HIV Pre-Exposure Prophylaxis and Postexposure Prophylaxis in Japan: Context of Use and Directions for Future Research and Action

Broader ethnographic study

In 2011 and 2012, we reviewed available epidemiological information on HIV at regional and national levels in Japan. This consisted primarily of MHLW surveillance reports with data from 1985 to 2011, ⁵⁷ plus a limited number of published epidemiological studies with regional and national level results. ^{58 –62} Measures of HIV occurrence reported in these sources were very small when compared with most countries globally and even within the Asia-Pacific region. ^{63 –71}

This could have indicated a very low HIV burden in Japan that was real, including remarkably low risk in the general population, or it could have meant that a significantly larger HIV epidemic was being obscured. The answer was not known with confidence. Also, due partly to a dearth of social science and qualitative research, little was known about how well available surveillance and epidemiological data captured the reality of HIV in Japan, and there was much to learn about why recent trends were occurring. These gaps in knowledge prompted us to conduct our broader study of HIV in Japan, which had two comprehensive aims: first, to determine how accurately official epidemiological data represented local HIV epidemics as understood and experienced on the ground, and second, to identify contextual factors that explained why the HIV epidemics were unfolding as they were.

We used a social science of medicine approach ⁷² to design our broader study, which was primarily ethnographic while also drawing upon epidemiology. This approach allowed for a flexible adoption of qualitative strategies from medical anthropology and sociology, while also enabling a role for epidemiology in informing the ethnographic procedures, analyses, and interpretation. This first involved 7 months of fieldwork in 2013 (January–August), conducted in two areas with the country's largest regional epidemics: urban Kansai and the Tokyo Metropolitan Area. In the Kansai region, this included four cities: Osaka, Kyoto, Kobe, and Nara. In the Tokyo Metro Area, the main field site was Tokyo, although we also collected some data in neighboring Kanagawa and Chiba prefectures. Data collection comprised observations, formal qualitative interviews, and document research, supplemented with emails.

Based on both epidemiology and theory, we chose key stakeholder groups in local HIV epidemics as priority populations on whom to focus our research. In our first month of fieldwork, we had collected and coded our initial ethnographic data, including field notes from observations and document research, and we had completed a close study of surveillance reports and unprocessed surveillance data publicly available from MHLW. ⁷³ The most recent MHLW data pointed to MSM as the primary population driving the national epidemic in Japan. In 2012, 72% of new HIV cases were through homosexual contact, almost all among men, ⁷⁴ which was likely an underestimate. ⁶¹ Transgender persons were not disaggregated in MHLW data, but we knew from previous research in Japan and globally that this population was highly vulnerable to HIV. ^75,76 The crossover risk between HIV and other sexually transmitted infections (STIs), including diagnosis and risk behaviors, was also well known. ⁷⁷

Our initial ethnographic fieldwork led us to discover significant intersections between HIV and violence across population groups, perpetrated by others and directed at the self. Guided by syndemics theory, ⁷⁸ research in Japan by Hidaka and others, ^{79 –81} and studies in other countries, ^82,83 we predicted these intersections would be a major factor in explaining why Japan's HIV epidemics were unfolding as they were. In our own research in the United States, we had found that healthcare, mental health, and social service providers' daily interactions with persons living with or at elevated risk of HIV (e.g., MSM, transgender women, persons affected by violence, abuse, suicidality, or nonsuicidal self-harm) uniquely positioned them to discern patterns in the groups they served. ⁸⁴

Considered together, this information allowed us to select seven main categories of stakeholders on whom to subsequently focus our observations and qualitative interviews: (1) interventionists (healthcare, mental health, and social service providers; community organizers) working directly on the HIV problem; (2) MSM and transgender women; (3) interventionists working with MSM or transgender women; (4) persons living with HIV; (5) persons diagnosed with an STI or presenting for HIV or STI testing or counseling; (6) persons who had experienced or perpetrated violence, abuse, suicidality, or nonsuicidal self-harm; and (7) interventionists working with persons affected by violence, abuse, suicidality, or nonsuicidal self-harm.

We used a combination of ethnographic mapping, purposive sampling, and theoretical sampling. For observations, we used ethnographic mapping to locate sites in local communities where people from these stakeholder groups congregated in clusters. We then used theoretical sampling to fill gaps in the logic of findings that emerged while in the field, which needed to be further explored by observations of additional sites and persons, or by repeated observations of the same sites and persons. Observations of the target groups often led also to observations of persons who fell outside those groups when it was clear that relevant data could be captured.

For the qualitative interviews, we used purposive sampling to recruit from the seven target groups. Concurrently, we used theoretical sampling to fill gaps in our emergent understanding of the data, which required additional participants or needed to be challenged with negative cases until data saturation was exceeded. For document research, we similarly used purposive sampling to locate documents with HIV content, and theoretical sampling to acquire other content relevant to regional HIV epidemics that filled gaps in tentative, evolving findings.

We understood the importance of other populations that have faced increased HIV risk at various points in the history of Japan's national epidemic, including sex workers, drug users, young adults, and foreigner nationals. ^7,56,75 Because we already had seven primary stakeholder categories, and to the keep the study feasible, these were not priority groups for sampling in our study. However, we anticipated that we would encounter people from these groups in our interactions with the seven main stakeholder groups. This occurred, and we observed and interviewed sex workers, drug users, young adults, and foreign nationals, who often qualified for inclusion in the study because they concurrently represented one of the seven main stakeholder categories (e.g., a cisgender female sex worker who was also an HIV interventionist, and an MSM drug user).

To augment validity and reliability, we used in-person member checking with study participants in Japan to confirm, challenge, and alter our evolving understanding of the data. ⁸⁵ We continued follow-up member checking through email (n = 9 persons) from the United States through December 2014.

