Saturday, 13 January 2007

Inadequate anaesthesia in lethal injection for execution

Inadequate anaesthesia in lethal injection for execution
The Lancet September 24, 2005; v. 366, n. 9491, p. 1073
When asked by editors of The Lancet to assess for publication “Inadequate anaesthesia in lethal injection for execution” by Leonidas Koniaris and colleagues (Apr 16, p 1412),1 I expressed strong support for the article because it contained the largest series of post-mortem lethal injection thiopental concentrations published to date. Furthermore, the finding that many inmates had low serum thiopental concentrations, and were possibly awake during execution, was new and vital information that could profoundly affect the public discourse on lethal injection.
However, after more research, I became concerned that the statement that “21 (43%) [inmates] had [thiopental] concentrations consistent with consciousness” may be erroneous because of a lack of equipoise in the study. In their zeal to “prove” that thiopental concentrations during execution were low, Koniaris and colleagues may have erred in their reporting of the crucial measurement of the elapsed time between the moment of death and the retrieval of blood samples, stating that the samples were collected the “same day or next day”. In fact, a graph provided to reviewers, but not included in the paper, suggests that most samples were obtained 12 or more hours after death. This graph clusters nine samples exactly 1 day after death, and 15 or more at about 0·5 days, suggesting that these times were rounded off. Most importantly, only two samples seem to have been obtained within a few hours of execution.
The elapsed time is critical because thiopental—a lipid-soluble and ultra-short-acting anaesthetic agent—redistributes into fat and muscle, even after death. In addition, a lethal injection is a unique clinical event, in that death occurs within a few minutes of injection of a large bolus of this drug, therefore a steady-state is not present. Under these circumstances, post-mortem serum concentrations are not reliable if a substantial amount of time has elapsed, because the high concentration of drug in the blood rapidly diffuses across a concentration gradient into the surrounding tissues after death. To state that “thiopental concentrations did not fall with increased time between execution and blood sample collection… consistent with data showing that thiopental is quite stable in stored human plasma” is erroneous since few samples were taken within the first few hours after death. Furthermore, there is a huge difference between the behaviour of thiopental in a corpse (where it diffuses out of the blood and into tissues in the body) and in a test tube of serum (where it has nowhere else to go). Other studies, not cited by Koniaris and colleagues, suggest that post-mortem serum thiopental concentrations in thiopental-caused deaths are lower in blood than in tissue2,3 and could be unreliable.4
Although Koniaris and colleagues' conclusion that lethal injection has “led to the unnecessary suffering of at least some of those executed” is probably true, it is not supported by the data presented. Clearly, public review of lethal injection is warranted for several reasons, but so is more careful scrutiny of how and when post-mortem blood samples are obtained.
I declare that I have no conflict of interest.


1. Koniaris LG, Zimmers TA, Lubarsky DA, Sheldon JP. Inadequate anaesthesia in lethal injection for execution. Lancet 2005; 365: 1412-1414. Abstract | Full Text | Full-Text PDF (66 KB) | CrossRef
2. Bruce AM, Oliver JS, Smith H. A suicide by thiopentone injection. Forensic Sci 1977; 9: 205-207. MEDLINE | CrossRef
3. Yasuda T, Yamaba T, Sawazaki K, et al. Postmortem concentrations of thiopental in tissues: a sudden death case. Int J Legal Med 1993; 105: 239-241. MEDLINE | CrossRef
4. Janik R, Becker J, Zollner W, Manz J. Thiopental concentrations in the whole blood of pregnant and non-pregnant patients in the first minutes following injection: forensic aspects. Anasthesiol Intensivmed Notfallmed Schmerzther 1991; 26: 132-136. MEDLINE

a. Children's Hospital, 700 Children's Drive, Columbus, OH 43205, USA
Dr. Groner – mail:

No comments: