Burden of cervical lesion in Ethiopia: Systematic review and meta-analysis

Though cervical cancer is largely preventable, it is still the second most common female cancer globally and the leading cause of cancer deaths among females in African. Though many efforts has been done to study the burden of the disease in Ethiopia, there are still fragmented primary studies reported cervical lesion. Hence, this systematic review and meta-analysis aimed in estimating the pooled prevalence of cervical cancer and its trend in Ethiopia. Methods: This systematic review and meta-analysis was conducted using available data. PubMed, Web of Science, SCOPUS, Science Direct, Google Scholar, African Index Medicus (AIM), African Journals Online databases and Ethiopian University research repositories were searched following the Preferred Items for Systematic review and Meta-analysis (PRISMA) guideline. STATA 15 statistical software was used to analyse the data. The quality of included studies was assessed using the Joanna Briggs Institute (JBI) quality appraisal tool for meta-analysis. Heterogeneity between studies was assessed using Cochrane Q test and I2 test statistics based on the random effects model. A random effect model was computed to estimate the pooled prevalence of cervical lesion in Ethiopia. In addition, the trend of cervical lesion in the country was depicted. Result: Fifteen studies with a total of 25,348 participants were included in the analysis. The overall pooled prevalence of cervical lesion was 16.36 (95 CI: 10.97-21.75). The subgroup analysis by region showed the highest prevalence of cervical lesion at the Southern Nations and Nationalities Peoples Region (19.65%; 95% CI: 15.51 – 23.80). The trend of cervical lesion showed an increased pattern overtime from 1992 to 2018 in the meta-regression analysis. Conclusion: One among six of the study participants had cervical lesion. The trend also showed that there is still an increased cervical lesion in the country. Best practices in achieving high vaccination coverage shall be adapted from other successful countries.

Though many efforts has been done to study the burden of cervical lesion, still there are fragmented primary studies reported the outcome of interest. Different previous studies and literatures showed variations in the past and across different geographical areas in the country, Ethiopia. Hence, this systematic review and meta-analysis aimed in estimating the pooled prevalence of cervical cancer and its trend in Ethiopia. In addition, the nding will provide an insight to decision and policy makers on how to strengthen the existing CC prevention and control strategies. Or device a new way of disease prevention and control mechanisms.

Reporting
The Preferred Reporting Items of Systematic Reviews and Meta-Analysis (PRISMA) checklist guideline was used to report the result of this systematic review and meta-analyses. In addition, PRISMA Flow chart was utilized to show the selection process of studies for a systematic review and meta-analysis (37).

Searching strategies
This systematic review and meta-analysis was conducted to estimate the pooled prevalence of cervical cancer and its' trend in Ethiopia. To conduct this study, all potentially relevant articles, grey literatures, and government reports were meticulously searched. We searched PubMed, Web of Science, SCOPUS, Science Direct, Google Scholar, African Index Medicus (AIM) and African Journals Online databases for all available studies using the following search terms: "Cervical cancer", "Cervix neoplasm", "Cervix lesion", "Pap Smear Positive", "Cervix precancerous lesion" and "Ethiopia". Search string was developed using "AND" and "OR" Boolean operators. In addition to this, grey literatures were searched from research repository online library and a secondary search technique known as "footnote chasing" was utilized to identify additional articles from the included articles.
The search was carried out from August 28 to October 10, 2019 and all articles published until October 10, 2019 were included in the review.

Inclusion and exclusion criteria
This systematic review and meta-analysis encompassed studies conducted only in Ethiopia using English language published from 1992 to 2018. Research articles that reported the prevalence of cervical cancer, published in the scienti c journals, grey literatures were included in the review. Studies focused on the assessment of knowledge, attitude, and practice towards CC without the outcome of interest of this study, program evaluation studies, studies with only abstracts, case studies, qualitative studies and citations without full-text were excluded.

Operational de nition
Cervical lesions: Lesions that are visible and de ned, after the application of visual inspection with acetic acid (VIA)/visual inspection with Lugol's iodine (VILI) (38). They have borders and degrees of acetowhitening, vessel patterns, and margins. Precancerous areas are not "dark red and puffy" as the caption suggests (39).

Data extraction
Three authors (AS, TD and YB) independently searched the studies and extracted all necessary data using a standardized data extraction format using Microsoft Excel. The extracted parameters were: primary author, publication year, region where the study was conducted, method of assessment, study design, age range of the study participants, mean age of the study participants, response rate, sample size, prevalence of cervical cancer, and quality of each study. Then, four authors (GD, AE and YW) checked the data extraction process. Finally, authors (BZ, EA and WA) participated in approving and resolving the disagreements.

