Study design and setting
A systematic review and meta-analysis were conducted to assess knowledge, attitude, and practice of cervical cancer screening and associated factors among women aged 15–49 years in Ethiopia. Ethiopia is located in the horn of Africa. It is bounded by Eritrea to the north, Djibouti, and Somalia to the east, Sudan and South Sudan to the west, and Kenya to the south[16].
Search Strategies
The Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guideline was used to confirm the scientific accuracy [17]. We searched articles without any limit to the date of publication and study design from the major databases such as PubMed/MEDLINE, Web of Science, EMBASE, CINAHL, Google Scholar, Science Direct and Cochrane Library. EndNote (version X7) reference management software was used to download, organize, review and cite the articles. This review included all articles published in peer-reviewed journals, that were written in the English language. [18]. Also, Gray literature of observational studies and official websites of international and local organizations and universities were searched. The search of the literature was conducted from the 1st of September, 2019 to the 1st of December, 2019. All papers published until the 1st of December, 2019 were considered. The Electronic databases were searched by medical subject headings (MeSH) terms and using the following keywords: “Knowledge”, “Attitude”, “Practice”, “cervical cancer”, “and 15–49 years old women”, “barriers”, “determinants”, “associated factors and Ethiopia”. We used the search terms independently and/or in combination using Boolean operators like “OR” or “AND”.
Eligibility Criteria
Inclusion criteria
Study area
Studies conducted only in Ethiopia were considered.
Population
Only studies involving reproductive age women (15–49 years old women) were included.
Publication condition
Both published and unpublished articles were included.
Study design
All observational study designs (i.e., cross-sectional) reporting the knowledge, attitude, and practice of screening cervical cancer among reproductive age groups of women were included.
Language
Articles reported only in English language were considered.
Exclusion criteria
Articles, which were not fully accessible and not reported the outcome of interest, after at least two-email contact with the primary authors, were excluded.
Measurement of outcome variables
The main outcome of this study was to assess the pooled levels of knowledge, Knowledge, Attitude, and Practice of Cervical Cancer Screening and Associated Factors.
Data Extraction And Quality Assessment
Data were extracted using a standardized data extraction spreadsheet format in Microsoft Excel. The data abstraction format includes author/s name, year of publication, study region, study design, sample size, the prevalence of knowledge, prevalence of attitude and prevalence of cervical cancer practice. All published and unpublished primary studies in the English language, conducted at the community or health institution level from 2015 to December 2019 were included. The data were extracted by three independent authors (YM, AAS, and AA). Joanna Briggs Institute (JBI's) critical appraisal checklist for prevalence studies was used to assess the quality of included studies [19]. Additionally, a modified version of the Newcastle-Ottawa Scale (NOS) was used to assess the methodological quality of studies for cross-sectional studies [20]. Newcastle –Ottawa scale criteria include representativeness of the sample, response rate, measurement tool used, comparability of the subject, appropriateness of the statistical test used to analyze the data. Based on the Newcastle –Ottawa scale criteria, two authors (YM and AA) independently assessed the quality of each article. Whenever necessary, another reviewer (AAS) was involved and any discrepancy was resolved through discussion and consensus.
Data Processing And Analysis
For farther analysis, we imported the data to STATA version 14.0 statistical software after extracting the data using Microsoft Excel format. We identify the heterogeneity between the studies using Cochrane's Q statistics (Chi-square), inverse variance (I2) and p-values [21]. As the test statistic showed there is significant heterogeneity among the studies (Fig. 2A, 2B and 2C) as a result a random-effects meta-analysis model was used to estimate the DerSimonian and Laird's pooled effect. In the current meta-analysis, arcsine-transformed proportions were used. The pooled proportion was estimated by using the back-transform of the weighted mean of the transformed proportions, using arcsine variance weights for the fixed-effects model and DerSimonian-Laird weights for the random-effects model [22]. Egger’s and Begg’s tests, with the p-value, less than 5% was used to declare the presence of publication bias [23]. Also, using forest plot, point prevalence, as well as 95% confidence intervals, was presented. In this plot, the size of each box indicated the weight of the study, while each crossed line refers to a 95% confidence interval. For the second outcome, a log odds ratio was used to determine the association between determinant factors and cervical cancer screening among 15–49 years old women.
