Volume 174 Issue 7

Healthcare The management of testicular cancer in Victoria, 1988-1993 Guy C Toner, Greg J Neerhut, Max A Schwarz, Vicky J Thursfield, Thomas F Sandeman, Graham G Giles and Ross M Snow, for the Urology Study Committee of the Victorian Co-operative Oncology Group MJA 2001; 174: 328-331 For editorial comment, see Boyer & Stockler Abstract - Methods - Results - Discussion - Acknowledgements - References - Authors' details - - More articles on Oncology Abstract Objectives: To evaluate the patterns of care and management of testicular cancer in Victoria. Design and setting: Retrospective analysis of all cases of testicular cancer in Victoria from 1988 to 1993 identified through the Victorian Cancer Registry. Main outcome measures: Description of patient characteristics, staging investigations, initial management, and outcome. Results: 667 eligible cases of testicular cancer were identified and questionnaires were returned for 633 of these patients (94.9% response rate). There were 357 (56.4%) patients with pure seminoma; 271 (42.8%) with non-seminomatous germ cell tumours, 3 (0.5%) with stromal tumours, and 2 (0.3%) with other tumours. The median age was 32 years (range, 0-80 years). Preoperative marker levels were not available for 8% of patients, and initial staging was considered inadequate in 6%. Surveillance programs used for patients with Stage I disease were considered inadequate in most. Relative survival at five years was 99% for patients with seminoma and 91% for non-seminoma. Conclusions: There was considerable variation in the investigation, treatment, and follow-up of these patients, which is likely to have resulted in unnecessary morbidity. Clinical practice guidelines should be developed and implemented to promote optimal management. Testicular cancer is uncommon but has a significant effect on the community, as it typically develops in young men. The median age of incidence is approximately 30 years.1 Excellent survival rates are expected with the use of surgery, radiation therapy, and chemotherapy.2 Cure is expected in about 80% of patients with metastatic and relapsed disease.2There is a strong evidence base available to guide management, with relatively uniform practice at major centres internationally. Some aspects of management are peculiar to this malignancy. These include: the strong reliance on serum tumour markers for management,3,4 the use of surveillance in patients with Stage I disease, with anticipated cure by salvage therapy in those who relapse,5 the expectation of cure of metastatic disease with a brief, intensive course of chemotherapy,6 and the use of surgery to resect residual masses at metastatic sites after chemotherapy.7 Because of the potential for cure, the multidisciplinary nature of management and the availability of high quality evidence to guide practice, the management of testicular cancer is expected to provide an excellent assessment of the quality of cancer management in a community. However, there are no comprehensive data on the management of testicular cancer in the community in Australia. Methods The Urology Study Committee of the Victorian Co-operative Oncology Group undertook a survey of the management of testicular cancer in Victoria. A population-based sample was identified from the cancer registry. The period 1988 to 1993 was chosen to allow an adequate sample size and follow-up for the survey. A questionnaire was developed covering aspects of presentation, diagnosis, staging investigations, therapy, and follow-up. In January 1996, after approval by the Ethics Committee of the Anti-Cancer Council of Victoria, the questionnaire was sent to the treating doctors of all patients with testicular cancer in 1988-1993 identified in the cancer registry. Questionnaires were subsequently sent to other doctors identified as participants in the patients' care. Neither the doctors' nor the patients' names were identified to third parties. At the end of the survey, the data were entered into a database and a de-identified file was used for analysis. Classification of testicular cancer Germ cell tumours were classified as seminoma or non-seminoma. Seminoma included only those cases with pure seminoma in the histological specimens and a normal serum α-fetoprotein level. Cases where any other histological component (such as embryonal carcinoma, yolk sac or endodermal sinus tumour, choriocarcinoma, or teratoma) was identified, or in which the serum α-fetoprotein level was elevated, were classified as non-seminoma. Analysis Analysis was conducted using the SPSS statistical package8 for descriptive statistics and χ2 tests for significance of associations for categorical data. Survival analysis was performed using the RelSurv Version 2.09 software package. First, relative survival proportions were computed separately for each prognostic indicator. This method adjusts the survival proportions for the other causes of mortality associated with age using Victorian life-tables. Second, a multivariate analysis (proportional hazards regression) was performed, with variables entered stepwise into the regression model in descending order of significance in the univariate analyses. Improvement to the model was tested after each addition and the variable was removed if the improvement was not significant (P < 0.05). Quality