- Email: [email protected]
Might physicians be restricting access to liver transplantation for patients with alcoholic liver disease?
Journal of Hepatology 51 (2009) 707–714 www.elsevier.com/locate/jhep
Might physicians be restricting access to liver transplantation for patients with alcoholic liver disease?q Vale´rie Perut1, Filome´na Conti2, Olivier Scatton2, Olivier Soubrane2, Yvon Calmus2, Gwenae¨lle Vidal-Trecan1,3,* 1
Unite´ de Gestion des Risques et Qualite´, Groupe hospitalier Cochin-Saint Vincent de Paul, AP-HP, Paris, France 2 Unite´ de Transplantation He´patique, Faculte´ de Me´decine, Universite´ Paris Descartes, Groupe hospitalier Cochin-Saint Vincent de Paul, AP-HP, Paris, France 3 Departement de Sante´ Publique, Faculte´ de Me´decine, Universite´ Paris Descartes, 24, rue du Faubourg Saint Jacques, 75014 Paris, France
Background/Aims: In France, the most common cause of cirrhosis is excessive alcohol consumption. Post-transplant survival rates in patients with alcoholic liver disease (ALD) are at least as good as those seen with other indications. However, fewer of these patients are found on the waiting list. To understand the reasons for this discrepancy, it was decided to examine physicians’ attitudes concerning the allocation of deceased donor liver allografts. Methods: Using a standardized postal questionnaire, 1739 physicians were asked to allocate 100 liver transplants to two competing groups of patients who were equivalent except for the cause of their cirrhosis (i.e. alcohol-related or primary biliary cirrhosis). A composite score was then used to assess their attitude regarding the behavior of alcoholics and their responsibility for their illness. Results: Among the 475 respondents (response rate: 27.3%), 55.2% allocated fewer than 50 transplants to ALD patients. This lower rate was independently associated with factors such as being a general practitioner (odds ratio [OR] = 3.2, 95% confidence interval [95%CI] = 1.8–5.9), a misinterpretation of ALD patients being equivalent to others (OR = 1.8, 95%CI = 1.1–3.0) or unfavorable attitudes regarding alcoholics (OR = 4.0, 95%CI = 1.7–9.5, to OR = 126.8, 95%CI = 34.0–472.1). Conclusions: Greater information and education of physicians may improve access to liver transplantation for ALD patients. Ó 2009 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. Keywords: Liver transplantation; Liver diseases; Alcoholic; Access to health care
1. Introduction
Received 19 September 2008; received in revised form 16 March 2009; accepted 29 April 2009; available online 28 May 2009 Associate Editor: P.-A. Clavien q The authors who have taken part in this study declared that they do not have anything to disclose regarding funding or conflict of interest with respect to this manuscript. * Corresponding author. Tel.: +33 1 58413146; fax: +33 1 58412678. E-mail address: [email protected] (G. VidalTrecan). Abbreviations: ALD, alcoholic liver disease; LT, liver transplantation; UK, United Kingdom; GP, general practitioner; US, United States; OR, odds ratio.
Alcohol consumption in France is the leading cause of morbidity and mortality related to liver disorders [1]. Liver transplantation (LT) is currently the only treatment available for end-stage alcoholic liver disease (ALD). The post-transplant survival of ALD patients, with or without a relapse of alcoholism, is at least as good as in patients with other liver diseases [2]. Nevertheless, in view of the frequency of ALD in France, the proportion of these patients on the LT waiting list could be higher (ALD patients accounted for 72.6% of patients aged 25–75 years registered for liver-related deaths in 2005 [1], whereas they only represented 24.0% of new registrations on the LT
0168-8278/$36.00 Ó 2009 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.jhep.2009.04.018
708
V. Perut et al. / Journal of Hepatology 51 (2009) 707–714
waiting list in 2007 [3]). We share the opinion of other authors [4] who have considered that registration on the waiting list should be solely dependent on recognized criteria reflecting medical need (such as the MELD score) [5] regarding this scarce, expensive, rationed and yet life-saving resource. Two studies [6,7] performed in the USA found that ALD was associated with a lower rate of LT than other causes of liver disease. Other studies [8–11] performed in English-speaking countries have described a public reluctance to finance LT for ALD patients. According to Ubel et al. [12], the decision to allocate fewer organs to ALD patients reflects a belief that they are ‘‘responsible for their own illness” or ‘‘less worthy of transplants than others”. To our knowledge, the attitudes of physicians regarding the access of ALD patients to LT has only been investigated by one study performed in the United Kingdom (UK) [10], during which a sample of gastroenterologists and general practitioners (GPs) was asked to rank their priorities for the allocation of liver transplants to patients with controversial criteria. They assigned the lowest priorities to ALD patients and prisoners, as did the general public [10]. In France, any physician can refer patients with endstage liver disease to a hospital specialist who is authorized to register patients on the national transplant waiting list. We hypothesized that, consciously or not, ‘‘first-line” physicians might be restricting access to LT for ALD patients (e.g. by not referring patients to the authorized team at all, or by referring them when they have reached a critical stage at which transplantation is contraindicated). Using a postal questionnaire, the aim of this study was thus to examine whether physicians managing liver disease patients might be restricting the access of ALD patients to LT on the basis of criteria unrelated to medical need, but rather reflecting their perception of the ‘‘self-induced” nature of alcoholic disease, associated with a potential risk of return to a pattern of alcoholic consumption and noncompliance. 2. Materials and methods 2.1. Study design A cross-sectional study was performed from October 2005 to March 2006 on a sample of physicians, using a standardized postal questionnaire. To explore their attitudes regarding the access of ALD patients to LT, they were asked to allocate, under a hypothetical scenario, a limited number of liver transplants to two competing groups of patients who were equivalent in every characteristic except for the cause of their liver disease, i.e. alcohol-related or primary biliary cirrhosis.
2.2. Setting This study involved physicians working in public hospitals, i.e. University Hospitals within the Paris Public Hospital Network (Assistance Publique-Hoˆpitaux de Paris, AP-HP) (38 hospitals) and non-teaching hospitals (30 hospitals), as well as physicians caring for outpatients in the Paris region.
To be registered on the LT waiting list, patients must be referred by a physician (hepatologist, gastroenterologist or other specialists, seldom general practitioners) to a hospital physician or surgeon authorized by the French agency responsible for organ procurement and transplantation (Agence de Biome´decine). After clinical examination, an analysis of the mandatory investigations required to determine suitability for liver transplant listing, and calculation of a liver score specific to France (i.e. the ‘‘ScoreFoie” [13], derived from the Model for End-stage Liver Disease (MELD score) [5]), a multidisciplinary team (liver surgeons, hepatologists, anesthetists, psychologists) will decide whether to register the patient or not. The MELD score is the principal parameter taken into account when allocating liver grafts in France, but the ‘‘ScoreFoie” used to prioritize the patients in each blood group also includes parameters related to hepatocellular carcinoma, the presence of liver diseases other than cirrhosis, and the distance between the donor center and recipient center. A full six months without any alcohol consumption is usually required for ALD patients to be registered on the waiting list and then undergo LT [14].
2.3. Study population The questionnaire was circulated to 1753 physicians. Fourteen forms were returned as undeliverable, so that 1739 physicians were included in the study. The AP-HP physicians database, the Health Ministry database for medical staff in non-teaching hospitals in the Paris region, and the France Telecom phone book were used to identify all hospital and private hepatologists–gastroenterologists and also to select a random sample of GPs from the 6059 listed. The questionnaire was sent to all AP-HP physicians working in gastroenterology or hepatology departments (n = 343), including LT physicians (n = 32), senior hepatologists–gastroenterologists or internists in non-teaching hospitals (n = 91), private hepatologists–gastroenterologists (n = 214) and the sample of GPs (n = 992). Occupational physicians (n = 99) attending a continuing training meeting were also given the questionnaire.
2.4. Data collection The questionnaire, together with an introductory letter and a postage-paid reply envelope, was sent out in October 2005, followed by a reminder questionnaire in March 2006. The questionnaire was circulated to the occupational physicians during their meeting, and then collected. No direct or indirect nominative information was collected, as required by the CNIL (Commission Nationale de l’Informatique et des Liberte´s).
