Section 2 -

The Introduction of Liquid Based Cytology into an Organised Screening Programme Peter A Smith Consultant Cytopathologist and Clinical Director of Pathology RoyalLiverpool University Hospital Liverpool UK

Background The cervical screening programme in the United Kingdom is part of the National Health Service (NHS) and some insight into the nature and organisation of UK health care, and of the history of the programme, is essential background to the consideration of the introduction of liquid based cytology (LBC). Since its establishment in 1948, the fundamental aim of the NHS has been to provide a comprehensive health service to all the UK population "free at the point of delivery". While currently much criticised, from many points of view the NHS has been outstandingly successful, and the role of the NHS is deeply embedded in British society. Its existence and mode of organisation have meant that private healthcare use has been limited, and of that, much has been by use of private healthcare facilities within the NHS. The result is that the truly independent private sector in UK healthcare is small, and the NHS has been effectively a monopoly employer of doctors and paramedical professionals. There is no large scale supply of private pathology laboratory services. This position may now be changing as a result of current government policy, dissatisfaction with the NHS, increasing affluence and increasing public expectation, but the historic situation is what has largely influenced the implementation of LBC in the NHS Cervical Screening Programme (NHSCSP). Since the 1970s the basic functional unit the NHS in secondary care has been the District General Hospital (DGH), usually serving a population of around 300,000 – 400,000 and providing all main acute services, with a few more specialised services confined to regional specialist units. Generally, all DGHs have their own laboratories providing a comprehensive range of services, including cellular pathology. Although recently there has been some consolidation or centralisation of smaller laboratories, in general the cervical cytology service is still provided locally from the DGH; currently some 150 NHS laboratories provide a cervical cytology service in England . The NHS is of enormous political importance within the UK , and is inevitably used as a political "football". Current political considerations may impact significantly on NHS matters. One such is the view that there should be no "postcode lottery" i.e. individuals should be able to receive the same service from the NHS whether they live in central London or the most remote Scottish island. Another is that local Health Services should be under local control with services commissioned by Primary Care Trusts, and it is currently considered "politically incorrect" for central government to direct what must happen; rather it is considered appropriate for the Department of Health to recommend or suggest what it thinks should happen, to the 28 regional Strategic Health Authorities and 330 local Primary Care Trusts. Such political views may have paradoxical or conflicting effects on the NHS! With an overall UK population of 59 million, the screening programme is costed at £175 million per annum, and any decision significantly increasing expenditure for the programme as a whole will require a government backed decision. A further complication is that the NHS in England is under the control of the Department of Health whilst in Northern Ireland , Scotland and Wales the Health Service is controlled by the relevant national office or assembly. Cervical screening in the UK started during the 1960s, but generally with little success; by the mid-1980s deaths from invasive cancer of the cervix in the UK overall were unchanged. At that time an increase in cervical cancer deaths was expected by the current decade. Recognition of the problems at this time led to the institution of a properly organised screening programme from 1988. The main problem was the failure to cover an adequate proportion of the population at risk, but lack of failsafe follow up of women with abnormal smears, lack of appropriate training for all grades of staff and lack of quality assurance in the programme were also significant . The result is a now highly successful, quality assured programme, the running of which is very closely regulated and the performance of laboratories and individuals working in the programme, especially cytotechnologists, is subject to continual monitoring. Evaluation and Implementation of LBC With the ready availability of LBC technologies in the later 1990s, there was consideration of their applicability in the NHSCSP. There were several papers indicating that LBC could increase sensitivity and specificity of the test and result in a reduced inadequate sample rate, although the quality of the relevant publications has been criticised [1, 2]. It was felt that in the UK the potential benefit of using LBC would lie in improvement in laboratory productivity and reduction of the inadequate sample rate rather than improvements in test sensitivity and specificity . In 1998/9 the inadequate rate 10th- 90th centile range from the NHSCSP (English)laboratories was 5.8-12.9%; in a programme with 4.9 million tests annually, there was the prospect of considerable saving of repeat inadequate tests, and the associated anxiety and inconvenience for women. There has also been a chronic laboratory staffing problem leading to unacceptable long turnaround times. There were also potential benefits for ancillary testing on the same sample, and future automation of the screening process. The National Institute for Clinical Excellence (NICE) felt that there was sufficient evidence to expect that LBC would be beneficial to the programme, but recommended in June 2000 that implementation pilot studies be carried out. These pilot studies were to be combined with HPV testing in the triage of low grade abnormalities. The manufacturers of two systems, ThinPrep® (Cytyc® Corporation, Marlborough, MA, USA) and SurePath ™ (Tripath Imaging® Inc., Burlington , NC , USA ) applied for their systems to be piloted