- Journal List
- Ann Surg
- v.243(1); 2006 Jan
- PMC1449959
Fatal and Nonfatal Hemorrhagic Complications of Living Kidney Donation
Abstract
Objective:
After anecdotal reports of severe hemorrhage from failure of surgical clips to sustain closure of renal artery stumps in live donor nephrectomies were received, this study was designed to identify specific surgical techniques that are associated with an increased risk of failure to control bleeding and might represent opportunities to improve patient safety.
Background:
Preventing complications for living kidney donors must be paramount in addressing end-stage renal failure through living kidney donation. Major hemorrhage from technical failure, albeit an infrequent occurrence, can cause significant, yet preventable, morbidity or death. Open and laparoscopic approaches to living kidney donation use several vascular control methods, some of which may be more prone to failure and life-endangering hemorrhage than others.
Methods:
To define hemorrhagic complications of living kidney donation, a survey was sent to all 893 surgeon-members of the American Society of Transplant Surgeons. Descriptive and bivariate analyses were used to ascertain study participant characteristics, most frequently used vascular control techniques, and incidence of events (death, transfusion, reexploration or conversion to open nephrectomy, or contralateral [remaining kidney] renal failure). Outcomes of hemorrhage and comments by respondents were sought as were data from other sources.
Results:
In 213 surveys returned (24%), 66 and 39 episodes of arterial and venous hemorrhage were reported, respectively. Among arterial control problems, 2 resulted in donor death and 2 resulted in renal failure; 19 episodes required transfusion. Open conversions in laparoscopic nephrectomy or late reoperations for hemorrhage were reported for 29 cases. Locking and standard clips applied to the renal artery were associated with the greatest risks.
Conclusions:
Significant hemorrhagic complications occur with living kidney donation in both open and laparoscopic approaches. Loss of arterial control jeopardizes donor life and health, especially when it occurs in the postoperative period. Vascular transfixion provides the best vascular control of major vessels.
Living kidney donation in the United States has grown from 2851 cases in 1993 to 6474 cases by 20031 due to prolonged waiting time for deceased donor kidneys,2 availability of minimally invasive donor nephrectomy procedures,3–6 and significantly better recipient outcomes with live donor renal transplantation.7–10 While living kidney donation appears to be relatively low risk for the donor, anecdotal and published reports of death attributed to living kidney donation are extant. In a survey of 234 renal transplant programs, Matas et al found 2 donor deaths and 1 case of persistent vegetative state, all from hemorrhagic shock, in 10,828 cases.11 In the largest single center experience, the University of Minnesota reported on 2540 living donors with complication rates of 0.2% (major) and 8% (minor).10 Other surgical complications of living kidney donation are documented, but with little consistency in reporting.12–15 Thankfully, few centers have sufficient experience with donor death to report on mortality causes or trends.
In 2003, 3 of the authors (A.L.F., T.G.P., L.E.R.) were made aware of 3 previously unreported severe hemorrhagic events, 2 of which resulted in donor death; all were related to failure of surgical clips applied to the donor renal artery. As we reviewed these cases, concerns arose that such problems might not be sporadic, yet no published report could be found dealing specifically with hemorrhagic complications of living kidney donation. We felt, therefore, that it was imperative to investigate a potentially preventable serious risk to living kidney donors, and we elected to survey members of the American Society of Transplant Surgeons to further define the risks of hemorrhage in living kidney donation.
METHODS
The American Society of Transplant Surgeons (ASTS) is the only organization in America principally devoted to issues that affect transplant surgeons, their institutions, and the patients whom they serve. The majority of members are board-certified surgeons with additional training or experience in transplantation and a record of publication in transplantation science. The membership has been surveyed on a number of occasions regarding professional standards, regulatory matters, and issues germane to the Society and to organ transplantation in the United States.
Study Population and Data Collection
An ASTS-approved, 33-question survey instrument to assess transplant surgeon knowledge and experience with specific techniques of vascular control in the conduct of live donor nephrectomy was mailed to all ASTS members in October 2003. Questions posed included details about member training in open or laparoscopic donor nephrectomy, as well as the techniques currently used to control the renal artery and vein during donor nephrectomy. The member was asked about failure of any technique to prevent severe hemorrhage in the surgeon's own case, direct knowledge of vascular control failure with the member as recipient surgeon and a non-ASTS member (who would not have received the survey) procuring the kidney, and primary knowledge of other vascular control failures in living donors. Also, detailed outcomes of the above were sought. Data from the 3 index cases about which the authors were contacted were included. Response options served to eliminate duplication. Participants were asked to rate the safety of 11 techniques of major blood vessel control in open and laparoscopic cases on a 5-point Likert scale (ranging from 1 “not safe” to 5 “extremely safe”). Finally, permission for follow-up contact was asked unless anonymity was desired.
Clavien Classification of Surgical Complications
Hemorrhagic events were classified using the method of Dindo et al,16 which relies on treatment needed to address the complication. Grade I represents any deviation from the normal postoperative course without the need for pharmacologic or invasive intervention. Grade II requires pharmacologic or blood transfusion therapy. Grade III requires invasive intervention either a) not under general anesthesia, or b) under general anesthesia. Grade IV is a life-threatening complication involving a) single organ dysfunction (including dialysis) or b) multiorgan dysfunction. Grade V is death.