2.5. Questionnaire The questionnaire was adapted for use by a population of French physicians from that used in a survey on a general population sample in the USA [11]. The fictitious patients were equivalent in terms of all characteristics (i.e. compatibility for transplant, survival time with and without liver transplant, duration of illness and period on the waiting list, treatment compliance, effectiveness of therapy, cost, age, health status) except for the cause of the condition, i.e. ALD or autoimmune disease. For the sake of clarity, no patients without psychosocial contraindications (e.g. ongoing substance use, severe personality disorder) were included. Three questions were asked regarding the allocation of deceased donor organs for LT. The first concerned the allocation of a liver transplant to one of two patients: one with ALD, the other with primary biliary cirrhosis. The second focused on the allocation of a 100 liver transplants among patients with the two different causes for their cirrhosis. Thirdly, after putting forward a series of medically unfounded arguments, the second allocation question was repeated to test the influence of these arguments on the allocations made by the physicians. A set of 10 questions developing these arguments (e.g. preventable, controllable, voluntary or immoral nature of alcoholism) was then asked in order to evaluate the physicians’ attitudes concerning alcoholics. The physicians were also asked to provide information on their social and demographic characteristics, the
V. Perut et al. / Journal of Hepatology 51 (2009) 707–714 consumption of psychotropic agents by them or their relatives, and their understanding of the scenario. The questionnaire had previously been tested in a sample population of 21 physicians.
2.6. Definition of variables The allocation of 100 liver transplants to ALD patients was classified as ‘‘0–49 transplants” or ‘‘50 transplants or more”. The professional status of physicians was divided into salaried employment only, private practice only, or a combination of salaried employment and private practice. If the physician’s response indicated that he or she was neither a salaried employee nor a private practitioner, this was deemed to be missing data. Workplaces (i.e. where the letter was sent) were classified as hospital, occupational physicians, private hepatologists–gastroenterologists and GPs. The medical specialty was classified as liver transplant physicians, hepatologist–gastroenterologists, GPs and occupational physicians. The consumption of sleeping ts and tranquilizers was defined as at least one tablet in 30 days, tobacco consumption as at least one cigarette a day, alcohol consumption as at least three glasses of alcohol per week and cannabis consumption as at least 10 times within the past year. Responses to questions regarding the consumption of psychotropic agents by physicians or their relatives were combined in two composite scores. The first was categorized as at least one personal consumption (i.e. positive) or not, and the second as at least one consumption by the physician or a relative (i.e. positive) or not. The attitudes of physicians were evaluated using a 5-point, agree– disagree, Likert-type dichotomized scale, with middle scores being interpreted as disagreement [11]. Responses were combined to obtain a composite score of unfavorable attitudes regarding alcoholics [11]. The equivalence of two patients was evaluated using one question containing two items: ‘‘equivalent (yes or no)” or ‘‘different (yes or no)”. If a physician answered ‘‘yes” to both ‘‘equivalent” and ‘‘different”, the answer was considered as missing. Their interpretation of the scenarios was assessed from their understanding that the two groups of patients only differed in terms of the cause of their liver disease and their final outcome (i.e. death of patients not receiving liver transplant).
709
2.7. Statistical analysis Data were analyzed using standard descriptive statistics (mean, standard deviation (SD), and frequency). A correlation between the mean of liver transplants allocated to ALD patients, before and after presentation of the opinion arguments, was evaluated using the Spearman test. The concordance of the allocation of ‘‘0–50 transplants” or ‘‘50 transplants or more” to ALD patients, before and after presentation of the opinion arguments, was evaluated using the Kappa test. Bivariate analyses were performed using the v2-test or Fisher’s exact test. A logistic regression model was used to identify factors independently associated with the allocation of fewer than 50 transplants to ALD patients. Backward stepwise procedures were applied so that the final model only included factors providing a significant explanation of outcomes. Associations were expressed as odds ratios (ORs) with their 95% confidence intervals (95% CIs). Statistical significance was set at 0.05. SPSS Software for WindowsTM version 14.0 was used for statistical analysis.
3. Results 3.1. Characteristics of responding physicians and their attitudes concerning alcoholics Of the 1739 physicians included, 526 returned the questionnaire (response rate: 30.2%). Of these, 34 did not answer the main question (i.e. allocation of liver transplant) and the workplace of 17 was unknown. These 51 questionnaires were considered as invalid and were thus excluded, so that the answers from 475 physicians (27.3%) were analyzed. The distribution of workplaces and response rates is shown in Fig. 1. Among the physicians included in the study, 61.7% were men
1753 questionnaires circulated
14 questionnaires undeliverable 1739 physicians included
526 questionnaires returned response rate=30.2%
475 valid responses valid response rate=27.3%
1206 private physicians included 282 valid responses valid response rate=23.4%
533 salaried physicians included 193 valid responses valid response rate=36.2%
343 AP-HP 109 valid reponses valid response rate=31.8%
32 liver transplant physicians 17 valid responses valid response rate=53.1%
91 non-teaching hospitals 13 valid responses valid response rate=14.3%
99 occupational physicians 71 valid responses valid response rate=71.7%
214 private gastroenterologists 57 valid responses valid response rate=26.6%
311 other physicians of AP-HP 92 valid responses valid response rate=29.6%
Fig. 1. Origin of responses as a function of the workplace of physicians.
992 private general practitioners 225 valid responses valid response rate=22.7%
710
V. Perut et al. / Journal of Hepatology 51 (2009) 707–714
Table 1 Comparison of the characteristics of responding physicians as a function of the allocation of liver transplants to alcoholic liver disease patients. Variables
n/N
%
0–49 Transplants
%
50 or more transplants
%
Male gender Age 25–34 35–44 45–54 55 and + Professional status Salaried employee only Private practice only Salaried employee and private practice Work location Hospitala Occupational physician Private general practitioner Private hepatologists–gastroenterologists Medical specialty Liver transplant physician Hepatologist–gastroenterologist General practitioner Occupational physician Personal consumption Sleeping tablets and tranquilizers Tobacco Alcohol Cannabis At least one psychotropic agent Alcohol consumption by relatives Personal consumption or by a relative No misinterpretation of the scenario
293/475
61.7
171/262
65.3
122/213
57.3
34/474 92/474 219/474 129/474
7.2 19.4 46.2 27.2
11/261 57/261 117/261 76/261
4.2 21.8 44.8 29.1
23/213 35/213 102/213 53/213
10.8 16.4 47.9 24.9
169/460 215/460 76/460
36.7 46.7 16.5
63/251 135/251 53/251
25.1 53.8 21.1
106/209 80/209 23/209
50.7 38.3 11.0
122/475 71/475 225/475 57/475
25.7 14.9 47.4 12.0
50/262 32/262 152/262 28/262
19.1 12.2 58.0 10.7
72/213 39/213 73/213 29/213
33.8 18.3 34.3 13.6
17/475 156/475 231/475 71/475
3.6 32.8 48.6 14.9
8/262 69/262 153/262 32/262
3.1 26.3 58.4 12.2
9/213 87/213 78/213 39/213
4.2 40.8 36.6 18.3
60/469 67/467 202/472 2/470 245/470 160/468 316/473 256/475
12.8 14.3 42.8 0.4 52.1 34.2 66.8 53.9
36/260 26/259 110/260 2/259 130/259 95/261 174/262 121/262
13.8 10.0 42.3 0.8 50.2 36.4 66.4 46.2
24/209 41/208 92/212 0/211 115/211 65/207 142/211 135/213
11.5 19.7 43.4 0.0 54.5 31.4 67.3 63.4
a
p-Value 0.075 0.02
<0.001
<0.001
<0.001
0.45 0.003 0.81 0.5 0.35 0.26 0.84 <0.001
Hospital = AP-HP and non-teaching hospitals.
