Webb and Repatriation Commission - [2001] AATA 633

DECISION AND REASONS FOR DECISION [2001] AATA 633

ADMINISTRATIVE APPEALS TRIBUNAL )

) No V1998/1350

VETERAN'S APPEALS DIVISION          )
           Re      GLORIA WEBB
  Applicant
           And    REPATRIATION COMMISSION
  Respondent

DECISION

Tribunal Mr J. Handley, Senior Member Mr A. Argent, Member Dr A. Shanahan, Member

Date22 June 2001
PlaceMelbourne

Decision      The decision under review is set aside and in substitution IT IS DECIDED that the death of Andrew Rupert Webb was war-caused.
  ...........Sgd. Mr J. Handley.............
  Senior Member
CATCHWORDS
Veteran's Affairs – Operational Service – smoking by service conceded – death from non Hodgkins lymphoma – whether related to smoking – decision set aside.

REASONS FOR DECISION

22 June 2001 Mr J. Handley, Senior Member Mr A. Argent, Member Dr A. Shanahan, Member

This application has a long history.
The claim is brought by Gloria Jean Webb who is the widow of the late Andrew Rupert Webb who died on 17 May 1985 when he was then aged 60 years. The cause of death was certified to be-
"lobar pneumonia – 2 days
Histiocytic infiltration of reticuloendothelial system – 7 months".
The application was made to the respondent on 15 January 1992 for a pension. The respondent decided on 12 June 1992 that the death of Andrew Rupert Webb (deceased) was not war-caused. An appeal against that decision was lodged with the Veterans Review Board on 8 June 1993 and on 19 October 1994 the VRB affirmed the decision made by the respondent.
Mrs Webb then instituted proceedings in this Tribunal by application V1994/1139. The application was contested and the review was heard on 14 June 1996. On 22 November 1996 the Tribunal decided that the decision under review should be affirmed.
The applicant's solicitor subsequently lodged an appeal in the Federal Court (vg 761 of 1996) and on 24 October 1997 Finklestein J decided that the appeal be allowed, the decision of this Tribunal be set aside and the application be remitted to the Tribunal for hearing by a different member. (1997 1130 FCA).
The respondent lodged an appeal against the decision of Finklestein J in the Federal Court and on 5 November 1998 Tamberlin, Finn and Marshall JJ decided that the appeal be dismissed and the application be remitted for determination by a differently constituted Tribunal. (1998 1411 FCA).
The present decision arises out of a hearing, which was conducted on 18 & 19 October 2000. Having regard to the date of the primary decision by the respondent, Statements of Principles are not applicable and the application is to be decided by reference to s.8(1)(i) and s.120(i) and (iii) of the Veteran's Entitlements Act 1986.
An issue, which emerged in the previous AAT hearing, was the controversy as to the cause of death. Indeed the Tribunal decided, having regard to the T-documents lodged, the exhibits tendered during evidence and the medical evidence given from doctors at the hearing, that "no positive diagnosis of the disease which caused the death of Mr Webb has ever been established". Slides of diseased cells were sent to Great Britain and the United States in an attempt to identify the cause of death and the consensus seems to have been that the deceased suffered from non-Hodgkin's lymphoma (nHl).
Evidence was given at this hearing that the deceased took up cigarette smoking during service, which became a lifelong habit. The hypothesis pursued was that there was a relationship between cigarette smoking and nHl.
In preparation for the hearing, which has given rise to this decision, the relevant slides taken during autopsy were reviewed by a number of specialists. Further opinions were obtained and it was agreed between the parties that the deceased did suffer from nHl. For the purposes of these proceedings the respondent conceded that there was an association between service and cigarette smoking. The respondent did not concede the association between smoking and nHl.
At the commencement of this hearing Mr DeMarchi on behalf of Mrs Webb raised another hypothesis namely an association between malaria and nHl. It was said that the deceased suffered from malaria during service and subsequently developed nHl.
Mr Hanks appeared on behalf of the respondent. A number of documents were received into evidence and will be referred to in these reasons. Evidence was given by Mrs Webb, Doctor Parkin, Professor Fox and Professor Peach.
The veteran was born on 2 May 1925 and enlisted in the AIF a few days after his 18th birthday, on 17 May 1943. He was trained as a signals linesman and served in New Guinea from 14 October 1944 to 8 July 1945. During his Army service he was admitted to hospital with dengue fever 12-24 April 1945 and malaria 20-31 August 1945, 27 February – 5 March 1946, 30 May – 5 June 1946 and 31 August – 7 September 1946. He was discharged on 13 November 1946.
Gloria Jean Webb
Mrs Webb is the widow of the late Andrew Rupbert Webb. Some of her evidence concerned the deceased's smoking habits, which were no longer in issue because of the concession made by the respondent.
Mrs Webb said that her husband suffered from malaria in service and continued to suffer from it after he was discharged. She said he had a "yellow" appearance and was consuming atebrin, an anti-malaria medication. Mrs Webb said that her husband consumed this medication for some years after discharge. She recalled that he would suffer from high temperature associated with malaria and that she would comfort him by cold sponging.
In 1964 or 1965 Mrs Webb said that her husband had a skin cancer removed from his lower lip by Doctor Graham, his local General Practitioner. Mrs Webb associated the cancer with the consumption of rolled cigarettes. She said that her husband developed a "blister" on his lower lip, which was believed to have an association with cigarette paper adhering to his lip.
John Desmond Parkin
Doctor Parkin is the Director of Laboratory Medicine at the Austin Repatriation Medical Centre. He has held that position for five years. Previously he was the Chairman of Pathology at the Repatriation General Hospital and visiting haematologist at the Austin Hospital. He has been a specialist haematologist for 30 years.
Doctor Parkin provided a report dated 29 November 1999 received into evidence as Exhibit B. The report largely summarised research papers concerned with whether there was any association between nHl and smoking. At page 3 of the report Doctor Parkin reported-
"There are a number of papers which record a statistical association between smoking and nHl. Conversely there are papers which have failed to show such an association.