of care In an attempt to assess the quality of care provided, a number of performance measures were developed retrospectively by the working group after review of the initial data analysis. All patients were expected to have had preoperative measurement of serum α-fetoprotein and human chorionic gonadotropin levels. Adequate staging investigations were considered to be computed tomography (CT) scans of the abdomen and chest (or chest x-ray) for non-seminoma and Stage II and III seminoma. Staging of Stage I seminoma was expected to include the above or a lymphangiogram instead of the CT scan. Adequate surveillance for Stage I non-seminoma was deemed to be at least six measurements of tumour marker levels and four CT scans of the abdomen in the first 12 months of follow-up. We arbitrarily defined these performance measures, but we believe they represent reasonable standards of care at the time of the study. Results There were 689 patients with testicular cancer identified in the Victorian Cancer Registry for the six-year period 1988 to 1993. There was a trend of increasing incidence, with the age-standardised incidence per 100 000 men for the six consecutive years increasing from 3.7 to 5.3, with an average annual rate of increase of 7.5%. Twenty-two patients were excluded for reasons including residence outside Victoria, diagnosis outside the study period, and incorrect diagnosis. Questionnaires were sent to the treating doctor for all 667 eligible patients, and replies were received for 633 (94.9% response rate). Patient characteristics Patient characteristics are shown in Box 1. There was a slight predominance of seminoma. In most patients, disease was confined to the testis at diagnosis; non-seminomas were more likely to present at an advanced stage. There were 64 (10%) patients with cryptorchidism and 2 (0.3%) with bilateral tumours at presentation. Location of care The 633 patients were treated at 93 hospitals. Eighty-six of these hospitals treated fewer than five patients with testicular cancer per year, and 71 treated fewer than two per year. One hundred and seventeen (19.4%) patients were managed exclusively at centres that treated fewer than five patients with testicular cancer per year, and 111 patients (17.5%) were treated exclusively in private practice. Initial investigation Staging investigations at diagnosis included abdominal CT scan in 591 men (93.4%), thoracic CT in 270 (42.7%), chest x-ray in 496 (78.4%), and lymphangiography in 234 (37.0%). There was a time trend for less use of lymphangiography (P = 0.02) and greater use of CT scanning of the chest (P = 0.005) during the period of the survey. An abdominal CT scan or lymphangiogram was not performed in 26 patients (4.1%), and a chest CT scan or chest x-ray was not performed in 33 (5.2%). Serum tumour markers were not measured at diagnosis in 16 patients (2.5%) and were recorded as "don't know" in 44 (7.0%). Management Initial management of the patients is shown in Box 2. Thirty-four patients with seminoma and 103 patients with non-seminoma were managed with surveillance. No patients were managed with primary retroperitoneal lymph node dissection for Stage I or II non-seminoma. Thirty-one (4.9%) patients did not receive initial chemotherapy or radiation therapy and did not appear to enter an active surveillance program. There was considerable variation in the follow-up of patients on surveillance, which was inadequate for many patients according to our criteria. For patients with seminoma, investigations during the first year of surveillance did not include an abdominal CT scan in 12 (36%), did not include a chest x-ray or chest CT scan in 17 (55%), and included fewer than four serum tumour marker measurements in 25 (76%). Similarly, for patients with non-seminoma, no abdominal CT scan was performed in the first year in 6 (6%), no chest imaging in 31 (30%), and fewer than four measurements of serum marker levels in 26 (26%). Three of the 33 patients with seminoma managed with surveillance relapsed, and one of these patients died. Twenty-seven of 102 patients with Stage I non-seminoma managed with surveillance relapsed (one of whom died), with a median time to relapse of nine months. Overall relative survival for these 102 patients was 99%. Survival Median follow-up was 87 months. The five-year relative survival proportion for all men was 95% (95% CI, 94%-97%). There were significant differences in survival according to histology, age, and stage. These factors remained significant in multivariate analysis (Box 3). There was no significant difference in survival by date of diagnosis (1988-1990 v 1991-1993) or location of treatment (categorised according to number of patients seen). The relative survival according to tumour type and stage is shown in Box 4. The relative survival at five years for patients with seminoma (all stages) was 99% (95% CI, 98%-100%). For patients with non-seminoma, relative survival at five years was Stage I, 98% (95% CI, 97%-100%); Stage II, 92% (95% CI, 84%-100%); Stage III, 69% (95% CI, 56%-82%). The overall relative survival for Stage II and III patients was 84% (95% CI, 76%-91%). Quality of care Only 68% of patients completed all the quality-of-care measures that we considered appropriate (Box 5). There was a correlation between the successful completion of all