(Table 1), 46.2% were aged between 45 and 54 years, and 47.4% were general practitioners in private practice. The composite score concerning consumption by the physician or by a relative of psychotropic agents was positive for 66.8% of physicians. The scenario was interpreted correctly by 53.9% of physicians: 60.1% understood that both patient groups were equivalent except for the cause of their liver disease, and 95.1% were aware that untransplanted patients would die. The opinion regarding alcoholics expressed by 60.2% of physicians was that patients should be blamed for the health consequences of alcohol consumption; 14.1% did not express any unfavorable response (Table 2). 3.2. Allocation of liver transplants to ALD patients In response to the initial question on deceased donor organ allocation, 53.0% of physicians favored allocation to the patient with primary biliary cirrhosis (Table 3). For the second question, 55.2% allocated fewer deceased donor organs for transplantation in ALD patients. After the physicians had read the opinion arguments, this latter proportion diminished slightly to 51.8%. The concordance between the physicians’ allocation of transplants
to ALD patients, before and after reading the opinion arguments, was strong (Kappa = 0.93; 95%CI = 0.90– 0.96). The physicians allocated a mean of 36.9% liver transplants to ALD patients before reading the opinion arguments and 37.6% afterwards. In both cases, this differed significantly from the 50% that might have been expected (p < 0.001). The positive correlation between the two means was strong (r = 0.96; p < 0.001). 3.3. Physician characteristics and attitudes as a function of liver transplant allocations to ALD patients Younger physicians, those who were in salaried employment only, physicians working in a hospital, and hepatologist–gastroenterologists more frequently allocated 50 liver transplants or more to ALD patients (Table 1). By contrast, GPs, private physicians and those who had misinterpreted the scenario allocated fewer than 50 liver transplants to these patients. Tobacco consumption was more common amongst physicians who allocated 50 liver transplants or more to ALD patients. The more unfavorable the attitude towards alcoholics (as estimated by the composite opinion score), the less physicians allocated liver transplants to ALD patients (Table 2). The proportions of salaried
V. Perut et al. / Journal of Hepatology 51 (2009) 707–714
711
Table 2 Comparison of attitudes expressed by responding physicians as a function of the allocation of liver transplants to alcoholic liver. Variables
n/N
%
0–49 Transplants
%
50 or more transplants
%
p-Value
(a) Alcoholism is a disease. People should not be blamed for health consequences caused by alcohol (b) Alcoholics can be good citizens and contribute much to society despite their alcoholism (c) Alcohol is a legal substance. We should not punish those who drink (d) People with problems caused by primary biliary cirrhosis are not personally responsible for their health problems related to these factors (e) The decision to drink alcohol is a personal choice, and people should take responsibility for their choice (f) There is enough information out there for people to know the risks of drinking. If they caused themselves health problems they knew what they were doing (g) All patients’ lives are equally valuable, regardless of what caused their disease (h) People can avoid becoming alcoholics (i) Alcoholics who currently drink have control over any health problems due to their alcoholism (j) Being an alcoholic is voluntary Composite opinion score No response unfavorable to alcoholics 1–3 Responses unfavorable to alcoholics 4–6 Responses unfavorable to alcoholics 7–9 Responses unfavorable to alcoholics
187/470
39.8
45/260
17.3
142/210
67.6
<0.001
313/469
66.7
146/260
56.2
167/209
79.9
<0.001
209/465
44.9
69/258
26.7
140/207
67.6
<0.001
461/474
97.3
254/261
97.3
207/213
97.2
0.93
253/471
53.7
178/260
68.5
75/211
35.5
<0.001
214/469
45.6
165/258
64.0
49/211
23.2
<0.001
386/470
82.1
179/258
69.4
207/212
97.6
<0.001
223/466 130/467
47.9 27.8
160/255 85/257
62.7 33.1
63/211 45/210
29.9 21.4
<0.001 0.005
100/466
21.5
83/257
32.3
17/209
8.1
<0.001 <0.001
65/461 166/461 161/461 69/461
14.1 36.0 34.9 15.0
8/254 54/254 128/254 64/254
3.1 21.3 50.4 25.2
57/207 112/207 33/207 5/207
27.5 54.1 16.0 2.4
Table 3 Allocation by physicians of a single liver transplant to one of two patients or of 100 liver transplants to alcoholic liver disease (ALD) patients. Allocation Allocation of a single transplant to The ALD patient The patient with primary biliary cirrhosis One or other of the patients
n/N
%
95% CI
3/468 248/468
0.6 53.0
0.1–1.9 48.4–57.6
217/468
46.4
41.8–51.0
shown). After adjustment for confounding factors (Table 4), being a GP, misinterpreting the scenario or having an unfavorable attitude towards alcoholics were independently and positively associated with the allocation of fewer than 50 liver transplants to ALD patients than to others.
4. Discussion
Allocation of 100 transplants to ALD patients before reading the arguments 0 1–49 50 51–99
28/475 234/475 195/475 18/475
5.9 49.3 41.1 3.8
4.0–8.4 44.7–53.9 36.6–45.6 2.3–5.9
Allocation of 100 transplants to ALD patients after reading the arguments 0 1–49 50 51–99
29/467 213/467 207/467 18/467
6.2 45.6 44.3 3.9
4.2–8.8 41.0–50.3 39.8–49.0 2.3–6.0
employees (p < 0.001), smokers (p = 0.011), or physicians who had not misinterpreted the scenario (p = 0.002) were all higher in the event of the least unfavorable responses regarding alcoholics (data not
This study, based on a fictitious scenario, enabled an analysis of attitudes regarding LT for ALD patients in a sample of 475 French physicians. Only a few studies performed in English-speaking countries [8,9,11] have to date considered the attitudes of the general population regarding transplantation for ALD patients, and only one [10] questioned a sample of physicians. Ours is the first French study to examine this issue, in which we questioned all private, university and public nonteaching hospital hepatologist–gastroenterologists in the Paris region. The objective of our study was to determine whether physicians might restrict access to LT for ALD patients because of their attitude regarding alcoholics and not to examine the impact of medical or psychological co-morbidity issues. The
712
V. Perut et al. / Journal of Hepatology 51 (2009) 707–714
Table 4 Characteristics and attitudes of responding physicians associated with the allocation of fewer than 50 transplants to alcoholic liver disease patients (N = 440). Variables Age 25–34 35–44 45–54 55 and + Professional status Salaried employee only Private practice only Salaried employee and private practice Work location Hospitala Occupational physician Private general practitioner Private hepatologists–gastroenterologists Tobacco consumption No Yes Misinterpretation of the scenario No Yes Composite opinion score No response unfavorable to alcoholic persons 1–3 Responses unfavorable to alcoholic persons 4–6 Responses unfavorable to alcoholic persons 7–9 Responses unfavorable to alcoholic persons a
Crude OR
95% CI
p-Value
Adjusted OR
95% CI
p-Value
1 3.4 2.4 3.0
1.5–7.8 1.1–5.2 1.4–6.7
0.004 0.025 0.007
1 2.8 3.9
1.9–4.3 2.2–6.9
<0.001 <0.001
1 1.2 3.0 1.4
0.7–2.1 1.9–4.7 0.7–2.6
0.580 <0.001 0.307
1 1.6 3.2 1.2
0.7–3.4 1.8–5.9 0.5–2.8
0.227 <0.001 0.643
1 0.5
0.3–0.8
0.004
1 2.0
1.4–2.9
<0.001
1 1.8
1.1–3.0
0.014
1 3.4 27.6 91.2
1.5–7.7 12.0–63.6 28.2–294.7
0.003 < 0.001 <0.001
1 4.0 27.4 126.8
1.7–9.5 11.1–67.6 34.0–472.1
0.002 <0.001 <0.001
Hospital = AP-HP and non-teaching hospitals.
