Crown Research Institute’s publication on nicotine in tobacco inadequate
Publication of research data subsequently found to be flawed—notice of research results to interested parties—adequacy of subsequent retraction—inclusion of contextual material with media release—relationship between an Ombudsman's jurisdiction and issues involving scientific techniques and the course of research—matter of administration—Ombudsmen Act 1975, s 13
In 1996 and 1997 a Crown Research Institute carried out a number of tests on the constituents of New Zealand tobacco products, including measuring the nicotine content of cigarettes. Initial results were published in May 1997, although further work was planned. Subsequent testing suggested that the initial results overestimated nicotine content by a factor of 1.47 and a retraction was published in November 1997.
The complainant expressed concern that publication of erroneous test results was very damaging to the industry. It noted that the May 1997 paper attracted considerable attention and that the Institute had issued a media release to publicise the results. The complainant also advised that the Institute did not provide the industry with a copy of the article for comment prior to publication.
Before commencing the investigation, there was an issue of jurisdiction to address. The Institute submitted that "it is inappropriate for the Ombudsmen to consider complaints concerning matters of science rather than administration". Section 13 of the Ombudsmen Act 1975 provides that it is a function of the Ombudsmen to investigate complaints about actions or decisions ‘relating to a matter of administration.’
The view was taken that in a constitutional context, functions are of three kinds: legislative, judicial, or administrative. Where the functions are not legislative or judicial, it follows that they will be administrative, and an organisation performing such functions will be acting, for the purposes of the Ombudsmen Act, in relation to a matter of administration. In any event, scientific technique or the rationale behind the research was not being questioned. The issues related to the timing of publication, the notice given to the industry, the adequacy of the retraction, and the content of a media release made at the time of publication. These are administrative actions.
There were four main grounds of complaint. The first alleged that the publication of the test results in May 1997 was unreasonable at that time because concerns about the validity of the data had been expressed, and a review had been commenced but not concluded. The complainant drew attention to a number of documents which, in its view, suggested that the authors of the paper held concerns about the accuracy of the test results prior to publication.
In response, the Institute stated that at the time the article was published, neither the authors nor the reviewers had any reason to believe that the test data was flawed. The Institute noted that the process of scientific understanding relies as much on refuting old research as it does on conducting new work. It observed that this does not reflect on the quality of the original work, but is indicative of the gradual expansion of knowledge. With respect to the documents identified by the complainant, the Institute stated that the writers were not casting doubt on the work already undertaken but were identifying steps which could be taken to evaluate the results further.
The decision on when to publish research data involves the exercise of professional judgement. It is not the role of an Ombudsman to determine when in the life of a research project data should be published. In order to sustain this ground of complaint, it would need to be shown that the Institute had exercised its professional judgement in an administratively unreasonable manner. As there was no evidence that, at the time of publication, it was known the results overstated nicotine content, inquiries into this ground of complaint were discontinued.
The second ground of complaint contended that it was unreasonable for the Institute to have published the test results without first providing the industry with a copy of the article in advance, particularly in circumstances where there was risk to the legitimate interests of the industry parties.
In this case, the combined effect of the high nicotine measurements and probability of damage to the industry arising from publication of the results suggested that advance notice would have been appropriate. Whilst the Institute drew attention to concerns it had that legal proceedings would be used to stop or delay publication of the research, the counter view that it would be unreasonable not to give advance notice to affected third parties also had to be weighed. In particular, there was a concern that affected parties might otherwise lose the right to have legitimate concerns considered by a Court. Accordingly, the view was formed that the industry should have been provided with a copy of the article prior to publication.
The third ground of complaint questioned the adequacy of the Institute's retraction published in November 1997. The Institute submitted that considerable care was taken to explain the nature and extent of the flaws in the original article. The retraction clearly stated that nicotine content and concentration had been overestimated. Similar procedures were adopted for both articles, in that they were published in the same journal and a media release was made in each case. Accordingly, the view was formed that the further retraction sought by the complainant was not justified.
The fourth ground of complaint arose from the Institute's choice of analytical procedures. The industry test methods set out in the Smoke-free Environments Act 1990 require that smoke yield tests be used as the basis of measurements of nicotine content. However, the Institute's research measured nicotine content as obtained by solvent extraction. The question thus arose of whether it would have been reasonable to include with the media release a statement explaining any factors such as the analytical method, which might influence the conclusions drawn. This being particularly pertinent when a media release is made to a non expert audience.
In this case, the view was formed that it would have been reasonable for the Institute to have concluded its audience was comprised largely of non experts and to have included with the media release a statement which drew attention to the differing methodologies, and any other factors which might influence the conclusions drawn. Those receiving the media release would then have been in a position to treat the results with appropriate caution or seek further advice before acting on them.
The remedy sought by the complainant related to the methodology and processes adopted by the Institute in future testing of the nicotine content of cigarettes. No such recommendation was found to be necessary or appropriate. The findings of the investigation did not give rise to concerns about systemic issues that might warrant a recommendation to change Institute procedures.
This case note is published under the authority of the Ombudsmen Rules 1989. It sets out an Ombudsman’s view on the facts of a particular case. It should not be taken as establishing any legal precedent that would bind an Ombudsman in future.