Statistical Analysis
We used descriptive analyses to ascertain characteristics and experience of study participants, their views regarding safety of vascular control techniques (open and laparoscopic), and reports of hemorrhage associated with specific control techniques. Reported events were assessed as to the frequency of arterial hemorrhage and associated rates of transfusion, reexploration, or conversion of laparoscopic to open procedures, or death. In bivariate (χ2) and ANOVA analyses, we determined if hemorrhage rates were associated with the type of control technique (ie, transfixion versus nontransfixion), and whether the interval (time) to hemorrhage was associated with technique. Analyses were performed using STATA (release 8.0; StataCorp, College Station, TX) statistical software.
Other Sources
The FDA was contacted to determine safety concerns and standards regarding surgical clips and stapling devices and their approval for renal artery control. We reviewed recommendations of company representatives regarding instruments used in both laparoscopic and open procedures for control of major blood vessels. Information in package inserts was documented to determine the published indications and contraindications for various surgical appliances used to control major blood vessels.
Additionally, information was sought from the National Practitioner Data Bank, the Defense Research Institute (an international organization concerned with the defense of civil actions), and a number of insurance companies as to any specific data regarding morbidity and mortality in living kidney donation. The FDA Web site was searched using the words “surgical clip,” and the information returned was reviewed for any cases mentioning nephrectomy or living donor surgery. There was no information specific to living donor nephrectomy from these sources.
RESULTS
Of 893 surveys, 213 (24%) were returned. The 118 (55%) surgeons trained in performing both open and laparoscopic nephrectomy outnumbered the 76 (36%) respondents trained in only open technique, and a single surgeon was trained only in laparoscopic donor nephrectomy. Only 4 (2%) reported training in neither technique, and 14 (7%) did not respond regarding donor nephrectomy training. Respondents had a mean experience of 11.5 years (range, 0.5–33 years) as an attending kidney transplant surgeon, and had performed a mean of 12.5 (range, 0–80) open and 26 (range, 0–100) laparoscopic donor nephrectomies in the prior 12 months. Specific techniques used for renal artery control in open and laparoscopic nephrectomy are detailed in Table 1. In open cases, the most common techniques were a simple tie plus suture ligature (85 surgeons, 40%) and oversewing (52 surgeons, 24%), while in laparoscopic cases it was the GIA stapler (64 surgeons, 30%) and multiple locking clips used by 39 surgeons (18%).
There were 66 and 39 episodes of arterial and venous hemorrhage, respectively. The specific failed techniques associated with the 66 episodes of arterial hemorrhage are detailed in Table 2. Passage of suture material or metal staple through the vascular tissue was defined as transfixion, while nontransfixion was defined as suture, metal, or polymer material placed solely around the vascular tissue. Considered collectively, there were 45 (68%) reports of arterial hemorrhage linked to nontransfixion techniques and 21 (32%) reports in cases of transfixion techniques.
Two deaths occurred from arterial hemorrhage related to failure of multiple nonlocking clips after completion of the operation (Table 3). Two cases of contralateral renal failure, one of which also involved cardiac arrest, occurred; both were associated with failure of multiple nonlocking clips applied to the renal artery. In 29 cases (44%), the respondent was able to specifically recall whether or not blood was transfused; in 19 patients, transfusion was undertaken. It was known in 55 (83%) of the 66 cases of arterial hemorrhage whether reexploration occurred after closure of the surgical incision, or conversion of a laparoscopic technique to open procedure had been required. In 29 of these cases, reexploration or conversion was required. While surgically troublesome, venous hemorrhage was not associated with life or organ endangerment.
Opinions about the safety of specific techniques for arterial closure in open and laparoscopic cases are summarized in Table 4. The majority of respondents felt that all techniques of arterial closure involving tissue transfixion aresafe. Nonlocking clips, whether used singly or multiply, were not deemed safe by the majority for both open and laparoscopic nephrectomy. Single locking clips were considered “not safe” in open nephrectomy by 58% of respondents and by 61% regarding laparoscopic cases. In contrast, multiple locking clips were considered to be “very safe” by 40% of respondents for open cases and by 52% for laparoscopic cases.
Table 5 depicts the modified Clavien classification of events stratified by type of hemorrhage, closure technique, and use of clips. Arterial events were associated with more severe outcomes than venous events (P = 0.05). When all hemorrhagic events were included, nontransfixion techniques were associated with more severe outcomes compared with transfixion techniques (P = 0.01). When compared with all other techniques, clips were statistically significantly associated with worse complications of all hemorrhagic events, with nonlocking clips most likely to be associated with class 4b and 5 complications (P = 0.001).
Of great interest is the package insert for the Auto Suture* Endo Clip* disposable clip applier (United States Surgical Corporation).17 This clip appliance, used by certain respondents in our survey, is accompanied by an FDA-approved package insert which states, “Do not use the ENDO CLIP* disposable clip appliers on the renal artery, iliac artery, or other vessels upon which metal ligating clips would not normally be used.”
