol substantially similar to the protocol we uphold today would not create a risk that meets this standard.”  Id.

II. Analysis

 This court reviews the district court's findings of fact for clear error and conclusions of law de novo.  Overton Distribs., Inc. v. Heritage Bank, 340 F.3d 361, 366 (6th Cir.2003).   Here, the issue of whether the protocol exposes the inmate to a substantial risk of serious harm is a question of law.   See Emmett, 532 F.3d at 300 (considering the same issue with respect to Virginia's lethal injection protocol post-Baze:  “․ Emmett has failed as a matter of law to demonstrate a substantial or objectively intolerable risk that he will receive an inadequate dose of thiopental․”).

Thirty-five states and the Federal government use lethal injection as their primary method of execution.   See Death Penalty Information Center, Facts About the Death Penalty 1 (2009), available at http://​www.​deathpen​altyinfo.​org/​Fact ​Sheet.​pdf. At least 30 states, including Tennessee, use the same three drugs:  sodium thiopental, pancuronium bromide, and potassium chloride, in varying amounts.  Baze, 128 S.Ct at 1527.

According to the Tennessee protocol, two paramedic technicians, members of the “IV Team,” insert catheters into each of the inmate's arms.   These catheters run from the inmate's arms to the execution room, where the executioner administers the drugs into one of the lines (the other is a back-up line).   The executioner injects each drug intravenously, with saline flushes in between the different drugs.   The drugs and saline flushes are administered in 50 cubic centimeter syringes.   The following amounts of each drug are used:  5 grams of sodium thiopental, administered in four syringes;  100 milligrams (1 mg/mL) of pancuronium bromide, administered in two syringes, and 100 milliliters (of a 2 mEq/mL) of potassium chloride, administered in two syringes.   The protocol also describes the role of each drug:  the sodium thiopental depresses the central nervous system, causing “sedation or sleep, depending on the dose,” the pancuronium bromide is a muscle paralytic that “will assist in the suppression on breathing and ensure death,” and finally the potassium chloride “causes cardiac arrest and rapid death.”   After all of the injections are administered, the executioner closes the IV line and opens up the drip chamber, and signals the warden that all of the syringes have been administered.   After a five-minute waiting period, a physician is brought in to pronounce the inmate's death.

As in Baze, 128 S.Ct. at 1526, the inmate here concedes that if the protocol were followed perfectly, it would not pose an unconstitutional risk of pain, and argues instead that maladministration of the sodium thiopental would result in a severe risk of pain from the subsequent drugs that could go undetected.  Harbison, 511 F.Supp.2d at 884.   The district court invalidated the Tennessee protocol on several bases:  failure to check for consciousness before the pancuronium bromide is administered, inadequate selection and training of personnel, and failure to provide for tactile monitoring of the IV lines during the administration of the drugs.  Id. at 884-92.   The Supreme Court, however, considered these risks under the Kentucky protocol, and found they did not constitute a substantial risk of serious harm.  Baze, 128 S.Ct. at 1537-38.   In addition, the Court rejected the failure to adopt a one-drug protocol as a basis for finding the current protocol unconstitutional.  Id. at 1535.   Given the direction in Baze that a protocol substantially similar to Kentucky's would not create a risk that violates the constitutional standard set forth in the Court's opinion, Tennessee's protocol must be upheld because Baze addressed the same risks identified by the trial court, but reached the conclusion that they did not rise to the level of a constitutional violation.

A. Failure to Check for Consciousness

 The district court first concluded that the amended protocol was deficient because it did not provide a proper procedure for ensuring that the inmate was unconscious before administering the pancuronium bromide.   Harbison, 511 F.Supp.2d at 884.   The court noted that other states required the execution team to determine if the inmate is still conscious before proceeding with this step.  Id. at 884-85.   The Tennessee protocol review committee also had recommended that procedures be put in place to ensure that the inmate was unconscious at this step.  Id. at 885-86.   Possible methods for determining consciousness included lightly brushing the eyelashes, lifting up an arm, or pinching a nipple.   Despite this recommendation, these safeguards were not adopted in the amended protocol.  Id. at 886.   Instead, the prison warden who is in the room with the inmate, and the executioners who would be able to see the inmate through a one-way glass window, monitor the prisoner visually during the execution process, which the State believed to be a sufficient safeguard.  Id. The district court disagreed, holding that the failure to check for consciousness greatly enhanced the risk that the inmate would suffer unnecessary pain.  Id.

Baze, however, rejected the necessity of the procedures relied on by the district court.   It noted at the outset that because a proper dose of sodium thiopental would render any check for consciousness unnecessary, “[t]he risks of failing to adopt additional monitoring procedures are thus even more ‘remote’ and attenuated than the risks posed by the alleged inadequacies of Kentucky's procedures designed to ensure the delivery of thiopental.”  Baze, 128 S.Ct. at 1536 (citing Hamilton v. Jones, 472 F.3d 814, 817 (10th Cir.2007) (per curiam);  Taylor v. Crawford, 487 F.3d 1072, 1084 (8th Cir.2007)).   While the plaintiffs in Baze argued that the state needed to adopt certain steps to ensure the prisoner's unconsciousness, including some of the same tests suggested by Harbison, the Court concluded that a visual inspection of the inmate by the warden was sufficient to protect the inmate's Eighth Amendment rights.  Id. at 1536-37.

B. Inadequate Selection and Training of Personnel

 Second, the district court determined that Tennessee's amended protocol was deficient because it did not contain adequate procedures for selecting and training the personnel involved in executions.  Harbison, 511 F.Supp.2d at 886-91.   The court noted that one of the members of Tennessee's current execution team had a history of drug and alcohol addiction and psychological disorders, which raised questions about the screening process for members of the execution team.  Id. at 887-88.   Further, the court found that the executioners did not have sufficient training.   Two of the executioners received approximately forty hours of training on the insertion of IV catheters, but not on setting up IV lines, administering drugs through the IV lines, or monitoring the IV lines during the injections.  Id. at 887.   The third executioner only received four hours of training on IV insertion.   Id. at 887.   Warden Bell, the only person in the room with the inmate, testified his training was viewing executions in Texas, visiting an execution site in Indiana, and talking with other states about the process.  Id. The executioners and Warden Bell also participate in monthly practice sessions where they inject saline solutions into volunteers.  Id. Medical experts at the trial identified several problems that could occur with IV catheters and lines during their use, including slippage of the catheter, stopcocks used to set the directional flow of the IV turned in the wrong direction, and injection of the wrong drug.  Id. at 888-90. When questioned, the members of the execution team could not identify these potential problems.  Id. at 890.   The district court held that the inability of the executioners to identify these problems demonstrated the likelihood that an accident would occur at some point which increases the inmate's risk of unnecessary pain.   Id. at 891.

The Supreme Court in Baze, however, reached a different conclusion when faced with a similar question.   While the Baze plaintiffs argued that the executioners were inadequately trained, the Court noted that Kentucky required team members to be certified medical professionals who engage in regular practice sessions.  Baze, 128 S.Ct. at 1533-34.   Further, Kentucky's protocol required the team to run primary and backup IV lines for the lethal injection process and to prepare two sets of the drugs before the execution begins.  Id. at 1534.   The warden and deputy warden monitored the IV lines during the execution for any signs of difficulties.  Id. The Court determined that these redundant measures ensured that the prisoner would still receive an adequate dose of sodium thiopental if problems should arise with one of the lines.  Id. Tennessee's amended protocol contains similar requirements.   Two certified paramedics insert the catheters into the inmate's arm.  Harbison, 511 F.Supp.2d at 886.   The execution team conducts monthly practice sessions where they inject saline solution into volunteers.  Id. at 887.   The warden also monitors the execution to safeguard against potential problems.  Id. at 886.   In addition, although medical experts testified that the State should employ an expert to advise the execution team on the mixing of the lethal injection chemicals, id. at 896, Baze rejected this requirement, noting that this task was not difficult.  Baze, 128 S.Ct. at 1533.

C. Failure to provide for tactile monitoring of the IV lines

 Third, the district court held that Tennessee's amended protocol failed to provide procedures for adequately monitoring the administration of drugs.  Harbison, 511 F.Supp.2d at 892.   The court noted that the IV lines were only monitored visually, by looking through a one-way window and by looking at a video screen in the executioner's room.  Id. at 891.   An expert opined that monitoring the IV lines by touching or palpating the insertion site would be far more effective in preventing potential problems.  Id. at 891-92.   Based on this testimony, the district court concluded that the sole use of visual monitoring increased the prisoner's risk of unnecessary pain.  Id. at 892.

