One News – Gisborne pathologist – misdiagnosis – inaccurate interpretation of statistics; unfair to pathologist
Standard G1 – not inaccurate – no uphold
Standard G4 – not unfair to report the errors – no uphold
This headnote does not form part of the decision.
The inquiry into the misdiagnosis of cervical smears in Gisborne was the subject of a report on One News broadcast on TV One on 7 January 2000 between 6.00–7.00pm. The item reported that the pathologist’s error rate was 86%.
Stuart Slater complained to Television New Zealand Ltd, the broadcaster, that the report was inaccurate and unfair to the pathologist. In his view, an attempt should have been made to provide a neutral, informed commentary against the allegations made.
TVNZ responded that its report accurately represented the figures released by the Health Funding Authority and were attributed to it. It therefore declined to uphold the complaint that the report was inaccurate. Further, TVNZ said it did not consider itself to be in a position to judge whether the figures were unfair to the pathologist.
Dissatisfied with TVNZ’s decision, Mr Slater referred the complaint to the Broadcasting Standards Authority under s.8(1)(a) of the Broadcasting Act 1989.
For the reasons given below, the Authority declines to uphold the complaint.
The members of the Authority have viewed a tape of the item complained about and have read the correspondence which is listed in the Appendix. On this occasion, the Authority determines the complaint without a formal hearing.
A retired pathologist from Gisborne was the focus of an inquiry after it was alleged that he had misdiagnosed cervical cancer smears. In an item broadcast on One News on 7 January 2000 between 6.00–7.00pm, it was reported that the Health Funding Authority had found that the pathologist’s error rate was 86%.
Stuart Slater complained to TVNZ that the report was inaccurate and was unfair to the named pathologist. He said that Australian pathologists asked to re-read the slides had found that of about 2000 slides so checked, the Gisborne pathologist had failed to diagnose abnormalities in 103 of the 780 cases where the Australians had alleged abnormalities were present.
Mr Slater contended that it was inaccurate and unfair to say that the pathologist’s error rate was 86%. Further, he noted, no effort had been made to provide a neutral (as opposed to balancing) informed commentary.
To illustrate this further, Mr Slater suggested that the Australian pathologists were likely to set their criterion for judgment very conservatively, knowing that they were being required to check on another’s work. As a result, he said, they would diagnose more abnormalities, and inevitably there would be a higher false positive rate. In his view, it was false and unfair to imply that the Australian results were correct, and that the Gisborne pathologist’s mistaken and discreditable. A comparison could only be made, he said, if the Australians managed to achieve fewer false negatives while not increasing the false positive rate.
Mr Slater asked TVNZ to correct the item and to apologise to the pathologist. He also suggested it hire a "good mathematical statistician" to assist it in preparing its news bulletins.
TVNZ assessed the complaint under standards G1 and G4 of the Television Code of Broadcasting Practice. Those standards require broadcasters:
G1 To be truthful and accurate on points of fact.
G4 To deal justly and fairly with any person taking part or referred to in any programme.
In TVNZ’s view it was not its job to rule on whether the method by which the 86% figure had been reached was statistically sound. Its role, it continued, was to report accurately and fairly the findings of the Health Funding Authority (HFA) which had been released that day. It reported that it had checked the source material and noted that the information given to viewers accurately reflected the figures released by the HFA. TVNZ pointed out to Mr Slater that no news service, no matter how large, had the luxury of having on its staff experts in all intellectual disciplines to assess independently the information supplied to it, particularly in the context of same-day turnaround news reports.
TVNZ said it accepted Mr Slater’s belief that the figures did not stand up when subjected to statistical scrutiny. However, it said it believed his argument was with the HFA rather than with TVNZ. In this case, it continued, TVNZ had only been the messenger, and those figures had been provided by the HFA. As far as standard G1 was concerned, TVNZ found no breach as the report accurately reflected the content of the figures released that day, and they were accurately attributed to the HFA.
