Abstract
Schoolgirls in two Ugandan districts were recently vaccinated against human papillomavirus that causes most cervical cancer. This cross-sectional comparative study used mixed research methods to assess influence of human papillomavirus vaccination on adolescents’ worrisome thoughts about being vaccinated and psychological distress. Vaccination predicted worrisome thoughts among the recently vaccinated (adjusted odds ratio: 1.65, confidence interval: 1.13–2.41; p = 0.01). Vaccination predicted distress (1.75, confidence interval: 1.09–2.82; p = 0.02), particularly among those recently vaccinated (1.92, confidence interval: 1.27–2.89; p = 0.001) and those who experienced worrisome thoughts (1.80, confidence interval: 1.06–3.07; p = 0.02). Parental communication mitigated distress (0.50, confidence interval: 0.35–0.72; p = 0.000).
Keywords
Introduction
Every year, about 500,000 women develop cervical cancer, which is associated with certain types of human papillomavirus (HPV) and about 274,000 die from the disease globally (World Health Organization (WHO), 2007). Two HPV prophylactic vaccines, Cervarix and Gardasil, have turned out to be 90 percent effective in safely preventing HPV types 16 and 18 infections, which together account for about 70 percent of worldwide cervical cancer cases (Program for Appropriate Technology in Health (PATH), 2009). The HPV is primarily sexually transmitted (Lenselink et al., 2008). The HPV vaccines are mostly targeted on young adolescent girls and they are most effective if given to HPV-unexposed girls. Research has captured diverse fears about HPV vaccination of young adolescent girls. Some of the fears are related to safety and unknown side effects; the possibility of jeopardizing the future fertility of vaccinated girls; and the misunderstanding in some environments that the vaccinations are meant to control fertility or reduce the population of certain groups (Agosti and Goldie, 2007; Katahoire et al., 2008; PATH, 2009; Remes et al., 2012). Moreover, many children loathe injections because they are painful, frightening, and psychologically stressful (French et al., 1994). Fear of HPV vaccination is known to be associated with significant distress among targeted adolescents (Bernard et al., 2011). Previous research on psychological distress in Ugandan adolescents mainly focused on war-and HIV/AIDS-affected children (Atwine et al., 2005; McMullen et al., 2012; Musisi and Kinyanda, 2009; Okello et al., 2007). The possibility of psychological distress in adolescents facing potentially fear-provoking HPV vaccinations has not been studied in Uganda. The psychosocial paradigm of stress–distress relationship holds that external stressors, if unchecked, do disrupt an individual’s psychological equilibrium and induce physiological and/or psychological responses in the form of distress (Ensel and Lin, 1991). In this study, we considered HPV vaccinations and the associated worries to constitute stressful events, which are threatening and undesirable rather than life readjustment due to change per se (Vinokur and Selzer, 1975). Research has demonstrated a consistency in the relation between life stress (life events) and distress (Ensel and Lin, 1991; Muhwezi et al., 2007). From 2008 to 2011, cohorts of adolescent girls from primary schools in Ibanda and Nakasongola districts, Uganda were vaccinated annually against HPV. The vaccination was part of a demonstration project by the Uganda Government and PATH to evaluate different HPV vaccine delivery strategies. A school-based HPV vaccine delivery strategy was adopted in Ibanda district targeting girls enrolled in primary grade 5 (P5). In Nakasongola district, the HPV vaccine was delivered during the routine Child Days Plus (CDP) program, targeting girls of at least 10 years. Each eligible girl was to receive three doses of the vaccine administered in Month 1, Month 2, and Month 6 (PATH et al., 2010). The purpose of this study was to assess adolescent girls’ psychological distress in Ibanda district, following HPV vaccination. Based on the psychosocial paradigm of stress–distress relationship, it was hypothesized in the research for this article that girls vaccinated against HPV compared to unvaccinated girls were more likely to have had worrisome thoughts about vaccination and to experience high psychological distress.
Methods
Design
This comparative, mixed methods cross-sectional study was conducted between November and December 2011. Girls vaccinated against HPV were drawn from Ibanda district while unvaccinated participants for comparison were selected from Mbarara district.
