Oireachtas Joint and Select Committees

Wednesday, 18 December 2019

Joint Oireachtas Committee on Health

Royal College of Obstetricians and Gynaecologists Independent Expert Panel Review into Cervical Screening: Discussion

Professor Henry Kitchener:

I thank the joint committee for the invitation to appear today. I welcome the opportunity to engage with it on the findings of our review. I am joined today by Dr. Patrick Walker, deputy lead assessor of the RCOG review. We would like to take this opportunity to thank all of the women or their next of kin who consented to participate in this review. We would like to acknowledge the contribution of the cytology laboratory at University Hospital Monklands in Scotland and other consultant cytopathologists, the members of the expert panel, including our two lay representatives, and our project manager at the RCOG for assistance throughout this process.

After a complex and demanding exercise, the college had met the objectives defined in the terms of reference with the completion of 1,308 individual reports for women or next of kin by October 2019 and the publication of our aggregate report on December 3 2019. The timing of the final report was dictated by the requirement to ensure that all the women or next of kin had received their individual report prior to the aggregate report being published. At this point we wish to state how acutely aware we are of the impact that screening failures have had on the lives of affected women and their families. We would also like to recognise the 103 women in our review who have died, and will neither be able to read their individual reports nor the final RCOG report. We hope the individual reports will provide some service to next of kin.

The primary purpose of the individual reports was to provide consenting women or next of kin with a transparent and independent analysis of their CervicalCheck cytology slides prior to their diagnosis of cancer, along with an explanation of the findings and an independent conclusion regarding the clinical implications of any discordant result. Throughout this process, we felt it was of the utmost importance to obtain input from participating women with regard to the content, tone and style of letter that would convey their individual reports. In May 2019, we met a group of women and one next of kin to discuss our proposed style of letter. We received broad approval and suggestions, which we adopted. The group also conveyed a strong message that any women who so wish should have access to an appointment with a health professional to receive their reports. We were in full agreement and immediately conveyed this to the HSE.

The primary purpose of the final report was to aggregate the findings of the cytology slide review that directly compared the RCOG readings of 1,659 slides with those of CervicalCheck. By way of background, the slide review panel was comprised of screeners from the University Hospital Monklands cytology centre in Scotland as well as consultant staff actively working in the English NHS cervical screening programme. The role of the screeners was to identify those slides considered to be negative, that is, showing no abnormality. All slides considered to show any abnormal results were reviewed by a consultant. If there was disagreement with the CervicalCheck result, a second consultant read the slides. If there was disagreement between the first and second consultants, the slides would be viewed by a consensus panel over a multiheaded microscope to arrive at a final result.

With regard to process, those reading the slides were always blind to the CervicalCheck result. However, for reasons laid out in the report, seeding the slides blindly among routine day-to-day laboratory work would have been impracticable. We believe our methods were thorough and produced reliable results. In both the report and in explanatory notes to each participant, we have acknowledged that in a slide review of this type the detection of abnormalities will be enhanced. Our role was not to assess the performance of the CervicalCheck laboratories but to inform women whether or not the review had found that abnormal cells had been missed. In those cases where there was discordance between the CervicalCheck and the RCOG review, the college sought to determine whether this had impacted on the clinical outcome in terms of failure to prevent cancer or diagnose it at an earlier stage.

This was done by the expert panel determining the likely quantum of delay in the diagnosis of cancer engendered by missing abnormal cells, or undercalling cells in terms of grade, and applying this quantum to a set protocol of time intervals and stage diagnosis.

The RCOG expert panel felt the individual letters should include not only the results of the slide review but also the effect any discordant reading had on a delay in diagnosis. In the majority of cases, the expert panel considered there was no such effect but in half of the discordant cases the panel concluded there had been a missed opportunity to prevent or diagnose at an earlier stage. We applied scrutiny to those cases where prolonged colposcopy management was identified and found that in a quarter of such cases colposcopy management was suboptimal, sometimes further complicated by discordant cytology and sometimes not.

