Background

Progress in the prevention and treatment of cervical cancer has been made but challenges still exist in developing countries [1, 2]. Low-resource settings are faced with challenges of financial resources, poor and none existent health infrastructure (laboratories, cervical cancer screening centres), lack of technology and few qualified health personnel mainly due to the brain drain [1]. These challenges coupled with lack of proper epidemiological data, lack of knowledge and inadequate information on cervical cancer in low-resource settings have created a major public health issue that is threatening to derail the progress made under the millennium development goal (MDG) number five; reducing maternal mortality and achieving universal access to reproductive health.

The adverse effects of HIV in most low-resource settings have increased the burden of cervical cancer [2,3,4,5]. Like most opportunistic infections, HIV-seropositive women are at higher risk of HPV infection due to their immune compromised status with a risk 2 to 12 times more when compared to HIV-negative women [3,4,5].

Since current treatment modalities for precancerous lesions and cervical cancer are based on the stage of the lesion and available resources, the associated poor outcomes of treatment among HIV-seropositive women in developing countries may be due to a lack of optimal treatment regimen [6]. Most developing countries lack skilled surgeons to carry out radical surgery for cervical cancer and this has left HIV-seropositive women with cervical cancer with few treatment options. In cases where surgeons are available, surgery is expensive and out of reach of many, who happen to be poor [7]. In developing countries especially sub-Saharan Africa, many women with cervical cancer have no access to radiotherapy, further limiting their treatment options.

However, little or no information exists that has shown that any of the treatments are more effective than others when it comes to treating cervical cancer in HIV-seropositive women. In sub-Saharan Africa, treatments like radiation therapy and other surgical procedures are not fully utilised because of lack of equipment and qualified personnel, hence little has been documented on which treatment procedures are being used for cervical cancer in HIV-seropositive women [8]. There is lack of evidence-based guidelines and strategies for screening, vaccination against HPV, prevention and treatment of cervical cancer in HIV-seropositive women in most developed countries [9, 10]. Coupled with this, there is little rigorous evidence on the global epidemiology of the treatment of cervical cancer in HIV-seropositive women [8]. Therefore, we aimed to review the different treatment methods being used to treat cervical cancer in HIV-seropositive women in developing countries.

Methods/design

The PRISMA-P statement (see attached PRISMA-P statement) guided the development and reporting of this protocol [11] whilst the systematic review will be reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [12].

Protocol registration

This review protocol is registered in PROSPERO database (registration number: CRD42017054676, [13], https://www.crd.york.ac.uk/PROSPERO/display_record.php?RecordID=54676).

Studies’ eligibility criteria

Studies will be included if

  • Cervical cancer treatment methods for HIV-positive women (such as chemotherapy, radiation therapy, surgery, cryotherapy and targeted therapy among others)

  • Cervical cancer treatment methods and HIV are considered being independent and outcome variables

  • Published in peer-reviewed journals and grey literature (conferences, dissertations, government health reports)

  • They were done in or for countries or regions that are considered developing by the United Nations [14]

  • They are observational study designs (retrospective cohorts, prospective cohorts, cross-sectional and case-control) or randomised controlled trials

  • Studies done across developed and developing countries, the team will follow the guidance provided by Mapanga and colleagues [15]. The review team will extract results from the developing countries where it is possible and will contact study authors for more information if it is not available.

Studies will be excluded if they are describing cervical cancer in general, their samples are unrepresentative (non-parametric tests as alluded to in the previous protocol, will be used to determine unrepresentative samples) or if they are reviews [15]. No studies will be excluded because of the length of the follow-up period; instead, follow-up periods will be used to assess the quality of the study outcomes [15]. Non-English language studies, reports and dissertations will also be sought as part of the search strategy and translation of data will be performed by a volunteer where feasible.