Quality of the included studies
To appraise the quality of the included articles we used the Joanna Briggs Institute (JBI) quality appraisal tool adapted for studies reporting prevalence data, cross-sectional and case-control studies (40). The following items were used to appraise cross-sectional studies: [1] inclusion criteria; [2] description of study subject and setting; [3] valid and reliable measurement of exposure; [4] objective and standard criteria used; [5] identi cation of confounder; [6] strategies to handle confounder; [7] outcome measurement; and [8] appropriate statistical analysis. In addition, JBI Systematic Reviews checklist for case control studies (41) using ten items were used to appraise one case control study (42). Studies considered low risk whenever tted to 50% and or above quality assessment checklist criteria's (40,43). Using the aforementioned quality appraisal tools, no study was excluded (Table 1).

Data analysis
Data extraction were compiled using Microsoft Excel format and analyzed using STATA version 15 statistical software.
Heterogeneity across studies was checked using the inverse variance (I 2 ) and Cochran Q statistics and the cut-offs of 25%, 50%, and 75% were used to declare the heterogeneity as low, moderate, and severe respectively (44,45).
As the preliminary output of the test statistics revealed a signi cant heterogeneity among studies (I 2 > 70%, P< 0.05), a random effects model was used to estimate the pooled prevalence of Cervical cancer (CC) with 95% con dence interval (CI). Subgroup analysis was also performed among regions and Human Immuno Virus (HIV) status in relation to outcome variable. Funnel plot, Egger and Begg tests at 5% signi cant level were employed to assess publication bias (46,47).

Results
Description of the identi ed studies Until October 10, 2019 we identi ed a total of 109 articles using different data bases and University research repository. Sixty one (55.9%) of these identi ed articles were excluded because of similarity and duplication. Among the remaining 48 articles, 27 articles were excluded because the outcome was not clearly measured. Six articles were excluded due to exclusion criteria. Finally, 15 studies ful lled the eligibility criteria and included in the nal systematic review and meta-analysis ( Figure 1).

Overview of included studies
Among the total 15 studies included to the current review, ve of them were from Addis Ababa City Administration (17,42,(48)(49)(50). Five studies were included from the Southern Nations Nationalities and People Region (SNNPR) (51)(52)(53)(54)(55). Except one study (42), all the rest included studies were conducted using cross sectional study design. Most of the studies used VIA as the measure of assessment tool for the outcome variable. The total participants in the included studies were 25,348 with sample ranged from 189 to 16,632. The response rate of all the included studies were greater than 90% and the quality score of the studies ranged from 62.5 to 100% ( Table 2).

Publication bias
To assess publication bias, both the funnel plot and the Egger's test was conducted in the meta-analysis. The visual examination of the funnel plot exhibited a symmetric distribution of studies ( Figure 2). In addition to the funnel plot, Egger's regression test was (β = -0.0061, SE=0.07, P=0.87) showed that no evidence of publication bias for the included studies.

Prevalence of Cervical Lesion
Fifteen studies with a total sample size of 25348 study subjects were included to the current review. From these, the lowest and the highest prevalence were found to be 1.56% (48) and 28.4% (17). The lowest prevalence was reported from SNNPR whereas the highest prevalence were reported from a studies done in Addis Ababa. Overall, the pooled prevalence of cervical lesion in Ethiopia was 16.36 (95 CI: 10.97-21.75) (Figure 3).
Since the I 2 static test for heterogeneity indicated that the studies differed signi cantly (I 2 = 96.0%, p< 0.05) and because theoretically we expected differences among studies due to different reasons, we tted a DerSimonian and Laird random effect model (61,62) to estimate the pooled prevalence of cervical lesion. Studies that showed the largest weight were Plezer et al (48) that showed 7.12 and Tadesse B. et al (59) that showed 6.80 weight. Whereas Birara et al showed a relatively smaller weight which is 6.45 ( Figure 3).
The sub-group analysis by region showed the highest pooled prevalence of cervical lesion was found in SNNPR, 19.65% (95% CI: 15.51 -23.80) followed by Addis Ababa City Administration in which the pooled prevalence was 15.10 (95% CI: 4.77-25.44).
In addition, sub-group analysis was done using HIV status of study participants. The highest pooled estimate of cervical lesion was found in those study participants whose sero-status is unknown, 17.27 (95% CI: 5.12-29.41). Whereas the pooled prevalence cervical lesion of those study participants whose status was con rmed to be positive was 15.27 (95% CI: 9.06 -21.48) ( Table 3).
The sub-group analysis showed as there is presence of heterogeneity across the studies. Hence, to identify the source of heterogeneity, we conducted meta-regression and sensitivity analysis. The meta-regression analysis was conducted using the following study covariates: publication years, HIV status, sample size, and study area. However, the results showed that none of these variables were a statically signi cant source of heterogeneity. We also carried out sensitivity analyses to observe the in uence of each study on the overall effect size. No study signi cantly affected the overall pooled estimate of cervical lesion.