measures and the number of cases treated at an institution (P < 0.001; data not shown). Discussion Many published series from major cancer treatment centres internationally have demonstrated that excellent outcomes are now the expected result of testicular cancer management. However, less detailed information is available in population-based series. Our survey was designed to assess management patterns and results in the population of Victoria. The 94.9% response rate to the questionnaire ensured that the results are representative of the entire community. Studies in Scotland,10 Norway11 and New York12 have indicated higher survival proportions for patients treated in centres treating a greater number of cases of testicular cancer compared with patients treated in a "community" setting. A recently published clinical trial performed in 49 European centres found the relative risk of death was 1.85 (95% CI, 1.16-3.03) for patients with metastatic poor-prognosis non-seminoma treated at centres entering fewer than five patients on the study compared with those treated at centres entering a greater number.13 We found that management in Australia has been dispersed increasingly to many individual clinicians and hospitals. In this series, testicular cancer was managed in 93 centres; 71 of these centres treated fewer than two patients per year, and 86 fewer than five per year. Yet, we did not find a significant difference in survival between these centres and ones that treated more patients. Our results show a pleasing 95% relative survival of testicular cancer patients when compared with a matched cohort from the same population. Survival in other reported series includes a 95.7% relative survival in the United States Surveillance, Epidemiology and End Results (SEER) review for a similar time period,14 and a 99% cancer-related survival in a community survey in southern Norway.11 The stage at diagnosis and survival outcomes in the SEER review are very similar to ours. The 69% relative survival at five years for Stage III non-seminoma is lower than expected. However, it is difficult to compare this figure with published results, as not all details of prognostic factors4 are available. The 7.5% per year increase in incidence over the six-year period of the survey is consistent with a continuation of the threefold increase in Victoria between 1950 and 1980.15 Similar dramatic increases in incidence have been reported in other Western countries,1 but the cause remains uncertain.16 Survival figures alone are an inadequate assessment of treatment. The aim of management is to achieve cure with the least morbidity. Patients who are initially managed poorly may be cured by salvage therapy, but at the price of greater toxicity. The resulting morbidity may be severe and potentially includes infertility, sexual dysfunction, and an increased risk of secondary malignancy. Our survey identified a number of problems. There was a wide variation in patterns of practice, particularly in patients with early-stage disease and those undergoing surveillance. Thirty-two per cent of patients failed to complete five simple measures of quality of care. More complex measures of quality of care, such as the extent of disease at relapse after initial surveillance or the quality and type of radiation therapy and chemotherapy, were not assessed in this survey. Surveillance after orchidectomy of Stage I seminoma, as an alternative to radiation therapy, was initially described in 1988,17 but we did not consider it to be accepted management during the period of the survey. Thus, we were surprised by the number of Stage I seminoma patients managed by surveillance. These patients were managed at multiple centres and there is no evidence that they were entered into a clinical trial. We were particularly concerned about inappropriate schedules for follow-up when surveillance was initiated. Published results of surveillance programs have emphasised the need for strict and frequent follow-up and investigation to ensure that any relapse is detected at an early stage.18 There are no published data to indicate less rigorous programs of surveillance maintain the same proportion of cure. Many patients with Stage I disease were also lost to follow-up or refused therapy. Young men are often not well motivated to continue regular and extensive follow-up. Therefore, it is essential to ensure such patients are fully informed about the options for therapy and the potential risk of unnecessary morbidity and mortality if an appropriate program of management is not followed. We believe the development and implementation of management guidelines for testicular cancer in our community may help to reduce the problems we identified. Acknowledgements We gratefully acknowledge the support of the Anti-Cancer Council of Victoria, the Urological Study Committee of the Victorian Co-operative Oncology Group, the Victorian Cancer Registry, and the urologists and oncologists who participated in the survey and completed the questionnaires. References McKiernan JM, Goluboff ET, Liberson GL, et al. Rising risk of testicular cancer by birth cohort in the United States from 1973 to 1995. J Urol 1999; 162: 361-363. Bosl GJ, Motzer RJ. Testicular germ-cell cancer. N Engl J Med 1997; 337: 242-253. Bosl GJ, Geller NL, Cirrincione C, et al. Serum