Accept or refuse? Factors influencing the decision-making of transplant surgeons who are offered a pancreas: results of a qualitative study

In Germany, 163 pancreases from donors after brain death were transplanted in the year 2010 [11] (transplantation from donors after circulatory death is not allowed in Germany). These pancreas transplants were performed in 24 centers; the three hospitals with the highest volume transplanted 13, 14, and 16 pancreases per year, respectively. 14 of the centers transplanted more than five pancreases per year [11]. In all of these 14 relative ‘high volume’ centers, experienced transplant surgeons are on call to decide upon the acceptance or refusal of whole organ donor pancreases when offered by Eurotransplant. In most centers, 2–3 surgeons are specialized in pancreas transplantation.

Semi-structured, face-to-face interviews were conducted with surgical consultants from all 14 ‘high volume’ pancreas transplant centers between May and November 2011). The interview partners (n = 14) met the following inclusion criteria:

Purposive sampling was used to recruit the sample. Two transplant surgeons from transplant centers with less than five pancreas transplants per year were interviewed first in order to pre-test the interview guide; they provided comments on the questions (for the final interview guide see Table 1). These interviews were not included in the analysis. All participants agreed to being interviewed and recorded. The transcript, analysis and presentation of data did not enable identification of individual participants or their respective centers.

The interviews covered factors that influence the transplant surgeon’s decision to accept or refuse a pancreas. If donor-related medical factors such as age, BMI or length of ICU stay were named by the interview partner, it was asked whether certain limits would apply for these variables, and what these limits were. If non-medical criteria were not mentioned spontaneously, e.g. staff shortage, these aspects were asked explicitly. One of the authors (KPD) conducted all the interviews, each of which was audio-taped and lasted for 20–70 minutes.

Interview tapes were transcribed verbatim. The transcripts were de-identified and continuously numbered (Interview Partner = IP 01-IP 14), so none of the researchers, except the interviewer, could link the answers back to the interviewed transplant surgeon or the respective transplant center. The transcripts were examined using thematic content analysis [12]. Key themes that spanned specific questions and topics were identified, and were used to organize the material [13]. In the course of the data analysis, the initial categories, distinguishing between medical and non-medical factors influencing the decision, evolved into more sophisticated coding structures and additional categories, including e.g. concepts of trust, patient’s prognosis, and strategic aspects. To enhance the validity of the findings, the interview transcripts were read and coded independently by two authors (KPD and JL); deviant cases and contradictory data were analyzed with particular attention [14].

No authorized ethical approval was needed as per the ethics committee of the University of Regensburg (reference number 11-160-0192).

From the interview results we could gain insight into the decision-making process of transplant surgeons who are offered a pancreas: The main factors that influence the decision (e.g. pancreas macroscopy, donor medical history) are relatively consistent throughout the interviews. However, when analyzed in depth, there are extreme differences between the surgeons’ assessment and general handling of pancreas offers. Firstly, while nearly all of interviewed surgeons list the same medical donor factors as being relevant for pancreas selection (above all donor age, time of ICU stay, and lab results), the assessment of these factors varies substantially between the surgeons; no standardized or consistent cut offs exist. For example, some surgeons told the interviewer that they had no doubts about accepting the pancreas of a 55-year old if the other parameters were favorable, whereas others claimed not to accept the pancreas of a donor over 40.

Secondly, the macroscopic appearance of the donor organ and the technical quality of the pancreas recovery are the predominant factors which influence the decision. In Germany, the transplanting surgeons usually do not perform the pancreas recovery as well, so the macroscopy is described to them on the telephone. The transplant surgeons differed significantly as to whether or not they relied on the donor surgeon’s assessment of the macroscopic pancreas quality. Some prefer to have the organ shipped to their center even if it has been described as having poor quality. Others trust the donor surgeon’s judgment – some surgeons do so in every case and others only if they know and have confidence in the respective donor surgeon.

Non-medical factors play a minor role, but certainly exist; above all, capacity problems can lead to the refusal of pancreases in some centers. Interestingly, the sample was split on the legitimacy of occasional staff shortages.

It became clear in the interviews that the relatively benign prognosis of the patients on the pancreas waiting list may induce surgeons to wait for excellent (‘flawless’) organs. This restrictive policy is in contrast to some surgeons’ preference to accept pancreases in a very permissive way – even if turned down frequently before and having been described as macroscopically poor – so that they can inspect and evaluate the organ personally.

