Between 1996 and 1999 we screened just under 6000 women recruited from a peri-urban settlement 20 kilometeres outside Cape Town, an impoverished area known as Khayelitsha where two-thirds of the residents live in informal housing. We recruited previously unscreened women aged 35 – 65 years. We compared conventional Pap smears to three other tests:

• Direct Visual Inspection or DVI
• Testing for Human Papillomavirus (HPV) Cervicography

While DVI and HPV testing proved to be equivalent and/or superior to Pap smears for identifying lesions, both tests had a very low specificity – that is, there were many false positive cases, compared to Pap smears. This meant, that if we were to link screening with immediate treatment of women with positive tests we would end up doing the following:

• Treating 1 in 5 screened women
• And more importantly, over-treating large numbers of women who had a positive test but did not have a true cervical cancer precursor.

We designed a randomised control trial of screening and treating women based on DVI and HPV testing. 7200 women were recruited between 2000 and 2003 and randomly assigned to one of three groups: Treatment performed by a nursing sister at a primary care level using cryotherapy if DVI positive, treatment with cryotherapy if HPV positive or to no treatment regardless of the result of the DVI or HPV test.

All women were seen at one month post treatment to assess for complications of treatment, including untreated women so we had a comparison group, and again 6 monthly for 36 months for a colposcopic assessment, histological sampling and treatment, if necessary. All colposcopies were performed by myself and trained medical officers on-site working out of adapted second hand shipping containers located at two primary health care clinics in Khayelitsha. This study is still ongoing but our first analysis of the data has shown that women screened with DVI and treated with cryotherapy had a 40% reduction of cervical cancer precursors at 6 and 12 months post treatment compared to untreated women. Women in the HPV arm of the study had an 80% reduction compared to untreated women. We performed just under a 1000 cryotherapies and had only one serious complication in a woman suffering from advanced HIV who had a bleed post cryotherapy and required admission to hospital for a blood transfusion. She subsequently recovered.

These data proved to us that screening and treating women, albeit in the confines of a stringently controlled and monitored research study, is safe, feasible and effective. We also found that the women experienced the process as being highly acceptable. Another indicator of acceptability is that our follow up rates have been extremely high, with 98% of women returning for their one month interview, over 86% for the 6 month examination and around 80% for the 12 month examination. These high return rates are exceptional for large clinical studies and have been made possible partly by having a dedicated team of community health workers visiting patients in their homes to remind them to return for their visits. In addition, we have provided a very patient centred service, which we acknowledge may not always be possible in the ‘real world’, although certainly we should be aspiring to provide this quality of service to our patients.

We now wish to upscale The Khayelitsha Cervical Cancer Screening project to a large group of women to evaluate a number of ongoing questions which include:

• How screening and treating would work outside of the controls inherent in research studies i.e. in field conditions, including training of nursing sisters, instituting on- going quality control, counselling and providing other medical services to our patients, integrating screening into primary health care structures etc.

• Does screening and treating women immediately increase the HIV seroconversion rate in this group of women? To answer this question we need to screen and treat
56 000 women. Our data from the 7200 women in this study showed no increase in HIV seroconversion in treated versus untreated women, but the study did not have sufficient numbers to be absolutely sure that this is not the case when large numbers of women are screened.

There is little doubt that preventing infection of the cervix with an effective HPV vaccine could eliminate cervical cancer worldwide. It is important to note however, that the vaccine is still in the process of clinical trials, that the vaccines have been developed by two multi-national drug companies at considerable cost and its availability to poor countries at an affordable price is a long way off. In the meantime, hundreds of thousands of poor women will continue to develop cervical cancer unless implementable and sustainable prevention programmes are put in place. Further, establishing screening programmes will enable the development of the infrastructure that will be required by mass HPV vaccine programmes once they become available.

Another aspect of this research is that the project has provided a wide range of health services to women whose access to health care is extremely limited and often of poor quality, particularly at primary care level. Screening is considered an activity offered to ‘well women' however, in poor communities there are very few ‘well women' and most women come for screening because of other problems, both related and unrelated to their health. The project has made it a priority to engage with women's non-research related problems and to actively promote access to and information around issues such as contraception, HIV counselling and testing, domestic violence (which is alarmingly common) and access to the State health and social services. In addition, we have developed a package of culturally appropriate health education materials, including a making a 30 minute film in the ‘soap-opera' formula which clearly explains the nature of screening and the many barriers women face. We have also developed a radio-play and have tested the impact of these initiatives in community-based trials and shown them to be effective tools.

We have formed strong links with the many community-based organisations working in the area, including with traditional healers, who also avail themselves of our services.

Also of importance is that over 40 staff employed by the project over the years have been trained and acquired new and unique skills. These have included doctors, nurses, community health workers, project administrators, office managers and data capturers. They have been taught the rigours of research methodology, the principles of good clinical practice and the multiple levels of interaction and engagement required to provide high quality research and service to patients.

The project represents an amalgam of high level internationally recognised clinical research, conducted through a very productive ‘first and third' world collaboration, however, located in the community of women who will benefit most from the outcome of this research. Service delivery at all levels has been an integral part of the study which has been performed at primary care clinics, using rudimentary resources very effectively. It is a model that can be adapted for any resource restricted environment.

Professor Lynette Denny
Secretary Treasurer, AORTIC