Predicting cervical screening and HPV vaccination attendance of Roma women in Hungary: community nurse contribution is key

As reported by the World Health Organization, Human papillomavirus (HPV) is an established cause of cervical cancer and ranks # 4 globally of all cancers in women [1]. Screening and vaccination are therefore key to prevent the disease. However, screening and vaccination participation remain a challenge, especially for women of ethnic minorities [2]. While ethnic groups were reported to have become more active in HPV vaccination uptake, they are still lagging behind in follow-through [2]. We also recognize that cultural HPV awareness and knowledge have a strong influence on vaccination attitudes and outcomes [3, 4]. Besides ethnic affiliation and personal knowledge, religious orientations were also identified as either constraints or enablers of vaccination [5, 6]. Riza et al. [7] confirmed that socio-economic background and education were important facilitators of Pap smear testing and HPV vaccination. They argued that Roma ethnic origin was associated with misguided beliefs concerning cervical screening and the HPV vaccine. Jackson et al. [8] described cultural aversion against HPV vaccination within travelling Roma communities which may also be part of a larger public mistrust [9]. A number of studies stressed that concerns over vaccine safety was another major barrier of getting vaccinated [10, 12, 13]. Income (access to vaccine and healthcare) and actual logistics of obtaining the vaccine and seeing a healthcare professional also prevented seeking HPV vaccination [11]. Getrich et al. [14] reasoned that favorable healthcare provider attitudes were essential in the decision to get screened and immunized. However, when healthcare staff, particularly female healthcare professionals, were surveyed, studies reported lower than expected HPV vaccination alertness and patient immunization encouragement by healthcare staff [15‐17]. To achieve better screening and vaccination coverage, the quality of the patient-provider relationship between members of the Roma community and nurses and doctors was considered critical for prevention and immunization success [18]. Why this relationship may be so essential was underscored by Ilisiu et al. [19] who observed non-ethic women choosing hospital services for HPV screening whereas the majority of female Roma sought attention from general practitioners. Finally, Vu et al. [20] underlined that research should investigate how clinicians’ views and actions towards minorities influence HPV testing and vaccination to help develop or improve culturally sensitive interventions.

In conclusion, while global HPV screening and vaccination rates have increased for ethnic groups, there are still barriers to fully utilizing preventive services. Obstacles have to do with cultural norms and expectations but are also related to socio-economic reasons. Academics emphasized conducting research that looks at the quality of interactions with health professionals and their impact on screening attendance. Therefore, this research aimed to explore causes that predict past HPV screening and vaccination attendance in Roma women in Hungary. Since cultural and healthcare provider contexts vary and may differ across countries, authors developed their own, culture specific instrument for the purposes of this research. Indicators revealed in the literature review informed and guided item development.

The actual research design was based on a cross-sectional survey with the intent to identify factors associated with past participation in HPV screening/vaccination. Data was collected between July–September 2019. Potential participants received a mail contact and called researchers to indicate their willingness to join. Local community nurses were contacted and trained by the research team to identify participants using inclusion/exclusion criteria. The same nurses contacted participants, handed over the research instrument and supported subjects in responding to items if requested. To avoid respondent bias, nurses only explained the research instrument but did not partake in responding. Participants forwarded surveys postal paid to the research team in sealed and unidentifiable envelopes. Final data entered was anonym and researchers could not trace it back to the original respondent.

Table 1 displays sample characteristics. The average age of our sample was 36.4 years (SD 11.3). A total of 9% lived alone while 81% lived with some significant other (husband, relatives or children). In terms of education, 61.4% had less than or equal to primary school degree, 33.4% held high school degree and 5.2% graduate or postgraduate. Of the total sample, only 62.5% had active employment. As for financial status, 36.1% said their financial situation was below average, 62% thought they had average income, and 1.9% reported above average standards of living. Finally, 17.4% of women attended at least one cervical screening and HPV vaccination in the past. When asked about the first time it happened, the average age was 21.1 (SD 6.9) years, 14 years of age being the youngest and 54 the oldest.

Considering self-reported health of participants, average health rating (on a 5-point scale) was 2.96 (SD 0.96). The average score on the full screening attitude scale was 47.00 (SD 7.49).