Nested study

In meeting the broader study's second aim, we discovered a general lack of awareness in 2013 among HIV interventionists and community members regarding biomedical HIV prevention technologies, including PrEP and PEP. This prompted us to conduct the present nested analysis on PrEP and PEP, for which we used the original 2013 data and newer data collected between August 2013 and July 2016. This entailed an expansion of document and web-based data sources, further email correspondence with key informants, and additional fieldwork in Tokyo in April, May, and July 2016, and in Osaka in July 2016.

Observation was a full-time method and consisted of both unobtrusive and participant observation. Unobtrusive observation was conducted covertly to capture naturally occurring behavior, including social interactions among persons observed, while reducing reactivity based on knowledge of being observed. It involved no direct interaction between us and the persons observed. Participant observation was overt and included verbal interaction between us and observed persons; this frequently took the form of informal, conversational interviews in natural environments. Both forms of observation also allowed us to investigate key aspects of the physical and sociocultural environments in which observed persons were embedded. Oral informed consent was acquired from persons observed during participant observation, and no informed consent was acquired from persons observed during unobtrusive observation.

Over the combined study periods in 2013–2016, we engaged in observations at (1) public health centers (observed persons included managers, program coordinators, public health nurses, community health educators, citizens utilizing HIV testing and counseling, and other health and social services); (2) an infectious disease ward, outpatient clinic, pharmacy, and other areas at a public city hospital, and a comprehensive, private medical clinic with a focus on HIV and STIs (observed persons included physicians, nurses, a pharmacist, a medical intern, front office and other staff, and patients seeking HIV/STI testing, counseling, and treatment); (3) community-based organizations (CBOs) and community centers focused on HIV (directors, program coordinators, HIV case managers, testing counselors, social workers, other paid staff and volunteers, and clients utilizing HIV testing, counseling, health education, and programs for HIV-positive persons); (4) street-based HIV outreach events and sexual and gender minority (SGM) community events (community and event organizers, and attendees); (5) streets of SGM neighborhoods, a gay sex club, and SGM bars, restaurants, and book/DVD/adult shops (MSM, transgender women, and other SGM community members, and business owners, employees, and customers); (6) heterosexual brothels, love hotels, red light districts, and bars (sex workers and their clients, other sex industry employees, and bar owners, employees, and customers); (7) an intimate partner violence shelter (shelter case managers and hotline workers); (8) election campaign events and political demonstrations (politicians, candidates, and event attendees); and (9) other common urban locations such as city streets, government buildings, museums, hotels, apartment complexes, shops, train stations, trains, airports, cafes, and universities (immigrants and foreign nationals; youth; drug and heavy alcohol users; perpetrators, witnesses, and victims of violence, vandalism, and other crimes; police officers; security guards; civilian neighborhood crime patrols; a mental health counselor; and academics and students). In total, this yielded 461 pages of field notes derived from observations of 178 people.

We conducted formal, semistructured qualitative interviews in Japanese (n = 27) or in a combination of Japanese and English (n = 5), according to participant preference, with 32 of the persons observed in the study. Interviewees provided written informed consent. They were from at least one of the seven target groups, had a mean age of 40.9 years, and were 40.6% cisgender female, 9.4% transgender female, 89.6% ethnically Japanese, 87.5% Japanese national, 53.1% sexual minority, 65.6% employed full time, 75.0% with at least a university education, and 71.9% HIV negative.

We acquired additional data through email correspondence with nine key informants, eight of whom were from the original 2013 interview sample and one person who was new in 2016: a public health center employee working in infectious disease control near Tokyo. Across the study periods in 2013–2016, we asked open-ended questions generally addressing preventive behaviors and specifically about the following aspects of PrEP and PEP: general awareness, specific knowledge, perceptions, availability, cost, access, prescription and use by providers and community members, and government approvals.

We collected and analyzed 321 documents and web-based data sources (n = 315 in Japanese; n = 6 in English), current through June 2016, which pertained directly to HIV, PEP, or PrEP in Japan, or were immediately relevant. These were produced by government institutions, nongovernmental organizations and CBOs, other civil society groups, media corporations, other businesses, and individual community members (Table 1). We acquired these data sources in person during fieldwork, and through databases (e.g., the National Diet Library database and the five highest circulation newspapers in Japan) and wider Internet searches using the English acronyms “PEP” and “PrEP,” plus multiple Japanese translations, including terms for “pre-exposure prophylaxis,” “postexposure prophylaxis,” “pre-exposure (oral) medication,” “postexposure (oral) medication,” “pre-exposure,” “postexposure,” and “exposure.”

Study protocols were approved by the institutional review board at the first author's university in the United States and the university institute in Japan where he was affiliated. Protocols complied with the social–behavioral modules of human subjects research training by the Collaborative Institutional Training Initiative. ⁸⁶ Participants' confidentiality was protected through several measures. Only we knew participants' names and identifying information. No research assistants had contact with personally identifiable information, and we used only participant identification numbers in study-related documents. Interview audio files and transcripts, observation field notes, documents and web-based data sources, and emails were shared exclusively between the two authors through password-secured cloud storage and backed up to the first author's password-secured laptop, or were kept in hard copy in a locked file cabinet.

Drug approvals and overall use

The pharmaceutical regulatory body and ultimate decision-maker for drug approvals in Japan is MHLW. After restructuring of the regulatory system as mandated in the Revised Pharmaceutical Affairs Law of 2001, MHLW has consulted mainly with two incorporated administrative agencies: the Pharmaceutical and Food Safety Bureau, and the Pharmaceuticals and Medical Devices Agency (PMDA), which document drug reviews and approvals.

Approval of pharmaceuticals for prevention of any disease has been rare. In PMDA documents dating back to 2004, we found only a few examples, including drugs for prevention of influenza, herpes, hepatitis B, and malaria. Although several ARV drugs have been approved for treatment of HIV, no drugs have been approved for HIV prevention. In addition, MHLW has not officially endorsed either PrEP or PEP as general strategies.