questionnaire, derived from an US questionnaire [11] previously used and tested in the general population, was tested for its readability and acceptability in a sample population of French physicians. More than half of those responding would have allocated fewer than 50 out of the 100 available grafts to ALD patients. Only 14.1% of the physicians expressed the opinion that alcoholics were not responsible for their own illness or that alcoholism was not morally reprehensible. The allocation of fewer than 50 grafts to ALD patients was independently and positively associated with characteristics such as having an unfavorable attitude towards alcoholics and ALD, being a GP, or misunderstanding the scenario. This sample of physicians was similar to the French physician population with respect to gender and age [15]. The rate of smokers, lower than expected [16,17], may have reflected an under-reporting of tobacco consumption (which is increasingly negatively connoted), whereas the rate of alcohol consumers, higher than expected, may have been due to the sample composition (older men) [16,18] and to French tolerance of this behavior. In the same way as a study performed in the UK [8], the findings confirmed that a large proportion of physicians would restrict the access of ALD patients to transplantation. One previous study [6] performed in a medical center had shown that even if ALD patients satisfied the guidelines for referral, they underwent fewer
LTs. Others performed in the UK [9,10], USA [11,12] and Hong Kong [8] found similar results in the general population. However, during the present study, a smaller proportion of physicians (5.8% vs. 35.0%) denied access for ALD patients to LT, when compared with a sample of the general population in the USA [11]. Another USA study [12] found that the general population wished to restrict access to care for patients perceived as being responsible for their health problems. The allocation of liver transplants to ALD patients was also prioritized lower by a North American Center of Ethics [19] and a Department of Philosophy [20]. However, other US authors refuted these statements [21]. In the context of the present survey, and similar to findings in a general population sample [11], the more unfavorable was their attitude towards alcoholics (according a composite opinion score), the less physicians recommended the allocation of transplants to ALD patients, independently of other factors. Despite their medical training, and as found by another study on physicians [22], only 40% of physicians considered that alcoholism is a disease whose medical consequences cannot be blamed on patients, and 53% considered that drinking alcohol in a pattern deemed as dangerous was a personal choice for which patients were responsible. Being a GP was the second independent factor associated with the allocation of fewer liver transplants to
V. Perut et al. / Journal of Hepatology 51 (2009) 707–714
ALD patients. Since GPs are the primary care physicians responsible for referring or not referring patients to a specialist, who in turn send the patient to a transplant center, this additional step may limit access to LT. The third independent factor affecting the access of ALD patients to LT was a misunderstanding of the scenario; although it explicitly portrayed both patient groups as being medically identical, half of the physicians considered them to be different. By confirming the results of a previous study [11] performed in a general population sample, the misunderstanding that ALD patients could be equivalent to others was significantly associated with an unfavorable opinion towards alcoholics and with the allocation of fewer transplants to ALD patients. This suggests that the prior submission of medical arguments was ineffective if these arguments did not match the physicians’ attitudes. Other, but not independent, factors affecting the access of ALD patients to LT, were being a smoker or being a younger or salaried physician. A clearer understanding of addiction amongst smokers, better training for younger and salaried physicians, and more significant exchanges between salaried physicians might explain this more equitable choice. Smokers and salaried physicians also had fewer unfavorable attitudes regarding alcoholics. This study had certain limitations. First, the scenario was fictitious and physicians knew that their choice would not affect a real patient. Second, if we had compared ALD patients with patients other than those with primary biliary cirrhosis, the results might have been different. However, the introduction to the hypothetical scenarios in the questionnaire clearly explained that both patient groups were equivalent in terms of all characteristics (i.e. compatibility for transplant, survival time with and without liver transplant, duration of illness and period on the waiting list, treatment compliance, effectiveness of therapy, cost, age, health status) except for the cause of liver disease. Third, and as often is the case in opinion surveys [23], responding physicians probably tended to give the less controversial answer, i.e. the equal distribution of transplants between the two patients. Fourth, fewer than one-third of physicians receiving the questionnaire returned it completed. Nevertheless, this sample remained representative of French physicians, and the response rate was similar to that seen with other postal surveys not involving any monetary incentive [24,25]. However, it was difficult to interpret the alcohol consumption of physicians because the questionnaire did not focus on alcoholic behavior but rather on a regular consumption of alcohol, as defined by the French Observatory on Drugs and Addictions [16]. For this reason, alcoholics could not be identified in our population of physicians. During this study, we found that French physicians would restrict the access of ALD patients to LT. This
713
suggests that regardless of recognized criteria, this restriction could be exercised by a physician at any point in the continuity of care leading to LT. In order to complete and compare these results, we plan to extend this study to physicians in other European countries. The education of physicians, through more intensified training on alcoholic disease, might improve access to LT. Information should focus on the equivalent medical status of ALD patients, their favorable post-transplant results, and the fact that alcoholism is a disease for which patients are not responsible or blameworthy. Initiating and developing healthcare networks might also improve the care of alcoholic patients. Acknowledgements We are greatly indebted to the Agence de la biome´decine, which funded this survey. We would like to thank all the physicians who responded to our survey, without whom this research would not have been possible. References [1] Centre d’e´pide´miologie sur les causes me´dicales de de´ce`s. Detailed data on death causes from 1979 to 2005. Available from: http:// www.cepidc.vesinet.inserm.fr. [2] Poynard T, Naveau S, Doffoel M, Boudjema K, Vanlemmens C, Mantion G, et al. Evaluation of efficacy of liver transplantation in alcoholic cirrhosis using matched and simulated controls: 5-year survival. Multi-centre group. J Hepatol 1999;30: 1130–1137. [3] The Agence de la biome´decine. Annual activity report 2007. Available from: http://www.agence.biomedecine.fr. [4] Cohen C, Benjamin M. Alcoholics and liver transplantation. The Ethics and Social Impact Committee of the Transplant and Health Policy Center. JAMA 1991;265:1299–1301. [5] Wiesner R, Edwards E, Freeman R, Harper A, Kim R, Kamath P, et al. Model for end-stage liver disease (MELD) and allocation of donor livers. Gastroenterology 2003;124:91–96. [6] Julapalli VR, Kramer JR, El Serag HB. Evaluation for liver transplantation: adherence to AASLD referral guidelines in a large veterans affairs center. Liver Transpl 2005;11: 1370–1378. [7] Tuttle-Newhall JE, Rutledge R, Johnson M, Fair J. A statewide, population-based, time series analysis of access to liver transplantation. Transplantation 1997;63:255–262. [8] Chan HM, Cheung GM, Yip AK. Selection criteria for recipients of scarce donor livers: a public opinion survey in Hong Kong. Hong Kong Med J 2006;12:40–46. [9] Ratcliffe J. Public preferences for the allocation of donor liver grafts for transplantation. Health Econ 2000;9:137–148. [10] Neuberger J, Adams D, MacMaster P, Maidment A, Speed M. Assessing priorities for allocation of donor liver grafts: survey of public and clinicians. BMJ 1998;317:172–175. [11] Wittenberg E, Goldie SJ, Fischhoff B, Graham JD. Rationing decisions and individual responsibility for illness: are all lives equal? Med Decis Making 2003;23:194–211. [12] Ubel PA, Baron J, Asch DA. Social responsibility, personal responsibility, and prognosis in public judgments about transplant allocation. Bioethics 1999;13:57–68. [13] Calmus Y, Jacquelinet C. New rules for distributing liver grafts. Gastroenterol Clin Biol 2008;32:585–588.
714
V. Perut et al. / Journal of Hepatology 51 (2009) 707–714
[14] Consensus conference: indications for liver transplantation, January 19 and 20, 2005, Lyon-Palais Des Congres Liver Transpl 2006;12:998–1011. [15] Annual report of national observatory of health occupational 2005. Available from: http://www.sante.gouv.fr/ondps/synthese_ rapport200.pdf. [16] French Observatory on Drugs and Addictions. Drugs and drug addictions-indicators and trends 2002. Available from: http:// www.ofdt.fr. [17] Josseran L, King G, Guilbert P, Davis J, Brucker G. Smoking by French general practitioners: behaviour, attitudes and practice. Eur J Public Health 2005;15:33–38. [18] INPES. Barometer of health general practitioners 1998/1999. Available from: http://www.inpes.sante.fr/70000/dp/00/dp000119.pdf. [19] Moss AH, Siegler M. Should alcoholics compete equally for liver transplantation? JAMA 1991;265:1295–1298. [20] Glannon W. Responsibility, alcoholism, and liver transplantation. J Med Philos 1998;23:31–49.
[21] Benjamin M, Turcotte JG. Ethics, alcoholism and liver transplantation. In: Lucey MR, Merion RM, Beresford TP, editors. Liver transplantation & the alcoholic patient. Cambridge: Cambridge University Press; 1994. p. 113–130. [22] Rosta J. Is alcoholism a self-induced disease? A survey among doctors in Aarhus, Denmark and in Mainz, Germany. Nord J Psychiatry 2004;58:219–222. [23] Fisher RJ, Katz JE. Social-desirability bias and the validity of selfreported values. P&M 1999;17:105–20. [24] Streiff MB, Dundes L, Spivak JL. A mail survey of United States hematologists and oncologists: a comparison of business reply versus stamped return envelopes. J Clin Epidemiol 2001;54:430–432. [25] Whiteman MK, Langenberg P, Kjerulff K, McCarter R, Flaws JA. A randomized trial of incentives to improve response rates to a mailed women’s health questionnaire. J Womens Health (Larchmt) 2003;12:821–828.