Once again, Baze compels a different conclusion.   While the plaintiffs in that case raised concerns about the adequate monitoring of IV lines, the Court did not find the risk to rise to the level of a constitutional violation because the warden and deputy warden were in the execution room with the prisoner to visually monitor for signs of any problems.  Baze, 128 S.Ct. at 1534.   Medical experts in Baze testified that identifying signs of possible infiltration occurring at the IV site would be very obvious to the average person because of the swelling that would result.  Id. Similarly, Tennessee's amended protocol requires the warden to be in the execution room in order to guard against problems, and the IV line also is monitored visually by other execution team members by video camera and through a one-way window.   Harbison, 511 F.Supp.2d at 892.

D. Failure to adopt alternative procedures

 Finally, the district court held that the State's failure to adopt a “one-drug” protocol and other safeguards demonstrated cruel and unusual punishment.  Id. at 898.   The Tennessee protocol review committee recommended that the State begin to use the one-drug protocol because it was simpler and provided a lower risk of error.  Id. at 895.   However, Commissioner Little rejected this recommendation due to concerns about being in the forefront on this issue and potential political ramifications.  Id. at 896.1  The district court concluded that, by rejecting the one-drug protocol, the Corrections Department knowingly disregarded an excessive risk of pain to the prisoner.  Id. at 895-96.

While Harbison argues that Tennessee's rejection of the “one-drug” protocol reflects deliberate indifference to the likelihood of his suffering severe pain, the Baze Court determined that a state's failure to adopt the one-drug protocol did not violate the prisoner's Eighth Amendment rights since the comparative efficiency of that method was not well-established.  Baze, 128 S.Ct. at 1535.   Although the one-drug protocol was not specifically proposed to the state trial court in Baze, and therefore the Court did not have any findings of fact on the effectiveness of the one-drug alternative, it noted how the continued use of the three-drug protocol cannot be seen as an “objectively intolerable risk” in light of the fact that no other state adopted it.  Id.

 In addition, the district court's finding that the failure to adopt other safeguards, such as medical professionals to show the executioners how to mix the sodium thiopental and additional checks for consciousness after the sodium thiopental, Harbison, 511 F.Supp.2d at 896-98, requires reversal under Baze. Baze spoke to the general standard for evaluating proposed alternative procedures and held that the Eighth Amendment requires the plaintiff-inmate to first establish a substantial risk of serious harm before offering an alternative that is feasible, readily implemented, and that significantly reduces a substantial risk of severe pain.  128 S.Ct. at 1532.   Only if the inmate shows the State refused to adopt such an alternative “without a legitimate penological justification for adhering to its current method,” will such a refusal violate the Eighth Amendment.  Id.

III.

For these reasons, we vacate the decision finding the Tennessee lethal injection protocol violative of the Eighth Amendment, and remand to the district court to vacate the injunction barring the State from executing Harbison and to enter judgment consistent with this decision.2

VACATED and REMANDED.

The majority concludes that Tennessee's lethal injection protocol is “substantially similar” to the Kentucky protocol deemed constitutional in Baze v. Rees, --- U.S. ----, 128 S.Ct. 1520, 170 L.Ed.2d 420 (2008), and that the district court erred in granting judgment in favor of Harbison.   At first glance, the majority opinion is straightforward:  the majority marches through the standard set forth in Baze, contrasts the Baze plurality's findings with the district court's findings, and holds that Harbison failed to satisfy the Baze standard.   At closer inspection, however, it becomes obvious that this approach is flawed, both legally and analytically.   Because the district court issued its opinion before the Baze decision, the district court never had the opportunity to receive and consider evidence in light of the Baze standard, and it never rendered a judgment as to whether Tennessee's protocol complied with Baze. By failing to provide the district court with an opportunity to consider Tennessee's protocol in light of Baze, the majority effectively usurps the district court's role as a factfinder and decides an issue never presented to the district court:  whether there are material differences between Kentucky's and Tennessee's lethal injection protocols.   As a court of appeals, we are obligated to provide the district court with the first opportunity to receive evidence and rule on this question.   Because I would remand this case for an evidentiary hearing in light of Baze, I respectfully dissent.