Negative studies do not necessarily detract from the hypothesis that smoking is associated with nHl. NHL is a heterogeneous group of disorders and it is likely that different study populations will contain a different spectrum of nHL. For example the study of Herrington (1998) establishes an association between smoking and follicular lymphoma. Follicular Lymphoma is a NHL, which is often of small cell type. Conversely the negative study of Nelson (1997) relates to high and medium grade lymphomas which would not usually include small cell nHl.
Furthermore the environment will differ between studies in ways which may be difficult in the causation of nHl but which are presently not capable of being identified. For example the study of Brown (1992) selects white males in Iowa and shows a positive association of nHl with smoking. The study of McLaughlan (1995) involves US veterans and shows a negative association of nHl with smoking".
Doctor Parkin then identified 13 studies of which he said 6 were "positive" and 7 were "negative". He concluded at page 6-
"1. The ex member had a disorder with an unresolved diagnosis. Many of the features of this disorder are consistent with nHl.
2. The ex member commenced smoking during service.
3. Thirteen studies are presented in this review which examine the relationship between nHl and smoking. These studies vary in the spectrum of nHl, which is studied.
4. Six studies demonstrate a significant association between nHl and smoking.
5. A reasonable hypothesis can be made that there is a causal relationship between nHl and smoking.
6. The epidemiological data presented does not establish scientific proof of a causal relationship between nHl and smoking".
In evidence Doctor Parkin said that he was aware of a report prepared by a Professor Peach of 9 October 2000 and a report of Doctor Buchanan of 25 September 2000. Doctor Buchanan is an anatomical pathologist at the Royal Melbourne Hospital and who, in conjunction with Doctor Byron Collins observed slides prepared following autopsy which were stained "using the immunoperoxidase technique" which was apparently not available in 1985. Doctor Buchanan reported that the slides showed that the deceased suffered an intravascular lymphoma which he thought was more likely to be a T cell lymphoma but the immunoperoxidase staining suggested a histiocytic lymphoma. Doctor Buchanan reported that the only way to firmly establish the diagnosis was by molecular biology (genetic studies).
Professor Peach reported (in response to the report of Doctor Buchanan), in terms of the epidemiological evidence for a causal relationship between smoking and nHl that little was to be achieved by differentiating between a T cell or a hystiocytic lymphoma.
Doctor Parkin agreed with this analysis. He said there was no precise classification of nHl and distinguishing between a T-cell or hystiocytic lymphoma would achieve little in the interpretation of the epidemiological studies.
Doctor Parkin said that subsequent to his report of 29 November 1999 he had identified another six studies examining the association between nHl and smoking. He said therefore that he was aware of 19 studies, which had examined this association, some of which examined it specifically, and others which examined it with a number of other variables. He concluded that 10 of the 19 studies found a relationship between smoking and nHl or an association between a sub-group of nHl and smoking. He also said that several of the studies show a 'dose response relationship'. Fifty per cent of the studies (the remaining 9) he said failed to show a relationship between smoking and nHl. He thought that the papers published subsequent to his report of 29 November 1999 were similar in nature to the papers, which existed at the time of this report, and the conclusions were also similar.
Having regard to the 19 studies which he was aware of, Doctor Parkin said that they do not "scientifically resolve the question of the relationship between smoking and nHl". In fact he said that the epidemiological studies "were never going to be able to resolve this relationship because lymphoma is still inadequately classified and inadequately understood". Nonetheless he said "if the object of the exercise is to establish a reasonable hypothesis which to me means something that is reasonable and worthy of further scientific investigation, then I think this would be the basis for a reasonable hypothesis".
In response to specific questions from Mr DeMarchi, Doctor Parkin said that the "positive" studies did "point to" the association between smoking and nHl. He said that 50% of the studies reached that conclusion and where persons were heavy smokers, the risk of nHl was more substantial (dose relationship). In terms of an analysis of the sub-groups being analysed, he said an association was drawn between smoking and follicular lymphoma. Whilst acknowledging that a reasonable hypothesis did exist Doctor Parkin said that that standard did not constitute scientific proof.
In further discussion concerning the association between smoking and follicular lymphoma Doctor Parkin said that where there has been a "translocation" in the blood, of smokers, "this translocation has a very, very high association with follicular lymphoma which in several studies has been shown to be associated with smoking. So, I mean, again, this isn't scientific proof, but it is far from a fanciful relationship".
By reason of what Doctor Parkin described as the "inadequate" classifications of nHl, it was his opinion that the conclusions reached by Doctor Buchanan, namely that the deceased either suffered a "T-cell or a histiocytic lymphoma" came within the umbrella of nHl.
In cross-examination Doctor Parkin was taken to a report prepared by Doctor Byron Collins, a consultant forensic pathologist dated 3 November 1999. In that report Doctor Collins said that "it is apparent the current consensus of opinion is that there is an increased incidence of follicular lymphoma in addition to other conditions affecting the haemopoietic and lymphopoietic systems in individuals who are cigarette smokers. Unfortunately the late Mr Webb did not suffer from this particular type of lymphoma". When asked whether he agreed with that opinion Doctor Parkin said whilst acknowledging that there was evidence relating cigarette smoking to follicular lymphoma there was "no adequate classification of non-Hodgkin's lymphoma, so I think it is very hard to draw any very strong conclusions about sub-typing particularly as we don't know what type of lymphoma really that the ex-member had only that the ex-member had a non-Hodgkin's lymphoma". Doctor Parkin said he did not necessarily "agree or disagree with the opinion of Doctor Collins, because he said sub-typing and classifications were "quite unreliable". He said this was a proposition to which Professor Peach would agree and that most of the research material concerning nHl were based on "quite crude classifications".