Participants
Primary school education in Uganda occurs over 7 years, beginning with Primary 1 (P1) and ending with Primary 7 (P7). The participants in this study were girls enrolled in Primary 5 (P5) and Primary 6 (P6). Vaccinated P6 study participants had completed their vaccination 13 months before, while their P5 colleagues completed vaccination 1 month prior to the date of data collection. During data analysis, the latter were considered as recently vaccinated. Using multi-stage sampling, 16 selected schools in Ibanda district and the same number in Mbarara district were stratified by rural–urban location. In Ibanda district, all assenting P5 and P6 girls present on the day of the survey that had received at least one dose of the HPV vaccine were recruited. Similar criteria were followed for the unvaccinated comparison group in Mbarara district. A sample size of 800 girls was determined for the survey using the formula for sample size calculation to compare two proportions in terms of risk (Pandey, 1999). Of the 800 girls targeted, 777 (444 in Ibanda district and 333 in Mbarara district) participated. There were 23 non-responders including those who were absent from school and those who did not return their parents’ consent forms. Five focus group discussions (FGDs) each with 8–12 vaccinated girls in P5 or P6 were conducted one per school in Ibanda district, involving five schools that had not been sampled for the survey. FGD participants aged 13–16 years were selected purposively with the help of teachers, targeting self-confident and uninhibited girls who could talk freely in a group. Girls who participated in FGDs had to have been vaccinated against HPV so as to capture their observations and experiences during vaccination.
Data collection
A survey questionnaire and FGDs were used to collect data. Selected girls were assembled in a room on the day of the survey. A pre-tested and translated self-administered questionnaire was used. Each girl filled the questionnaire under close supervision of the first author (A.K.T.) and research assistants who also explained to them the purpose and procedure of the survey. Data were collected first in Ibanda district before Mbarara district. FGDs were conducted by A.K.T. and a research assistant after survey data collection in Ibanda district.
Measures
Independent variables were socio-demographic characteristics (age, class attended, kind of school, location of school, having parents, and perceived parental communication) and HPV vaccination status. Dependent variables were worrisome thoughts about being vaccinated and level of psychological distress. Moderator variable was knowledge of HPV and HPV vaccine. Each variable was assessed through responses to a set of questions or statements. The variables were analyzed as dichotomous.
HPV vaccination status was assessed by asking respondents whether they had ever been vaccinated against HPV. Worrisome thoughts were assessed by asking respondents whether they had ever felt very worried about having been vaccinated or going to be vaccinated.
Psychological distress was assessed using the self-administered Strengths and Difficulties Questionnaire (SDQ) for children aged 3–16 years (Goodman, 1997). The SDQ has 25 items, some positive and others negative. The items are divided between five scales each with five items: emotional symptoms, conduct problems, hyperactivity/inattention, peer relationship problems, and prosocial behavior. The first four scales are added together to generate a Total Difficulties Scale (TDS) based on 20 items with scores ranging from 0 to 40. A cut-off score of 20 is usually used to identify likely “cases” with non-specific mental health disorders. In this study, TDS scores ranged from 1 to 30. In order to make our findings comparable with other studies (Abbo et al., 2013; Kinyanda et al., 2013; Moscardino et al., 2012) that used the SDQ to assess adolescent psychological distress in the same country, Uganda, we adopted a cut-off score of 16 to indicate adolescents’ high psychological distress. Scores 1–15 were banded to denote low psychological distress, whereas scores 16–30 were banded to denote high psychological distress.
Knowledge was assessed through responses to 13 factual statements based on information from the print materials used for community sensitization prior to HPV vaccination. Perceived parental communication was assessed by asking respondents how often their parents or guardians talked to them about their future and the dangers of pre-marital sex.
Data analysis
Quantitative data were analyzed for 769 girls. Data for eight girls were excluded from the analysis when their age (17–19 years) was found to be above 16 years, the upper age limit for the version of SDQ used in this study. Data were entered and cleaned using EpiData, and then analyzed using SPSS v. 16.0 (SPSS Inc., 2007). Univariate analysis was done to generate frequencies and percentages for categorical variables. Means and standard deviations (SDs) were generated for continuous variables. The independent samples t-test was the statistical test used to compare the continuous variables. Since our sample was clustered by schools, we conducted complex sample analyses by applying sampling weights to the statistics (means, percentages, and odds ratios (ORs)) to take care of the effect of the sampling design. At bivariate level, we evaluated relationships between independent and dependent variables using the Pearson χ2 test to determine statistical significance of observed differences. We performed complex samples logistic regression to test the hypothesis that girls vaccinated against HPV were more likely to report worrisome thoughts about vaccination and high psychological distress. In the analysis, we compared proportions of girls vaccinated against HPV with proportions of unvaccinated girls. Relevant ORs with corresponding confidence intervals (CIs) set at 95 percent were determined. Level of statistical significance was set at 0.05.