The complexity of the PrEP and PEP situation in Japan begins here, because despite a lack of explicit approvals, MHLW's actions have implied at least tacit approval of one application: PEP for occupational exposures in healthcare workers. The clearest evidence of this was in multiple updated versions (2013–2016) of two documents produced with direct MHLW support: guidelines for clinical treatment of HIV and a training manual for nurses working in HIV. The clinical treatment guidelines listed PEP only, not PrEP, and PEP was only recommended for exposures by healthcare workers:

If possible, initiation of post-exposure prophylaxis drugs within two hours is considered important. … An “HIV post-exposure manual” should be incorporated into each medical institution's hospital-acquired infections manual. … Having ARV drugs in each facility is beyond their scope. … so a system must be established such that when an exposure occurs, ARV drugs from outside the medical institution can be quickly accessed.

Truvada, which MHLW approved in March 2005 for HIV treatment, was recommended in the guidelines in combination with raltegravir (brand name Isentress) as the first-line regimen for occupational PEP only.

The training manual for nurses addressed all three applications: occupational PEP, n-PEP, and PrEP. By first discussing PrEP in 2013, this manual was comparatively ahead of its time as a training document for healthcare providers. However, the information was limited to a brief definition, and that Truvada had been approved by the US FDA; there was no mention of PrEP use in Japan. The manual detailed occupational PEP for exposures such as needlesticks, but it also specified that “PEP for non-occupational exposures is at the patient's own burden of expense.” This was the only mention of n-PEP that we found among professional guidelines, manuals, and other similar documents. The statement illustrated two salient matters: first, that providing both occupational PEP and n-PEP to patients was possible; second, that the parties responsible for paying depended on the type of exposure.

Legality, insurance coverage, and prescribed versus off-label use

Regarding the first matter (i.e., that providing occupational PEP and n-PEP was possible), we identified distinctions among three elements: officially approved uses of ARV drugs by MHLW, prescribed use by providers and patients, and the legality of preventive ARV use. Despite the lack of explicit MHLW approval, participants reported that ARV drugs for PEP were being prescribed and used in Japan, although at very low levels: rarely for occupational PEP and very rarely for n-PEP. All such use was technically “off-label” (i.e., outside approved indications); however, off-label use also was not against the law in Japan. We found no language in any laws, regulatory documents, or professional guidelines that explicitly prohibited ARV drugs from being prescribed or used for prevention, even for nonoccupational exposures.

Regarding the second matter (i.e., parties responsible for payment), in our analysis of PMDA and MHLW documents, we determined that because approved indications had not been expanded for any ARV drug to include prevention, such preventive uses were not mandated for coverage by health insurance. Whereas our overall data showed that lack of explicit MHLW approval curtailed ARV prescriptions for prevention significantly, our participants indicated that it was actually the lack of insurance coverage coupled with the high cost of HIV drugs that presented the largest obstacles to accessing ARV drugs for prevention. This was mainly a problem for the two applications that would protect persons outside the healthcare field: n-PEP and PrEP. It was less of a concern for occupational PEP, which, as outlined previously, had implicit MHLW support.

For example, when we asked whether it was possible for someone who had unprotected sex to acquire PEP, a public health center employee working in infectious disease control answered this way:

It is possible to get PEP after consultation at a medical facility, in cases of non-needlestick accidents. However, for sexual exposures, it is different from needle accidents in that it is your own expense. Each medical facility might have different operating procedures regarding prescription of PEP—whether they will prescribe it or not … so we recommend you ask the facility in advance.

We heard a fairly wide range of estimates for cost. For example, a pharmacist estimated that for n-PEP, “A required medical examination fee, plus prescription and pharmacy dispensing fees and 28 days of Kaletra (lopinavir/ritonavir) or Truvada, would cost approximately $412 and $945, respectively.” For PrEP specifically, neither prescriptions nor use was reported by any of our participants. A private clinic physician discussed this issue:

As far as I know, no one in Japan is on PrEP. The main reason is the very high cost. As you know, in Japan, medical insurance cannot be applied for preventive medicine, including PrEP. If you take PrEP for one year in Japan, you must pay more than US $10,000!

A newsletter for persons living with HIV put the cost of PrEP a bit higher: “As of November 2015, there is no insurance coverage in Japan, and the full amount will be out-of-pocket: about ¥120,000 per month ($1130/month or $13,560/year).”

Although no study participants reported off-label use of ARV drugs for PrEP, our analysis of web-based data gave some indication that this was occurring. For example, a website post by an anonymous foreign resident in Tokyo—a gay man—conveyed frustration with the barriers to obtaining PrEP, and his solution:

My Japanese boyfriend contracted HIV earlier this year. … I found out about PrEP in the US online, but we were told by my boyfriend's doctor that Truvada is not covered for PrEP by Japan's health insurance system. I couldn't afford to pay for Truvada in full, so I decided to buy the Truvada generic … for about $100 per month via a Japanese online pharmacy (manufactured by Cipla in India). … I am just angry and frustrated that I didn't know about PrEP earlier, and that it is not being promoted in countries like Japan. … I am angry because I believe I could have convinced my boyfriend to go on it and could have protected him from infection.

Comments on PrEP and PEP from a study interviewee, a Japanese HIV activist, resonated with the foreign resident's website post, suggesting that similar measures were being considered by others:

Unless you are a certain type of medical personnel, or you personally import the HIV medicines, those methods [PrEP and PEP] are not available in Japan at present. I think it is also hard to consult doctors [about PrEP and PEP] in medical institutions.

By characterizing the legal situation optimistically and the insurance and access situations as more problematic, another pharmacist interviewee captured the essence of these concerns:

Prophylactic administration [of medication] is not covered by insurance, but whatever the reason, use is legally possible. Whether it is because of sex with a partner or you're a drug user or whatever, it is legally possible. Although it's possible that the doctor will be prejudiced and not write a prescription.