tumor markers in patients with metastatic germ cell tumors of the testis. A 10-year experience. Am J Med 1983; 75: 29-35. International Germ Cell Cancer Collaborative Group. International Germ Cell Consensus Classification: a prognostic factor-based staging system for metastatic germ cell cancers. J Clin Oncol 1997; 15: 594-603. Read G, Stenning SP, Cullen MH, et al. Medical Research Council prospective study of surveillance for stage I testicular teratoma. Medical Research Council Testicular Tumors Working Party. J Clin Oncol 1992; 10: 1762-1768. Einhorn LH. Testicular cancer as a model for a curable neoplasm: the Richard and Hinda Rosenthal Foundation Award Lecture. Cancer Res 1981; 41: 3275-3280. Toner GC, Panicek DM, Heelan RT, et al. Adjunctive surgery after chemotherapy for nonseminomatous germ cell tumors: recommendations for patient selection. J Clin Oncol 1990; 8: 1683-1694. SPSS Inc. SPSS-X user's guide. 3rd ed. New York: McGraw-Hill; 1988. Hedelin G. RELSURV: a program for relative survival. Technical report of the Department of Epidemiology and Public Health, Faculty of Medicine. Strasbourg, France: Louis Pasteur University, 1995. Harding MJ, Paul J, Gillis CR, Kaye SB. Management of malignant teratoma: does referral to a specialist unit matter. Lancet 1993; 341: 999-1002. Hernes EH, Harstad K, Fossa SD. Changing incidence and delay of testicular cancer in southern Norway (1981-1992). Eur Urol 1996; 30: 349-357. Feuer EJ, Frey CM, Brawley OW, et al. After a treatment breakthrough: a comparison of trial and population-based data for advanced testicular cancer. J Clin Oncol 1994; 12: 368-377. Collette L, Sylvester RJ, Stenning SP, et al. Impact of the treating institution on survival of patients with "poor-prognosis" metastatic nonseminoma. European Organization for Research and Treatment of Cancer Genito-Urinary Tract Cancer Collaborative Group and the Medical Research Council Testicular Cancer Working Party. J Natl Cancer Inst 1999; 91: 839-846. Ries LAG, Kosary CL, Hankey BF, et al. SEER cancer statistics review, 1973-1996. Bethesda, MD: National Cancer Institute, 1999. Stone JM, Cruickshank DG, Sandeman TF, Matthews JP. Trebling of the incidence of testicular cancer in Victoria, Australia (1950-1985). Cancer 1991; 68: 211-219. Bergstrom R, Adami HO, Mohner M, et al. Increase in testicular cancer incidence in six European countries: a birth cohort phenomenon. J Natl Cancer Inst 1996; 88: 727-733. Horwich A, Peckham MJ. Surveillance after orchidectomy for clinical stage I germ-cell tumours of the testis. Prog Clin Biol Res 1988; 269: 471-479. Boyer MJ, Cox K, Tattersall MH, et al. Active surveillance after orchidectomy for nonseminomatous testicular germ cell tumors: late relapse may occur. Urology 1997; 50: 588-592. (Received 28 Jun, accepted 7 Nov, 2000) Authors' details Peter MacCallum Cancer Institute, Melbourne, VIC. Guy C Toner, MD, FRACP, Head, Medical Oncology; Thomas F Sandeman, MD ChB, FRACR, Radiation Oncologist. Geelong Hospital, Geelong, VIC. Greg J Neerhut, MB BS, FRACS(Urol), Urologist. Alfred Hospital, Melbourne, VIC. Max A Schwarz, MB BS(Hons), FRACP, Head, Medical Oncology Unit; Ross M Snow, MB BS, FRACS(Urol), Head, Urology Unit. Anti-Cancer Council of Victoria, Melbourne, VIC. Vicky J Thursfield, BSc, GradDip(Applied Stats), Information Manager, Cancer Epidemiology Centre; Graham G Giles, PhD, Director, Victorian Cancer Registry. Reprints will not be available from the authors. Correspondence: Associate Professor G C Toner, Head, Medical Oncology, Peter MacCallum Cancer Institute, St Andrew's Place, East Melbourne, VIC 3002. gtonerATpetermac.unimelb.edu.au Make a comment 1: Patient characteristics Seminoma Non-seminoma Total* Number of patients 357 271 633 Stage I (confined to testis) 295 (83%) 158 (58%) 457 (72%) II (retroperitoneal nodes) 50 (14%) 61 (23%) 112 (18%) III (more extensive) 12 (3%) 52 (19%) 64 (10%) Median age (range)† 35 (17-80) 27 (0-76) 32 (0-80) * Includes three patients with stromal tumours and two with "other" histology. † The difference in age between men with seminoma and non-seminoma was significant (P<0.001). Back to text 2: Principal initial management after orchidectomy Surveillance Radiation therapy Chemotherapy Other Seminoma (Stage) I (confined to testis) 33 239 6 17 IIa,b (RP nodes 1 25 6 1 IIc,d (RP nodes >5cm) 0 2 15 0 III (more extensive) 0 2 9 1 Non-seminoma (Stage) I (confined to testis) 102 5 39 12 II (RP nodes) 1 0 60 0 III (more extensive) 0 0 52 0 Patients who received more than one form of therapy (eg, radiation and chemotherapy) were classified to the most significant form of treatment. Patients in the "other" category were reported as lost to follow up, 7 (22.6%); refused therapy, 16 (51.6%); received other therapy, 6 (19.4%); and not stated, 3 (9.7%). RP nodes=retroperitoneal nodes. Back to text 3: Multivariate analyses of prognostic factors for survival Non-seminoma (n=266) All patients (n=633) 5-year survival RR (95% CI) P 5-year survival RR (95% CI) P Morphology Seminoma 99% 1.0 Non-seminoma 91% 6.2 (1.9-20.0) Other 65% 51.9 (8.5-319) Age 93% 1.0 96% 1.0 30-49 years 92% 1.4 (0.5-3.6) 0.039 98% 1.5 (0.6-3.7) 0.002 ≥50 years 61% 6.0 (1.9-19.3) 78% 9.8 (4.4-32.0) Stage I 98% 1.0 99% 1.0 II 92% 3.6 (0.8-17.0) 96% 2.4 (0.7-8.6) III 69% 15.3 (4.4-52.8) 73% 11.9 (4.4-32.0) Multivariate analysis of survival for seminoma was not appropriate given the small number of deaths in this group. RR=relative risk. Back to text Back to text 5: Measures of quality of care Measure Eligible Completed Preoperative measurement of tumour markers 633 584 (92%) Adequate staging investigations 633 595 (94%) Radiation therapy for Stage I seminoma 295 239 (81%) Appropriate surveillance for Stage I non-seminoma 102 22 (22%) Chemotherapy for Stage IId, III seminoma and Stage II, III non-seminoma 132 127 (96%) All of the appropriate measures completed in each case 633 430 (68%) Back to text