and three sites were used, two using ThinPrep® and one using SurePath™. The pilots ran in 2001-2 and 250,000 samples were examined (and about 12500 HPV tests performed). The results were compared with those from the preceding period using conventional cytology. The effects on laboratory costs including equipment, consumables and staff productivity were included, as well as evaluation of the cost effectiveness of HPV triage and the psychological effects on women of inadequate smear test results and positive HPV tests. There was a comprehensive training programme for laboratory staff, described briefly below, and smear takers. Separate, similar pilot studies were undertaken in Scotland and Wales, but only used the ThinPrep® system. The report of the English pilot study [3] was submitted to NICE, which reported in October 2003 [4].It was estimated that LBC implementation would be cost neutral in that the increased costs of processing equipment and consumables would be offset by the saving of some 300,000 episodes in primary care because of the need for far fewer repeat tests (although these are not realisable savings in terms of transfer of money from one budget to another). There would also be improved laboratory productivity, and implementation of HPV testing would be cost effective. There was evidence that LBC was at least as sensitive as conventional cytology but no clear evidence of an overall improvement in sensitivity for the detection of low grade or high grade abnormalities. It was recommended that LBC should be implemented in England and that either system could be used. During 2004 the regional cytology training schools would be trained and then LBC training would be rolled out to all laboratories in 2005-8.Choice of system was left to local decision and negotiation after a framework agreement with the suppliers by the NHS Purchasing and Supplies Agency. In my own area the Cheshire and Merseyside Strategic Health Authority, SurePath was selected because the costs quoted were slightly, but significantly, less than for ThinPrep. In other areas, additional factors such as availability of laboratory staff may be relevant. Overall, the choice has been just over 50% of laboratories selecting ThinPrep. The expense of the equipment has meant that sample processing has been centralised in some areas with transport of prepared slides back to the home laboratory for reporting, on a "hub and spoke" basis. £10 million was allocated for rollout in first two years (equipment, training, consumables), with subsequent funding to be provided by the annual budget uplift of the Primary Care Trusts. Training laboratory staff is a massive exercise involving a three day theoretical and practical training course for all medical and technical laboratory staff at one of the national cytology training schools, followed by an assessment of reading of 200 cervical samples, and a test set of 20 normal and abnormal slides. Screeners are then assessed for cumulative sensitivity in reading (a minimum of) a further 200 slides. The time required has to be fitted in with other laboratory commitments at a time when there are significant national shortages of both medically qualified pathologists and technical laboratory staff. Smear taker training has been organised on a regional basis. On Merseyside it consisted of a half day course including anatomy and physiology of the cervix, use of the Cervex-Brush® (Rovers Medical Devices BV, the Netherlands) and an overview of the two approved LBC systems from the suppliers. Inadequacy rates are dependent on the criteria used to assess adequacy. Because of the 3 and 5 yearly screening intervals applied in the programme, there is a view that the adequacy criterion of 5,000 squamous cells per LBC preparation used in the Bethesda System is not strict enough. National standard adequacy criteria have yet to be agreed. Pending such agreement, in Liverpool we are using 15,000 squamous epithelial cells. Cells are counted over a number of representative fields according to a standard protocol. The main performance indicators are of high grade abnormalities as a percentage of all adequate samples, low grade abnormalities as a percentage of all adequate samples, and of inadequate samples as a percentage of all samples. Performance after LBC implementation is clearly of prime interest, but unforunately other changes in the programme taking place concurrently mean that other facors than LBC may affect them. These are the change from an optional 3 or 5 yearly screening interval from 20-64 to 3 screening for all women from 25-49 and then 5 yearly screening from 50-64. It will probably be several years before the nationally base performance indicators stabilise. Current Position To date, LBC has been introduced to the whole of Scotland , about half the laboratories in England, and the whole of Wales . The main results have been that the inadequate rate has fallen to low levels as expected. Referral for colposcopy for three consecutive inadequate tests has become a rare event. Laboratory productivity has improved. Laboratory workloads have fallen as a result of the need to repeat far fewer inadequate tests, and laboratory screening backlogs have largely disappeared; to date therefore, the implementation of LBC to the NHS CSP has fulfilled its promise to deliver significant benefits to the programme. References: Moseley RP, Paget S. Liquid-based cytology: is this the way forward for cervical screening? Cytopathology 2002; 13: 71-82 Davey E, Barratt A, Irwig L, Chan SF, Macaskill P, Mannes P et al. Effect of study design and quality on unsatisfactory rates, cytology classifications and accuracy in liquid-based versus conventional cervical cytology: a systematic review. Lancet 2006; 367: 122-132 Moss SM, Gray A, Marteau T, Legood R, Henstock E, Maissi E. Evaluation of HPV/LBC Pilot Studies. Report to the Department of Health (Revised October 2004). Available at : www.cancerscreening.nhs.uk NHS National Institute for Clinical Excellence. Guidance on the use of liquid-based cytology for cervical screening. London, October 2003. Technology Appraisal 69 Available at: www.nice.org.uk