Systematic literature review of risk factors for cervical cancer in the Chinese population

Abstract

Objectives:Human papillomavirus is the necessary cause of cervical cancer, in particular the human papillomavirus-16/18 strains, which have been detected in ~70% of all cervical cancer cases worldwide. This study aims to assess whether other cofactors, which might be specific for the Chinese population, are involved in the development of cervical cancer. These findings may support the future direction of cervical cancer prevention.Study Design:Systematic literature review.Methods:The following databases were searched: MEDLINE, MEDLINE-IN-PROCESS, EMBASE, China National Knowledge Infrastructure, Wanfang Data and Chongqing VIP Information. The target population were adolescents or adults from mainland China. All observational studies irrespective of intervention or comparator reporting risk factors for cervical cancer were included. The Newcastle-Ottawa Scale was used to assess study quality. The impact of each outcome was reported in numerical terms.Results:A total of 2,676 articles were screened. A total of 21 articles met the inclusion criteria. All studies were case-controlled designs mostly conducted in hospitals of South-Eastern China. A total of 18 studies reported lifestyle behaviours as significant influencing factors in the development of cervical cancer. Sexual behaviour, gestational factors, screening history, disease history and socio-demographics status were reported as significant risk factors for cervical cancer development.Conclusion:This review provides an up-to-date insight of current cervical cancer risk factors in China. Due to the heterogeneity of the results, further evaluation is recommended to determine the association of these risk factors to the overall risk of cervical cancer.

Risk factors for cervical cancer in women in China: A meta-model

“…The original systematic review grouped risk factors into 6 main categories and 44 sub-categories. 10 For the present analysis, risk factors were further stratified into more detailed categories to allow meaningful comparison of risks and prevalence. This resulted in 105 different groups of risk factors, which we refer to in the remainder of the article as 105 “risk factors” for simplicity (Supplemental Appendix Figure 2).…”

“…A random-effects model was chosen because of the heterogeneity across studies observed in the systematic review. 10,12 The random-effects model considered both the variance within studies ( R i ) and the variance between studies ( R i k ) to determine the overall risk R i . The variance within studies relates to the log-study risk, and thus the random-effects model aggregated the transformed log-risks of each study to determine an overall log-risk.…”

“…First, relevant data and risk factors were extracted from the case-control studies identified by a systematic literature review. 10 Second, the risk factor information extracted from the case-control studies was reviewed by a panel of Chinese experts, who were asked to select the most important or most plausible risk factors for CC, based on their insights. Third, the final meta-model was developed based on the risk factors selected by the experts.…”

“…The main risk factors were socio-demographic (age and education level), life-style behavior (dietary consumption, smoking status, and personal hygiene), sexual behavior (number of partners (for the woman and her partner), number of marriages, age at sexual debut, and age at first marriage), gestational factors (age at first pregnancy, total number of pregnancies, and contraceptive method), and screening and disease history (cervical screening, gynecological diseases, family disease history, and other diseases). 10 Persistent infection with HPV is a necessary cause of CC, detected in more than 99% of cases. 8 Risk factors for disease are commonly evaluated using case-control studies, in which the frequency of suspected risk factors or other attributes is compared between people with the disease of interest (cases) and people without that disease (controls), providing odds ratios (OR) for the various factors.…”

“…OR) could not be combined, for example, because different data were reported (such as coefficient of risk equation vs OR) or different statistical models were used. 10 Thus, the objective of this analysis was to explore a novel method of considering and combining information on multiple risk factors from several studies to predict the risk of CC for any woman's profile and assess the overall contribution of each risk factor to the overall risk of CC. The technique chosen was to develop a meta-model, which combines several regression equations into a unified model to present the relationship between the overall estimated effect and the various risk factors in order to estimate the probability that an adult (aged 18-85 years) woman in Mainland China is at risk of CC based on her characteristics at the time of assessment.…”