Search strategy

The search strategy of the online databases will be based on the criteria developed in the previous review protocol by Mapanga and colleagues [15]. MEDLINE (1966–present) and Embase (1980–present) will be searched via the OVID interface as indicated in Table 1. In addition, PubMed, Cochrane and CINAHL (1961–present) will be searched using a combination of the following keywords: cervical cancer, treatment, developing countries (the geographical search concept will be extended to include country names of developing countries), invasive cervical cancer, HIV, management of cervical cancer, chemotherapy, surgery and radiation. In addition, health databases which cover developing countries (3ie Systematic Reviews, WHO library and databases, World Bank website) and databases and websites containing on-going research (such as WHO ICTRP and cliniccaltrials.gov) will also be searched for relevant literature. Proximity operators, Boolean logic operators and truncation commands (see Table 2) will be used as suggested in the previous review protocol as well as conducting preliminary search trials [15]. To search for additional and relevant papers, reference and citation tracking will be conducted as indicated in the PRISMA flow diagram (Fig. 1).

Table 1 MEDLINE and Embase search strategy
Full size table
Table 2 Techniques to be used in the online databases search
Full size table
Fig. 1
figure 1

PRISMA review flowchart

Full size image

Study selection

Two independently working reviewers (from among WM, TC and SF) will merge the results of the electronic database, citation and reference searches as well as screen for abstracts. An additional full-text screening form (see Additional file 1) is going to be used to identify and select potential papers for the review. The two reviewers will make sure the papers are double screened and document reasons that they have for exclusion [15]. The three reviewers (WM, TC and SF) will solve all disagreements and other issues around the screening process through discussions.

Data extraction

The data extraction process and choice of indicators to be extracted from the included studies will be guided by the process indicated in the previous review [15]. There will be double data extraction by WM, TC and SF and the team will solve disagreements and discrepancies through discussions. An additional data extraction form (see Additional file 2) which will be used to extract data from the identified studies will be pretested and adjusted accordingly. The team will extract the following data from included studies: title of the study, study setting, publication year, study design, exposures, sample size, outcomes, descriptive statistics, risk/odds ratios, and confounders, results of linear and logistic regression.

Quality assessment

The assessment of the quality of included studies will be guided by a modified version of the Newcastle-Ottawa Quality Assessment Scale (see Additional file 3) [16]. Quality of studies will be assessed according to the criteria suggested by Mapanga and colleagues [15]. All included studies will be assessed based on their study design that was used to measure cervical cancer treatment, validity of key findings (if study is describing cervical cancer treatment methods or a comparison of treatment methods), follow-up period and sample representativeness.

Randomised controlled trials will be assessed according to the criteria in Table 3, whilst observational studies with a control group will be assessed according to Table 4, observational studies without control groups will be assessed according to Table 5 and the quality of the studies’ outcomes will be assessed according to Table 6.

Table 3 Randomised clinical trials quality assessment checklist
Full size table
Table 4 Observational studies with a control group quality assessment checklist
Full size table
Table 5 Observational studies without a control group quality assessment checklist
Full size table
Table 6 Outcome measures’ quality assessment checklist
Full size table

Two reviewers among WF, TC and SF, will examine the quality and relevance of the extracted data by scoring each study from zero to five and through discussions; disagreements and discrepancies will be resolved [15].

Data management

Literature search results will be saved in their respective database user accounts and citation records will be uploaded to EndNote Software, a reference package which facilitates the management of references and bibliographies.

Synthesis

Results of this review will be synthesised using both the narrative synthesis and meta-analysis as described in the planned review [15]. Narrative synthesis will be used for descriptive analysis whilst random-effects aggregate data meta-analysis will be used to combine all numerical findings from the included studies. Assessment of bias will also be analysed using the meta-analysis and funnel plots will be produced using the RevMan software and statistical significance at 95% using t test will be inferred. Higgins and Thompson’s I2 statistic is going to be used to assess heterogeneity, where a 0% will indicate no heterogeneity and increase in percentage will indicate increase in heterogeneity, which will be significant at p value of less than or equal to 0.05 [17]. STATA Statistical package is going to be used to run the meta-analysis and tables and forest plots will be used to present estimates.

Reporting

The PRISMA statement will guide this systematic review and its findings [12].

Discussion

A systematic review of the published literature to identify different treatment methods that are currently being used to treat cervical cancer in HIV-seropositive women in developing countries will be undertaken. The review will investigate the different treatment methods of cervical cancer for HIV-seropositive women in developing countries and sources of heterogeneity within the studies. The evidence generated from this review will be used to address the gap that exist in this area as well as provide a basis for future research, cervical cancer policies and cervical cancer interventions. The strengths and limitations for this review will be considered and the review findings will be discussed in the context of other reviews and evidence that are relevant.