Doctor Parkin was then taken to a report prepared by Professor Peach dated 8 May 2000. It transpired that Doctor Parkin had not ever read that report nor had a copy of it ever been provided to him. The report is detailed concerning survey methodology and criteria and a summary of a number of research papers. There was then discussion between the Tribunal and counsel and Dr Parkin as to whether an opportunity should be given to him to read the report before he continued in cross-examination. Doctor Parkin volunteered that he was a haematologist and not an epidemiologist (as was the qualification of Professor Peach) and that he thought it unreasonable to expect him only to be able to digest the report in a relatively short time. He said it was his practice when commenting upon epidemiological matters to consult with other epidemiologists. Doctor Parkin also said that the report of Professor Peach concerned issues of research design and design flaws, which he understood Professor Peach to report as affecting the survey outcomes.
We decided to stand down for 20 minutes to permit Doctor Parkin to peruse the report. On resumption Doctor Parkin said that he "didn't have a strong feeling that it (the report) addressed a causal relationship". He said it appeared that Professor Peach had referred to a particular survey methodology (Austin Bradford Hill). Doctor Parkin said he was familiar with this methodology and had read textbooks written by Bradford Hill. Doctor Parkin said that he agreed with Professor Peach that the studies involving population and hospital based control studies were an appropriate epidemiological study and were a common basis for epidemiological research.
The witness also acknowledged that Professor Peach had reported that of 17 case studies, 8 had a "positive" outcome and 9 had a "negative outcome". He also acknowledged that Professor Peach had reported that with respect to the 8 positive studies there were shortcomings in the design of the study. He also noted that the negative studies were "poorly controlled" and he drew no conclusions as to the design flaws in the negative studies. Doctor Parkin therefore was of the opinion that the conclusions reached by Professor Peach were "incomplete" because "he has only addressed in the summary the design failures of the positive studies". Doctor Parkin said that had Professor Peach made a balanced summary of the research material he would have referred to the design failures of both the positive and negative studies.
In re-examination Doctor Parkin said that whilst accepting some of the validity of the criticism of Professor Peach of the positive studies, his opinion as to the reasonableness of the hypothesis between smoking and nHl had not altered. He said that 50% of the studies "published in the scientific literature in referee journals show an association between smoking and non-Hodgkin's lymphoma". He said he acknowledged that Professor Peach had argued that there was a design fault in the studies referred to, but he thought that "all of the studies have design faults".
With respect to the hypothesis connecting malaria with nHl, Doctor Parkin had answered questions in cross-examination concerning a report that he wrote in 1994 associating malaria with African Burkitt's lymphoma. In that report he said that the deceased did not suffer from that disorder and so far as he was aware this was the only link to malaria. In answers to questions from Mr DeMarchi in re-examination Doctor Parkins said that Burkitt's lymphoma was "not common in non African environments and in general terms there was no association between that type of lymphoma and malaria accept in tropical and sub-tropical Africa".
Richard Fox
Professor Fox is the Director of the Clinical Haematology and Medical Oncology Departments at the Royal Melbourne Hospital. He provided two reports dated 11 May 1999 and 10 May 2000. He provided reports for the previous proceedings and gave evidence in the previous proceedings.
With respect to the hypothesis advanced in this application as to a connection between service, malaria and nHl, Professor Fox interpreted a number of research papers as establishing no "link". He said that the presence of nHl in Victoria is the highest incidence in the world and that malaria was one of the most common diseases in the world. He said there has been extensive research data and no link was found. He said the potential association has been studied "extremely carefully" but an association between nHl and malaria has not been established. The closest association involves malaria and Burkitt's lymphoma, which he described as "childhood non Hodgkins lymphoma". He said there was an "aetiological role" between malaria and Burkitt's lymphoma and the association between malaria and Burkitt's lymphoma had given rise to other studies where an association between malaria and nHl had been explored. No such connection, he said, in the literature existed.
With respect to the hypothesis concerning smoking and nHl, Professor Fox agreed with the evidence of Doctor Parkin that with follicular lymphoma there is a chromosomal translocation, with respect to T-14 – 18 which was described by Professor Fox as being associated with a genetic abnormality.
With respect to the evidence of Doctor Parkin that an association was found in approximately 50% of the applicable research papers between smoking and nHl, Professor Fox agreed that some of the papers did raise a positive association between smoking and the risk of follicular lymphoma, but they did not find an association between increased duration or intensity of smoking. Professor Fox did not agree with the methodology adopted by Doctor Parkin in his interpretation of the research data but he did note that the application brought by the widow in these proceedings did not concern an association between smoking and follicular lymphoma.
Professor Fox said that he was aware of a research paper authored by Herrington Freedman, which did show an increased risk of nHl in persons who did smoke for a period of between 30 and 39 years. However he interpreted that study as indicating that men over the age of 55 had a reduced risk of nHl. Professor Fox referred to a study by Zahm who, he said- "has done so much work in non-Hodgkin's lymphoma". He interpreted a study conducted by Zahm of 1177 cases of nHl where it was reported that there was "no association between non-Hodgkin's lymphoma and tobacco use". He regarded that report as being a "negative study". Professor Fox also referred to a report published by McLaughlin reviewing the incidence of cancer mortalities in United States veterans, who were smokers in 1995. Professor Fox noted that the report detailed 638 deaths from nHl but with no increased risk with respect to persons who did smoke.