Policies by healthcare institutions and government agencies

For persons in support of both occupational PEP and n-PEP, there were two additional reasons for optimism. First, data across sources indicated that in cases of occupational exposure, PEP was paid for as a work-related accident by employers, typically medical institutions such as hospitals or clinics. For healthcare workers, these institutional policies trumped the general rule of no insurance coverage. Second, based on interview and document data, there were two related situations in which n-PEP might be prescribed by physicians and could be paid for by parties other than the patient: cases of sexual assault, and when PEP recipients were otherwise vaguely defined as “victims.” For example, the private clinic physician distinguished such circumstances from exposure during consensual sex:

It might also be necessary for me to prescribe prophylactic HIV medication in cases of rape by someone who is HIV-positive or whose status is unknown. … However, when a patient says, “I want a prescription because I had unsafe sex,” except in cases of rape or other such special circumstances, I will not do it.

Our data indicated that if sexual assault crimes were reported to the police, the cost of n-PEP could be covered by public funds. We found evidence suggesting legal feasibility in government policies. For example, the Provision of Health Services and Welfare Services (Article 14) of the Basic Act on Crime Victims, states:

The national and local governments are to take necessary measures to provide crime victims with appropriate healthcare … according to their mental and physical conditions … to recover from psychological and physical trauma incurred in the crimes.

Article 8 of the Basic Act on Crime Victims mandated the government to produce a “Basic Plan for Crime Victims” to achieve measures listed in the Act. To date, there have been three versions of the plan. The most recent, the “Third Basic Plan for Crime Victims” from 2016, states:

The National Police Agency is to subsidize the cost borne by prefectural police for their financial support offered to sexual crime victims, including the fees for emergency contraception, abortion, initial diagnosis, medical certificate, sexually transmitted diseases and such, etcetera.

Based on the Basic Plan for Crime Victims, each prefecture has its own handbook with policies for supporting crime victims. For instance, Hyogo Prefecture, where Kobe is located, has the “Handbook on Support for Sexual Crime Victims, Hyogo Prefecture Version, May 2015.” The language in these policies is both inclusive and vague enough to accommodate n-PEP, although they mention neither ARV drugs for prevention nor HIV specifically. Our data suggest that some, although not all, providers, hospitals, social service workers, and prefectural police departments have been acting according to an interpretation accommodating payment of n-PEP after sexual assaults.

PrEP and PEP awareness

There were discrepancies both across and within participants' narratives on the topic of n-PEP after sexual assaults, which reflected a broader result: inconsistent but generally low levels of awareness of all three preventive ARV applications: occupational PEP, n-PEP, and PrEP. Awareness was typically low among participants who did not work in the health and human services fields in HIV, medicine, pharmacy, public health, and violence prevention. For example, we asked a transgender woman, who was a community member with a history of unprotected sex and chlamydia infection, about PrEP and PEP. She simply reported, “I haven't heard of them.” A conversation in 2013 among three counselors at an HIV CBO focused on awareness:

Interviewee 1: I have heard about PEP, but not PrEP. … PEP is a medicine; I know that. But probably almost nobody knows about it unless they are medical personnel. … I am the least experienced as a volunteer staff member, but by accident I happened to know … about PEP. As the person closest to the general public, I don't think many people know about these.

Thus, whereas PEP awareness was low, collective knowledge specifically about PrEP was lower. In late 2015, a statement by one of the HIV CBO staff members, quoted previously, confirmed this was the case more than 2 years after our first interview. It also reflected a typical response from community member participants: “I do not know if PrEP is being used in Japan. I don't know where to obtain PrEP and PEP, their cost, or who pays for them.”

The above quotes also illustrate that there was inconsistent knowledge even among persons working in the health and human services fields. Most participants who had any awareness of PEP reported only its occupational application.

Six participants working in health and human services did also identify use after sexual assault and general “victimization,” but a comparison of their narratives demonstrated a lack of clarity, including a range of understanding about what was “approved.” For example, a public health physician who worked as a section manager in a public health center stated that government approval for PEP included both healthcare workers and rape victims. This contrasted with an infectious disease physician at a hospital, who stated once that “Japan has not yet approved PEP… other than for exposures among healthcare personnel” and, separately, that “PEP is limited to needlesticks and other blood exposures in medical institutions, and victimization such as rape.” One of the pharmacist participants inaccurately believed, “The government has approved Truvada tablets for prophylactic administration.”

Although few participants held this mistaken notion, most were unsure what was true. When we asked what people knew about PrEP and PEP, two case managers at a domestic violence shelter discussed several relevant points at length, including that they frequently encountered situations when PEP was clearly indicated, that foreign residents sometimes had a higher level of PEP awareness than Japanese citizens, and that there were few hospitals where patients could actually acquire ARV drugs for prevention:

Interviewee 1: When there has been sexual violence, a very high possibility of infection, and you want to take medicine immediately, you have to report it to the police; otherwise, the medicine won't be free … you have to pay for it. But there are few people who know that such medicines exist, and that they especially need them in cases of sexual violence.

Only one participant, the manager at a public health center, mentioned uses of n-PEP beyond those for healthcare exposures and sexual assault victimization: “There are treatments … in cases of contaminated needles for drug addicts … and police officers appear to be targets for PEP.”

Although no other participants mentioned use for IDUs and police officers, and we could find no confirmation or documentation of such uses, this interviewee had consulted two hospitals in her city before reporting.

Overall discourse on PrEP and PEP

The typically low level of PrEP and PEP awareness paralleled a general lack of discourse outside two limited spheres: HIV scientists and SGM activists concerned with Japan's HIV problem among MSM. Over the 2013–2016 study period, the discourses in those spheres grew, but they were not especially visible to the general public. Thus, PrEP and PEP discourses had not yet penetrated the community in a significant way in writing, visual advertisements, or promotional campaigns, and we found limited evidence of even oral discourses among regular community members.

In 2013, there was almost no discourse on PrEP and PEP in Japan, except for some discussion of occupational PEP—probably because of its longer, implicitly sanctioned use in the country. For example, thinking back on our first interview with her, a pharmacist recalled, “I didn't know about PrEP and PEP until you asked me about it in 2013.” Overall, there was no culture of acceptance for PrEP and n-PEP in the current medical and public health spheres to support their use. An infectious disease doctor explained:

We don't have a system for it yet. Also, we don't have a culture of PrEP and PEP; it is not rooted here yet. … To be honest, I want to offer both PrEP and PEP, but there is no hospital that does that. And from here, it is very close to the area where MSM hang out. If they learn that we prescribe PrEP and PEP in this hospital, so many patients would rush here. We don't have an infectious disease specialist during the night shift, and it would be difficult for a doctor on duty to make a judgement. … The system is not ready yet. That is the current problem.