DISCUSSION
Nephrectomy for living kidney donation is one of the few major extirpative procedures in which a healthy person has a major organ removed for the benefit of someone else. There is a general sense in the professional transplant community that this operation is quite safe.10,11,15,18–20 However, even with fully evaluated, appropriately selected, and entirely healthy persons, major surgery with transection of named vessels may be attended by life-threatening technical mishap.3,6,11–13 With both open and laparoscopic techniques, various surgical appliances may be chosen to control major vessels, and the study goal was to determine how vascular control techniques and related complications might be prevented.
The survey instrument, as with any such tool, has substantial shortcomings. We cannot know how often any specific technique is applied to renal vascular control in living donor operative procedures, open or laparoscopic. Furthermore, a surgeon may use several techniques depending upon the circumstances of anatomy as it unfolds in dissection. The best that one can hope for in a survey as reported herein is some sense about how surgeons recall their experiences, and relate their technical experiences to patient safety. The 213 responding surgeons provided a rich compendium of information including experiences of some non-ASTS donor team members. The high rate of transfusion and need for reexploration or open conversion substantiates that reported hemorrhagic events were, indeed, severe. So, despite the survey methodology, we maintain that it has rendered specific information about a series of low-frequency, yet high-impact, events.
An issue raised by several respondents was the shearing of vessels, principally the renal artery, when localized arterial disease or a fairly thin arterial wall predisposed to a tear at or near the aorta regardless of vascular control method, although shearing seemed to occur more often from the application of clips. In addition, the lack of postoperative pain control was an important cause of delayed hemorrhage in at least 1 case. At the time of nephrectomy, the intraoperative donor blood pressure was approximately 110/60; the patient had severe postoperative pain, the blood pressure went to 180/120, and hemorrhage occurred. At reexploration, a silk ligature was found to be intact but, obviously, not on the arterial stump.
While the FDA Web site did not specifically indicate problems with donor nephrectomy, the manifold appliance failures (from use of clips in open surgery to specific laparoscopic instrument failure), indicate that virtually all appliances from all manufactures may have failures such as misfire, incomplete closure, missing parts, and other problems.21,22 Certainly, the development of the locking clip, thought to be an advance regarding secure use on arterial structures, must mean to some persons that the standard nonlocking clips had some undesirable features. In both the FDA material and our own survey, however, both standard ligaclips and locking clips failed a number of times, particularly when applied to a major artery.
While the package insert for certain devices admonishes use on major vessels, it may be that surgeons sometimes should not adhere strictly to instruction from the device manufacturer. Reasons for this include the unanticipated circumstances in operative procedures that often occur, the situation of advancing new uses to appliances designed for specific but not broad tasks, and the natural inventiveness that many thoughtful surgeons apply to the benefit of their patients each day. Nonetheless, at least one particular device was not to be used on the renal artery, and there were at least 13 failures of this or a similar device from our survey results, with 12 occurring in the delayed intraoperative or postoperative time periods. Accordingly, we think that the package inserts are correct when asserting that clips should not be applied for primary control of the renal artery. Locking clips do not offer a safety advantage; they were associated with arterial hemorrhage in 8 cases in the delayed intraoperative or postoperative time periods.
Basic surgical principles are fundamental to the safe conduct of operative procedures. While the survey information cannot be considered comprehensive, we think it is apparent that the mishaps reported by our own colleagues occur more often in arterial control with nontransfixion versus transfixion techniques. Experienced surgeons clearly understand the difficulties of arterial hemorrhage, and most have been taught at one time or another that the use of transfixion control of some sort is more secure than simple ligature or other sort of external nontransfixion method. While not necessarily true for small vessels or those with relatively low pressure, there is little doubt that ligature and transfixion together add safety and are thought to do so by surgeons responding to our survey. Clearly, the stapling devices, which transfix vessels, serve the same purpose as suture transfixion.
While live donor nephrectomy remains a safe operation, there is still the ever-present potential for serious risk. The unfortunate deaths and significant morbidity of volunteer living donors, giving a perfectly good organ for the benefit of someone else, must compel transplant professionals to find every possible way to prevent such outcomes. To that end, the survey results from members of the ASTS indicate that: 1) surgical mishap and life-threatening hemorrhage appear to be associated more frequently with surgical clips than with other methods of arterial and venous control; 2) the perception that suture or staple transfixion of the renal artery is the safest and most appropriate way to manage the living kidney donor renal vasculature is agreed upon by the majority; and 3) postoperative pain control to prevent severe hypertension in the immediate postoperative period is an important adjunct to control delayed hemorrhage. Applying such principles to the living kidney donor may reduce the already low risk to life-endangering hemorrhage for these patients who offer a living gift to their recipients.
Footnotes
Reprints: Amy L. Friedman, MD, Department of Surgery, Yale University School of Medicine, FMB 112, 333 Cedar Street, New Haven, CT 06520. E-mail: ude.elay@namdeirf.yma.
Facebook
Twitter
Google+