Qualitative data were transcribed, word processed, and entered into ATLAS.ti GmbH (1993) for coding and analysis. Codes were developed based on the broad themes identified from the data. The codes were used to retrieve segments of the data. Memos were written describing the patterns and variations in the different segments of retrieved data. Data matrices were drawn out of the retrieved data sets according to the plan of analysis. Qualitative data were used to interpret quantitative data.
Ethical issues and approval
The research was approved by the Higher Degrees Research and Ethics Committee of the then Makerere University Medical School and the Uganda National Council for Science and Technology. Additional approvals were obtained from relevant local government and school authorities. Prior to the interviews, written assent and consent were obtained from the targeted girls and their parents, respectively.
Results
Socio-demographic characteristics of respondents
The overall age range of girls was 9–16 years. The mean age of vaccinated girls (13.4 years, SD = 1.36) and unvaccinated girls (13.3 years, SD = 1.32) was similar. Table 1 compares vaccinated and unvaccinated girls in terms of socio-demographic characteristics. Vaccinated girls were significantly more likely to perceive high parental communication (crude OR: 1.74, CI: 1.00–3.05; p = 0.05).
Socio-demographic characteristics of vaccinated and unvaccinated respondents.
HPV: human papillomavirus; CI: confidence interval; ORs: odds ratios.
For all socio-demographic variables, second column odds are divided by third column odds.
Column figures do not add up to 329 due to a missing case.
Significant at p ⩽ 0.05.
Bivariate analysis
Worrying about vaccination
Table 2 shows respondents’ likelihood to report worrisome thoughts about being vaccinated according to their background characteristics. Overall, fewer girls (270 out of 769 or 35%) responded affirmatively; the majority (65%) reported no worrisome thoughts about being vaccinated. Compared to unvaccinated girls, those vaccinated against HPV were more likely to report worrisome thoughts about vaccination and the difference was almost significant (crude OR: 1.74, CI: 0.90–3.23; p = 0.10). Compared to P6 girls, P5 girls were significantly more likely to report worrisome thoughts about vaccination (crude OR: 1.61, CI: 1.12–2.32; p = 0.01).
Relationship of the respondents’ characteristics with experience of worrisome thoughts about vaccination.
HPV: human papillomavirus; CI: confidence interval; ORs: odds ratios.
For all respondents’ characteristics, second column odds are divided by third column odds.
Column total does not add up to 499 due to a missing case.
Significant at p ⩽ 0.01.
Psychological distress
Table 3 shows respondents’ likelihood to report high psychological distress according to their background characteristics. Consistent with expectation, more girls vaccinated against HPV (173 out of 440 or 39.3%) compared to unvaccinated girls (91 out of 329 or 27.7%) had high psychological distress and the difference was statistically significant (crude OR: 1.70, CI: 1.04–2.77; p = 0.04). Compared to P6 girls, P5 girls were significantly more likely to be highly distressed (crude OR: 2.05, CI: 1.31–3.21; p = 0.003). Girls who perceived high parental communication were significantly less likely to be distressed (crude OR: 0.52, CI: 0.37–0.70; p = 0.001). Consistent with expectation, a significant proportion of girls with worrisome thoughts about vaccination (66 out of 270 or 24.8%) compared to girls without such thoughts (79 out of 499 or 15.2%) were highly distressed (crude OR: 2.01, CI: 1.18–3.45; p = 0.01).
Relationship of the respondents’ characteristics with level of psychological distress.
HPV: human papillomavirus; CI: confidence interval; ORs: odds ratios.
For all background variables, second column odds are divided by third column odds.
Row figures do not add up to 632 due to a missing case.
Significant at p ⩽ 0.05.
Significant at p ⩽ 0.01.
Significant at p ⩽ 0.001.