Mainstream media and web coverage

There was very little coverage of PrEP and PEP in the mainstream media or on the Japanese language Internet. Our searches revealed that there was no single, agreed-upon term for either PrEP or PEP in Japanese, with multiple translations used even on Japanese government websites and documents. In the few cases that PrEP or PEP were reported in the mainstream media, the articles tended to be about global studies outside Japan, and especially the US FDA's approval of Truvada for PrEP in 2012. The latter was covered by several media outlets, including the Nikkei Shinbun, Asahi Shinbun, Reuters' Japanese website, The Japan Times, and on websites supported by the Japan Foundation for AIDS Prevention Studies, an organization contracted by MHLW.

Other examples included a summary of recent studies of Truvada's efficacy for PrEP on a Japanese health encyclopedia website, Medical Tribune, which also appeared in the newspaper Yomiuri Shinbun, and an article about controversies surrounding PrEP in the United States in an online Japanese news website, Gigazine. We found nothing in the mainstream media discussing PrEP and PEP use in Japan itself.

Discourse among scientists and activists

Among HIV scientists and SGM activists, there was growing interest and support, especially around PrEP for MSM. Yet the discourse in these spheres remained nascent in 2016, with significant reservations and concerns about remaining barriers. An HIV activist participant described, “The views of HIV activists vary so much, and the argument around implementation of n-PEP and PrEP has just begun recently.” Likewise, a former HIV CBO staff member stated, “Regarding PrEP, the extent to which I've heard about it is that the topic has recently come up among some MSM, and I've heard of its relationship to the MSM community.” In 2013, an infectious disease physician mentioned the activities of an organization of Japanese HIV scientists:

Among the directors of the Japanese Society for AIDS Research, there is movement toward promoting PrEP and PEP; and they are encouraging the individual members of the Society and the government on this. … There may be some change in the near future.

There was indeed subsequent movement by this group. In late 2015, the Society held its annual meeting, where significant information about PrEP was presented. We examined the content of the annual meeting and a previous, related press conference using publicly available documents and videos, and the conference website. The main theme of the conference was “prevention,” which was described as the area where Japan lagged furthest behind several other countries. Of the five main presentations in the conference program, three included the latest research on PrEP, presented by very prominent scientists working in HIV and pharmaceuticals internationally. Another point in favor of PrEP was the description of a new epidemiological study to better estimate HIV incidence in Japan's MSM, and the Society's apparent plans to use the data to inform decisions around PrEP adoption and rollout. This aligned with interview data suggesting relatively high and increasing PrEP awareness in this circle of Japanese scientists.

Discordantly, however, comments by a Society leader also signaled a questionable understanding of where greatest risk lay and, therefore, how PrEP could be most effectively applied. The statements resonated with a history of xenophobic responses to HIV dating to the beginning of Japan's epidemic and were not consistent with current epidemiology on transmission risk comparing Japanese nationals and foreign nationals in the country. Discussing a slide in his presentation (“Needs in Japan Moving Forward”), he referred to Shinjuku Ni-Chome, which is the main SGM neighborhood in Tokyo:

PrEP for prevention in persons at high risk of infection: we especially need to consider this, because the Olympics are coming to Tokyo in five years. For male homosexuals from all over the world, Shinjuku Ni-Chome is the number-one place they want to go. It is thought that many will probably go there. How to prevent infections at that time is a problem that we need to consider.

Most of the discourse in Japan on preventive ARV strategies among scientists and activists was in the scientific literature and in conference documents, with a small amount also in magazines and community newspapers targeting MSM and persons living with HIV. Japanese authors have written about PrEP and PEP in all these formats, and similar to the mainstream media, most of the content has been about PrEP efficacy as demonstrated in trials and its approval by the US FDA. For example, Tanuma ⁴⁷ referred to international research on PrEP: “The effectiveness of pre-exposure prophylaxis is also becoming clear. Used in place of preventive vaccines, which have not yet been successfully developed, such effective antiretroviral drug techniques are important and show promise as infection blocking measures.”

Appeal of PrEP and PEP moderated by concerns

We found no unreserved endorsements by Japanese scientists or activists of PrEP or n-PEP for use in Japan. Instead, as mentioned previously, we found indications of PrEP's and PEP's appeal in these groups, tempered by significant reservations.

Some reservations resulted from concerns about pressing HIV issues that needed to be resolved before PrEP and PEP could be adopted on a large scale. These included (1) early testing for HIV; (2) treating all persons already living with HIV early in the clinical course of infection; (3) more surveillance and more epidemiological and behavioral research to inform prevention generally; (4) more and better research on MSM, including studies to determine who will need PrEP; (5) evaluations of current HIV prevention programs, which focus on condom promotion, to assess whether they are sufficient, and adapting them to PrEP; (6) elimination of HIV stigma and prejudice against people living with HIV; (7) involvement of various stakeholder groups in prevention planning and vigorous discussion within the community; (8) increasing community understanding about what PrEP is, and its efficacy; (9) assessment of existing outreach networks and creation of new networks by community centers that will lead awareness campaigns for PrEP; and (10) advocating for insurance coverage for PrEP and n-PEP to create political “understanding.”

Other reservations stemmed from concerns about negative consequences if PrEP and PEP were actually scaled up. These included (1) misperceptions by those who used PrEP that it would provide thorough protection, leading to increased unprotected sex; (2) concerns about drug resistance; and (3) a perceived danger that encouraging PrEP use in serodiscordant couples would pressure the HIV-positive partner to come out to the HIV-negative partner about their serostatus, resulting in negative social consequences.