Might physicians be restricting access to liver transplantation for patients with alcoholic liver disease?q Vale´rie Perut1, Filome´na Conti2, Olivier Scatton2, Olivier Soubrane2, Yvon Calmus2, Gwenae¨lle Vidal-Trecan1,3,* 1
Unite´ de Gestion des Risques et Qualite´, Groupe hospitalier Cochin-Saint Vincent de Paul, AP-HP, Paris, France 2 Unite´ de Transplantation He´patique, Faculte´ de Me´decine, Universite´ Paris Descartes, Groupe hospitalier Cochin-Saint Vincent de Paul, AP-HP, Paris, France 3 Departement de Sante´ Publique, Faculte´ de Me´decine, Universite´ Paris Descartes, 24, rue du Faubourg Saint Jacques, 75014 Paris, France
Background/Aims: In France, the most common cause of cirrhosis is excessive alcohol consumption. Post-transplant survival rates in patients with alcoholic liver disease (ALD) are at least as good as those seen with other indications. However, fewer of these patients are found on the waiting list. To understand the reasons for this discrepancy, it was decided to examine physicians’ attitudes concerning the allocation of deceased donor liver allografts. Methods: Using a standardized postal questionnaire, 1739 physicians were asked to allocate 100 liver transplants to two competing groups of patients who were equivalent except for the cause of their cirrhosis (i.e. alcohol-related or primary biliary cirrhosis). A composite score was then used to assess their attitude regarding the behavior of alcoholics and their responsibility for their illness. Results: Among the 475 respondents (response rate: 27.3%), 55.2% allocated fewer than 50 transplants to ALD patients. This lower rate was independently associated with factors such as being a general practitioner (odds ratio [OR] = 3.2, 95% confidence interval [95%CI] = 1.8–5.9), a misinterpretation of ALD patients being equivalent to others (OR = 1.8, 95%CI = 1.1–3.0) or unfavorable attitudes regarding alcoholics (OR = 4.0, 95%CI = 1.7–9.5, to OR = 126.8, 95%CI = 34.0–472.1). Conclusions: Greater information and education of physicians may improve access to liver transplantation for ALD patients. Ó 2009 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. Keywords: Liver transplantation; Liver diseases; Alcoholic; Access to health care
1. Introduction
Received 19 September 2008; received in revised form 16 March 2009; accepted 29 April 2009; available online 28 May 2009 Associate Editor: P.-A. Clavien q The authors who have taken part in this study declared that they do not have anything to disclose regarding funding or conflict of interest with respect to this manuscript. * Corresponding author. Tel.: +33 1 58413146; fax: +33 1 58412678. E-mail address: [email protected] (G. VidalTrecan). Abbreviations: ALD, alcoholic liver disease; LT, liver transplantation; UK, United Kingdom; GP, general practitioner; US, United States; OR, odds ratio.
Alcohol consumption in France is the leading cause of morbidity and mortality related to liver disorders [1]. Liver transplantation (LT) is currently the only treatment available for end-stage alcoholic liver disease (ALD). The post-transplant survival of ALD patients, with or without a relapse of alcoholism, is at least as good as in patients with other liver diseases [2]. Nevertheless, in view of the frequency of ALD in France, the proportion of these patients on the LT waiting list could be higher (ALD patients accounted for 72.6% of patients aged 25–75 years registered for liver-related deaths in 2005 [1], whereas they only represented 24.0% of new registrations on the LT
0168-8278/$36.00 Ó 2009 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.jhep.2009.04.018
708
V. Perut et al. / Journal of Hepatology 51 (2009) 707–714
waiting list in 2007 [3]). We share the opinion of other authors [4] who have considered that registration on the waiting list should be solely dependent on recognized criteria reflecting medical need (such as the MELD score) [5] regarding this scarce, expensive, rationed and yet life-saving resource. Two studies [6,7] performed in the USA found that ALD was associated with a lower rate of LT than other causes of liver disease. Other studies [8–11] performed in English-speaking countries have described a public reluctance to finance LT for ALD patients. According to Ubel et al. [12], the decision to allocate fewer organs to ALD patients reflects a belief that they are ‘‘responsible for their own illness” or ‘‘less worthy of transplants than others”. To our knowledge, the attitudes of physicians regarding the access of ALD patients to LT has only been investigated by one study performed in the United Kingdom (UK) [10], during which a sample of gastroenterologists and general practitioners (GPs) was asked to rank their priorities for the allocation of liver transplants to patients with controversial criteria. They assigned the lowest priorities to ALD patients and prisoners, as did the general public [10]. In France, any physician can refer patients with endstage liver disease to a hospital specialist who is authorized to register patients on the national transplant waiting list. We hypothesized that, consciously or not, ‘‘first-line” physicians might be restricting access to LT for ALD patients (e.g. by not referring patients to the authorized team at all, or by referring them when they have reached a critical stage at which transplantation is contraindicated). Using a postal questionnaire, the aim of this study was thus to examine whether physicians managing liver disease patients might be restricting the access of ALD patients to LT on the basis of criteria unrelated to medical need, but rather reflecting their perception of the ‘‘self-induced” nature of alcoholic disease, associated with a potential risk of return to a pattern of alcoholic consumption and noncompliance. 2. Materials and methods 2.1. Study design A cross-sectional study was performed from October 2005 to March 2006 on a sample of physicians, using a standardized postal questionnaire. To explore their attitudes regarding the access of ALD patients to LT, they were asked to allocate, under a hypothetical scenario, a limited number of liver transplants to two competing groups of patients who were equivalent in every characteristic except for the cause of their liver disease, i.e. alcohol-related or primary biliary cirrhosis.
2.2. Setting This study involved physicians working in public hospitals, i.e. University Hospitals within the Paris Public Hospital Network (Assistance Publique-Hoˆpitaux de Paris, AP-HP) (38 hospitals) and non-teaching hospitals (30 hospitals), as well as physicians caring for outpatients in the Paris region.
To be registered on the LT waiting list, patients must be referred by a physician (hepatologist, gastroenterologist or other specialists, seldom general practitioners) to a hospital physician or surgeon authorized by the French agency responsible for organ procurement and transplantation (Agence de Biome´decine). After clinical examination, an analysis of the mandatory investigations required to determine suitability for liver transplant listing, and calculation of a liver score specific to France (i.e. the ‘‘ScoreFoie” [13], derived from the Model for End-stage Liver Disease (MELD score) [5]), a multidisciplinary team (liver surgeons, hepatologists, anesthetists, psychologists) will decide whether to register the patient or not. The MELD score is the principal parameter taken into account when allocating liver grafts in France, but the ‘‘ScoreFoie” used to prioritize the patients in each blood group also includes parameters related to hepatocellular carcinoma, the presence of liver diseases other than cirrhosis, and the distance between the donor center and recipient center. A full six months without any alcohol consumption is usually required for ALD patients to be registered on the waiting list and then undergo LT [14].
2.3. Study population The questionnaire was circulated to 1753 physicians. Fourteen forms were returned as undeliverable, so that 1739 physicians were included in the study. The AP-HP physicians database, the Health Ministry database for medical staff in non-teaching hospitals in the Paris region, and the France Telecom phone book were used to identify all hospital and private hepatologists–gastroenterologists and also to select a random sample of GPs from the 6059 listed. The questionnaire was sent to all AP-HP physicians working in gastroenterology or hepatology departments (n = 343), including LT physicians (n = 32), senior hepatologists–gastroenterologists or internists in non-teaching hospitals (n = 91), private hepatologists–gastroenterologists (n = 214) and the sample of GPs (n = 992). Occupational physicians (n = 99) attending a continuing training meeting were also given the questionnaire.
2.4. Data collection The questionnaire, together with an introductory letter and a postage-paid reply envelope, was sent out in October 2005, followed by a reminder questionnaire in March 2006. The questionnaire was circulated to the occupational physicians during their meeting, and then collected. No direct or indirect nominative information was collected, as required by the CNIL (Commission Nationale de l’Informatique et des Liberte´s).