Logistic regression
Worrisome thoughts about vaccination and level of psychological distress
Variables entered in the first model of logistic regression to predict adolescents’ worrisome thoughts about vaccination were HPV vaccination status and class attended. As shown in Table 4, HPV vaccination did not predict worrisome thoughts about vaccination even after controlling for class attended. The odds of a vaccinated girl reporting worrisome thoughts increased but the difference was not significant (p = 0.09). Being in P5 predicted worrisome thoughts about vaccination even after controlling for HPV vaccination status (adjusted OR: 1.65, CI: 1.13–2.41; p = 0.01). Variables entered in the second model to predict level of psychological distress were the significant background variables in bivariate analysis and worrisome thoughts about being vaccinated. As shown in Table 4, vaccination against HPV (adjusted OR: 1.75, CI: 1.09–2.82; p = 0.02) and experiencing worrisome thoughts about vaccination (1.80, CI: 1.06–3.07; p = 0.02) each predicted high psychological distress even after adjusting for other variables in the model. High distress was more likely if the vaccinated girls were in P5 (1.92, CI: 1.27–2.89; p = 0.001). On the other hand, vaccinated girls were less likely to be psychologically distressed if they perceived high parental communication even after adjusting for other variables entered in the model (0.50, CI: 0.35–0.72; p = 0.000).
Predictors of adolescents’ worrisome thoughts about vaccination and psychological distress derived by complex samples logistic regression analysis.
HPV: human papillomavirus; CI: confidence interval; ORs: odds ratios.
Number of observations = 768.
Significant at p ⩽ 0.05.
Significant at p ⩽ 0.01.
Significant at p ⩽ 0.001.
Discussion
Worrisome thoughts about vaccination
Contrary to our expectation, HPV vaccination was not a strong predictor of adolescents’ worrisome thoughts about vaccination except among P5 girls (recently vaccinated). Our qualitative results showed that the introduction of the HPV vaccine had triggered several negative rumors that were likely to worry the targeted adolescents. Apparently, much of the reported worry especially concerning pain and possible physical harm from injection was anticipatory, which is a normal response to a perceived threat but an unnecessary psychological trauma if no real threat exists (Bernard et al., 2011). Qualitative findings indicated that the fear of pain and possible physical harm from injection was usually short-lived, suggesting that time was protective. Some girls were said to have missed the initial vaccine dosages out of fear only to accept later after realizing that those who had taken the vaccine had not been harmed. This behavior toward the HPV vaccine has previously been noted among both adolescents and their parents in other studies (Katahoire et al., 2013; Watson-Jones et al., 2012). This finding underscores a need for future HPV vaccination programs to provide for second chance of dose 1 for girls who initially avoid the vaccination out of fear but change their minds later after learning from the experience of their vaccinated colleagues that the vaccine is not harmful.
Psychological distress
As expected in this study, HPV vaccination predicted high psychological distress. Comparable studies that assessed adolescent psychological distress in Uganda using the SDQ and a cut-off TDS score of 16 to indicate clinically significant distress were conducted on adolescent samples drawn from largely high-risk populations in war-affected northern and north-eastern Uganda. Compared to the findings of those studies, the 39.3 percent prevalence of clinically significant distress among vaccinated adolescents in our study is clearly higher than the 18.3 percent among non-abducted adolescents but lower than the 51.2 percent among war-abducted adolescents in a war-affected district (Okello et al., 2007); it is higher than the 23.1 percent among males and 29.7 percent among females from two war-affected and two non-war-affected districts (Abbo et al., 2013); and it is greater than the 33.7 percent among never abducted adolescents but less than the 67.7 percent among child soldiers in a war-affected district (Moscardino et al., 2012). Our finding showing relatively high prevalence of distress among vaccinated girls is inconsistent with previous studies of children aged 7–11 years undergoing other vaccinations (cited in Cohen et al., 1999). It was observed from those studies that as children grow older they acquire more sophisticated and logical thinking processes and are able to understand the rationale for medical procedures. Consequently, they tend to experience decreased worry, distress, and pain (Cohen et al., 1999). However, our findings are largely in agreement with findings of a qualitative study of adolescent girls’ response to HPV vaccination under a school-based vaccine delivery arrangement (Bernard et al., 2011), which reported that fear and distress