The majority recasts the district court's evidentiary findings in light of criteria that the court never considered, presuming findings under Baze that the district court never made.   It does so in a cursory manner, with minimal attention to the Baze plurality's fact-specific analysis, summarily concluding at each juncture that any deficiencies in Tennessee's execution protocol had already been considered but rejected in Baze.

This analysis is unsustainable inasmuch as it undercuts the factual findings of both the district court and the Baze plurality.   For example, the majority concludes that the failure to check for an inmate's consciousness under the Tennessee protocol is not problematic because the Baze Court “concluded that a visual inspection of the inmate by the warden was sufficient to protect the inmate's Eighth Amendment rights.”   Slip op. at 538 (citing Baze, 128 S.Ct. at 1536-37).   In Baze, however, the plurality reached that conclusion only after it credited the testimony of medical experts who stated that the signs of IV problems, including infiltration, would be “ ‘very obvious,’ even to the average person, because of the swelling that would result.”  128 S.Ct. at 1534.   The majority engages in no discussion of how the district court in this case made the opposite factual findings.1  This is but one example of how the majority's misguided attempts to recast the district court's inquiry fail.

It is not unforeseeable that a three-drug protocol that is, at first glance, similar to Kentucky's protocol, could fail to meet the standard set forth in Baze. That determination would turn in large part, not on the state's written protocol, but rather on the way the protocol is implemented.   As Justice Stevens explained in his concurring opinion in Baze, the “debate about lethal injection as a method of execution” remains open, and “[t]he question whether a similar three-drug protocol may be used in other States ․ may well be answered differently in a future case on the basis of a more complete record.”  Id. at 1542-43.

This Court has a “heightened responsibility ․ to insist, even at the risk of delay, on having the fact-finding process carried out properly at the level intended rather than to assume, even indirectly, a fact-finding role.”   Lewis v. Bloomsburg Mills, Inc., 773 F.2d 561, 577 (4th Cir.1985).   This admonition is particularly appropriate in the instant case, where the district court is uniquely equipped to conduct fact-finding to determine whether Tennessee's execution protocol is “substantially similar” to Kentucky's protocol.   See Baze, 128 S.Ct. at 1537.   Moreover, because of the extensive testimony that has already been heard in this case, the district court is well-positioned to consider the record before it, to supplement the record, and to apply the facts of the case to the new standard enunciated in Baze.

Only after the district court's fact-finding is complete can the district court make a determination as to whether Tennessee's protocol can be carried out in accordance with the requirements set forth in Baze. And only then, if an appeal is pursued, would this Court be in a position to evaluate the district court's judgment.   By circumventing this process, the majority oversteps its role, and its instructions vacating the district court's opinion and injunction are unwarranted.   Instead, this Court should remand for an evidentiary hearing in light of Baze and provide the district court with the first opportunity to rule on whether Harbison can meet the Baze standard.   I therefore respectfully dissent.

FOOTNOTES

1.   We also noted in Workman that the committee concluded that using only sodium thiopental would slow down the death process, the effect of the required dosage would be less predictable as the sole drug inducing death, and there was no data or case studies from other states on the efficacy of the one-drug method.  Workman, 486 F.3d at 902-03 (citing Report on Administration of Death Sentences in Tennessee 8 (2007)).

2.   The dissent would remand for an evidentiary hearing in light of Baze to allow the district court to rule on whether Harbison can meet the Baze standard.   While recognizing this court's authority to fashion a remedy not requested by the parties, we note that neither party requested a remand in light of Baze in briefs or at oral argument.

1.   In the instant case, while the warden is also in the execution chamber, Tennessee's protocol does not require the warden to observe the lines, the site, or the inmate, (Joint Appendix (“J.A.”) at 75-76), and the district court found that the warden had no training or experience that would allow him to do so effectively.  (J.A. at 288.)   Moreover, medical experts testified that under the Tennessee protocol, the IV is inserted in an area in which swelling associated with infiltration would not be apparent.

SILER, J., delivered the opinion of the court, in which COOK, J., joined.  CLAY, J. (pp. 539-41), delivered a separate dissenting opinion.