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

Objectives: Cervical cancer is a leading cause of cancer-related mortality in women in China. This analysis is a quantitative evidence synthesis pooling information about each cervical cancer risk factor. Methods: A meta-model was developed to estimate the risk of cervical cancer for a woman aged 18–85 years in Mainland China based on her risk profile at the time of assessment. The meta-model was built using findings of a systematic literature review that identified 21 case–control studies reporting data on 105 groups of cervical cancer risk factors in Chinese women. Extracted risk factors were ranked, and 17 were selected by Chinese clinical experts for inclusion in the meta-model. Risk equations were developed for each selected study. Predicted risks for each study were dependent on the risk profile under consideration and study-specific risks were pooled to an overall risk estimate using a random-effects meta-analysis. Sensitivity analysis was conducted using 100 artificial patient profiles (in the absence of patient data). Results: Predicted risks for the 100 profiles suggested that the model had good face validity and could differentiate between high and non-high cervical cancer risk profiles. Conclusion: This innovative meta-model approach assesses cervical cancer risk in Chinese women from a holistic perspective and could be adapted for other diseases and settings.


Risk factors for cervical cancer in women
Appendix Figure 1. Flow chart of the systematic review and the literature update 4 Appendix Figure 2. Overview of the risk factors from previously published systematic review 1 and the literature update * Six categories are: socio-demographics, life style, sexual behavior and marital status, gestational risk factors, CC screening and gynecological disease, other factors. $ Each risk factor may correspond to different questions in different studies (i.e. education, below or above high school, below or above primary school, high school, college or university). # Individual entries of each risk factor by each individual question. ** Category "other factors" is removed. $$ Experts selected 18 risk factor groups, however, after the update post expert meeting, no studies were identified for one of the selected risk factor group, so only the remaining 17 risk factor groups were included in the final model. The missing risk factor is 'history of cervical treatment'. ## All risk estimates are listed in Appendix Tables 3-16.

Appendix 2. Selection criteria for the systematic literature review
Appendix organizations. If estimates were obtained from grey literature, the quality, validity and applicability of these estimates were discussed between both reviewers. 9 If no results were obtained from the first three tiers, the prevalence estimate of the CC risk factor in Chinese women was considered equal to the prevalence of the control group in the original case-control study from which the risk factor and the risk factor estimates had been derived (Tier 4). If risk factors were reported in multiple studies and no judgement could be made about which prevalence estimate was considered to be most representative, the average across studies was used.
If no results were available from the underlying case control study, e.g. due to lack of data about the prevalence in the control group, the search then focused on non-Chinese data sources (Tier 5), using the same approach as the first three tiers. Data sources were prioritized based on geographical and cultural proximity to China (e.g. data sources from the Asian region were preferred over European region estimates). If estimates were obtained from non-Chinese data sources, the quality, validity and applicability of these estimates were discussed between both reviewers.

Appendix 5. Random-effects meta-analysis
The total variance (Q-statistic) was calculated as follows: 2 where R ik is the mean risk for the patient i based on study k, and 2 is the associated variance. The expression log( ) − The degrees of freedom (df) = Number of studies -1 and thus the between study variance 2 can be calculated as: where : scaling factor to scale the weighted sum of squares Q (i.e. total variance) with Aggregating the log-study risks log( )′ in the random-effects meta-analysis, each logrisk was then weighted by an inverse-variance weighted method using weight * with * = 1 * where * = , 2 + 2 .
Using a back transformation, the overall risk R i is obtained by: and the corresponding variance by: The corresponding lower bound (LB) and upper bound (UB) of the 95% CI were calculated as follows:

Appendix 6. Ranking of initial risk factors
Appendix Figure 3. Top 30 risk factors ranked by number of studies in which risk factor was reported § HPV, human papillomavirus; § risk factor definitions see Appendix Table 3 Appendix