were commonplace. The study cited negative rumors or misinformation among the major factors behind the girls’ fear response. In our study, qualitative results show various negative beliefs the vaccinated adolescents had about HPV vaccination, which were largely based on rumors and misinformation. Different FGD participants heard rumors that those vaccinated could become barren or would face a greater risk of life-threatening childbirth complications, HPV vaccination was a disguised plot to gradually kill those vaccinated in unexplained ways, HPV vaccine causes cervical cancer that would ultimately kill the vaccinated person, and that those vaccinated would in future give birth to only twins. Some of the vaccinated girls worried that it would alter their menstrual cycles. The rumors, though not backed by scientific evidence, are potential sources of distress if they are perceived as real. The positive association between major stressors (such as negative life events) and psychological distress has been previously confirmed among adolescents in studies not related to HPV vaccination (Huang et al., 2009; McMullen et al., 2012; Qouta et al., 2007) and among adults (Jackson and Finney, 2002; Mak et al., 2005; Muhwezi et al., 2007; Tein et al., 2000). Life events are especially distressing if they are viewed as undesirable, uncontrollable, and unpredictable. Distress is predicted by the perception and the interpretation of events (as negative) rather than by the occurrence of events themselves (Jackson and Finney, 2002). In the context of HPV vaccination, it has been reported that being informed about and well prepared for vaccination may reduce the propagation of myths, rumors, and other fear-promoting factors (Bernard et al., 2011). Future HPV vaccination programs should include sensitization strategies to counteract the negative rumors about the vaccinations so as to mitigate fear and anxiety among targeted adolescents. In particular, future sensitization programs need to explore the strategy of peer involvement to mitigate targeted adolescents’ fear and anxiety. This suggestion is informed by this study’s qualitative findings showing that some vaccinated girls were mainly encouraged by their friends to get vaccinated against the HPV without fully understanding the vaccination, implying that peer influence may have mitigated some of the fears associated with the vaccination. Moreover, in the study by Bernard and colleagues (Bernard et al., 2011), girls’ concern about their possible reactions in front of peers seemed to amplify fear in some cases, yet having friends available in the vaccination room to support them by holding arms or talking to them had the reverse effect. Peer influence is a known positive and negative factor in shaping adolescent attitudes about HPV vaccine (Gamble et al., 2010; Katz et al., 2013).
High psychological distress was especially likely if the girls were recently vaccinated, consistent with findings of another study that reported high levels of distress among children during immunization in different settings (Cohen et al., 1997). High distress was also likely if the vaccinated girls reported worrisome thoughts about being vaccinated. This is in agreement with findings of the study (Bernard et al., 2011), which concluded that fear of HPV vaccination was a near universal experience among adolescents undergoing HPV vaccination in a school setting and was often associated with significant distress. Some girls in our study’s FGDs reported that while being prepared for HPV vaccination, they felt anxious that it would cause pain, bleeding, and irritation based on what they heard from their friends at school. Many reported experiencing pain and swelling at the injection site, which initially caused worries of probable long-term physical damage. These and other known adverse effects like dizziness and headache have been reported in other studies (Katahoire et al., 2013) but they are usually brief and non-serious (WHO, 2012). Since this study’s findings suggest that both distress and worry especially about pain and possible physical harm from injection were largely short-term phenomena, there is a need for a psychosocial component in future HPV vaccinations to identify and address the affected adolescents’ immediate psychosocial needs. Health workers who administer the vaccines should be trained in basic cognitive and behavioral interventions that have been found to be effective in mitigating pain and distress during immunizations such as child distraction, encouraging the child to adopt the stress-reduction behavior, imagery, preparation/information, modeling, rehearsal, and desensitization (Bernard et al., 2011; Cohen et al., 1997; French et al., 1994; Uman et al., 2008). Also, future sensitization prior to HPV vaccinations should clearly articulate the non-serious nature of the vaccine’s known adverse effects, their remedies, and their brevity so as to minimize worry and possible distress among the targeted adolescents.
