George Koliopoulos 1, Victoria N Nyaga 2, Nancy Santesso 3, Andrew Bryant 4, Pierre PL Martin-Hirsch 5, Reem A Mustafa 3, Holger Schünemann 6, Evangelos Paraskevaidis 7, Marc Arbyn 8.
1 Department of Obstetrics and Gynaecology, Elena Venizelou Maternity Hospital, Athens, Greece.
2 Unit of Cancer Epidemiology, Scientific Institute of Public Health, Brussels, Belgium.
3 Department of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, Canada.
4 Institute of Health & Society, Newcastle University, Newcastle upon Tyne, UK.
5 Gynaecological Oncology Unit, Royal Preston Hospital, Lancashire Teaching Hospital NHS Trust, Preston, UK.
6 Departments of Health Research Methods, Evidence, and Impact and of Medicine, McMaster University, Hamilton, Canada.
7 Department of Obstetrics and Gynaecology, Ioannina University Hospital, Ioannina, Greece.
8 Unit of Cancer Epidemiology, Belgian Cancer Centre, Sciensano, Brussels, Belgium

Review question We assessed studies comparing two tests to screen for cervical cancer: the HPV test (Human papillomavirus test) and the Pap test otherwise known as cervical smear or Papanicolaou test. The aim was to find out which test detects precancerous changes of the cervix more accurately.

Background The HPV and the Pap tests are tests that a doctor performs to check for the development of cervical cancer or precancerous changes to the cells of the cervix (called lesions). These lesions can develop into cervical cancer within about 10 to 20 years. The HPV test checks whether a woman has an HPV infection which may lead to cervical cancer. If the HPV test is positive, it may mean that there are precancerous changes in the cervix. There are many types of HPV tests. One of them is called the HC2 test. The Pap test checks for whether cells in the cervix are abnormal. Abnormal cervical cells that are tested as ‘low grade to high grade’ may mean that there are precancerous changes in the cervix that may lead to cervical cancer. One type of Pap test is ‘conventional cytology’ and another is ’liquid-based cytology’. Depending on the test, if it is positive a woman may need to have the cervix examined or could receive surgery to have the precancerous lesion removed.

Study characteristics We searched for all relevant studies up to November 2015. Forty studies compared the HPV test to the Pap test on over 140,000 women between 20 to 70 years old who attended for their routine cervical screening. The studies examined which test can detect precancerous cervical changes which are called cervical intraepithelial neoplasias (CIN 2 and CIN 3).

Quality of the evidence There were enough studies with enough women in them to allow us to draw conclusions. However, some of the results from the studies were different from each other. For example, tests were more accurate in studies in Europe than in Asia or Central or South America. Overall, the quality of the evidence was moderate to high.

Key results A perfect test would correctly say if a woman has precancerous changes or if a woman does not. But most tests are not perfect. This review found that for every 1000 women screened, around 20 women will have precancerous changes. The HPV test will correctly identify 18 of these women (but will miss 2 women). The Pap test will identify 15 of the women (but will miss 5 women). The women who are missed could develop cervical cancer. For every 1000 women screened, there will be 980 women who will not have precancerous changes. The HPV test will correctly identify 881 women (but 99 women will be incorrectly told that they have a lesion). The Pap test will correctly identify 885 women (but 95 will be incorrectly told that they have a lesion). Women who are incorrectly told that they have a lesion may have their cervix examined or may receive surgery unnecessarily.

Source Cochrane Database of Systematic Reviews 2017, Issue 8. Art. No.: CD008587. DOI: 10.1002/14651858.CD008587.pub2.