A Longitudinal KAP Study on HPV Immunised Adolescents’ in Malaysia

Abdul Nazer Ali1*, Jong Shi Jie1, Sunil Kumar Prajapati1*, Mohd. Zahid Iqbal1, Nazer Zulfikar Ahmed2 and Thamir M. Alshammari3 1Faculty of Pharmacy, AIMST University, Malaysia; abdul.nazerali16@gmail.com, jsj_6118@hotmail.com, clinical.sunil@gmail.com, drmmziqbal@gmail.com 2Cognizant Technology Solutions India Private Limited, Hyderabad – 500019, Telangana, India; n.zulfik@gmail.com 3Department Pharmacy Practice, University of Hail, Saudi Arabia; thamer.alshammary@gmail.com

connecting to Cervical Cancer 2 .The major risk factors include; history of STDs, multiple sex partners, sex at young age and promiscuous male partners 3 .
Centre for disease control reported, among 38,793 females and males diagnosed annually with HPVrelated cancers, 7.4/100,000 were Cervical Cancer 4 .Among the Malaysian women aged above 15 years, 2145 are diagnosed and 621 die due to Cervical Cancer every year with an annual economic burden of RM312 million towards prevention 5,6 .Even with the national HPV immunization and CC screening in place, the death rates are still not decreased to desired levels due to lack of resources, awareness and infrastructure in rural areas 7 .
Prevention through early detection and treatment, proper condom use and limited sexual partners can reduce risk 8 .A Thailand study reported 74% male and 70% female students aged 12-15 years were not aware about screening for CC 8 .Furthermore, an England study indicated little knowledge about HPV, the vaccine and need for screening among 50% of vaccinated girls 9 .A number of studies focus on female's knowledge of HPV and vaccine, but paucity among males does prevail 10 .Lack of knowledge towards HPV infection and vaccine is the main concern for its high prevalence 11 .
The outcome measures were to assess the Knowledge, Attitude and Perception (KAP) 12 levels among HPV vaccinated (immunised) secondary school adolescents and to estimate the impact of educational intervention on KAP over time.

Method and Material
This longitudinal study used convenience sampling with a single factor time and KAP as dependent variable using the same questionnaire at three time points.The study populations included were HPV vaccinated secondary school students (at-least 1 dose), aged 13, 14 and 16 years (as permitted by Ministry of Secondary Education, Malaysia) and those participated in all three phases.All other ages between 9 and >18 years, incomplete survey forms, not HPV vaccinated and those participated in pilot studywere excluded.The study was conducted between April and December, 2014.
In phase-1, survey forms (questionnaire) were distributed in class room setting and completed forms were retrieved followed by distribution of educational pamphlet.The phase-2 survey was conducted after two weeks.The phase-3 survey was conducted between three and six months (0, .5 and 3-6 months).
The sample size was calculated based on prevalence of adolescent population in Kedah state, using Raosoft on-line sample size calculator.The estimated sample size was calculated at 95% CI, 5% margin of error and 50% response distribution.The required sample size was N = 384, further 30% added for drop-outs and rounded off to 500 to achieve recommended sample size (N = 500).

Development of Questionnaire and Educational Pamphlet
The educational pamphlet was developed to address informations related to HPV infection, cervical cancer, its screening and vaccination 13,14 .Questions in the knowledge section included: definition; epidemiology; progression of infection; signs and symptoms; diagnosis; risk factors; prevention and treatment of HPV infection and related cancer 15,16 .

Validation of the Questionnaire
The initial draft of 30 item questionnaire, was content validated by six experts, face validated by potential respondents and subjected to Known-Group's validity.The online readability test confirmed easy readability and understanding ability among 13 years old.A testretest study was done among two groups and Spearman's correlation showed moderate to strong correlation between pre-and post-test respectively.The Mann-Whitney test indicated the total score differences were significantly higher for post-test than pre-test.Further, the Intra-class Correlation Coefficient (ICC) for the KAP questionnaire showed good to excellent reliability with significant correlations 17 .Finally, Cronbach's alpha test was conducted among potential respondents and two knowledge based questions removed due to negative correlations.The alpha value for the final, 28 item questionnaire was .947[Mdn = 31(3)] with strong positive correlation and excellent reliability 13 .

Translation of Questionnaire and Educational Pamphlet
The developed questionnaire and pamphlet was forward and backward translated with the help of linguistic specialists from a private university.The translated Malay version was cross matched, verified and back translated to English to confirm clear and easy understanding.
The inter-rater reliability test (Cohen's κ) was run to determine agreement of translation between the two versions and a strong agreement was observed.