Perceiving high parental communication mitigated adolescent girls’ distress associated with HPV vaccination. This finding supports results of studies (Jackson et al., 1998; Raveis et al., 1998), unrelated to HPV vaccination, which show positive correlation between parent–adolescent communication and adolescent psychological well-being. Our finding is in agreement with the study of adolescent girls undergoing HPV vaccination (Bernard et al., 2011), where the responsible nurses reported that parents who discussed HPV vaccination with their adolescent girls helped to prepare them for the experience. It is likely that in our study, the parents/guardians of vaccinated adolescents who communicated highly with them included HPV vaccination in their discussion with the adolescents and prepared them for the vaccination, leading to parental communication’s mitigation of the adolescents’ psychological distress. Future HPV vaccination programs should invest adequately in sensitization of parents and encouraging them to talk about vaccination with their adolescents so as to make vaccination a positive experience for the adolescents.
Study limitations
There was a possibility of recall bias among P6 girls who were interviewed 1 year after vaccination. This was minimized during assessment for psychological distress by asking respondents how things had been in the last 6 months. One vaccination cohort (P6) was studied 1 year after completion of vaccination while the other (P5) was studied 1 month after vaccination. This was a limitation because some of the psychological reactions to the vaccinations may have subsided or cleared by the time of field data collection for this study. Moreover, girls who may have developed clinically significant psychological distress during vaccination were denied opportunity for early detection and treatment. Future vaccination programs need to include psychological distress assessment to further evaluate time pattern of clinically significant distress and to enable the affected girls to get timely therapeutic intervention. The age design of the study whereby the same age cohort was simultaneously studied poses a limitation, whereby cause–effect relationships between variables could not be evaluated. Longer term follow-up studies involving more than one age cohort will be needed in future to be able to assess adolescents’ long-term psychological reactions to HPV vaccinations. A major limitation of the study was the assessment of worrisome thoughts by asking respondents whether they had ever felt very worried about having been vaccinated or going to be vaccinated without delving into aspects such as content and degree of worry. All respondents who responded affirmatively to the question were treated as homogenous during data analysis, which may not be true in reality. This limited the ability of the study to make recommendations for focused clinical interventions to deal with adolescents’ anxiety associated with the vaccination. Furthermore, by not specifying worry content, the study left a possibility that the worry reported by respondents could have been induced by some of the topics covered in the study questionnaire such as cervical cancer and dangers of pre-marital sex and the future, which may be independently related to worry about HPV vaccination. Future studies should adopt standard instruments for assessment of worrisome thoughts whose findings would provide firm basis for recommendation of clinical interventions to deal with the affected adolescents’ anxiety. Selection of FGD participants with the help of their teachers targeting girls sharing certain characteristics may have biased the qualitative results since those girls could have shared attitudes associated with shared characteristics. The effect of this was minimized by organizing FGDs in five different schools. The main strength of this study is its relatively large sample size that increased the power of the study tests.
Conclusion
We set out to assess the influence of HPV vaccination on adolescents’ experience of worrisome thoughts about vaccination and levels of psychological distress. HPV vaccination predicted experience of worrisome thoughts only among recently vaccinated adolescents. Several HPV vaccination-related fears were reported qualitatively among vaccinated girls. HPV vaccination predicted high psychological distress, more so if the adolescents had completed their vaccination doses recently, and had experienced worrisome thoughts about being vaccinated. Perceiving high parental communication mitigated psychological distress among adolescent girls vaccinated against HPV. This study makes several recommendations for future HPV vaccination program design and research based on findings, related literature, and study limitations.
Footnotes
Acknowledgements
We recognize Professor F. Mirembe and Dr C. Banura, Principal Researcher and Senior Researcher, respectively, in the Young Women’s HPV Project for reading the research work and advising at the different stages.
All data for this research are freely accessible and are contained in the manuscript. Further data are not able to be publicly shared as the data were collected without consent for public sharing by the Higher Degrees Research and Ethics Committee of the Makerere University Medical School (now College of Health Sciences; REC REF 2011-200) and the Uganda National Council for Science and Technology (REF SS 2478). Requests for further data should be addressed to Andrew Kampikaho Turiho, the Corresponding Author on
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
This study was principally funded by the Uganda National Council for Science and Technology through the Millennium Science Initiative/College of Health Sciences (CHS) Young Women’s HPV Project based in CHS, Makerere University. The authors received other financial support from the Carnegie Corporation of New York—Makerere University Next Generation of African Academicians (NGAA) Project (2012). The study also received training support by grant no. 5R24TW008886 supported by OGAC, NIH, and HRSA.