Dataset for meta-analysis and metaregression of published outcomes data on definitive radiotherapy with or without chemotherapy for locally advanced cervical cancer in patients with contraindications to cisplatin

Published: 25 July 2022| Version 4 | DOI: 10.17632/r9x8h9h98r.4


We performed a systematic review and meta-analysis to examine the treatment outcomes in patients with locally advanced cervical cancer with contraindications to cisplatin, that were managed with definitive radiotherapy with or without chemotherapy. We then performed a metaregression to examine the effects of clinical and treatment variables on outcomes. We identified eligible studies and extracted relevant data using a standardized data extraction template. Table 1 summarizes the characteristics of the population or subgroup (including the nature of contraindication to cisplatin) and the intervention groups (radiotherapy, brachytherapy, and/or chemotherapy regimens) per each eligible study. Table 2 summarizes our assessment of the risk of bias for each study using templates that were based on the CASP Critical Appraisal Checklists. Tables 3 and 4 summarize the outcomes (tumor response, survival, compliance rates, and toxicity) for each intervention group. Table 5 consolidates numerical data from Tables 1, 3, and 4, suitable for meta-analyses and meta-regression. Table 6 is a composite database that contains the main template used in the preparation of the other datasets for the proportional meta-analyses and metaregression, given the variables and outcomes of interest. Table 7 is a composite database that contains the datasets used for the comparative meta-analyses. Figures 1 to 5b are Forests plots for pooled outcomes according to the intervention categories. Figures 6a to 6f are Forest plots for pooled outcomes according to the use or non-use of chemotherapy and nodal boost.


Steps to reproduce

1. Do a systematic search to identify relevant studies to your research question and objectives. 2. Screen each study by examining the title and abstract; review the full text to determine final eligibility. 3. Extract data using a standardized and piloted data-extraction template. 4. Tabulate the data. For proportional meta-analyses, outcome rates will be necessary. For comparative meta-analyses, risk ratios, odds ratios, or hazard ratios will be necessary. 5. For the meta-analyses, use MetaXL v5.3. The software allows for both proportional and comparative meta-analyses. 6. When pooling outcomes from different study designs, use the quality effects model. This model allows assigning quality index (Qi) scores to each study. 7. In the case of heterogeneity, MetaXL allows for sub-group and sensitivity analyses. 8. In the case of analyses that include >10 studies, MetaXL allows for the generation of both funnel plots and Doi plots to evaluate publication bias. 9. For the metaregression, use the MetaXL to generate the metaregression dataset. The metaregression dataset could then be entered into Stata v16.1. 10. Stata allows for both categorical and continuous metaregression, depending on the nature of the variable of interest.


University of Santo Tomas Hospital


Chemotherapy, Cervical Cancer, Renal Failure, Multivariate Regression, Clinical Oncology, Brachytherapy, Pelvic Radiotherapy, Meta-Analysis, Elderly Patient, Frailty, Treatment of Locally Advanced Disease