Scoring Grades and Scoring Pattern
The scoring grades were adopted from Original Bloom's cut off points, 80-100% correct response was good, 60-79% <60% was poor 12 .The sum of total knowledge score and total attitude and perception score represented total KAP score.One mark for each correct response of 18 knowledge items and the 10 attitude and perception domains were assessed using five point Likert scale, 1 mark for negative marking and 5 marks for positive marking.

Ethical Considerations
The research proposal along with the study instruments was submitted to AIMST University Human Ethical Committee and the ethical clearance was obtained.Further, permission from: Ministry of Secondary Education, Kuala Lumpur, Malaysia and District Education Officer, Kedah state was obtained before initiating the study.

Statistical Analyses of Data
The data analysis was performed using IBM SPSS Statistics for Windows (SPSS version 23).Descriptive statistics for categorical variables was computed using frequency and percentage.Numerical data (distribution not normal) presented as median (IQR).The Chi-square test used for associations.The inferential statistics using Friedman's test and pair-wise application of Wilcoxon signed rank test with Bonferroni correction was used to estimate the precise differences.The significance level was set at .05 for all tests and p values <.05 considered significant.All percentages displayed in text or parentheses are with no decimal places.

Response Rate
A total of 13,000 survey forms were distributed among vaccinated adolescents and the overall response rate was 23% (2928/13,000).The drop-outs were mainly due to absenteeism, unavailable for all three study phases, incomplete survey forms or violated the inclusion criteria.

KAP Score Differences among Socio-Demographic Variables
The minimum possible KAP score was 10 and maximum 68.The KAP scores were categorised as poor, moderate and good for each phases as summarized in Table 2.

The Overall KAP Score Comparison at Three Time Points
The median KAP score at ph1 was 47 (12), with 58%  A post hoc, Wilcoxon signed rank test on the three pairs of contrasts was done to compare the differences in medians between pairs controlling for type-I error (significance level at p<.017) across comparisons, the median KAP score difference and Bonferroni corrected 'Z' score at ph2 was significantly greater than ph1, [15(4) > 9( 6

Effect of Educational Intervention for Knowledge Gain (ph1 vs. ph2)
The pre-test vs. post-test (ph1 vs. ph2) results for the 18 items knowledge assessment are presented in Figure 1.
The correct responses increased in all 18 items from preto post-test and showed highly significant differences in KAP scores using McNemars test (p<.001).The effect of 'Educational Pamphlets' was estimated from the difference in percentages of knowledge score between pre-and post-test which represents the knowledge gained due to intervention.The overall median knowledge score at ph1 and ph2 showed significant differences, 9(6) and 15(4) respectively, with a significant improvement of 6 points.

Effect of Educational Intervention for Knowledge Retention (ph1 vs. ph3)
Figure 2 presents the result of pre-and post-test, 3 to 6 months after intervention (ph1 and ph3).The number of correct responses increased in all 18 knowledge based items and showed highly significant differences in KAP scores which represent knowledge retained after 3 to 6 months of intervention.The overall median knowledge score showed significant differences, 9(6) and 15(7) at ph1 and ph3 respectively which represents knowledge retained.

Impact of Education Intervention (Pamphlet) on Total KAP Score
A comparison of Knowledge, Attitude and Perception and KAP scores were done to estimate the impact of   intervention as tabulated in Table 3.In ph1, the overall knowledge score was poor (63%), attitude/perception was neutral (53%) and KAP score was moderate (58%).In ph2, the knowledge score was good (64%), attitude/ perception was positive (56%) and KAP score was good (67%).It also observed a decrease in percentage of poor category at ph2 reflecting knowledge gain.The overall total KAP scores between ph1 and ph2 showed statistically significant differences and indicated the median KAP score was preferred more in ph2 (Mdn = 57), with Z score = 40.4,p<.001.

Response to K-Items
The knowledge retained was studied after three to six months of intervention (ph3).In ph3, the knowledge score was good (64%), attitude/perception was positive (44%) and KAP scores was good (59%).The difference in knowledge score between ph1 and ph3 showed an increase which reflected the knowledge retained after three to six months.The overall total KAP scores between ph1 and ph3 showed statistically significant differences and indicated the median KAP score preferred ph3 (Mdn = 56) over ph1 (Mdn = 47), Z = 31, p<.001 (Table 3).