2.5. Questionnaire The questionnaire was adapted for use by a population of French physicians from that used in a survey on a general population sample in the USA [11]. The fictitious patients were equivalent in terms of all characteristics (i.e. compatibility for transplant, survival time with and without liver transplant, duration of illness and period on the waiting list, treatment compliance, effectiveness of therapy, cost, age, health status) except for the cause of the condition, i.e. ALD or autoimmune disease. For the sake of clarity, no patients without psychosocial contraindications (e.g. ongoing substance use, severe personality disorder) were included. Three questions were asked regarding the allocation of deceased donor organs for LT. The first concerned the allocation of a liver transplant to one of two patients: one with ALD, the other with primary biliary cirrhosis. The second focused on the allocation of a 100 liver transplants among patients with the two different causes for their cirrhosis. Thirdly, after putting forward a series of medically unfounded arguments, the second allocation question was repeated to test the influence of these arguments on the allocations made by the physicians. A set of 10 questions developing these arguments (e.g. preventable, controllable, voluntary or immoral nature of alcoholism) was then asked in order to evaluate the physicians’ attitudes concerning alcoholics. The physicians were also asked to provide information on their social and demographic characteristics, the
V. Perut et al. / Journal of Hepatology 51 (2009) 707–714 consumption of psychotropic agents by them or their relatives, and their understanding of the scenario. The questionnaire had previously been tested in a sample population of 21 physicians.
2.6. Definition of variables The allocation of 100 liver transplants to ALD patients was classified as ‘‘0–49 transplants” or ‘‘50 transplants or more”. The professional status of physicians was divided into salaried employment only, private practice only, or a combination of salaried employment and private practice. If the physician’s response indicated that he or she was neither a salaried employee nor a private practitioner, this was deemed to be missing data. Workplaces (i.e. where the letter was sent) were classified as hospital, occupational physicians, private hepatologists–gastroenterologists and GPs. The medical specialty was classified as liver transplant physicians, hepatologist–gastroenterologists, GPs and occupational physicians. The consumption of sleeping ts and tranquilizers was defined as at least one tablet in 30 days, tobacco consumption as at least one cigarette a day, alcohol consumption as at least three glasses of alcohol per week and cannabis consumption as at least 10 times within the past year. Responses to questions regarding the consumption of psychotropic agents by physicians or their relatives were combined in two composite scores. The first was categorized as at least one personal consumption (i.e. positive) or not, and the second as at least one consumption by the physician or a relative (i.e. positive) or not. The attitudes of physicians were evaluated using a 5-point, agree– disagree, Likert-type dichotomized scale, with middle scores being interpreted as disagreement [11]. Responses were combined to obtain a composite score of unfavorable attitudes regarding alcoholics [11]. The equivalence of two patients was evaluated using one question containing two items: ‘‘equivalent (yes or no)” or ‘‘different (yes or no)”. If a physician answered ‘‘yes” to both ‘‘equivalent” and ‘‘different”, the answer was considered as missing. Their interpretation of the scenarios was assessed from their understanding that the two groups of patients only differed in terms of the cause of their liver disease and their final outcome (i.e. death of patients not receiving liver transplant).
709
2.7. Statistical analysis Data were analyzed using standard descriptive statistics (mean, standard deviation (SD), and frequency). A correlation between the mean of liver transplants allocated to ALD patients, before and after presentation of the opinion arguments, was evaluated using the Spearman test. The concordance of the allocation of ‘‘0–50 transplants” or ‘‘50 transplants or more” to ALD patients, before and after presentation of the opinion arguments, was evaluated using the Kappa test. Bivariate analyses were performed using the v2-test or Fisher’s exact test. A logistic regression model was used to identify factors independently associated with the allocation of fewer than 50 transplants to ALD patients. Backward stepwise procedures were applied so that the final model only included factors providing a significant explanation of outcomes. Associations were expressed as odds ratios (ORs) with their 95% confidence intervals (95% CIs). Statistical significance was set at 0.05. SPSS Software for WindowsTM version 14.0 was used for statistical analysis.
3. Results 3.1. Characteristics of responding physicians and their attitudes concerning alcoholics Of the 1739 physicians included, 526 returned the questionnaire (response rate: 30.2%). Of these, 34 did not answer the main question (i.e. allocation of liver transplant) and the workplace of 17 was unknown. These 51 questionnaires were considered as invalid and were thus excluded, so that the answers from 475 physicians (27.3%) were analyzed. The distribution of workplaces and response rates is shown in Fig. 1. Among the physicians included in the study, 61.7% were men
1753 questionnaires circulated
14 questionnaires undeliverable 1739 physicians included
526 questionnaires returned response rate=30.2%
475 valid responses valid response rate=27.3%
1206 private physicians included 282 valid responses valid response rate=23.4%
533 salaried physicians included 193 valid responses valid response rate=36.2%
343 AP-HP 109 valid reponses valid response rate=31.8%
32 liver transplant physicians 17 valid responses valid response rate=53.1%
91 non-teaching hospitals 13 valid responses valid response rate=14.3%
99 occupational physicians 71 valid responses valid response rate=71.7%
214 private gastroenterologists 57 valid responses valid response rate=26.6%
311 other physicians of AP-HP 92 valid responses valid response rate=29.6%
Fig. 1. Origin of responses as a function of the workplace of physicians.
992 private general practitioners 225 valid responses valid response rate=22.7%
710
V. Perut et al. / Journal of Hepatology 51 (2009) 707–714
Table 1 Comparison of the characteristics of responding physicians as a function of the allocation of liver transplants to alcoholic liver disease patients. Variables
n/N
%
0–49 Transplants
%
50 or more transplants
%
Male gender Age 25–34 35–44 45–54 55 and + Professional status Salaried employee only Private practice only Salaried employee and private practice Work location Hospitala Occupational physician Private general practitioner Private hepatologists–gastroenterologists Medical specialty Liver transplant physician Hepatologist–gastroenterologist General practitioner Occupational physician Personal consumption Sleeping tablets and tranquilizers Tobacco Alcohol Cannabis At least one psychotropic agent Alcohol consumption by relatives Personal consumption or by a relative No misinterpretation of the scenario
293/475
61.7
171/262
65.3
122/213
57.3
34/474 92/474 219/474 129/474
7.2 19.4 46.2 27.2
11/261 57/261 117/261 76/261
4.2 21.8 44.8 29.1
23/213 35/213 102/213 53/213
10.8 16.4 47.9 24.9
169/460 215/460 76/460
36.7 46.7 16.5
63/251 135/251 53/251
25.1 53.8 21.1
106/209 80/209 23/209
50.7 38.3 11.0
122/475 71/475 225/475 57/475
25.7 14.9 47.4 12.0
50/262 32/262 152/262 28/262
19.1 12.2 58.0 10.7
72/213 39/213 73/213 29/213
33.8 18.3 34.3 13.6
17/475 156/475 231/475 71/475
3.6 32.8 48.6 14.9
8/262 69/262 153/262 32/262
3.1 26.3 58.4 12.2
9/213 87/213 78/213 39/213
4.2 40.8 36.6 18.3
60/469 67/467 202/472 2/470 245/470 160/468 316/473 256/475
12.8 14.3 42.8 0.4 52.1 34.2 66.8 53.9
36/260 26/259 110/260 2/259 130/259 95/261 174/262 121/262
13.8 10.0 42.3 0.8 50.2 36.4 66.4 46.2
24/209 41/208 92/212 0/211 115/211 65/207 142/211 135/213
11.5 19.7 43.4 0.0 54.5 31.4 67.3 63.4
a
p-Value 0.075 0.02
<0.001
<0.001
<0.001
0.45 0.003 0.81 0.5 0.35 0.26 0.84 <0.001
Hospital = AP-HP and non-teaching hospitals.