With respect to other material Professor Fox had reviewed and researched he concluded that there was "no association between cigarette smoking and nHl with the possible exceptions that are not easy to put together on the paper with the follicular low grade non Hodgkins lymphoma".
In cross-examination by Mr DeMarchi Professor Fox agreed that the nomenclature for nHl was not precise and "can be extremely difficult in some cases". Professor Fox agreed that the deceased died from a histiocytic nHl and possibly a T cell nHl, but that it was not a follicular lymphoma. He agreed that a histiocytic nHl fitted within the nomenclature of nHl. He also agreed that the research studies examining the connection between smoking and nHl did not, in the majority, deal with classification.
Professor Fox had not read the report of Doctor Parkin, who recorded that a number of studies did show an association between smoking and nHl. Whilst he was dismissive of some studies, which on a statistical basis showed no risk or increased risk of nHl in persons who did smoke, he agreed that some of the reports referred to by Doctor Parkin did show a "statistically significant increase in nHl" in the case of persons who did smoke. He disagreed however that "nearly ½ of those studies" showed an association between smoking and nHl, because of the interpretation that he (Professor Fox) placed on the data and an interpretation of the data.
When asked to comment on the alleged failure of Professor Peach to critically analyse the research studies which did not support an association between smoking and nHl, Professor Fox said that it did not appear to him that Professor Peach had found anything that was "wrong" with the studies, that he (Professor Fox) thought that the studies were "appropriate" but he deferred to the opinion of Professor Peach who he regarded as an "expert epidemiologist". It was his opinion that on balance the research studies which supported the hypothesis of connecting smoking with nHl were those concerning men under the age of 50 with follicular lymphoma.
Mr DeMarchi took Professor Fox to a study published by Namboodiri and Harris in 1991 ('Hematopoietic and Lymphoproliferative cancer among male veterans using the veterans Administrative Medical system' – refer p. 218 of Exhibit A) which showed a "clear increasing trend in non-Hodgkin's lymphoma risk with age was apparent with a median age of onset of 60 years The rates for the veterans were significantly higher than those for the general population for ages younger than 65". Professor Fox agreed that the deceased died at the age of 60 and whilst he agreed that the Namboodiri survey was of veterans of varying age groups from Vietnam, Korea, World War II and World War I, he said one of the tables concluded that the "relative risks of haemopoietic and lymphoproliferative neoplasms were highest amongst veterans 45 years of age or younger, Vietnam era". He said there was a discussion with smoking but it was linked only to leukaemia.
Professor Fox and Mr DeMarchi then entered into a long discussion with respect to the malaria hypothesis. The discussion extended to chromosomal cell changes, immuno-suppression and renal disease. That discussion may be usefully summarised by the following extract from the transcript at page 72 and 73-
"If non-hodgkin lymphoma is caused by the arrangement of chromosomes that particular process may be a factor in the development of this veterans non-Hodgkin's lymphoma?
Mr DeMarchi I can't see any possible link whatsoever in what you are talking about".
The study I have just drawn your attention to Professor, doesn't it indicate that paracitic infection by malaria ….. ?
Which study is this?
The one that I have just ….. ?
Abu Shakra or Wosencraft?
Abu Shakra. Does it indicate that malaria can act as a trigger?
Trigger for what?
Auto immune activity?
Certainly yes. But what has that got to do with non Hodgkin's lymphoma?
You see no connection at all?
None
No possible connection?
No".