Guy C Toner · Greg J Neerhut · Max A Schwarz · Vicky J Thursfield · Thomas Sandeman

The Profession Should doctors wear white coats? Paul R Harnett MJA 2001; 174: 343-344 For editorial comment, see Van Der Weyden The wearing of white coats by hospital doctors is becoming a rarity, making it difficult for patients to identify doctors from other hospital staff. I asked patients with cancer whether they thought that doctors, both junior and senior, should wear white coats. Only a minority disapproved. Asking the patients - Patients' responses - Discussion - Acknowledgements - References - Authors' details Make a comment - - - More articles on Administration and health services - More articles on General medicine The white coat is a well established symbol of the medical profession. In more than 90 medical schools in the United States its symbolism is recognised by formal ceremonies at which students are awarded the "right" or distinction of wearing a white coat to emphasise the humanistic values of medicine.1,2 Among the many recent changes in the way medicine is practised, one little-studied change has been the addition of "white-coated" doctors to the "endangered species" list in many hospitals. Although some relatively isolated pockets of this species remain, their numbers continue to decline. In many hospitals they are now so rare as to prompt comment, if not amusement, when observed by chance "in the wild". Subtle societal changes have been proposed as a reason for the demise of white-coated doctors -- modern-day patients may find the use of a white coat to be an inappropriate status symbol or a barrier to effective communication. In particular, in paediatric practice it has been suggested that white coats may be an impediment, although studies have failed to confirm this.3,4 I investigated the phenomenon of the disappearing white coat doctors in my environment by asking oncology patients about their attitudes to doctors and white coats. I was prompted to do this by remarks from patients indicating that identification of doctors and nurses in hospitals was an issue of concern. Asking the patients Patients attending the oncology outpatient clinics of three Sydney hospitals in September and October 1999 completed a short questionnaire about whether junior and senior doctors in hospitals should wear white coats. Patients could "agree", "disagree", or respond "I don't care". They could give reasons for their choice either by agreeing with suggested options ("for identification purposes", "looks more professional" or "for reasons of hygiene"), or by free-text response. Patients' responses Questionnaires were returned by 180 patients (47 men, 113 women [20 did not specify their sex]; age range, < 20-90 years [32 did not give their age]). Only 153 patients responded to questions about senior doctors, which were on the reverse side of the page. Junior doctors 106 respondents (59%; 95% CI, 52.8%-67.2%) agreed that junior doctors should wear white coats; 17 disagreed (9%; 95% CI, 4.8%-13.2%), and 57 patients (32%; 95% CI, 25.2%-38.8%) responded " I don't care". The same general pattern was seen for both men and women. Reasons for agreeing (46 respondents): "for identification purposes"(41); "looks more professional" (36); and "reasons of hygiene" (6). Reasons for disagreeing (31 respondents): white coat a barrier to communication between doctor and patient (10); wearing a coat was a choice for the individual doctor (6); and avoiding additional cost. Senior doctors 61 respondents (40%; 95% CI, 32.2%-47.8%) answered that senior doctors should wear white coats; 32 disagreed (21%; 95% CI, 14.5%-27.5%) and 60 (39%; 95% CI, 31.3%-46.7%) said "I don't care". Reasons for agreeing (28 respondents): white coats look more professional or assist in identification (22); and hygiene (3). Reasons for disagreeing (34 respondents): not wearing white coats distinguishes senior doctors from junior doctors, and white coats distance doctors from patients (11). Patients who preferred junior doctors to wear white coats were also likely to prefer senior doctors in white coats. The proportion of patients disagreeing with doctors' wearing white coats reduced significantly with advancing patient age (P = 0.01, by logistic regression analysis). Discussion A clear majority of respondents thought that junior doctors should wear white coats (only 9% disagreed), but the situation was less clear for senior doctors, with roughly equal numbers of respondents either agreeing that senior doctors should wear white coats or indicating they didn't care. Our results contrast with the current practice in many Australian hospitals (including those I work in) where junior doctors no longer wear white coats. Many respondents who supported