Study Selection And Data Extraction
This systematic review and meta-analysis have been reported by the preferred reporting items for systematic reviews and meta-analyses (PRISMA) statement [24]. Initially, 455 articles related to KAP of cervical cancer screening were found. Of these, 168 duplications and 263 unrelated articles were excluded. Second, from the remaining 24 potential articles, 18 met eligibility for the review and included in the analysis. Six full-text articles were excluded due to did not meet outcomes of interest or study area (Fig. 1).
Characteristics of included studies
Eighteen studies with a total sample of 9, 897 women (aged 15–49 years old) were included in this systematic review and meta-analysis. The main characteristics of the included studies are described in Table 1.
Table 1
Characteristics of the included studies in the meta-analysis in Ethiopia, 2019
SN | Author | Year | Region | Study area | Study design | Sample size | Quality Score |
1 | Shiferaw et al | 2018 | Addis Ababa | Addis Ababa | cross-sectional | 581 | High |
2 | Getachew et al | 2019 | Addis Ababa | Addis Ababa | cross-sectional | 520 | High |
3 | Kalau et al | 2018 | Amhara | Debre Berhan | cross-sectional | 578 | Moderate |
4 | Kasa et al | 2018 | Amhara | Finote Selam | cross-sectional | 735 | Moderate |
5 | Mesfin et al | 2017 | Addis Ababa | Addis Ababa | cross-sectional | 508 | High |
6 | Tsegaye et al | 2018 | SNNP | Hawassa | cross-sectional | 380 | High |
7 | Gebrie et al | 2015 | Addis Ababa | Addis Ababa | cross-sectional | 275 | High |
8 | Worku et al | 2018 | Oromia | Bishoftu | cross-sectional | 845 | High |
9 | Assefa et al | 2015 | Amhara | Dessie | cross-sectional | 660 | High |
10 | Zelalem et al | 2018 | Amhara | Gonda | cross-sectional | 267 | High |
11 | Geremew et al | 2018 | Amhara | Finote Selam | cross-sectional | 1137 | Moderate |
12 | Bulto et al | 2019 | Oromia | Shoa | cross-sectional | 423 | Moderate |
13 | Dulla et al | 2017 | SNNP | Hawassa | cross-sectional | 367 | High |
14 | Atif et al | 2019 | Oromia | Jimma | cross-sectional | 154 | High |
15 | Tilahun et al | 2019 | Oromia | Wollega | cross-sectional | 805 | Moderate |
16 | Aweke et al | 2017 | SNNP | Hadiya | cross-sectional | 583 | High |
17 | Assefa et al | 2019 | SNNP | Hawassa | cross-sectional | 342 | High |
18 | Nigussie et al | 2019 | Oromia | Jimma | cross-sectional | 737 | Moderate |
Meta-analysis
Level of knowledge, attitude, and practice towards cervical cancer screening among 15–49 years old aged women in Ethiopia
Overall, the pooled level of good knowledge about cervical cancer screening among 15–49 years old women in Ethiopia was 40.37% (95% CI: 30.09–50.66). The lowest was (9.60%) and the highest was (86.90%) (Fig. 2A). The pooled level of good attitude towards cervical cancer screening among 15–49 years old women in Ethiopia was 58.87% (95% CI: 42.69–75.05). The lowest was (37.00%) and the highest was (94.10%) (Fig. 2B). The pooled prevalence of cervical cancer screening practice among 15–49 years old women in Ethiopia was 14.02% (95% CI: 9.67–18.37). The lowest was (2.10%) and the highest was (40.10%) (Fig. 2C).