(Table 1), 46.2% were aged between 45 and 54 years, and 47.4% were general practitioners in private practice. The composite score concerning consumption by the physician or by a relative of psychotropic agents was positive for 66.8% of physicians. The scenario was interpreted correctly by 53.9% of physicians: 60.1% understood that both patient groups were equivalent except for the cause of their liver disease, and 95.1% were aware that untransplanted patients would die. The opinion regarding alcoholics expressed by 60.2% of physicians was that patients should be blamed for the health consequences of alcohol consumption; 14.1% did not express any unfavorable response (Table 2). 3.2. Allocation of liver transplants to ALD patients In response to the initial question on deceased donor organ allocation, 53.0% of physicians favored allocation to the patient with primary biliary cirrhosis (Table 3). For the second question, 55.2% allocated fewer deceased donor organs for transplantation in ALD patients. After the physicians had read the opinion arguments, this latter proportion diminished slightly to 51.8%. The concordance between the physicians’ allocation of transplants
to ALD patients, before and after reading the opinion arguments, was strong (Kappa = 0.93; 95%CI = 0.90– 0.96). The physicians allocated a mean of 36.9% liver transplants to ALD patients before reading the opinion arguments and 37.6% afterwards. In both cases, this differed significantly from the 50% that might have been expected (p < 0.001). The positive correlation between the two means was strong (r = 0.96; p < 0.001). 3.3. Physician characteristics and attitudes as a function of liver transplant allocations to ALD patients Younger physicians, those who were in salaried employment only, physicians working in a hospital, and hepatologist–gastroenterologists more frequently allocated 50 liver transplants or more to ALD patients (Table 1). By contrast, GPs, private physicians and those who had misinterpreted the scenario allocated fewer than 50 liver transplants to these patients. Tobacco consumption was more common amongst physicians who allocated 50 liver transplants or more to ALD patients. The more unfavorable the attitude towards alcoholics (as estimated by the composite opinion score), the less physicians allocated liver transplants to ALD patients (Table 2). The proportions of salaried
V. Perut et al. / Journal of Hepatology 51 (2009) 707–714
711
Table 2 Comparison of attitudes expressed by responding physicians as a function of the allocation of liver transplants to alcoholic liver. Variables
n/N
%
0–49 Transplants
%
50 or more transplants
%
p-Value
(a) Alcoholism is a disease. People should not be blamed for health consequences caused by alcohol (b) Alcoholics can be good citizens and contribute much to society despite their alcoholism (c) Alcohol is a legal substance. We should not punish those who drink (d) People with problems caused by primary biliary cirrhosis are not personally responsible for their health problems related to these factors (e) The decision to drink alcohol is a personal choice, and people should take responsibility for their choice (f) There is enough information out there for people to know the risks of drinking. If they caused themselves health problems they knew what they were doing (g) All patients’ lives are equally valuable, regardless of what caused their disease (h) People can avoid becoming alcoholics (i) Alcoholics who currently drink have control over any health problems due to their alcoholism (j) Being an alcoholic is voluntary Composite opinion score No response unfavorable to alcoholics 1–3 Responses unfavorable to alcoholics 4–6 Responses unfavorable to alcoholics 7–9 Responses unfavorable to alcoholics
187/470
39.8
45/260
17.3
142/210
67.6
<0.001
313/469
66.7
146/260
56.2
167/209
79.9
<0.001
209/465
44.9
69/258
26.7
140/207
67.6
<0.001
461/474
97.3
254/261
97.3
207/213
97.2
0.93
253/471
53.7
178/260
68.5
75/211
35.5
<0.001
214/469
45.6
165/258
64.0
49/211
23.2
<0.001
386/470
82.1
179/258
69.4
207/212
97.6
<0.001
223/466 130/467
47.9 27.8
160/255 85/257
62.7 33.1
63/211 45/210
29.9 21.4
<0.001 0.005
100/466
21.5
83/257
32.3
17/209
8.1
<0.001 <0.001
65/461 166/461 161/461 69/461
14.1 36.0 34.9 15.0
8/254 54/254 128/254 64/254
3.1 21.3 50.4 25.2
57/207 112/207 33/207 5/207
27.5 54.1 16.0 2.4
Table 3 Allocation by physicians of a single liver transplant to one of two patients or of 100 liver transplants to alcoholic liver disease (ALD) patients. Allocation Allocation of a single transplant to The ALD patient The patient with primary biliary cirrhosis One or other of the patients
n/N
%
95% CI
3/468 248/468
0.6 53.0
0.1–1.9 48.4–57.6
217/468
46.4
41.8–51.0
shown). After adjustment for confounding factors (Table 4), being a GP, misinterpreting the scenario or having an unfavorable attitude towards alcoholics were independently and positively associated with the allocation of fewer than 50 liver transplants to ALD patients than to others.
4. Discussion
Allocation of 100 transplants to ALD patients before reading the arguments 0 1–49 50 51–99
28/475 234/475 195/475 18/475
5.9 49.3 41.1 3.8
4.0–8.4 44.7–53.9 36.6–45.6 2.3–5.9
Allocation of 100 transplants to ALD patients after reading the arguments 0 1–49 50 51–99
29/467 213/467 207/467 18/467
6.2 45.6 44.3 3.9
4.2–8.8 41.0–50.3 39.8–49.0 2.3–6.0
employees (p < 0.001), smokers (p = 0.011), or physicians who had not misinterpreted the scenario (p = 0.002) were all higher in the event of the least unfavorable responses regarding alcoholics (data not
This study, based on a fictitious scenario, enabled an analysis of attitudes regarding LT for ALD patients in a sample of 475 French physicians. Only a few studies performed in English-speaking countries [8,9,11] have to date considered the attitudes of the general population regarding transplantation for ALD patients, and only one [10] questioned a sample of physicians. Ours is the first French study to examine this issue, in which we questioned all private, university and public nonteaching hospital hepatologist–gastroenterologists in the Paris region. The objective of our study was to determine whether physicians might restrict access to LT for ALD patients because of their attitude regarding alcoholics and not to examine the impact of medical or psychological co-morbidity issues. The
712
V. Perut et al. / Journal of Hepatology 51 (2009) 707–714
Table 4 Characteristics and attitudes of responding physicians associated with the allocation of fewer than 50 transplants to alcoholic liver disease patients (N = 440). Variables Age 25–34 35–44 45–54 55 and + Professional status Salaried employee only Private practice only Salaried employee and private practice Work location Hospitala Occupational physician Private general practitioner Private hepatologists–gastroenterologists Tobacco consumption No Yes Misinterpretation of the scenario No Yes Composite opinion score No response unfavorable to alcoholic persons 1–3 Responses unfavorable to alcoholic persons 4–6 Responses unfavorable to alcoholic persons 7–9 Responses unfavorable to alcoholic persons a
Crude OR
95% CI
p-Value
Adjusted OR
95% CI
p-Value
1 3.4 2.4 3.0
1.5–7.8 1.1–5.2 1.4–6.7
0.004 0.025 0.007
1 2.8 3.9
1.9–4.3 2.2–6.9
<0.001 <0.001
1 1.2 3.0 1.4
0.7–2.1 1.9–4.7 0.7–2.6
0.580 <0.001 0.307
1 1.6 3.2 1.2
0.7–3.4 1.8–5.9 0.5–2.8
0.227 <0.001 0.643
1 0.5
0.3–0.8
0.004
1 2.0
1.4–2.9
<0.001
1 1.8
1.1–3.0
0.014
1 3.4 27.6 91.2
1.5–7.7 12.0–63.6 28.2–294.7
0.003 < 0.001 <0.001
1 4.0 27.4 126.8
1.7–9.5 11.1–67.6 34.0–472.1
0.002 <0.001 <0.001
Hospital = AP-HP and non-teaching hospitals.