In separate commentaries about PrEP adoption among MSM in a magazine for persons living with HIV, two HIV activist authors outlined most of these concerns. For example, the first author wrote, “If people need assistance in deciding whether and when to come out about their serostatus to their partners, who will be responsible for providing that assistance?” He was also skeptical of people “at international conferences who say in a loud voice that ‘This is the only path forward in prevention.’” The second author stated, “It will be a challenge to deliver PrEP in this complex set of conditions,” and he warned against ignoring these problems and “suddenly introducing PrEP.” However, the activists also expressed guarded support. For example, the first author concluded, “PrEP will be an important opportunity to rethink the way prevention awareness for MSM is undertaken.”

This generally conformed to reservations expressed by the participants in our study. In addition to the health insurance issue discussed previously, their concerns focused on development of drug resistance by HIV and increased unprotected sex (i.e., risk compensation). For example, an infectious disease physician spoke about resistance:

[PrEP and PEP] are good things, but from our position, we are concerned about the problem of drug resistance. We're worried that when people take the drugs, then do not take them, repeatedly like that, the drugs might not function when that person is really infected in the future. … I don't think it is only us, but also many other infectious disease doctors who think resistant virus would be a problem if [PrEP and PEP] are too widely used in Japan.

One of the pharmacists in the study, who worked at a public hospital and had many HIV patients, also described concerns about resistance. When asked about PrEP and PEP use to prevent infection in cases of high-risk unprotected sex, he stated:

We don't do that. … I haven't even heard of those. … [but] I understand it. … I think it could prevent infections. … But, there is the problem of resistance. It's scary. All things considered, it is better that it not be used, I think. … If you want to have it before having sex, you should not take it; but if you get blood all over you, I think it's OK to take it.

Regarding risk compensation, the director and a counselor at an MSM community center focusing on HIV prevention shared their views:

Interviewee 1: Well, there is one thing that we have to be careful about. For example, you think you won't get infected as long as you take the drug. To avoid this kind of misunderstanding, that someone can have unprotected sex as long as they take that drug, it is very important to disseminate accurate information.

Concerns about risk compensation sometimes interacted with incomplete knowledge about biomedical prevention strategies, resulting in decisions to hide available information from people who could substantially benefit. For example, an HIV counselor, who worked at an NGO serving exclusively HIV-positive clients, withheld information about TasP and how it could reduce transmission risk to HIV-negative partners. He was unlikely to recommend PrEP, which could further reduce risk:

Interviewee: We tell them to use condoms.

In addition, a private clinic physician's comments resonated with activists' concerns about the forced “outing” issue feared to result from PrEP in serodiscordant couples:

I have never heard the voices of citizens in Japan advocating, “PrEP and PEP should be covered by insurance!” … I also have heard little from patient groups and the medical profession. Why is that? Partners of HIV-positive persons must want PrEP to be covered by insurance. The reason they do not raise their voices is that if they did, they would be publicly outing their partner as HIV-positive.

Scarcity of visual representation

During our fieldwork at multiple sites in urban Kansai and the Tokyo Metro Area in 2013 and 2016, we observed no visual presence of PrEP or PEP in the form of public posters, flyers, pamphlets, TV commercials, Internet banners, or other advertisements and promotions. In 2013, this occurred in a broader context of waning government and public attention to HIV, overall, since 2009. One manifestation of this was a reduction in the volume and visibility of HIV prevention advertisements and social marketing campaigns. Our field notes from 2013 provide an example. This excerpt was completed after a week of walking slow grid patterns during daytime and evening hours in one city's main SGM neighborhood, and documenting visual data with mobile phone cameras.

We detected almost no change in the visual representation of PrEP and PEP in April 2016, as field notes from the same neighborhood indicated:

Visited first gay DVD/magazine/toy shop on [name of street redacted] between 8:30 and 9:00 pm … one male customer, no posters or anything related to HIV/AIDS or PEP/PrEP … [Name of shop 2 redacted] is a gay DVD shop: two solitary male customers, no posters or anything related to HIV/AIDS or PEP/PrEP. [Name of shop 3 redacted] sells DVDs, books, magazines, sex toys, snacks; two solitary male customers, no posters or anything related to HIV/AIDS or PEP/PrEP. [Name of shop 4 redacted]: book and magazine store; all of the magazines wrapped in plastic. One solitary visitor, no posters or anything related to HIV/AIDS or PEP/PrEP. [Name community center redacted]: Inside the space, there was a variety of pamphlets, newsletters, and community newspapers on the table and the shelf at the entrance. One community newspaper had a feature on PrEP. There were no posters or flyers related to PrEP or PEP. In [name of neighborhood redacted], there were no ads or posters related to PEP or PrEP.

In one of the shops already noted, we purchased recent issues (2015–2016) of a popular magazine for MSM. In one issue, similar to academic documents that we examined, there was a short piece discussing PrEP's availability in North America, but it did not refer to Japan specifically.

In July 2016, we again noted no change in the visual representation of PrEP and PEP: we observed nothing publicly inside bars, sex clubs, community centers, MSM DVD/magazine/sex toy shops, or on the street. However, we discovered a positive development signaling that the trend of reduced general HIV prevention advertisements and social marketing campaigns since 2009 might have begun a reversal. Updated field notes from the same SGM neighborhood focused on this:

On the wall of a landing outside the door of [name of community center redacted], the same “safer sex” poster featuring the condom and the mostly naked man is still there, but it now has a lot of company. Two walls are now covered with posters, most of which promote either condom use or HIV testing, including a new service by which clients can receive a free HIV home test kit from the community center. … On the side of a building at one of the main intersections of [name of street redacted], there is now a huge sign with two men embracing and a message in both Japanese and English encouraging regular HIV testing. … These changes are obvious and significant compared to all earlier field visits to this site dating back to January 2013, but I still saw nothing on PrEP or PEP.