The slide review formed the basis of providing women with individual reports. The thoroughness and quality of the review combined with full disclosure of the findings means women can have confidence in their individual reports. The expert panel made it absolutely clear that these reports are not a judgment with respect to medical negligence and that such a matter would require others to determine. An explanatory note to this effect was included with the individual reports. This compilation of individual reports disclosing the findings and implications, together with an aggregate report based on a point in time slide review of more than 1,000 women diagnosed with cervical cancer, is, as far as the authors can determine, unique in scale and scope. It represents a transparent approach that others can reference and with which future comparison can be made.

We chose to compare our slide review findings with the English audit, which is the only other published large-scale slide review conducted as part of a national cervical screening protocol. Both involved slides of liquid-based cytology and both involved the slides being reviewed in the knowledge that cancer had been the outcome. The intention was to determine whether the findings of the RCOG review represented an outlier or was in line with what might be expected. In the event, it was the latter. The findings of the RCOG review were striking in the sense that a high proportion of discordant readings were found, many with significant implications. However, this was not unexpected, not only because similar discordant rates were found in the English NHS cervical cancer audit but also because the natural history of cervical cancer indicates that it is highly probable the cervix would have shed abnormal cells for a number of years prior to the diagnosis of cancer, whether it was screen detected or not.

Our review, therefore, illustrates the limitations of cervical cytology rather than pointing to a screening service that falls below what might be expected. Data published by National Cancer Registry Ireland shows falling incidence and deaths from cervical cancer since CervicalCheck was established in 2008. These endpoints are the cardinal signs of effective cervical screening. The findings of our review back this up, with a high proportion of early-stage cancer identified in the review cohort, which indicates that screening undoubtedly saved the lives of many women in the review.

To address concerns expressed regarding the verification process of finalising individual reports, we have submitted a detailed statement to the committee. When these concerns were initially raised with us in October, we offered to meet participants and patient representatives and came to Dublin to outline the process in detail and answer any questions. In a complex and demanding exercise involving more than 1,000 women and 20,000 individual data items, including 3,300 smear results, we acknowledge there is capacity for error. As we note in the statement, in a very small number of cases, totalling fewer than five, it has been necessary to issue supplementary reports because new information came to light, and in two of these cases the conclusion was updated. While these isolated incidents are regrettable they in no way impact the overall conclusions or recommendations laid out in our aggregate report.

We would be concerned if unfounded criticism of our evidence-based report undermined trust in our findings and confidence in the CervicalCheck programme. This could have the effect of diminishing coverage, which could only place more lives at risk. It is very important that the benefits of cervical screening in the existing programme are accurately portrayed in terms of public health. Vaccination against human papilloma virus was introduced in Ireland for early adolescent girls in 2010 and for adolescent boys this year. As already demonstrated in Australia and Scotland, where screening begins at 20, with high coverage this could reduce considerably the incidence of abnormal smears and, therefore, pre-cancer in young women and will result in a further reduction in deaths from cervical cancer.

We believe the RCOG review has provided all eligible women with open disclosure of their slide review and has provided an aggregate picture of the overall result. Our conclusion that the CervicalCheck programme is working in terms of effectiveness, cytology and colposcopy is supported by incidents in mortality data published by National Cancer Registry Ireland, which demonstrates year-on-year falls since 2010 following the establishment of the CervicalCheck programme. The premier objectives in cervical screening are being met and our report should restore trust in cervical screening in Ireland. With high coverage, the CervicalCheck programme, which will soon incorporate primary HPV testing, combined with the impact of high coverage vaccination, should ensure that cervical cancer prevention in Ireland will become as good as it can be and cervical cancer deaths should eventually decline to the point where they become a rarity.

We were privileged to have been tasked with undertaking this important piece of work, which we hope will have benefited the women who participated in the review.