HIV/AIDS knowledge level, awareness of public health centers and … – BMC Public Health

Study scenario

This cross-sectional study was conducted using a self-administered, anonymous questionnaire survey.

We first developed an original questionnaire referring to previous studies [16,17,18,19]. Then, we considered methods to deliver the questionnaire to as many of the survey targets as possible so that they would respond to it. The actual data sampling method will be described in detail later. The obtained responses were statistically analyzed at SUMS.

The data collection period, which was originally scheduled to run from April 1 to August 31, 2021, was extended to March 31, 2022 because the survey’s implementation was affected by the restriction of outings due to the increase in the number of COVID-19 patients.

Response methods

The survey was conducted by combining a method of accessing and filling out a Google form via a QR code/URL included on a flyer (web survey method) and a method wherein completed questionnaires were to be directly returned to the university by mail (mail method). Respondents were instructed to choose one of the two methods to answer questions. For both the web and mail methods, a check box was provided at the beginning of the questionnaire to confirm the respondent’s consent.

Data source

Participants

The survey population consisted of 7,539 Brazilian nationals (as of December 31, 2019), excluding those under 18 and over 80 years of age. The target sample size (n’) was calculated using the following Eq. 1 (assuming 5% margin of error, 95% confidence level, and 50% response rate) and corrected by the population size (7,539) using Eq. 2, resulting in 366 respondents.

n; target sample size before finite modification.

n’; target sample size.

λ; reliability (= 1.96).

d; allowable limit of error (= 0.05).

p; response rate (= 0.5).

N; population (= 7,539).

Data sampling

For data sampling, we selected 28 places that are frequently visited by Brazilians, and the first author visited them to solicit their cooperation in the survey. These places included one restaurant, six Brazilian fresh food stores, four city halls in Shiga, two Brazilian nursery schools, two Brazilian schools, two recruitment agencies, one Brazilian Portuguese-Japanese translation agency, three Catholic churches, one welfare council, four international associations, one Japanese language school, and one business office that directly employed Brazilians.

The first author requested a meeting with the person in charge at each site and visited the places that granted permission, explaining in detail the intent and purpose of this study. In cases where it was judged that it would be easier to communicate with the site representative in Brazilian Portuguese (hereafter referred to as Portuguese) than in Japanese, the first author either visited while accompanied by an interpreter or used an interpreter via telephone or ZOOM interview. A total of 25 interpreting requests were made to three professional interpreters. All 28 sites agreed to participate in the study.

The first author asked the cooperating individuals at each site to post flyers in their facilities describing the intent, purpose, and method of the survey, to distribute questionnaires and flyers to their customers and employees or co-workers, and to encourage those who received the flyers to respond to the questionnaire. The first author also personally visited the six Brazilian fresh food stores and three Catholic churches and asked customers and worshipers to respond to the survey. In addition, the Shiga International Association was asked to publish an article about this survey in its institutional newsletter, and on its website with a link to the webpage for responses. With the cooperation of interpreting consultants at the city halls, teachers and clerks at schools and nursery schools, and recruitment agencies, flyers for the survey and a call for responses were sent out via social networking sites that each group uses as their own personal contact network.

Original questionnaire items and language

Question items included respondents’ age, sex, sexuality, length of stay in Japan, existence of roommates (including family members), employment status, coverage of a Japanese public health insurance, existence of their family doctor/regular doctor, area of residence, primary language of daily life, ability to speak and read Japanese, knowledge of HIV/AIDS, and knowledge of the HIV/AIDS treatment system in Japan.

Respondents were asked to select one of the following five levels of their Japanese speaking ability. Level 1 (S1); can make speeches and debate, Level 2 (S2); can do simultaneous and conference interpretation, Level 3 (S3); can hold daily conversation, Level 4 (S4); can hold simple conversation, and Level 5 (S5); cannot speak it at all.

They were also asked to select one of the following five levels of their Japanese reading ability. Level 1 (R1); can understand newspaper editorials and reviews, Level 2 (R2); can understand general newspaper articles, Level 3 (R3); can understand an overview of information from newspaper headlines, Level 4 (R4); can read simple Japanese characters, and Level 5 (S5); cannot read it at all.

The above “speaking” and “reading” level classifications were established by the authors with reference to the Japanese Language Proficiency Test [20] and Japanese elementary school education standards. The best level of Japanese speaking/reading ability is S1/R1 and the most inadequate is S5/R5.

Referring to previous studies [16,17,18,19], we set up four questions to assess the knowledge level of HIV/AIDS, which were designed to inform respondents that recent HIV/AIDS treatment enables reintegration into society, motivates them to take HIV testing and treatment, encourages them to prevent secondary infection, and makes them aware that HIV/AIDS is not just someone else’s problem but something that affects them, their families, and their loved ones. The more correct answers to the four questions, the more correct knowledge about HIV/AIDS the respondent had.

Although many Brazilians speak Portuguese as their first language, there are some young Brazilians who have not fully determined whether their first language is Portuguese or Japanese because they were born in Japan or came to Japan with their parents when they were young and were therefore educated in Japan [21, 22]. Therefore, the language used for each questionnaire item was written in both Portuguese and Japanese. The Japanese was translated into Portuguese as simply as possible so that those with at least a secondary education could read and understand the content. Translation between Japanese and Portuguese was requested from the Center for Multicultural Society Kyoto, a non-profit organization.

Analytical methods

The author loaded the responses from the web survey into Microsoft Excel, entered the mailed responses, and then tabulated the results by item.

A multiple logistic regression analysis was used to examine the factors related to “Knowledge of HIV/AIDS” and “Awareness of PHCs”. The dependent variable, “Knowledge of HIV/AIDS”, was binarized according to whether the respondents were aware of all four knowledge items (more knowledge) or three or fewer items (less knowledge). The dependent variable, “Awareness of PHCs”, was binarized according to the response to “Are you aware of the existence of PHCs?”.

For the independent variables to be included in the logistic regression analysis, we drew directed acyclic graphs (DAGs) [23] using the items of the basic attributes of the respondents in Table 1 and selected variables that were more strongly confounded by the two dependent variables. A DAG is a graph in which one-way arrows are used to represent known causal effects based on prior knowledge [23]. Regarding the number of independent variables, it was decided that there should be no more than 7 for each dependent variable, since there were 74 respondents for the question of knowledge level of HIV/AIDS for the smaller number and 76 respondents for the question of awareness of PHC for the smaller number.

We finally selected independent variables as follows; “age”, “sex”, “existence of roommate(s)”, “employment status”, “length of stay”, and “Japanese proficiency”. “Age” was categorized into four groups of ≤ 29 years old, 30 years, 40 years, and 50 years or older, and “length of stay in Japan”, was categorized into four groups according to quartile values: ≤5 years, 6 to 17 years, 18 to 25 years, and ≥ 26 years. Japanese proficiency was reclassified into three groups: advanced (combination of [S1 + S2 + S3] and [R1 + R2 + R3]), beginner (combination of [S4 + S5] and [R4 + R5]), and intermediate (other than beginner and advanced).

The independent variables had a variance inflation factor of less than 1.6. All variables were entered using a forced entry method. The goodness of fit of the model was tested using the Hosmer-Lemeshow test. Data were analyzed using the Statistical Package for Social Sciences software program (SPSS, IBM version 29), and the significance level was set at 0.05.