Meta-regression
Meta-regression was done to observe the trends of the prevalence of cervical lesion in Ethiopia using the reported prevalence and year of study. A signi cant upward trend in the prevalence of cervical lesion was observed from 1992 to 2018 (B =0.55, P=0.013) (Figure 4). However; there was no signi cant association between prevalence of cervical lesion and sample size of the included studies (B =−0.00023, P= 0.77).

Discussion
Cervical cancer ranks as the second most frequent cancer among women in Ethiopia. Awareness creation about the screening was being advocated in the country since 2008 (63). Despite this fact, very few women receive screening services (64). About 6,294 new cervical cancer cases are diagnosed annually in Ethiopia (estimates for 2018). Cervical cancer ranks as the 2nd leading cause of female cancer in Ethiopia (13).
The current systematic review and meta-analysis aimed in estimating the pooled prevalence of cervical lesion among Ethiopian women. The result reviled that the overall pooled prevalence of cervical lesion was 16.36% (95% CI: 10.97, 21.75). This nding is almost consistent with a study done in Rural Nigerian Women in which 16.6% had cervical lesion (65) and another study done in Tanzania (cross-sectional hospital-based, majority was HIV+) and Nigeria reveled 17% and 17.8% women had cervical lesion respectively (66,67).
This nding is higher than from a study done in Tanzania (systematic review) in which the overall prevalence of cervical lesion was 9.2% (68) and Zaria state Nigeria in which the overall prevalence of cervical lesion was 4.8 (69). Another study done in Kwara State, Nigeria (urban community) showed that 5% of study participants had cervical lesion (70). A community based screening in Turkey showed that a prevalence of 9.4% cervical lesion (71). A similar study in India on community-based cervical cancer screening program among women of Delhi obtained 4.67% cervical lesion (72).
The possible explanation for this discrepancy might be the mean age of marriage among the respondents in the aforementioned studies was relatively higher. As early marriage is one of the risk factor for having cervical lesion, this might augment the incidence and prevalence of the cervical lesion. The other explanation might be most of researches conducted in the aforementioned countries were from urban settings and they might have awareness and access to information this might lead them to have early screening for the disease. This nding is lower than a study done in Senegal in which 21.03% of study participants had cervical lesion (73). Another study done in Nigeria among HIV Positive women showed that 22.2% of women had cervical lesion (74). The possible elucidation for such discrepancy between the current nding and other comparable study ndings might be due to the difference in the Socio-demographic variations in the included study participants. A study from countries contained a data mostly collected mostly from the rural population and HIV positive women while in our study; both urban and rural settings were considered. The other possible explanation for the above variation could be due to the difference in study design.

Subgroup analysis
By region: The highest pooled prevalence of cervical lesion was seen in Southern Nation and Nationalities and People's Region (SNNPR) which was 19.65%. This nding was higher compared to studies conducted in Addis Ababa and Amhara region that revealed 15.10% and 14.35% respectively. The possible explanations for this variation might be accounted by variations in information dissemination across the regions for reproductive health women about the disease. Hence, those women who had not information might remain with the symptom/disease.

By HIV status:
It is known that HIV infection is one of the main risk factors for the development of cervical dysplasia. On this regard there are many reports on the association of HIV with increased risk of cervical dysplasia (67,74-76). However the current systematic review and meta-analysis in the subgroup analysis revealed that the highest cervical lesion (17.27%) was found in those study participants whose sero-status was unknown. This nding might be supported by the WHO report that stated the likely pattern of cervical lesions expected in a previously unscreened population of women (77,78). In addition to this, the possible explanation for this nding might be accounted by participants who were not screened for HIV might be exposed for other risk factors for cervical cancer. Plus to this, though their sero-status were not known at the time of cervical lesion screening, they might be previously know that their sero-status for HIV/AIDS as positive. Limitation: In the current systematic review and meta-analysis, we noticed some limitations. The pooled prevalence was determined using 15 studies from four different administrative regions in Ethiopia. Hence, it might not represent the whole country as studies in more than half of the regions not included in the current review.

Conclusion
The current study revealed that one among six of the study participants had cervical lesion. The trend also showed that there is still an increased trend cervical lesion in the country. Though it is late, Ethiopia launched HPV vaccine last year December 2018. This will play a great role in the ghting against the high prevalence of the disease. Awareness creation programs should be done using various social Medias. In addition, best practices in achieving high vaccination coverage shall be adapted from other successful countries. Competing interests: The authors declare that they have no competing interests Funding: No fund was obtained to carry out this study Authors' contributions: AS conceived and designed the research protocol. AS, TD and YB searched the studies and extracted all necessary data. GD, AE and YW checked the data extraction process. AS and TD led the revision and editing the manuscript and all the authors YB, GD, AE, BZ, EA and WA involved in revising and editing the nal version of the manuscript. All authors have read and approved the nal draft of the manuscript.