Assessment of knowledge about human papillomavirus vaccination among primary school girls in Arba Minch town, South Ethiopia, 2020. An institution-based cross-sectional study

Abstract Introduction: Cervical cancer is the second leading cause of cancer-related mortality among females in Ethiopia. The knowledge regarding human papillomavirus vaccination and its acceptability among adolescent girls affects the human papillomavirus vaccine uptake, however, the status of knowledge of the human papillomavirus vaccination among adolescent girls in Ethiopia, particularly in this study area is not well known. Therefore, this study aimed to determine the knowledge of human papillomavirus vaccination and associated factors among primary school girls in Arba Minch town, South Ethiopia, 2020. Methods: A school-based cross-sectional study in which 516 school girls participated was conducted on January 24, 2020. The study participants were selected by a simple random sampling technique. A pre-tested and self-administered questionnaire was used to collect the data. Data was entered to Epi_data version 4.6 and exported to SPSS Version 23 for analysis. The logistic regression model was used to identify the statistically significant variables for knowledge of the human papillomavirus vaccination. Results: The overall knowledge level of the study participants in this study was 71.7%. Their main information source was social media 41.74% followed by health care workers 29.69%. Conclusions: Nearly three-fourths of the study participants were knowledgeable about human papillomavirus vaccination. The knowledge about the human papillomavirus vaccination shows a positive association with age, education level, access to information sources, and parents' educational status. Keywords: Arbaminch town; cervical cancer; Ethiopia; human papillomavirus vaccine; girls; knowledge; schools

and access to information sources, and parents' educational status. The human papillomavirus (HPV) vaccine is a vaccine used to prevent genital warts, anal 53 cancer, cervical cancer, vulvar cancer, and vaginal cancer caused by certain types of human 54 papillomavirus (HPV) [1]. Cervical cancer is a global public health problem accounting for 55 almost 300,000 deaths annually. Eighty-three of new cases and 85% of related deaths occur in 56 resource-poor countries [2]. Cervical cancer is the most notable health problem of Sub-Saharan 57 Africa (SSA). Annually 34.8/100,000 women new cases and 22.5 / 100,000 women deaths were 58 occurred in this sub-region [2]. In Ethiopia cervical cancer is the second leading cause of cancer 59 mortality, every year, 7095 women are diagnosed and 4732 are dying from cervical cancer [3][4][5]. 60

84
Arba-Minch town, the capital city of Gamo zone, is 505km from Addis Ababa (capital city of 85 Ethiopia) and about 280 Km from Hawasa (capital city of southern Ethiopia). There are 1 86 hospital, 3 health centers, and 11 health posts providing health services. Arbaminch has 4 sub-87 cities and 11 kebeles. Arba Minch town has 55 KG schools from which 6 were public, 18 full 88 primary schools from which 8 were public, 9 secondary and preparatory schools from which 5 89 were public. A total number of students registered in 2020 in Arbaminch town were grade one up 90 to twelve 27,512 from which female students are 14,431, grade five up to eight 8,867 from which 91 female students are 4,758, [14], zonal and city health offices work jointly with schools to deliver 92 the vaccine in a school-based way after creating awareness by health extension workers and 93 other concerned bodies [15]. 94

95
A school-based cross-sectional study 96 Source population,

97
All girl students in primary schools of Arba Minch town, South Ethiopia 98 Those girls who were attending a primary school and in grade 5 up to 8 123

Exclusion criteria, 124
Those girls who were ill at the time of data collection 125 Sample Size Determination,

126
The sample size was determined by using a single population proportion formula with the 127 assumption of a 95% confidence level, 5% margin of error, by taking "p" (0.696) from similar 128 studies conducted among adolescent girls, design effect 1.5 and adding 10% non-response rate. 129 The final sample size was 537. 130 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 18, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 Sampling technique,

132
From total primary schools found in the city (in both private and public), 6 primary schools were 133 selected by simple random sampling without replacement through the lottery method. Then the 134 calculated sample size was allocated to each school proportionally, then to each grade level again 135 proportionally, after getting eligible adolescent girls for each grade level from their homeroom 136 teachers and relisted to prepare sampling frame to that grade level. while preparing the sampling 137 frame, a section of given students was listed in front of their name to finally get her, then study 138 units of a given grade level obtained by SRS through lottery method. Finally, selected study units 139 for given schools were requested to be in one place (to one of the sections at break time) and fill 140 the questionnaire through the guidance of data collection facilitators and supervisors. 141 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted October 18, 2021. ; https://doi.org/10. 1101/2021 Data quality control,

167
The training was given on clarification of some assessment tools, the aim of the study, time of 168 data collection, timely collection and reorganization of the collected data from respective 169 schools, and submission on due time. 170 The questionnaire was pre-tested with 5% of the sample size in Mirab Abaya primary school in 171 the Gamo zone which is not included in this study. In addition to appropriate recruitment and 172 training of data collectors, the quality of the data was monitored frequently both in the field and 173 during data entry. This was done in the field through close supervision of data collectors. All 174 filled questionnaire was examined for completeness and consistency during data collection. 175 176 177 178 179 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

