Study design
Secondary data analyses were conducted using the activity records of an NGO located in Aichi Prefecture, Japan.
Data source
The NGO, Medical Information Center, Aichi (MICA), has provided foreign residents with medical information and free monthly consultations since 1998. Supported by volunteer doctors, dentists, and nurses, as well as non-medical volunteers, the free consultations are provided in various locations in Aichi Prefecture, with each location being visited in rotation. The documentation for these consultation activities comprises a four-page form that the MICA staff members and volunteers complete for each attendee. The pages cover: 1) personal information, including sex, age, date of birth, nationality, occupation, years living in Japan, whether the attendee is covered under a Japanese public insurance scheme, medical history, family medical history, complaints and symptoms; 2) lifestyle, 3) data obtained through physical check-ups (height, weight, blood pressure, pulse rate, body temperature, and urine analysis results) and the results of the consultation with the doctor, including an overall assessment (i.e., the actions the attendee should take); and 4) oral check-ups and the results of consultations with a dentist, with an overall assessment (the actions the attendee should take). The form is available in 10 different languages (English, Portuguese, Spanish, Chinese, Korean, Vietnamese, Filipino, Indonesian, Burmese, and Thai), which ensures that most attendees can understand the questions. All of the information gathered on these forms is entered into a data-management program by MCIA staff members.
Participants
For the present study, anonymous data of all adults (aged 18 years or older) who participated in the free monthly consultation activities between 2012 and 2016 were extracted from the full dataset; this was performed under the terms of a data-use agreement between the first author and the NGO (MICA). We used 2012 as the start date for the data extraction because in 2012 the Japanese government made a legislative change that allowed foreign residents who lived in Japan for over three months to apply for National Health Insurance (previously, only foreign residents who had lived in Japan for over one year were eligible.) Cases were excluded if the age was unknown (four cases), nationality was Japanese (16 cases), the participant was a traveler (three cases), and if all records except attributes were missing (one case). Thus, of a total of 632 cases, 608 were used for the analyses.
Access to health care framework
There are various models and frameworks to consider access to health care [28–30]. For example, the model published by the Institute of Medicine in the United States [31] applies “use” of services as the intermediate indicators of its access to health care model. Andersen and colleagues [32, 33] have developed several frameworks over the years. In their frameworks, both individual characteristics and contextual characteristics influence health care access. Penchansky and Thomas [34] defined access as entry into or use of the health care systems and suggested specific dimensions, such as availability, accessibility, accommodation, affordability and acceptability, to describe the “fit” between the patient and health care system. Levesque et al. [30] suggested access as “opportunity” based on the review of previous studies. International surveys often use a question asking “unmet health care needs” [35], which are also often used in studies on nationals and migrants in Japan [22, 23, 36].
Variables measurement
Explanatory variables
All available participant characteristics included in the dataset, such as sex, age group, region of origin, occupation, whether he/she was covered by a Japanese public insurance scheme, and time living in Japan, were used as explanatory variables. Supply-side variables were not included in the dataset we used. Therefore, only attendee (demand-side) variables were included as explanatory variables in our analysis model.
Region of origin was defined based on the nationality recorded on the form and categorized based on United-Nations-recognized regions, unless the nationality was Brazilian, Chinese, Filipino, Vietnamese, or a high-income country that is defined by the World Bank. As Brazilian, Chinese, Filipino, and Vietnamese were the four major groups, they were treated as a single category. “Other Asia” featured 68 participants from South Asia (five countries), 34 from South-East Asia, except Vietnam and the Philippines, (three countries), two from West Asia (one country), and one from Central Asia (one country). Participants included in the “other” region were as follows: 35 from Latin America, except Brazil, (five countries), six from East Africa, three from West Africa, three from North Africa, one from Central Africa, two from Russia, and one from Melanesia.
In the data-collection form, attendee's occupation was determined using an open-ended, self-reported answer and, for the present study, we developed occupational categories based on these answers. If the type of factory work or construction work was recorded, or if the occupation was simply recorded as “factory,” the participant was categorized as engaging in “factory or construction work.” If part-time or contingent work (Arubaito, Paato, or Haken, in Japanese) was recorded as an occupation, the participant was categorized as having a “non-regular job.” “Trainee” and “intern” were combined to form “trainee/intern.” These three categories were then further grouped into the category “labor and non-regular,” because they often overlap and are used interchangeably in daily conversation. Professional workers, such as teachers, engineers, and company workers, were also combined because they are considered to represent the formal sector. Housewives and the unemployed were combined because these people usually stay home and, again, these terms are sometimes used interchangeably in daily conversation. “Other” included nine people engaged in “service,” seven helpers (care assistants), four people engaged in religion-related work, three salespeople, three businesspeople, three cooks, two refugees, two volunteers, one babysitter, and one diplomat. Two occupation records could not be categorized, and were also included in “other.”
Outcome variables
All cases for which there were any records on the third page of the form were defined as “had a medical consultation.” Two items were used as the outcome variables and considered as indicators of having barriers to health care: 1) “being advised to visit a medical facility” (attributed if the participant was advised to visit a medical facility); and 2) “being referred to a medical facility” (attributed if the participant was referred to a medical facility, with or without a referral letter). We assumed that participants who were advised or referred did not have easy access to the formal health-care system before attending the NGO’s consultations. Participants identified as 2) were part of 1), but were considered more serious cases.
Data analysis
First, participant characteristics (sex, age group, region of origin, occupation, whether he/she was covered by a Japanese public insurance scheme, and time living in Japan) were described. The frequency and the percentage of each category were calculated. Second, the two indicators of barriers to health care (“being advised to visit a medical facility” and “being referred to a medical facility”) were analyzed in terms of participant characteristics using chi-square tests. Third, log-binomial regression analyses were performed to calculate the prevalence ratios (PRs) of “being advised to visit a medical facility” and “being referred to a medical facility” for each participant characteristic. Variables that showed, in the chi-square tests, statistical associations for each of the two access indicators were included in the model. Furthermore, if the participants were referred to a medical facility with or without a referral letter, the recorded consequences were described. Stata SE version 12.1 (Stata Corp) was used for all statistical analyses.