the wearing of white coats for junior and senior doctors agreed that white coats looked more professional or assisted in identification. Interestingly, reasons related to infection control or hygiene were uncommon. Of the 20% of respondents who disagreed with white coats for senior doctors, most referred to white coats as being some sort of a barrier. As with all questionnaires, the data are potentially subject to bias. The number of questionnaires distributed was not recorded, making it impossible to exclude bias from an unbalanced sample. Patients who declined to complete the survey may have had a preference for, or against, the wearing of white coats. However, anecdotal evidence from hospital staff is that patient refusal to participate was rare. That the doctors at the hospitals surveyed do not wear white coats could influence patient responses. Perhaps patients did not express stronger support for senior doctors' wearing white coats for fear of displeasing their attending doctor. If so, our data would underestimate patients' preferences for senior doctors' wearing white coats. Many patients raised the issue of white coats with their doctor after completing the questionnaire. This feedback indicates that identification of hospital staff is a significant problem for patients, magnified by degrees of debility and dependence. Patients often have difficulty distinguishing the relative roles of staff wearing corporate uniforms (ie, nurses, clerical staff, kitchen staff, etc); name tags were difficult to read and insufficient for identification. Similar studies of patients' views of non-medical apparel in hospitals may be instructive. If, from the patients' perspective, the greatest usefulness of the white coat is its value as a rapid means of staff identification, then to simply recommend reintroduction of white coats may be premature, especially if an alternative, effective means of staff identification can be implemented. I do not believe that the results would be greatly different in other Australian settings, and recent reports from international clinical settings suggest qualitatively similar results.5,6 Our findings confirm that the clinical environment and medical interaction are complex, and that individual patients hold differing views. However, our data do not suggest that the clinical environment has changed to one that favours extinction of white-coated doctors. Acknowledgements Thanks are due to colleagues in the Department of Medical Oncology and Palliative Care at Westmead, Blacktown and Nepean Hospitals. The data management input by Shoma Barat is appreciated. Special thanks are due to the patients who offered their views for this study. Competing interests: None. References Branch WT Jr. Deconstructing the white coat [editorial]. Ann Intern Med 1998; 129: 740-742. Wear D. On white coats and professional development: the formal and the hidden curricula. Ann Intern Med 1998; 129: 734-737. McCarthy JJ, McCarthy MC, Eilert RE. Children's and parents' visual perception of physicians. Clin Pediatr (Phila) 1999; 38: 145-152. Matsui D, Cho M, Rieder MJ. Physicians' attire as perceived by young children and their parents: the myth of the white coat syndrome. Pediatr Emerg Care 1998; 14: 198-201. Menahem S, Shvartzman P. Is our appearance important to our patients? Fam Pract 1998; 15: 391-397. Ikusaka M, Kamegai M, Sunaga T, et al. Patients' attitudes toward consultations by a physician without a white coat in Japan. Intern Med 1999; 38: 533-536. Authors' details Department of Medical Oncology and Palliative Care, Westmead Hospital, Sydney, NSW Paul R Harnett, MB BS, FRACP, PhD, Director of Cancer Services, Westmead and Nepean Hospitals. Reprints will not be available from the author. Correspondence: Dr P R Harnett, Department of Medical Oncology and Palliative Care, Westmead Hospital, Westmead, NSW 2145. harnettprATwestgate.wh.usyd.edu.au Make a comment * Only 153 responses were received for the questions about senior doctors. Back to text

Paul R Harnett

Viewpoint Physician-assisted suicide in Oregon: why so few occurrences? Howard Wineberg MJA 2001; 174: 353-354 In the first three years that physician-assisted suicide (PAS) has been legal in Oregon, about two persons per month have taken medication to end their life. Most physicians are unwilling to prescribe the lethal medication. Because many terminally ill people are confined to their bed or home, the difficulty of finding a willing physician may have resulted in many abandoning the idea of using PAS. People living a long way from a large urban centre may be severely disadvantaged in their ability to obtain medication to end their lives. Legal requirements - Oregon - Using PAS - Not using PAS - Conclusion - References - Authors' details - - More articles on Ethics Physician-assisted suicide (PAS) has been legal in the state of Oregon, USA, for more than three years. The Oregon Health Division, the agency responsible for monitoring Oregon's Death with Dignity Act, has produced official reports documenting the number and characteristics of Oregonians who have used the provisions in the law to take medication to end their lives in the years 1998-2000.1-3 Only 70 people legally took medication to hasten their death during this period — in this article I examine some of the reasons why. Legal requirements