180
Data was entered to Epi data version 4.6 and exported for analysis to SPSS Version 23. 181 Descriptive analysis was made and measures of central tendency were also determined. In 182 scoring knowledge, one point was awarded for every correct answer and zero for incorrect or 'do 183 not know' responses. The total knowledge score was then be converted into percentages and the 184 level of knowledge of the respondents was classified based on their score and of overall 7 185 knowledge questions, study participants who score <50% (i.e. answered correctly ≤ 3 questions) 186 were labeled as having poor knowledge and those who answer 4 or above from 7 questions or 187 score 50% and above were labeled as having good knowledge. 188 Logistic regression was applied to see the association between dependent and independent 189 variables. Independent variables, which had an association in bi-variable analysis with p-value 190 <0.25 were entered into the multivariable logistic regression model. Independent variables with 191 p-value<0.05 in the multivariable logistic regression model were considered as statically 192 significant factors for knowledge. The results were presented as odds ratios (OR) with 95% 193 confidence intervals. An odds ratio with corresponding 95% CI was used to quantify the 194 association between a dependent variable and independent variables. The model fitness was 195 checked by using the software application of the Hossmare and Lame show test (>0.05). The 196 results were presented using text, and tables. 197 198 199 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted October 18, 2021. ;

200
Out of the 537 samples that were distributed, 516 study participants were interviewed in this 201 study, with a response rate of 96.1%. Majorities of the study participants were Gamo in ethnicity 202 (71.3%) and protestant Christians in religion (60.5%) ( Table 1). 203  . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted October 18, 2021. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted October 18, 2021. ; . These dissimilarities may be due to the 282 differences in culture and information access, differences in education institutions' structure, 283 sampling procedures and tools, and the cut-off points to define the level of knowledge. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted October 18, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 Familial educational status(those for whom, mothers educational status of secondary close to 296 fourteen fold and more than secondary close to twenty-two fold and fathers educational status of 297 more than secondary close to twofold have good knowledge about HPV vaccine than their 298 counterpart, AOR=13.60, 95%CI (5.69_32.53)P-value <0.001), 22.27, 95%CI(8.23_60.30), P-299 value<0.001) 2.18, 95%CI(1.09_4.35)P-value 0.03) respectively, this is similar with a study 300 conducted in Senegal [26]. It could be because of good communication between parents and their 301 daughters and that may aware them. 302 303 Information source about HPV vaccine has a statistically significant association with knowledge 304 of HPV vaccine, that is, those who obtained HPV vaccine information have close to nine-fold 305 more likely of having good knowledge about HPV vaccine than those who do not obtain HPV 306 vaccine information AOR=8.65, 95% CI(3.92_19.07), P-value <0.001) which is similar to a 307 study conducted in USA [19]. 308

309
As the study was conducted in a multidisciplinary way, the quality of the study was 310 unquestionable. In addition to this, a pretest was conducted and relevance, consistency, and 311 validity of the tool were checked. It was able to survey a random sample of female students, 312 thereby increasing the generalizability of the findings to other eligible students. This study 313 utilized a self-administered questionnaire which increases the likelihood of respondents 314 answering openly, hence increase internal validity. 315 316 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted October 18, 2021. ;

317
The limitation of this study was not included girls' parents in the study as they were decision-318 makers. 319

320
The overall knowledge level of study participants was 71.7%, which is still less than three fourth. 321 Factors positively associated with knowledge of HPV vaccine were age, Grade level, and 322 familial educational status, and information source. It is better if health care workers provide 323 extra information to increase girls' knowledge about the HPV vaccine. It is also better if 324 education institutions encourage girls' education as the increased level of girls' education is 325 associated with the increased level of HPV vaccine knowledge. 326

327
The data supporting the conclusions of this study were included in the text. 328 Funding Statement, 329 The research was fully funded by Arbaminch University. 330 ACKNOWLEDMENTS, 331 We want to forward our respectful gratitude to the Arbaminch city education bureau for their 332 cooperation in giving us important information. 333 We also want to forward our uncountable thank to directors and teachers of schools that were 334 selected for our study as well as data collectors and students who were participated in our study 335 by filling the questionnaires allocated to them. 336 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 18, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 Authors' contributions, 337 Eshetu Yisihak made substantial contributions to the conception and design of this study. Eshetu 338 Yisihak together with Feleke Gebremeskel made contributions to analyze and interpret the data. 339 Desta Markos, Gebresilasea Gendisha & Samuel Abebe made contributions by guiding the 340 collection of data and checking the consistency and completeness of the data. All authors were 341 responsible for drafting the manuscript. 342

343
Authors have no competing interest 344 345 Ethical Considerations, 346 Ethical clearance was obtained from the University of Arba Minch, institutional ethical review 347 board. Co-operation letter received from Gamo zone education office, and selected schools. 348 Ascent consent from homeroom teachers of adolescents and informed consent from adolescents 349 was obtained before the interview was commenced. The ascent was obtained from homeroom 350 teachers of adolescents through sending ascent requesting letters from the University of Arba 351 Minch to the schools selected for the study. Confidentiality was maintained by omitting their 352 name and personal identification. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 18, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021