Heterogeneity And Publication Bias
The presence of publication bias was examined using visual inspection of the funnel plot and Egger's test. Visual inspection of the funnel plot suggested asymmetrical distribution of included studies. We also observed publication bias in practice of cervical cancer screening using both Begg's and Egger's tests [25, 26] and these tests showed that there was statistical evidence of publication bias at p-value less than 0.05 and the funnel plot was asymmetry. By consideration of publication bias trim and fill meta-analysis [31] were done. However, based on this analysis no significant change was shown as compared with the main meta-analysis. Moreover, the result of sensitivity analyses using the random-effects model suggested that none of the studies included influence the overall estimate. Given that the result of this meta-analysis revealed statistically significant heterogeneity among studies, we performed a subgroup analysis by region to adjust and minimize heterogeneity (Fig. 3).
Meta-regression
Besides subgroup analysis and publication bias, univariate and categorical meta-regression analysis was performed undertaken by considering both continuous and categorical data for the included studies to identify sources of for heterogeneity for the pooled prevalence. In meta-regression analysis, Sample size, publication year and study regions were considered for each study. However, there was no statistical significance value from the meta-regression analysis (Table 2).
Table 2
Meta-Regression of Knowledge, Attitude, and Practice of Cervical Cancer Screening in Ethiopia, 2019
Domain | Variables | Characteristics | Coefficient | P-value |
Knowledge | Year | Publication | − .6734893 | 0.893 |
Sample size | Included | − .0029275 | 0.894 |
Region | Amhara | 3.45085 | 0.879 |
Addis Ababa | 6.844267 | 0.742 |
SNNP | 39.88295 | 0.099 |
Oromia | Reference | Reference |
Attitude | Year | Publication | 2.958676 | 0.662 |
Sample size | Included | − .0098279 | 0.785 |
Region | Amhara | -13.77482 | 0.338 |
Addis Ababa | 34.56641 | 0.111 |
SNNP | .7496539 | 0.960 |
Oromia | Reference | Reference |
Practice | Year | Publication | 1.221513 | 0.619 |
Sample size | Included | − .0293279 | 0.117 |
Region | Amhara | Reference | Reference |
Addis Ababa | 11.36227 | 0.274 |
Oromia | − .0293008 | 0.998 |
SNNP | 12.37763 | 0.237 |
The associated factor of the utilization of screening cervical cancer among reproductive age groups in Ethiopia
In this meta-analysis, the association between the practice of screening cervical cancer and the educational status of women had been seen using eight studies conducted in Ethiopia. The Mata-analysis of these eight studies revealed that the practice of screening cervical cancer among reproductive-age women was significantly associated with the educational status of women (OR: 3.10, 95% CI: 2.09–4.62). Specifically, the likelihood of cervical cancer screening practice was 3.10 times higher among literate women compared to their counterparts. The test result of this meta-analysis revealed a significant heterogeneity among eight studies (I2 = 78.0%, P ≤ 0.001). So the random effect meta-analysis model was used to see the association educational status with cervical cancer practice in Ethiopia (Fig. 3A). In this meta-analysis, women who had good Knowledge status about cervical cancer was found to be significantly associated with cervical cancer screening practice (OR: 5.87, 95% CI: 2.57–13.40). Reproductive age women who had good knowledge about cervical cancer screening were around six times more likely to practice cervical cancer screening compared to those women who did not hear about Cervical cancer screening (Fig. 3B). In this meta-analysis, the association between the practice of screening cervical cancer and women who had a positive attitude towards cervical cancer screening had been seen using six studies conducted in Ethiopia. The Mata-analysis of these six studies revealed that the practice of screening cervical cancer among reproductive-age women was significantly associated with women who had a positive attitude towards cervical cancer screening (OR: 4.20, 95% CI: 2.86–6.16) (Fig. 3C). This review revealed that the practice of screening cervical cancer among reproductive-age women was significantly associated with women who not being married (OR: 3.34, 95% CI: 1.09–10.22) (Fig. 3D).