Under Oregon's law,4 only Oregon residents who are aged 18 or over and have a terminal illness with a life expectancy of less than six months are eligible to request a prescription for medication to end their life. Patients must self-administer (swallow) the medication — euthanasia (involving a physician's active intervention) is not allowed. The physician's prognosis for the patient must be confirmed by a consulting physician and both physicians must determine that the patient is capable of making his or her own decision and does not have a mental health condition that impairs his or her judgement. The patient must make two oral and one written requests for the medication, and at least 15 days must elapse between the first and the final request. Physicians are not obligated to participate in PAS. Population and geography of Oregon Oregon occupies 96 000 square miles and borders the Pacific Ocean. Of the 3.3 million residents of Oregon, 1.4 million live in the Portland Metropolitan Area.5 Many of Oregon's counties are sparsely populated — half of the counties are east of the Cascade Range, yet this area contains only one city of at least 20 000 people.5 Many people living in eastern Oregon are a 5-7-hour drive from Portland or another metropolitan area. The number of people using PAS In the first three years of operation of Oregon's law (1998-2000), 70 people legally took medication to end their lives. Although the number increased from 16 in year one to 27 in years two and three, only about two people per month, on average, are using PAS,1 and only 0.07% of Oregonians die of PAS in a year.6 By comparison, around 545 people per month die of cancer in Oregon.7While the United States and the Netherlands are not directly comparable to one another, it is noteworthy that in the Netherlands about 2.4% of all deaths are from euthanasia and 0.3% from PAS.8 In Oregon, approximately 10% of all requests for PAS result in the person taking the medication to end their life, whereas in the Netherlands about a third of the requests result in death by euthanasia or PAS.9,10 In Australia, despite threats that physicians participating in euthanasia would be legally culpable, in the nine months that euthanasia was legal in the Northern Territory (from July 1996 to March 1997) four people died by this means,11 representing 0.7% of all deaths in the Northern Territory during that time.12 I must emphasise that my focus is on legal PAS in Oregon. Illegal PAS and euthanasia probably still occur, although their extent is unknown. Reasons for terminally ill people not using PAS One reason why so few people may have taken medication to end their lives is that an individual must wait at least 15 days after making the initial request to receive medication. This makes it difficult for people to use PAS on a whim. Terminally ill individuals have mentioned that the process one needs to go through to finally receive the medication can be quite burdensome (because of the waiting period, the need to get two physicians to confirm that the patient has a life expectancy of less than six months, and the requirement that the patient must get two people to witness the written request for the lethal medication).9 Approximately 30% of patients requesting a prescription may die before completing the requirements of the law.9Recent improvements in palliative care in Oregon13 may have allowed some people to die in relative comfort without having to hasten their death. Oregon has one of the highest rates of hospice admission and morphine usage per capita in the United States.14 It is estimated that about 45% of the patients for whom a substantive intervention is made will change their minds about wanting a prescription for a lethal medication.9 Just knowing that they can receive medication to end their life, if desired, is comfort enough for some individuals15 — the Oregon Health Division statistics show that 19 people who eventually died of their underlying illness had had the medication but not used it.1 Some people may be unable to swallow the medication themselves and thus can not use PAS.4 Others may fear that even if they swallow all the medication it may not kill them,16 and thus they do not attempt PAS. The above factors notwithstanding, I believe the most important reason for the limited use of PAS is that this service is not readily accessible to many Oregonians. Approximately 60% of the patients had to go to more than one physician before finding one who would write the prescription.1 The Oregon Health Division reported that only a fifth of physicians of control patients dying of similar terminal illnesses would have prescribed a lethal medication if asked,2 and, in a study of Oregon physicians, Ganzini et al9 found that only 16% of those asked actually wrote a prescription. The Veterans Affairs system, the Indian Health Services system and a major Catholic healthcare system do not allow PAS in their facilities.3 Some physicians willing to prescribe the medication have had difficulty finding a second physician to confirm the prognosis or a pharmacist willing to fill the prescription.9 Some physicians fear being penalised if they prescribe a lethal medication. In November 1997 Thomas Constantine, head of the US Drug Enforcement Administration, stated that if physicians prescribed a lethal