Hedley Peach

Professor Peach is the Professor of Public Health at the University of Melbourne. He holds a PHd in epidemiology, he has written two books on epidemiology and has been the author of in excess of 70 research studies concerning epidemiology.
Professor Peach prepared two reports dated 17 May 1999 and 8 May 2000. He also completed a handwritten letter, which was forwarded to the Tribunal shortly before the commencement of the hearing on 9 October 2000. Both reports and the letter were received into evidence.
Professor Peach described epidemiology as a comparison of groups of people and – for the purposes of the present case – identifying whether any relationship existed between smoking and nHl. Significantly he said epidemiology concerned the interpretation of data with respect to humans and associations in humans as opposed to research concerning animals. In some cases he said research data can permit conclusions to be drawn as to a causal relationship subject to the interpretation of the data.
Professor Peach said there were a number of methods adopted by epidemiologists in the collection of data. He referred to case control studies (which are either population based or hospital based or cohort studies). He referred to risk being measured by a numeric ratio and a protocol to determine whether a particular result could have been achieved by "chance". Professor Fox also referred to a methodology, which he said is universally accepted, devised by Bradford Hill who he described as "arguably one of the greatest statisticians in the world". He said the criteria developed by Bradford Hill involved consideration of chance, bias, consistency and biological plausibility being applied to test research studies.
Professor Peach was asked to comment on a part of the evidence of Dr Parkin who said his interpretation of relevant data revealed that 50% of the studies showed an association between smoking and nHl and 50%, which showed that there was no association. Professor Peach said that epidemiological evaluation of studies prohibited going 'straight to whether the results were positive or negative'. He said there needed to be a systematic examination of the studies to exclude those, which were "unreliable". To do so he said the Bradford Hill criteria is applied.
Fortunately Doctor Parkin and Professor Peach each refer to the same 19 research studies. Professor Peach said that having analysed these studies and then having excluded studies (which he regarded as being "unreliable") he then analysed the remaining studies, which he regarded as being "reliable". (A number of the studies which will be referred to in this decision are identified in an Appendix to the reports of Professor Peach of 17 May 1999 and 8 May 2000).
With respect to a report completed by Doll and Peto entitled "Mortality in relation to smoking: 20 years observations on male British Doctors", Professor Peach said that that report could be excluded, although it was a negative study. He said the report was deficient. In his report of 8 May 2000 Professor Peach said that this study had weaknesses because it excluded women, there was a bias and living persons were excluded.
Studies completed by Rogot and Murray entitled "Smoking and Causes of Death among US Veterans: 16 years of Observation" and a report of McLaughlin entitled "Smoking and cancer mortality among US Veterans : 26 year follow up," were described by Professor Peach as being the "same studies". Nonetheless he said that these reports should also be excluded because they failed to record persons who gave up smoking during the survey period and that it was biased in ascertaining deaths amongst smokers "in favour of an association between smoking and non-Hodgkin's lymphoma".
With respect to a study by Linet entitled "Is cigarette smoking a risk factor for nHl or multiple myeloma", Professor Peach said that this study did find a positive association between smoking and nHl but should be excluded because it was biased in favour of that association and "living cases" were excluded.
A study, he said, by Herrington and Friedman entitled "Cigarette smoking and risk of non-Hodgkin's lymphoma sub-types" should also, he said, be excluded because the association was found to be positive. However, he said the authors admitted that the result could have been by chance.
A study by Adami et al entitled "Smoking and the risk of leukemia lymphoma and multiple lymphoma" was regarded by him as a negative study but should be excluded because a change of smoking status was not taken into account. This study and the studies referred to earlier, were described as being "prospective studies". Professor Peach then analysed population based case control and hospital based case control studies. He said these types of studies were distinguished by surveys of members of the general public and therefore being a sample of "healthy people" (population based case control) whereas hospital based case control studies necessarily involved persons who are based in hospitals and who are not "healthy people".
With respect to the population based case control studies Professor Peach excluded a survey conducted by Williams and Horm entitled "Association of cancer sites with Tobacco and alcohol consumption and socio economic status of parents" because he regarded that survey as being biased against finding a relationship. He said a study by Paffenbarger Wing and Hyde entitled "Characteristics in youth predictive of adult onset malignant lymphomas, melanomas and leukemia's" should be excluded because it involved a survey of 80 patients only which he said was too small. A study by Brown et al entitled "Smoking and risk of non-Hodgkin's lymphoma and multiple myeloma" he also said should be excluded because although it found a positive association the data was collected by 'proxy' with respect to patients who had died. He described this as being a process of obtaining information from the living relatives of persons who had died. He said when the information collected by proxy was analysed "the positive result disappeared; it became non significant".
Professor Peach said in evidence that a study by Hoar et al entitled "Agricultural Herbicide Use and risk of Lymphoma and Soft Tissue Sarcoma" found no association, no bias and that study was "okay". However, in his report of 8 May 2000, Professor Peach said that there were weaknesses in the survey because of the non-inclusion of women and because the recall of tobacco consumption and diagnostic criteria had not been stated.
A study by Persson et al entitled "Some occupational exposures as risk factors for malignant lymphomas" should be excluded because of flaws in the study design, and because there was an emphasis on associating chemicals in the environment rather than smoking with cancers. A study by Nelson et al entitled "Alcohol, tobacco and recreational drug use and the risk of non-Hodgkin's lymphoma" was a study that he regarded as being reasonable and that "weight" could be put on it. He said there were adequate controls, there were a large number of persons considered and the diagnosis of cancers was based on histology. The weaknesses however in the survey were the recall of tobacco consumption. Nevertheless, he said the report did not show any statistically significant association between smoking and nHl.
A study by Miligi entitled "Occupational environmental and lifestyle factors associated with the risk of "hematolymphopoietec malignancies in women" should be excluded according to Professor Peach because it involved a study of women only, and there is – according to the witness – biological evidence suggesting that any relationship between smoking and nHl will differ between men and women. He said the survey did find a positive association but it did not examine evidence of causation.
A study by Freidman et al entitled "Relationship of cigarette smoking to non-Hodgkin's lymphoma among middle aged men" found a positive association between smoking and nHl but only in persons under the age of 45 and then for heavy smokers.