Insurance coverage, epidemiology, and the question of cost effectiveness

The general rule of no insurance coverage for preventive use of expensive ARV drugs might be the most complicated of these issues and raises the question of whether such strategies would be a cost-effective public health strategy in Japan. As discussed in the introduction, the WHO has recommended that HIV incidence ≥3 per 100 person-years (alternately but not quite accurately expressed as 3% incidence) be considered the conventional threshold at which offering PrEP would be potentially cost-effective. However, WHO also has stated that PrEP at lower incidence, especially in concentrated epidemics (e.g., in MSM), might still be cost-effective and make a significant impact. ¹

The latter qualification will be important to consider for Japan. In MHLW surveillance reports, new HIV cases are typically reported as counts, or as percentages of the total infected. MHLW tends not to report rates of new cases with the denominator representing all persons at risk of infection, and it is not clear that these newly reported HIV cases represent recent enough infections to comprise actual incidence.

This difference is important epidemiologically, because it is the incidence rate that measures risk in a population. MHLW does report rates of new cases by prefecture and age group, as noted previously. Rates were highest in Tokyo, ranging between 2.7 and 3.1 per 100,000 in 2013–2015, and next highest in Osaka, ranging from 1.7 to 1.9 per 100,000 during the same period. In 2013, the rate of new cases in all age groups was under 3.0 per 100,000, whereas 25- to 29-year-olds exceeded this rate in 2014 and 2015 at ∼3.5 and 3.2 per 100,000, respectively. ⁹ The closest the surveillance reports come to estimating the rate of new cases in the general population at the national level is the rate among blood donors, which ranged from 1.0 to 1.2 per 100,000 in 2013–2015, and is 0.7 per 100,000 so far in 2016. ^97,98 However, a key point is that these estimates are all per 100,000, not per 100, and thus fall far below the WHO's 3-per-100 rule of thumb.

MHLW surveillance reports indicate that that the greatest number and percentage of new cases have been among MSM, but true incidence rates in this group are not included. ⁸ Using MHLW surveillance data alone, Yoshikura ⁹⁹ calculated the rate of new cases among MSM by prefecture, which was highest in Tokyo at 0.39 per 100, followed by Osaka at 0.11 per 100. Estimates from epidemiological studies are scarce and not based on population-based samples, but they have found comparatively higher incidence. Itoda et al. ¹⁰⁰ reported incidence of 1.0 per 100 person-years in a sample of MSM seeking HIV and STI testing at a CBO in Yokohama, and Kimura et al. ¹⁰¹ reported 1.3% incidence in gay men attending two community events.

Although not a direct measure of risk, estimates of HIV prevalence among MSM over the past decade have ranged from 2.6% to 9.6% at testing sites in Tokyo, Osaka, and Nagoya, ^{102 –106} and 4.7% based on self-report survey data from Tokyo, Osaka, Kanagawa, Aichi, Fukuoka, and Okinawa prefectures. ¹⁰⁷ This is relatively high compared with national prevalence estimates in the general population of ∼0.02%. ^108,109

Populations other than MSM that typically have higher risk in concentrated epidemics include transgender persons, IDUs, sex workers, and the clients and sex partners of such populations. ¹¹⁰ Transgender persons are not disaggregated in MHLW data, and epidemiological studies are sparse. However, in a separate analysis, we found that transgender persons were a key group affected by social and health disparities related to HIV in Japan, and that poor mental health, substance use, and violence increased their vulnerability to HIV. ⁷

There are no surveillance numbers on sex workers in MHLW reports, and there has been little published epidemiological research on this group. The few studies to date have been limited to prevalence measures and have not included incidence. Most of these found no HIV in sex workers. ^{111 –114}

MHLW includes IDUs in its surveillance, and has reported very low occurrence in this group: only 0.2–0.3% of new HIV infections in 2013–2015⁹; however, a recent epidemiological study by Nishijima et al. reported a higher proportion: 2% of new infections. ¹¹⁵ Incidence rates for IDUs are not included in MHLW reports, and we found few epidemiological studies, which were limited to measures of prevalence ranging from 0% to 0.47%. ^116,117

Drug resistance and risk compensation: evidence from the literature

Most recent research on HIV drug resistance and risk compensation related to preventive ARV use has been in the context of PrEP, but its relevance to both occupational PEP and n-PEP is also evident. Predictions based on modeling have indicated that HIV drug resistance resulting from ARV use for treatment will far exceed that resulting from PrEP ¹ ; so far, the demonstrated risk of drug resistance resulting from PrEP has been low. ^1,15

Most published evidence also has not indicated that PrEP is associated with risk compensation in sexual behavior, such as decreased condom use or more sexual partners. ^1,15,118 However, there have been some mixed findings. Recently completed PrEP demonstration projects have reported increased rates of STIs among MSM, indicating risk compensation specifically in this group. ¹¹⁹ One study in Kenya and South Africa documented hypothetical intentions by women to reduce condom use and other HIV risk-reduction practices if they were to use PrEP. ¹²⁰ More optimistically, a recent meta-analysis of PrEP trials found that in open-label extensions with more intensive counseling, safer sex practices were maintained, ¹⁵ suggesting that counseling can be effective in preventing risk compensation.

Recommendations for future directions

The weight of the evidence on drug resistance and risk compensation might shift in years to come. Long-term monitoring of preventive ARV use, especially for PrEP, is needed—as with any medication. More studies in the future might show an association between risk compensation and PrEP and PEP use. However, in light of most recent evidence and despite valid concerns, there appears to be insufficient justification to object to full adoption and scale-up of PrEP and PEP, particularly in Japan's MSM population.