Table 2 presents outcomes of correlation analyses. In terms of the association between self-reported health and employment, those actively employed enjoyed better health. Education was also positively associated with health; greater education was linked to better health. Income however did not correlate with past screening attendance. Self-reported health however did not relate to past screening. Education, however, was positively associated with past attendance, greater education increased screening attendance. The fact that our respondent did not co-habit with someone had no influence on past screening attendance. Future screening attendance was positively associated with income and education; more income and greater education increased the potential of attending future screening (r_income = 0.17, p < 0.001, r_education = 0.19, p < 0.001).

Past negative screening experience was strongly correlated to doctors treating respondents racially unfair; the more unfair the personal conduct had been, the more past experience was appraised negatively. The opposite was true for community nurses; the more community nurses had been involved in screening, the less respondents reported negative experiences.

Finally, we preformed binary logistic regression predicting previous cervical screening attendance and HPV vaccination. The full model was significant (𝒳² = 121.96, p < 0.001, -2LL = 156.96) and resulted in an 87.9% correct classification and a good model fit. The model explained 44% of the variance in HPV screening/vaccination attendance. Table 3 presents outcomes of the logistic regression model. Variables that significantly predicted past cervical screening attendance were friends having cervical cancer, community nurse performing the check-up and vaccination, respondent feeling no shame, procedure being painful, respondent only financially motivated and had a negative prior experience, and less travel involved. Variables that made the biggest contribution to past screening attendance, expressed by the absolute value of the odds (odds ratio) in the model, were past negative experience (4.503), community nurse involvement (3.325) and feeling shameful (1,609).

The main objective of the current research was to predict past cervical screening attendance and HPV vaccination in a sample of Roma women. Past cervical screening attendance among Roma women was 17.4% which is not significantly lower compared to the general Hungarian female population (22.0–23.3%) [28]. However, when we compared our 3-year screening coverage to the general population, numbers fell unfavorably for our sample (7% vs 52.6%) [28].

As expected, we supported that self-reported health of respondents was positively correlated with greater income and education. We found that income and health were not linked to past screening attendance or HPV vaccination. Greater education, which we attributed to increased knowledge about health, was positively linked to past screening and vaccination. When considering our screening attitude assessment, both income and education was positively linked to screening attendance and vaccination (more income and better education increased turnout probability). These results were in support of Adjei Boakye et al. [3], Jeudin et al. [11], Marlow [4] and Riza et al. [7].

Our logistic regression model predicted past screening and vaccination with increased precision (88% correct classification). The greatest odds of attending screening/vaccination was attributed to ‘no negative prior experience’. Those who did not have such experience were 4.5 times more likely to seek screening/vaccination than those who had a negative past influence. In order of magnitude, community nurse performing the process was the second most influential variable. Those who favored the community nurse performing the check-up/vaccination were 3.3 times more probable to participate. The third most influential variable was ‘feeling no shame’. Those who agreed to ‘feeling no shame’ about the procedure were 60% more prone to attend screening/vaccination compared those who felt ashamed. All three findings confirm and support Getrich et al. [14] and Vu et al. [20] who argued that the quality of the patient-provider relationship was a critical aspect of screening and immunization preference.

Appraising the screening/vaccination process painful, receiving financial compensation to attend, and attendance involving a lot of travel decreased the odds of ‘no attendance’ by 50%, 40% and 41%, respectively. When respondents’ friends suffered of cervical cancer, the odds of screening attendance/vaccination dropped by 53%. The result is opposite to expectations and authors have no immediate explanations for this outcome. More in depth research is suggested to explore the background of this behavior.

Note that while community nurse involvement was a significant determinant of HPV screening/vaccination, ‘doctors acting racially unfair’ as well as ‘sampling performed by the physician in his (GP) office’ did not make any contribution to explaining attendance. The same was true for ‘preventive reasons’ and ‘cancer in the family’, which did not reach significance in our model. These outcomes may have to do with cultural norms, but further research is recommended to confirm the assumption.

Referring to the ratio of past cervical screening attendance, we conclude that our female Roma sample did not behave differently from the general population. We saw evidence that racial mistreatment negatively influenced cervical screening/vaccination participation, but when weighed together with other variables, racial mistreatment did not emerge as a significant predictor of screening attendance. Not feeling shameful about the screening procedure increased the odds of screening/vaccination participation. The positive role of the community nurse and her involvement in the procedure was strongly emphasized by respondents. Having no negative prior experience with screening increased the odds of participation the most. Cervical cancer of friends, procedural pain, financial motivation and travel distance all decreased odds of participation but to a much lesser extent.

In order to improve future screening and immunization, community nurses should play more central and advanced role in the organization and implementation of such services that specifically target Roma populations.