questionnaire, derived from an US questionnaire [11] previously used and tested in the general population, was tested for its readability and acceptability in a sample population of French physicians. More than half of those responding would have allocated fewer than 50 out of the 100 available grafts to ALD patients. Only 14.1% of the physicians expressed the opinion that alcoholics were not responsible for their own illness or that alcoholism was not morally reprehensible. The allocation of fewer than 50 grafts to ALD patients was independently and positively associated with characteristics such as having an unfavorable attitude towards alcoholics and ALD, being a GP, or misunderstanding the scenario. This sample of physicians was similar to the French physician population with respect to gender and age [15]. The rate of smokers, lower than expected [16,17], may have reflected an under-reporting of tobacco consumption (which is increasingly negatively connoted), whereas the rate of alcohol consumers, higher than expected, may have been due to the sample composition (older men) [16,18] and to French tolerance of this behavior. In the same way as a study performed in the UK [8], the findings confirmed that a large proportion of physicians would restrict the access of ALD patients to transplantation. One previous study [6] performed in a medical center had shown that even if ALD patients satisfied the guidelines for referral, they underwent fewer
LTs. Others performed in the UK [9,10], USA [11,12] and Hong Kong [8] found similar results in the general population. However, during the present study, a smaller proportion of physicians (5.8% vs. 35.0%) denied access for ALD patients to LT, when compared with a sample of the general population in the USA [11]. Another USA study [12] found that the general population wished to restrict access to care for patients perceived as being responsible for their health problems. The allocation of liver transplants to ALD patients was also prioritized lower by a North American Center of Ethics [19] and a Department of Philosophy [20]. However, other US authors refuted these statements [21]. In the context of the present survey, and similar to findings in a general population sample [11], the more unfavorable was their attitude towards alcoholics (according a composite opinion score), the less physicians recommended the allocation of transplants to ALD patients, independently of other factors. Despite their medical training, and as found by another study on physicians [22], only 40% of physicians considered that alcoholism is a disease whose medical consequences cannot be blamed on patients, and 53% considered that drinking alcohol in a pattern deemed as dangerous was a personal choice for which patients were responsible. Being a GP was the second independent factor associated with the allocation of fewer liver transplants to
V. Perut et al. / Journal of Hepatology 51 (2009) 707–714
ALD patients. Since GPs are the primary care physicians responsible for referring or not referring patients to a specialist, who in turn send the patient to a transplant center, this additional step may limit access to LT. The third independent factor affecting the access of ALD patients to LT was a misunderstanding of the scenario; although it explicitly portrayed both patient groups as being medically identical, half of the physicians considered them to be different. By confirming the results of a previous study [11] performed in a general population sample, the misunderstanding that ALD patients could be equivalent to others was significantly associated with an unfavorable opinion towards alcoholics and with the allocation of fewer transplants to ALD patients. This suggests that the prior submission of medical arguments was ineffective if these arguments did not match the physicians’ attitudes. Other, but not independent, factors affecting the access of ALD patients to LT, were being a smoker or being a younger or salaried physician. A clearer understanding of addiction amongst smokers, better training for younger and salaried physicians, and more significant exchanges between salaried physicians might explain this more equitable choice. Smokers and salaried physicians also had fewer unfavorable attitudes regarding alcoholics. This study had certain limitations. First, the scenario was fictitious and physicians knew that their choice would not affect a real patient. Second, if we had compared ALD patients with patients other than those with primary biliary cirrhosis, the results might have been different. However, the introduction to the hypothetical scenarios in the questionnaire clearly explained that both patient groups were equivalent in terms of all characteristics (i.e. compatibility for transplant, survival time with and without liver transplant, duration of illness and period on the waiting list, treatment compliance, effectiveness of therapy, cost, age, health status) except for the cause of liver disease. Third, and as often is the case in opinion surveys [23], responding physicians probably tended to give the less controversial answer, i.e. the equal distribution of transplants between the two patients. Fourth, fewer than one-third of physicians receiving the questionnaire returned it completed. Nevertheless, this sample remained representative of French physicians, and the response rate was similar to that seen with other postal surveys not involving any monetary incentive [24,25]. However, it was difficult to interpret the alcohol consumption of physicians because the questionnaire did not focus on alcoholic behavior but rather on a regular consumption of alcohol, as defined by the French Observatory on Drugs and Addictions [16]. For this reason, alcoholics could not be identified in our population of physicians. During this study, we found that French physicians would restrict the access of ALD patients to LT. This
713
suggests that regardless of recognized criteria, this restriction could be exercised by a physician at any point in the continuity of care leading to LT. In order to complete and compare these results, we plan to extend this study to physicians in other European countries. The education of physicians, through more intensified training on alcoholic disease, might improve access to LT. Information should focus on the equivalent medical status of ALD patients, their favorable post-transplant results, and the fact that alcoholism is a disease for which patients are not responsible or blameworthy. Initiating and developing healthcare networks might also improve the care of alcoholic patients. Acknowledgements We are greatly indebted to the Agence de la biome´decine, which funded this survey. We would like to thank all the physicians who responded to our survey, without whom this research would not have been possible. References [1] Centre d’e´pide´miologie sur les causes me´dicales de de´ce`s. Detailed data on death causes from 1979 to 2005. Available from: http:// www.cepidc.vesinet.inserm.fr. [2] Poynard T, Naveau S, Doffoel M, Boudjema K, Vanlemmens C, Mantion G, et al. Evaluation of efficacy of liver transplantation in alcoholic cirrhosis using matched and simulated controls: 5-year survival. Multi-centre group. J Hepatol 1999;30: 1130–1137. [3] The Agence de la biome´decine. Annual activity report 2007. Available from: http://www.agence.biomedecine.fr. [4] Cohen C, Benjamin M. Alcoholics and liver transplantation. The Ethics and Social Impact Committee of the Transplant and Health Policy Center. JAMA 1991;265:1299–1301. [5] Wiesner R, Edwards E, Freeman R, Harper A, Kim R, Kamath P, et al. Model for end-stage liver disease (MELD) and allocation of donor livers. Gastroenterology 2003;124:91–96. [6] Julapalli VR, Kramer JR, El Serag HB. Evaluation for liver transplantation: adherence to AASLD referral guidelines in a large veterans affairs center. Liver Transpl 2005;11: 1370–1378. [7] Tuttle-Newhall JE, Rutledge R, Johnson M, Fair J. A statewide, population-based, time series analysis of access to liver transplantation. Transplantation 1997;63:255–262. [8] Chan HM, Cheung GM, Yip AK. Selection criteria for recipients of scarce donor livers: a public opinion survey in Hong Kong. Hong Kong Med J 2006;12:40–46. [9] Ratcliffe J. Public preferences for the allocation of donor liver grafts for transplantation. Health Econ 2000;9:137–148. [10] Neuberger J, Adams D, MacMaster P, Maidment A, Speed M. Assessing priorities for allocation of donor liver grafts: survey of public and clinicians. BMJ 1998;317:172–175. [11] Wittenberg E, Goldie SJ, Fischhoff B, Graham JD. Rationing decisions and individual responsibility for illness: are all lives equal? Med Decis Making 2003;23:194–211. [12] Ubel PA, Baron J, Asch DA. Social responsibility, personal responsibility, and prognosis in public judgments about transplant allocation. Bioethics 1999;13:57–68. [13] Calmus Y, Jacquelinet C. New rules for distributing liver grafts. Gastroenterol Clin Biol 2008;32:585–588.
714
V. Perut et al. / Journal of Hepatology 51 (2009) 707–714
[14] Consensus conference: indications for liver transplantation, January 19 and 20, 2005, Lyon-Palais Des Congres Liver Transpl 2006;12:998–1011. [15] Annual report of national observatory of health occupational 2005. Available from: http://www.sante.gouv.fr/ondps/synthese_ rapport200.pdf. [16] French Observatory on Drugs and Addictions. Drugs and drug addictions-indicators and trends 2002. Available from: http:// www.ofdt.fr. [17] Josseran L, King G, Guilbert P, Davis J, Brucker G. Smoking by French general practitioners: behaviour, attitudes and practice. Eur J Public Health 2005;15:33–38. [18] INPES. Barometer of health general practitioners 1998/1999. Available from: http://www.inpes.sante.fr/70000/dp/00/dp000119.pdf. [19] Moss AH, Siegler M. Should alcoholics compete equally for liver transplantation? JAMA 1991;265:1295–1298. [20] Glannon W. Responsibility, alcoholism, and liver transplantation. J Med Philos 1998;23:31–49.
[21] Benjamin M, Turcotte JG. Ethics, alcoholism and liver transplantation. In: Lucey MR, Merion RM, Beresford TP, editors. Liver transplantation & the alcoholic patient. Cambridge: Cambridge University Press; 1994. p. 113–130. [22] Rosta J. Is alcoholism a self-induced disease? A survey among doctors in Aarhus, Denmark and in Mainz, Germany. Nord J Psychiatry 2004;58:219–222. [23] Fisher RJ, Katz JE. Social-desirability bias and the validity of selfreported values. P&M 1999;17:105–20. [24] Streiff MB, Dundes L, Spivak JL. A mail survey of United States hematologists and oncologists: a comparison of business reply versus stamped return envelopes. J Clin Epidemiol 2001;54:430–432. [25] Whiteman MK, Langenberg P, Kjerulff K, McCarter R, Flaws JA. A randomized trial of incentives to improve response rates to a mailed women’s health questionnaire. J Womens Health (Larchmt) 2003;12:821–828.