Professor Peach then discussed his conclusions with respect to hospital based control studies that he interpreted. A report by Cartright entitled "Non-Hodgkin's Lymphoma – case control epidemiological study in Yorkshire", he said was biased in favour of patients who had smoking related diseases. He thought that little emphasis could be placed on that report. A report by Franceschi and Tevani entitled 'The epidemiology of non-Hodgkin's lymphoma in the north east of Italy – a hospital base case controlled study' was also discounted because it apparently initially found a positive association between smoking and nHl but when the study was repeated and a greater number of cases and controls were included, the association "disappeared". He said "It's a positive study that became negative". A survey by Siemiatycki et al entitled "Associations between cigarette smoking and each of 21 types of cancer" produced a mixed result being a survey of both population and hospital control groups. He said the survey initially found a negative result, but later found a positive result when using another set of controls. Professor Peach concluded, "I think on the whole it did find a positive association". However he said from an epidemiological point of view, the report did not "go on to explore whether the association was causal or not". Professor Peach referred to a report by Stefani et al entitled "Tobacco, alcohol diet and risk of nHl" which was open to question because of the type of tobacco and type of cigarettes smoked by the persons who were surveyed. The survey was in South America and whilst it apparently concluded that there was an association between smoking and nHl, Professor Peach concluded "we have to ask questions about tobacco in Uruguay – I mean, how relevant is it?".
In his report of 8 May 2000 Professor Peach reported also on the "combined analysis of three population based case control studies" prepared by Zahm and Waddell. Professor Peach reported that these surveys found no association in men of any age group and when these surveys also analysed the survey conducted by Brown (refer earlier) it was said that the survey which was initially positive "became negative when they excluded the proxies".
Professor Peach also referred to other papers which were filed shortly prior to the commencement of the hearing and which were not included within his reports. One was a report by Holly et al, and another by Rachole and McGrath both being described by him as population based control studies which did not contain any study design faults and which he said produced a negative result.
Another report, by Baris and Zahm entitled "Epidemiology of Lymphomas" was submitted shortly prior to the commencement of the hearing. This report was published in the month before the hearing in a Journal entitled 'Current Opinion in Oncology'. It is a literature survey and significantly – he said for the purposes of this case – it refers to a survey by Perrson (refer earlier). Professor Peach said it is reported that Perrson has now completed two surveys, has combined the results and unlike the reported outcome of the first survey – he has now reported no association between smoking and nHl.
In the conclusion to his report of 8 May 2000, Professor Peach (page 12 – paragraph 5 reported)-
"The hypothesis that smoking is causally associated with nHl has been tested in the prospective population base case control studies and hospital base case control studies, the weight of evidence is against the hypothesis. The strongest evidence in favour of an association between smoking and nHl is indicative of an effect of heavy smoking among men under 50 years of age".
In his evidence at the Tribunal, Professor Peach reaffirmed that opinion and said he had no reason to modify it. He said on the balance of the evidence available from the reports that he has referred to earlier, most of the studies had found no association. He added "and certainly none of those studies have produced any convincing evidence of causation".
With respect to the alternative hypothesis advanced by the applicant, namely an association between malaria and nHl, Professor Peach said that he was aware that the deceased had suffered from malaria and that there had been a report prepared by Franceschi demonstrating "a significant association" with a past history of chronic infectious diseases including malaria. However, a survey by Cartright did not find that association, a survey by Tevani did find an association, and a more recent paper prepared by Tevani found the association only with respect to children who suffered malaria under the age of 10 years. Professor Peach said the survey by Franceschi was probably unreliable because the diagnosis of malaria was not checked against medical records and the deceased suffered from malaria as an adult.
In cross-examination Professor Peach said that many studies in epidemiology are "prone to biases from the very nature of the design" and the responsibility of an epidemiologist was to design a study without bias. In effect he said some studies had not been designed carefully. Additionally he said that studies and surveys could have defects because of unanticipated poor response rates or by reason of problems, which may occur during the research process.
Whilst acknowledging that the deceased did die from nHl he said that it was important for the Tribunal to "dissect the type of nHl for the purposes of the validity of the studies". He said that the working formulation of nHl was determined on the basis of morphology whereas a number of epidemiological studies on nHl were based on a diagnosis having its origin in the patient's histology. Nonetheless he acknowledged that a number of the studies in issue in these proceedings did show a "statistically significant" association between smoking and nHl. Professor Peach said that the results of those studies were unlikely to have been "due to chance", excepting studies where persons have "multiple comparisons".
Professor Peach discounted the Namboodiri and Harris study showing Veterans having a 20% higher incidence of nHl than the general population up to the age of 60 because although the veteran died from nHl at the age of 60, the association was reduced because he was aged 60. That is to say, the association became "less and less up until the age of 60 so he's really on that borderline". Additionally, he said that veterans had access to medical care, which was not available to the general population. This was in addition to the witness discounting veterans having successfully passed a medical examination prior to enlistment as evidence of good health. He said this was known as the "healthy worker effect" which he equated with a "healthy good soldier effect" where studies have shown that the effects of good health wear off after five or ten years following examination and those persons are then no different to other members of the public. Professor Peach explained that the likelihood of veterans having nHl diagnosed (compared to other members of the public) is because of regular medical surveillance.
With respect to the study by Barris and Zahm, Professor Peach said that the research concluded that there was either "no or a weak association" between smoking and nHl. He said that for a risk or an association to be statistically significant, a factor of two is universally used by epidemiologists to measure the significance or importance of risk. Whilst the Barris and Zahm study did not completely exclude the association between cigarette smoking and nHl, the witness said that the conclusions were not statistically significant or important. The witness excluded the Herrington and Friedman study, because the result achieved was by chance and it principally concerned the relationship between smoking and follicular lymphoma. The Linet study was also discounted because of defects in the research methodology, namely only persons who had died from nHl were the subject of research. He said the research design had an inherent bias of finding an association between smoking and nHl because persons who had not died were not surveyed. Nonetheless Professor Peach agreed that the survey was of a large population over many years and the deceased veteran in the present application was in a statistically significant risk group of heavy smokers. Whilst he agreed that a person in the heavy smoking category had a 300% more chance of death from nHl, Professor Peach was again quick to point out that the survey was flawed because of faults in the survey methodology. He suggested that a result of this type would not have been achieved had living smokers also been surveyed.