Discussion
This longitudinal study was conducted for repeated measures at 0, .5 and 3-6 months using educational pamphlet intervention among secondary school adolescents aged 13, 14 and 16 years in Kedah state, Malaysia.Another Malaysian study reported similar aged study population 18 .The results endorse females, Malay and rural located respondents dominated the study population.Though the overall response rate of 23%, (N = 2 928/13000) could have posed a degree of nonresponse bias, studies have reported low response rates (25%) for repeated measures in similar populations 19 .

Differences in Knowledge and KAP
The results showed that the KAP scores were significantly associated with all the demographic variables at all the three phases, except location at ph2.In 2010, Malaysia introduced the free national HPV immunization program for school girls aged 13 20 .This study is the first of its kind to investigate the KAP levels regarding HPV, five years after initiation of national HPV immunization program in Malaysia and among the HPV vaccinated adolescents, whereas a London based study reported a study on HPV knowledge after three years 21 .
This study found better knowledge scores among female respondents which were consistent with a study in Spain revealing (27%) poor knowledge among males 22 .This study recorded moderate knowledge scores regarding HPV transmission (61%) and its link to STD (61%).The results were consistent with few local studies, though lower than an Italian study (75%) [22][23][24] .This study also found, the adolescents had poor knowledge regarding the cause, vulnerability, consequences and treatment of HPV prior to intervention.Previous studies are in agreement with these findings 23 .The study found 29% adolescents from rural located schools had better knowledge on HPVs association with genital warts, comparable to a Spanish study 22 .Agreement of respondents to 'women infected with HPV are most likely to develop Cervical Cancer' was consistent (60%) with one Malaysian study 11 and lower than others (80% and 69%) [25][26][27] .Rashwans report of 30% awareness on the role of Pap smear in CC prevention was consistent with our findings (26%) 11 .Our study found 50% awareness on HPV vaccine, whereas, other studies observed either very low, 10% and 6% in Sweden 11,23 or moderately high awareness (78%) 25 .An interesting contradiction was noted while assessing the respondents' intention for cancer prevention.They generally had a positive attitude and most of the respondents (75%) were in favour of being vaccinated.

Impact of Educational Intervention
Intervention using pamphlets demonstrated, influencing perception and beliefs can substantially improve the knowledge level and significantly favour attitude towards prevention of HPV infections.A US study reported higher knowledge scores, 86% had higher intention to vaccinate compared to control among unvaccinated intervention group 34,35 .A Germany study reported 'balanced health information leaflet' can increase knowledge of HPV and vaccination uptake" 31 .This study supported the use of pamphlets and demonstrated significant knowledge gain among respondents after intervention.
In this study, the repeated administration of same questionnaires may have acted as a trigger to prompt for information search regarding the infection which might have further initiated discussion between near and dear about its prevention and related cancers.A Finnish study noted, the survey questionnaire by itself can be a stimulus to provide knowledge.According to Health Belief Model (HBM), perceived threat is a combination of susceptibility and severity.All studies investigating the effectiveness of interventions, irrespective of the methods, aimed at a common goal, either to protect from STDs or to improve sexual health.Nearly, all intervention studies have reported a positive outcome, whatsoever tool (lecture, pamphlet, video, community interaction, etc.) were used 30 .
The major strength of this study is the incorporation of 3 to 6-months follow-up evaluation on KAP.In addition, the comparatively large sample-size, inclusion of both genders with rural and urban locations provides enough power for evaluating the usefulness of pamphlet for this study.Continuous health-education is the best approach to prevent Cervical Cancer from early adulthood which mostly depends on regular 'booster' sessions 36 .Nevertheless, the strongest measures of successful interventions should target the long-term outcomes which can reduce Cervical Cancer incidence and related mortality.

Conclusion
The KAP scores were comparatively poor at baseline and educational initiatives taken proved to be appropriate in increasing the awareness and enhancing beliefs towards HPV prevention.The study identifies the need for 'Spaced, Repetition-Based Educational Intervention' providing clear, precise and required information about the implications of HPV infection, addressing misconceptions, its prevention, fears of screening and importance of HPV vaccination.The strengths of this study include the use of repeated measure strategy to investigate a population-based sample for knowledge gained through intervention and knowledge retained after education for a period of three to six months.

Fig. 2 .
Fig. 2. Comparison of knowledge retention, McNemar test based on binomial distribution shows p< .001for all 18 items.