medication to end a person's life it would be a violation of federal law.17 Then, in June 1998, United States Attorney General Janet Reno ruled that physicians in Oregon could legally write such a prescription.18 (This may be one reason why the number of people using PAS was higher in 1999 than in 1998.) However, the US Congress is now considering a bill that would make it illegal for physicians to prescribe a controlled substance to end a person's life. Consequently, some physicians may consider it risky to write such a prescription, particularly as to do so remains against professional guidelines.9,19 People residing in small cities, particularly those in eastern Oregon who are 75-400 miles from Portland or another metropolitan area, may be severely disadvantaged in getting access to PAS. Ganzini et al9 found that it was extremely rare for physicians practising in areas of fewer than 25 000 people (94% of Oregon's cities have fewer than 25 000 people5) to prescribe the medication. Physicians who do not have a large population base from which to draw patients may fear that if they participate in assisted suicide they could be the target of demonstrations outside their homes and offices, similar to those directed at physicians who perform abortions.20 Two-thirds of the physicians writing a lethal prescription expressed concern about reporting the fact to the Oregon Health Division.9 Because many terminally ill people are confined to their bed or home, the difficulty of finding a physician to prescribe the medication may result in many people abandoning the idea of using PAS as an end-of-life option. Conclusion Although PAS is legal for terminally ill people in Oregon, relatively few appear to have used medication to end their lives. The process of finding a physician willing to write the prescription, fulfilling all the legal requirements and finally receiving the medication can be time consuming, and perseverence is required. Once diagnosed with a serious or terminal illness, patients should probably start looking for a physician who is willing to prescribe a lethal medication if necessary. Without adequate planning for the possibility of using PAS, some terminally ill Oregonians may be unable to take medication to hasten their death. References Oregon's Death with Dignity Act: three years of legalized physician-assisted suicide. Portland, OR: Oregon Health Division, 22 February 2001. Sullivan AD, Hedberg K, Fleming DW. Legalized physician-assisted suicide in Oregon - the second year's experience. N Engl J Med 2000; 342: 598-604. Chin AE, Hedberg K, Higginson GK, Fleming DW. Oregon's Death with Dignity Act: the first year's experience. Portland, OR: Oregon Health Division, 18 February 1999. Haley K, Lee M, editors. The Oregon Death With Dignity Act - a guidebook for health care providers. Portland, OR: Oregon Health Sciences University, 1998. Wineberg H. Population estimates for Oregon: July 1, 1997. Portland, OR: Center for Population Research and Census, 1998. Center for Health Statistics. Oregon vital statistics annual report, 1997. Volume 2: Mortality. Portland, OR: Oregon Health Division, 2000. Center for Health Statistics. Oregon vital statistics county data 1998. Portland: Oregon Health Division, 2000. Van der Maas PJ, van der Wal G, Haverkate I, et al. Euthanasia, physician-assisted suicide, and other medical practices involving the end of life in the Netherlands, 1990-1995. N Engl J Med 1996; 335: 1699-1705. Ganzini L, Nelson HD, Schmidt TA, et al. Physicians' experiences with the Oregon Death with Dignity Act. N Engl J Med 2000; 342: 557-563. Van der Maas PJ, Van Delden JJM, Pijnenborg L, Looman CW. Euthanasia and other medical decisions concerning the end of life. Lancet 1991; 338: 669-674. Kissane DW, Street A, Nitschke P. Seven deaths in Darwin: case studies under the rights of the Terminally Ill Act, Northern Territory, Australia. Lancet 1998; 352: 1097-1102. Australian Bureau of Statistics. Deaths, Australia. Canberra: ABS, 1996, 1997. (Catalogue No. 3302.0). <http://www.abs.gov.au>. Lee MA, Tolle SW. Oregon's assisted suicide vote: the silver lining. Ann Intern Med 1996; 124: 267-269. Tolle SW. Care of the dying: clinical and financial lessons from the Oregon experience. Ann Intern Med 1998; 128: 567-568. Muskin PR. The request to die: role for a psychodynamic perspective on physician-assisted suicide. JAMA 1998; 279: 323-328. Groenewoud JH, Van der Heide A, Onwuteaka-Philipsen BD, et al. Clinical problems with the performance of euthanasia and physician-assisted suicide in the Netherlands. N Engl J Med 2000; 342: 551-556. Hill GK, Barnett J. Push is on to assess DEA clout on suicide. Oregonian November 13, 1997: D1, D5. Hogan D. Bills blocking assisted suicide continue to move. Oregonian July 25, 1998: B1, B3. Lee MA, Ganzini L, Brummel-Smith K. When patients ask about assisted suicide: a viewpoint from Oregon. West J Med 1996; 165: 205-208. Lee MA, Tolle SW. Oregon plans to legalise suicide assisted by a doctor: how much more open will the practice become? BMJ 1995; 310: 613-615. Authors' details Portland, Oregon, USA. Howard Wineberg, PhD, Private consultant. Reprints will not be available from the author. Correspondence: Dr H Wineberg, 1513 SE Oak Street, Portland, Oregon, USA 97214-1454. wineberghAThotmail.com Make a comment

Howard Wineberg