Although the phenomenon of pancreas under-utilization is well known in European and North American transplantation networks and has been described as a ‘major issue in pancreas transplantation’ [15], the analysis of the underlying reasons has been only superficial thus far. The strength of this study lies in the investigation of one of the pivotal steps in the allocation process: the transplant surgeon’s decision to use or not to use an offered pancreas. The main limitation of the study is that the interview partners were recruited by purposive sampling, and were limited to a number of 14 interview partners. Consequently, it cannot be assumed that the findings presented here are representative of the views of all pancreas transplant surgeons in Germany. However, the aim of qualitative studies is not to receive representative data, but to gain a deeper understanding of social and psychological processes. Using a qualitative study design allowed us to tap personal attitudes and experiences that are not readily expressed in response to survey questions [16], and that are key to understand the complex processes and conditions involved in making the decision. While the number of interviewed transplantation surgeons was relatively small, the selected interview partners represented all German transplantation centers that perform five or more pancreas transplantations per year, thus constituting a very balanced and almost representative sample. The use of the same interviewer for all participants, and two independent researchers for data analysis ensured quality control and minimal interpretive bias.

In order to explain the phenomenon of pancreas under-utilization, few quantitative studies analyzed donor characteristics of transplanted and/or discarded pancreases. Wullstein et al. compared donor profiles of accepted pancreas grafts versus grafts declined due to ‘medical reasons’. They found significant differences between both groups for cause of death, age, BMI, serum Lipase, alcohol abuse and history of smoking. They concluded that these aspects might be the most important reasons to refuse a pancreas [17]. These medical characteristics are largely overlapping with those named by the interviewed surgeons, although Wullstein et al. had not included the length of ICU stay in their analysis, which was a predominant aspect in the interviews. However, our analysis shows that the reasons to accept or refuse an offered pancreas cannot be pinpointed to single donor characteristics, but are far more complex. Wiseman et al. [18] analyzed multiple characteristics of donors whose pancreas was used, and compared them to donors whose pancreas was not used. This study found that a significant number of potentially suitable donor pancreases were not used, although no medical characteristics that precluded transplantation were identified. In the article’s discussion, Wiseman et al. speculated that ‘transplant centers … may be reluctant to accept pancreases that are not assessed by members of their own team or by colleagues who have demonstrated experience in pancreas transplantation.’ This notion by Wiseman, which was not backed by any other data yet, could clearly be confirmed by our study. Wiseman et al. also hypothesized that due to the relatively short waiting time, ‘the tendency for pancreas programs to decline use of a pancreas …with the hopes that an even more “optimal” donor could be forthcoming in the near future may limit pancreas utilization’. The finding of the interviews that some hospitals or surgeons display a stringent acceptance policy is consistent with this supposition.

The interviews shed light on some aspects that help explain the under-utilization of donor pancreases.

It became clear that transplant surgeons who decide to accept or refuse a pancreas offer always act on conflicting priorities: on one hand, the expectations of patients and hospital administration not to turn down a rare and precious donor organ; and on the other hand, the fear of endangering the patient’s health by accepting an organ that is not flawless. The latter is especially important in pancreas transplantation because unlike liver or heart transplantation, the patient’s condition is usually not life threatening, so there is less willingness to compromise. If many centers dictate a cautious acceptance policy, organs might be refused repeatedly. Consequently a ‘cascade effect’ [19] can ensue, because the refusal of one center might - consciously or unconsciously - increase the probability of further refusals, as the interview analysis suggests. This is especially critical if the allocation process is still ongoing when the organ has already been recovered. As a consequence, the extended ischaemic time may result in an increase of discarded organs; it can also lead to unequal access to donated pancreases [20]. Conversely, a very permissive acceptance policy that some interview partners displayed might also lower the pancreas utilization rate, because there is the risk involved that the organ cannot be placed anymore if its macroscopy is not considered favorable when the transplant surgeon inspects the organ personally in his center.

The interview results suggested that the assessment of medical donor characteristics is subjective, inconsistent and hardly standardized. This observation can also be explained by the fact that the interviewed surgeons base their cut-offs on varying evidence or customs. A more standardized approach in terms of cut-offs seems to be difficult, however, because one parameter needs to be assessed in conjunction with other risk factors. One has to discuss whether the allocation process would profit from a better standardization or an evidence-based approach. Particularly younger surgeons who have less experience might benefit from recommendations or guidelines, which can be developed by experts. However, one needs to consider that there is still incomplete evidence on whether or not a pancreas is suitable for transplantation, as recently pointed out by Magione et al. [21].

Within the relatively small community of surgeons who are experienced in pancreas transplantation, donor surgery and pancreas surgery, the aspect of trust seems to be an influential factor in the decision-making process: confidence in the donor surgeon and his or her capacity to a) assess the pancreas quality and b) to recover a pancreas without damaging it, and maybe even trust in the transplant center that had turned down the organ offer prior to one’s own decision. Better and more standardized training or perhaps more rigorous selection of donor surgeons could improve the expertise and thus boost confidence.