The Adami study was discounted even though on its face it produced a negative result. The witness discounted it because persons within the survey period had become non-smokers and there was therefore a bias in favour of a negative outcome.
Professor Peach continued to explain why the surveys were largely unreliable. This was because of either design faults or bias or results being achieved by chance. The observations he made were with respect to surveys, which were regarded as being both positive and negative. He also explained that some of the adopted definitions differed and used as an example the survey of Brown who defined a smoker as being a person who had consumed one cigarette per day for three months. Other surveys he said had different smoking definitions.
Whilst acknowledging that there were some surveys which on their face appeared to be positive, he dismissed the suggestion by Mr DeMarchi that a reasonable hypothesis exists because "none of the positive studies have actually demonstrated the cause; you know, a cause and effect relationship".
When asked to explain what he understood to be a reasonable hypothesis Professor Peach said-
"That it is supported by studies which are well designed; that there is an association which is significant and that the association shows features which would support or suggest a cause and effect relationship between smoking and non hodgkins lymphoma; in particular that it is a strong association, that is, is consistent between studies and within studies that there is a dose response relationship that the more you smoke the greater your risk and that it is biologically plausible".
With respect to the hypothesis connecting malaria and nHl, Professor Peach said that he was satisfied beyond reasonable doubt there is no such association. Epidemiologically, the only support for the proposition, he said, was in the case of persons who were exposed to malaria as a child. He said there was no material known by him demonstrating the deceased having malaria as a child. He said the survey by Cartright was deficient because medical records were not verified to confirm that the persons surveyed in fact did suffer from malaria. The Tevani survey, he said, was based on patients where their recall of having suffered malaria was not "tested" and the evidence of an association between persons who suffered malaria as a child and nHl could not be translated or transposed to the general population.
In re-examination Professor Peach said that on the balance of the studies, he could not support the hypothesis between smoking and nHl as reasonable. He said having excluded the "negative studies", the risks which emerged from the "positive studies" were "small, weak, relative risks".
Additionally he said there was no biological or epidemiological evidence to support the hypothesis.
Conclusion & Reasons For Decision
In East v Repatriation Commission 1987 74 ALR 518 at 533 – 534 (adopting and approving a decision of the Veterans Review Board in Stacey and subsequently a decision of this Tribunal in Re Dell and Repatriation Commission 1986 9 ALD 596, the Court decided-
"The addition of the word "reasonable" would however seem to imply that what is required is more than a mere hypothesis. In the opinion of the Board to be reasonable a hypothesis must possess some degree of acceptability or credibility – it must not be obviously fanciful, impossible, incredible or not tenable or too remote or too tenuous. For a reasonable hypothesis to be "raised" by material before the board we think it must find some support in that material – that is the material must point to; and not merely leave open the hypothesis as a reasonable hypothesis. At the same time however, a hypothesis may be reasonable without having been proved (either on the balance of probability or beyond reasonable doubt) to be correct as a matter of fact. Were it otherwise it would no longer be a hypothesis but would have been elevated to some higher status. Accordingly a connection asserted by a hypothesis to exist between death or incapacity and service may still be reasonable even though theoretical and it may be theoretical in either or both of at least two senses; by postulating a known medical fact but in circumstances not known to have definitely existed in the instant case or by postulating a medical principle which science is not yet able to definitely prove, but is unable to describe as unreasonable".
The High Court in Bushell v Repatriation Commission 1992 109 ALR 30 at page 35 said-
"…. the case must be rare where it can be said that a hypothesis based on the raised facts is unreasonable when it is put forward by a medical practitioner who is eminent in the relevant field of knowledge. Conflict with other medical opinions is not sufficient to reject a hypothesis as unreasonable. As we have earlier pointed out it is not the function of s.120(3) to require the Commission to choose between competing hypothesis or to determine whether one medical or scientific opinion is to be preferred to another. This does not mean however that in performing its functions under s.120(3) the Commission cannot have regard to the medical or scientific material, which is opposed to the material, which supports the veterans claim. Indeed the Commission is bound to have regard to the opposing material for the purpose of examining the validity of the reasoning, which supports the claim that there is a connection between the incapacity or death and the service of a veteran. But it is vital that the Commission keep in mind that that hypothesis may still be reasonable although it is unproved and opposed to the weight of informed opinion".
A Full Federal Court in Repatriation Commission v Bey 1997 79 FCR 372-3 reaffirmed and restated the status of East and Bushell and Byrnes v Repatriation Commission 1993 177 CLR 564.
In Owens v Repatriation Commission 1996 22 AAR 121, the Federal Court by majority decided that evidentiary material which tended to disprove a hypothesis would not of itself be sufficient to make the hypothesis untenable; proof to a level of probability connecting a medical condition and service was not a requirement of a reasonable hypothesis; an unproven theory supported by facts raised upon the material would be sufficient to constitute a hypothesis and a hypothesis will not be reasonable if it conflicts with a scientific principle or fact widely recognised as being true.
In the present application we had the benefit of evidence from Dr Parkin who is a prominent haematologist and from Professor Peach who is a prominent epidemiologist. Both considered the same research material and reached different conclusions as to whether a hypothesis was raised. Dr Parkin acknowledged that the opinions advanced by him could not be established at a scientific level or indeed upon the balance of probabilities.