With reference to standard G4, TVNZ said it did not consider itself to be in a position to judge independently whether the figures were unfair to the pathologist. As a further point, it noted that the HFA had not amended the figures since they were published and so it must be assumed that it was satisfied with the conclusions reached. It declined to uphold the complaint.
When Mr Slater referred the matter to the Authority, he began by taking issue with TVNZ’s defence that it was simply the messenger. In his view, that was a "feeble attitude to journalistic responsibility" because journalists had a responsibility to understand and question information given, especially when it was handed out from "reputable official sources".
TVNZ advised that it had no further comment to make.
The Authority begins by examining a full transcript of the item, which reads as follows:
Newsreader: The inquiry into the misdiagnosed cervical cancer smears in Gisborne has yielded more alarming results. Dr Michael Bottrill detected 103 high grade abnormalities after testing nearly 19 and a half thousand smears.
But since they’ve been retested by the Health Funding Authority 732 abnormalities have been found. That’s an 86% error rate, even higher than earlier detected. Breast specimens also read by Dr Bottrill are being rechecked with results expected by the end of the month.
Meanwhile the government’s announced it will fund legal representation for the women affected by the misreadings.
The item was accompanied by a graphic which repeated the exact numbers. It read:
Cervical Smear Tests
19475 so far read
103 detected by Dr Bottrill
732 detected by independent laboratory
86% error rate
Source: Health Funding Authority
There is some discrepancy between what Mr Slater said he heard and saw in the item and what was actually reported. First, the Authority notes, the item reported that the pathologist found only 103 abnormalities in total but that that the independent laboratory had found 732 in the 19,475 samples which had been retested. Mr Slater apparently understood the report to have said that the pathologist had failed to diagnose abnormalities in 103 cases. In fact, on the basis of the numbers cited in the report, he failed to diagnose abnormalities in 629 cases. If the number of true positives (as determined by the independent laboratory) was actually 732 out the total sample size of 19,475, then it is the Authority’s view that it was correct to say that as the pathologist failed to detect abnormalities in 629 of those cases, he therefore had an 86% error rate in relation to the positive samples identified.
The Authority acknowledges Mr Slater’s point that it was not made clear whether there were any false positives in the 732 positive results which the independent laboratory found. It accepts that if there were false positives in that group, then it would be incorrect to state that the pathologist’s error rate was 86%.
However, on the basis that the numbers were provided by the Health Funding Authority, the Authority is inclined to the view that the total number of true positives was 732 and that it was therefore not inaccurate to state that the pathologist’s error rate was 86%. The Authority therefore declines to uphold the complaint that the item was inaccurate.
Next it turns to the complaint that the report was unfair to the pathologist because of the imputations it made about his professional expertise. In dealing with this aspect of the complaint, the Authority notes the report arose in the context of the perspective of the women who had been incorrectly diagnosed and who had subsequently suffered serious health problems. From the point of view of women whose samples were incorrectly diagnosed, it was irrelevant that he had correctly diagnosed some samples which contained abnormalities. The story – and the subsequent inquiry – was concerned with those women who had suffered harm as a consequence of the failure to detect abnormalities in their cervical smears. On the basis of the information known at the time, the Authority does not consider it was unfair to the pathologist to report the error claimed by the HFA as a result of its retesting. It finds that viewers would have been generally aware that the inquiry was still proceeding and that any conclusions about the pathologist’s conduct would have to await the outcome. Accordingly, the Authority declines to uphold this aspect of the complaint.
For the reasons set forth above, the Authority declines to uphold the complaint.
Signed for and on behalf of the Authority
17 May 2000
The following correspondence was received and considered by the Authority when it determined this complaint:
1. Stuart Slater’s Complaint to Television New Zealand Ltd – 2 February 2000
2. Mr Slater’s Further Complaint to TVNZ – 13 February 2000
3. TVNZ’s Response to the Formal Complaint – 22 February 2000
4. Mr Slater’s letter to TVNZ – 5 March 2000
5. Mr Slater’s Referral to the Broadcasting Standards Authority – 20 March 2000
6. TVNZ’s Response to the Authority – 4 April 2000