Based on our analysis, we urge Japanese health officials to prioritize a national discussion, weigh empirical evidence, and strongly consider formal approval of ARV drugs for use in PrEP and in both occupational and n-PEP. Once approved, social marketing campaigns can be used to advertise these prevention strategies widely and increase awareness. There is precedent for such approaches to work well in Japan, as very successful campaigns for HIV testing in the years 2005–2008 demonstrated. ⁵⁶

Health officials should also consider policy changes necessary for Japan's combination of publicly funded social health insurance and private insurance to cover PrEP and n-PEP, thus making them both accessible. The existing surveillance and epidemiological data currently do not provide evidence that even populations at presumably higher risk in Japan meet WHO's threshold incidence of 3 per 100. However, in a separate analysis, we found problematic issues in the surveillance system that called into question some official epidemiological trends in Japan. ⁵⁶

This suggests that enhancements to MHLW's surveillance system are needed, including mechanisms to estimate incidence rates in the general population, and among MSM and transgender persons at a minimum. Updated epidemiological studies with population-based samples of MSM, transgender persons, sex workers, and the clients and sex partners of these groups are also essential. Once solid results from these activities are available, they can be used in additional studies to model the number of infections that could be averted using PrEP and n-PEP and any associated savings in healthcare costs. Such efforts would add clarity about the level of need for PrEP and n-PEP across populations, help to determine their level of cost-effectiveness, and potentially provide justification for insurance coverage.

New PrEP formulations, including intermittent use of pills and quarterly injections, could help improve adherence and achieve PrEP's full potential globally. ¹²¹ Multiple efficacy trials are currently underway in different phases. ^6,122 In Japan, these new formulations could reduce cost and barriers to implementation. ¹²³ Research on the acceptability of PrEP and PEP in specific populations in Japan, including MSM, transgender people, and serodiscordant couples, is needed and would do well to include assessments of new drug formulations in their assessments. A recent study by Chakrapani et al. in India ¹²⁴ could provide a template for such investigations.

Best practices for the integration of PrEP, PEP, and early ART recently have been recommended and can inform Japan's path forward. ^39,125 Inclusion of PrEP and n-PEP in normative guidelines produced or funded by MHLW (i.e., where MHLW's name is prominently displayed) could positively affect provider prescriptions to persons at substantial risk, ¹²⁶ at whatever threshold level Japan determines “substantial risk” to occur. For example, recent research has shown that the existence or absence of formal guidelines is a critical factor in providers' decisions to prescribe PrEP in multiple countries. ¹²⁷

It also will be important to consider the issue of trust in choosing scale-up plans. As Nagata and Raflzadeh-Kabe ¹²⁸ have discussed, the pharmaceutical industry has not enjoyed a high level trust from the Japanese public, after a number of scandals in past years. Although the degree of trust in physicians also has decreased, it remains very high. Doctors frequently only need to advise their patients of a benefit for the patients to agree to a clinical course of action.

Once ARV drugs are explicitly approved for PrEP and PEP by MHLW and covered by insurance, Japan can leverage its strengths in scale-up efforts. The same availability of highly skilled health professionals, technical and pharmacological innovation, and overall high-quality medical and public health infrastructures, which allowed for rapid adoption and scale-up of ARV drugs for treatment of persons living with HIV, ^56,129 can be applied to PrEP and PEP.

Several of our recommendations align with those offered by Lo ⁴² to encourage adoption of PrEP in countries throughout Asia. Scientists and activists in favor of PrEP and n-PEP adoption in Japan might also consider adapting some of his other recommendations, including preparing communication packages with clear messages of endorsement by WHO, UNAIDS, and the Asia Pacific Coalition on Male Sexual Health, and linking PrEP and n-PEP closely to early ART for treatment.

Theoretical considerations

Future research and action around PrEP and PEP in Japan would benefit from theoretical grounding. The results of this study suggest that diffusion of innovations theory ¹³⁰ would serve well to frame such efforts.

The large and rapidly expanding body of evidence demonstrating the effectiveness of ARV drugs for prevention, combined with international pressure generated by UNAIDS recommendations and consensus proclamations like that recently made by HIV scientists in Vancouver, ¹³¹ have not yet resulted in a tipping point in Japan. To express this in diffusion of innovations terms, it appears that an insufficient number of early adopter individuals have adopted PrEP and PEP to surpass the thresholds of other individuals, which, in turn, have precluded a critical mass at the system level. Critical mass occurs when enough individuals in a system have adopted an innovation such as PrEP or PEP so that their subsequent rate of adoption becomes self-sustaining. Thus, the bi-level phenomenon of individual thresholds and system-level critical mass are closely related. ^130,132

We have framed most of our discussion by scientific evidence that provides support for PrEP and PEP. However, future research might consider that because government regulatory agencies, prescribing providers, and patients seeking PrEP or PEP in Japan are people, they do not make decisions based purely on the weight of scientific evidence. Indeed, when decisions about health policy, prescribing, and health prevention behaviors are made, people frequently demonstrate that they can, as Spieldenner has phrased it, “resist the seductive totalizing power of empiricism.” ¹³³ Diffusion of innovations studies has shown that most people do not evaluate an innovation based on scientific studies, although such objective evaluations are not entirely irrelevant, especially to the first people to adopt. Rather, most individuals depend principally on a subjective evaluation of the innovation that is communicated to them by others who already have adopted it. ¹³⁰

The results of our study allowed us to see that these are fruitful areas for future investigation in Japan, but our qualitative design precluded our ability to test several of the implicit questions. Future studies whose designs are informed by a diffusion of innovations framework thus can pick up where we have left off. Moreover, the results of diffusion of innovations research on PrEP and PEP can guide future programs aimed at accelerating their adoption, including strategies to engage healthcare providers, following any policy enactment that might occur in Japan. ^127,134

Our study has limitations. Findings were derived from data collected in only two regions in Japan, and we used nonprobability sampling to recruit participants and acquire documents and web-based data. This limits generalizability of the results. Interview data were susceptible to reporting bias. Limitations to observational data included possible reactivity and social desirability bias. We took rigorous and multiple measures to minimize all such effects as described in the Methods section.

Regarding implications, beyond meeting the aims of our study, our motivation was to inform current and future ARV-based prevention strategies at a critical time in international conversations on this topic. This study presented an empirical account of PrEP and PEP and provides an ethnographic context to ongoing debates, current decision-making, and conversations in Japan. In addition, although we do not advocate generalization, our results can be applied also to other countries that are debating adoption or scale-up of PrEP or PEP, particularly high-income nations and countries in the Asia-Pacific region.