We note at paragraph 23 of these reasons that Dr Parkin concluded that a reasonable hypothesis was something, which he regarded as being worthy of scientific investigation. This may appear to demonstrate that Dr Parkin is not aware of what constitutes a reasonable hypothesis for the purposes of the Veterans Entitlements legislation. However, that is a decision for us to make. We note from his evidence that 6 of the 13 studies investigated by him demonstrated what he regarded as a "significant association between nHl and smoking". He also concluded that a "reasonable hypothesis can be made that there is a causal relationship between nHl and smoking". (paragraph 18).
Professor Peach concluded that there was no reasonable hypothesis supporting a connection between smoking and nHl. He said the "weight of evidence is against the hypothesis". In reference to the research studies discussed by him in his evidence, he concluded, "none of those studies have produced any convincing evidence of causation" (paragraph 64 & 65).
Again whilst we acknowledge the assistance Professor Peach gave us in the interpretation of the research data, whether there is material which points to a connection between cigarette smoking and nHl is a matter for us to decide. In reaching that conclusion we will not adopt a standard of having to apply "convincing evidence of causation".
In making the above comments we are not choosing between competing medical theories but rather demonstrating the manner in which each doctor approached the issue of whether a reasonable hypothesis exists. In deciding whether there is material raised which points to an association between smoking and nHl, we are mindful that the hypothesis may be unproved and may be "opposed to the weight of informed opinion" (Re Bushell). Additionally we are not persuaded that any preference should be given to the evidence of Professor Peach because he is an epidemiologist whereas Doctor Parkin is a haematologist. We regard both witnesses as being equally eminent and qualified to give the opinion that they did, yet at the same time recognising that each witness comes from a different area of medical speciality.
We thought the evidence of Professor Fox was of assistance in comprehending whether a reasonable hypothesis existed between smoking and nHl particularly because he, on a number of occasions, referred to the interpretation that an informed person would put upon the research material which was analysed by both Doctor Parkin and Professor Peach. We agree with that observation. It did seem to us that both Doctor Parkin and Professor Peach placed different emphasis and weight upon the research material in reaching the conclusions that they did. Professor Fox was candid enough to acknowledge that some of the research material relied upon by Doctor Parkin showed a "statistically significant increase in nHl in the case of persons who did smoke" (paragraph 41) yet he dismissed the studies because of the interpretation that he placed on the data.
We have concluded that the material raised does point to a reasonable hypothesis between smoking and nHl. We acknowledge the evidence of Professor Peach that at an epidemiological level the research studies may be flawed by factors of bias or chance or demography or gender or other features associated with design of the research methodology.
We could not exclude the studies of Linet entitled "Is cigarette smoking a risk factor for nHl or multiple myeloma" or the study by Brown entitled "Smoking and risk of non-Hodgkin's lymphoma and multiple myeloma". These studies (as examples) did show an association between smoking and nHl but which were dismissed by Professor Peach because of features peculiar to the design methodology or other factors referred to above. Nonetheless an association between smoking and nHl was demonstrated. We acknowledge that the relationship is probably a "medical principle which science is not yet able to definitely prove" (refer East) and it may also be "unproved and opposed to the weight of informed opinion" (refer Bushell). Nonetheless, we could not conclude that the hypothesis is obviously "fanciful, impossible, incredible or not tenable or too remote or too tenuous". We are satisfied that these studies alone demonstrate a degree of "acceptability or credibility" (refer East).
With great respect to Professor Peach we acknowledge that the particular standard that he thought needed to be achieved at an epidemiological level was one of scientific probability which at a scientific or medical level would be a sound basis to review and conclude upon the research data. For the purposes of the Veterans' Entitlements legislation no such standard exists.
Having regard to the concession made by the respondent that a connection does exist between service and smoking, we are satisfied pursuant to s.120(3) of the Veterans Entitlements Act that on the whole of the material before us, a reasonable hypothesis has been raised connecting nHl with service.
Pursuant to Section 120(1) of the Act, we are satisfied the deceased was engaged in operational service, that his death was war-caused and we are satisfied beyond reasonable doubt that there are sufficient grounds for making this determination.
It follows therefore that the decision under review should be set aside.
Insofar as the applicant raised a new hypothesis at this hearing, namely an association between service and malaria and nHl we are satisfied that the material does not raise a reasonable hypothesis.
We note that Dr Parkin dismissed the association having found that the only material known to him associating malaria with nHl was a particular type of lymphoma known as African Burkitt's lymphoma. He dismissed the association because that type of lymphoma is not common in non-African environments and there was nothing to suggest to him that the deceased had ever had any association with that type of lymphoma nor had ever been in tropical or sub-tropical Africa. Professor Fox dismissed the hypothesis with some vigour (refer paragraph 44). Professor Peach also dismissed the hypothesis, on the basis of research material and concluded that the only association that he knew between nHl and malaria was in the case of persons who suffer malaria under the age of 10 years. (refer also paragraph 76).
This hypothesis is regarded by us as being fanciful, impossible, incredible or not tenable or too remote or to tenuous. It does not posses any degree of acceptability or credibility. We dismiss the hypothesis under s.120(3) because on the whole of the material before us a reasonable hypothesis has not been raised connecting malaria with nHl.
I certify that the 102 preceding paragraphs are a true copy of the reasons for the decision herein of Mr J. Handley, Senior Member
Signed:         ...C. Irons...................................................
  Secretary
Date/s of Hearing  10 August 2000 & 18 October 2000
Date of Decision  22 June 2001
Counsel for the Applicant        Mr D. DeMarchi
Solicitor for the Applicant
Counsel for the Respondent    Ms J. Proimos
Solicitor for the Respondent