1. Introduction
Since the emergence of HIV, it has received extensive global attention. In the early years, limited medical technologies and widespread fear related to its high transmissibility and mortality contributed to significant stigmatisation and exclusion of people living with HIV (PLHIV) at both policy and societal levels (Ahmed et al., Reference Ahmed, Taylor, Lau, Lai, Deshan Diunugala, Louella, Villa, Freshwater, Averitt, Karris, Berry, Dee and Dubé2025; Duan and Xie, Reference Duan and Xie2026; Xie and Duan, Reference Xie and Duan2024a). These conditions indirectly reduced the willingness of PLHIV to voluntarily seek relevant treatment and hindered global efforts to eliminate this public health crisis (Babel et al., Reference Babel, Wang, Alessi, Raymond and Wei2021; Xie and Duan, Reference Xie and Duan2024b). In recent years, continuous advances in medical technology and the advent of antiretroviral therapy (ART) have transformed HIV from a highly lethal disease into a preventable and controllable chronic infectious condition. According to the principle of Undetectable equals Untransmittable (U = U), PLHIV who adhere to lifelong ART can maintain their viral load within a safe range. When viral suppression is sustained at an undetectable level for more than six months, usually below 200 copies per millilitre, the risk of sexual transmission becomes negligible or nonexistent (Okamoto et al., Reference Okamoto, Anam, Baptiste, Dukashe, Castellanos, Poonkasetwattana and Richman2024). The World Health Organization (WHO) further specifies the Treat All guideline, which recommends that all PLHIV initiate ART immediately after diagnosis without restrictions based on CD4 count or clinical staging (World Health Organization, 2016). Building on these developments, the Joint United Nations Programme on HIV and AIDS (UNAIDS) proposed the ‘95-95-95’ targets for 2025. These targets state that ninety-five percent of all PLHIV should know their HIV status, ninety-five percent of those diagnosed should receive sustained ART, and ninety-five percent of those receiving ART should achieve viral suppression, in order to ensure progress toward ending the AIDS epidemic by 2030 (UNAIDS, 2023). To expand access to HIV prevention, care, and treatment, WHO also recommends that national governments adopt policies that require domestic institutions to provide HIV care and treatment services at no cost (World Health Organization, 2005).
In response to these global calls, many governments have introduced policies that provide ART free of charge for PLHIV. However, numerous challenges persist. India established the National AIDS Control Organization (NACO) to implement free ART services through the public health system, yet disparities in the geographical distribution of treatment centres providing free antiretroviral drugs (ARVs) have affected both access and treatment adherence (Joglekar et al., Reference Joglekar, Paranjape, Jain, Rahane, Potdar, Reddy and Sahay2011). South Africa adopted a Treat All approach and offers free ARVs to all PLHIV with confirmed diagnoses, although the country continues to face challenges in effective monitoring and expanding coverage of ARVs (Kober and Van Damme, Reference Kober and Van Damme2004). In China, the ‘Four Frees and One Care’ policy, launched in 2003, stipulated that PLHIV who meet criteria for financial hardship are eligible to receive ARVs at no cost. In practice, financial hardship has not been used as an eligibility criterion, and free ARVs have been provided to all PLHIV. Due to concerns regarding fiscal burden, the government now places greater emphasis on incorporating newer ARVs into medical insurance policy, which require partial out-of-pocket payment from PLHIV. This shift has created challenges for updating the list of free ARVs covered under the ‘Four Frees and One Care’ policy (Xie and Duan, Reference Xie and Duan2024c). Given China’s vast territory and large population, the number of reported HIV cases remains immense even with a low national prevalence (Dou et al., Reference Dou, Luo, Zhao, Zheng and Han2023; Xie et al., Reference Xie, Chen and Duan2024). Therefore, examining China’s implementation of policies related to ARVs is of particular importance for further controlling the domestic HIV epidemic. At a broader level, such examination is also essential for advancing progress toward the UNAIDS goal of ending the global HIV epidemic by 2030.
Within China’s ‘Four Frees and One Care’ policy, the provision of free ARVs for PLHIV is mainly implemented by Centers for Disease Control and Prevention (CDC) and medical institutions. Differences in institutional goals shape their respective understandings of the policy and may influence implementation outcomes. Accordingly, this study addresses the following research questions: First, what differences exist between the CDC and medical institutions in their interpretations of the ‘Four Frees and One Care’ policy with regard to the provision of free ARVs? Second, what patterns of interaction emerge between the CDC and medical institutions during the implementation of this policy? Third, how does the policy overlap between the ‘Four Frees and One Care’ policy and health insurance policies shape differences in implementation preferences across institutions? Based on these research questions, this study takes CDC and medical institutions involved in implementing the ‘Four Frees and One Care’ policy as its primary research subjects. It aims to explore their differing interpretations of the policy, their interactions during the implementation process, and the sources of institutional variation in implementation preferences under the combined influence of this policy and medical insurance policies.
2. Literature review
2.1. The evolution of China’s free ARVs policy and current challenges
Since 2003, the ‘Four Frees and One Care’ policy has been implemented across many regions of China as a national-level HIV prevention, control, and healthcare protection initiative (Wu et al., Reference Wu, Sun, Sullivan and Detels2006). The introduction of this policy enabled the vast majority of PLHIV in China to receive standardised ART (Zhang et al., Reference Zhang, Pan, Yu, Wen and Zhao2005). In 2006, the State Council issued the Regulations on HIV/AIDS Prevention and Treatment, which formally incorporated the ‘Four Frees and One Care’ policy and signalled the development of a legalised framework for China’s free ARVs policy (Li et al., Reference Li, Gao and Zhang2024). By the end of 2006, this universal free ARVs policy had been implemented in all 31 regions of the country, resulting in 64% of PLHIV receiving free ARVs (Zhang et al., Reference Zhang, Haberer, Wang, Zhao, Ma, Zhao, Yu and Goosby2007). For a considerable period after the policy was introduced, eligibility for free ARVs required PLHIV to meet a specified CD4 cell count threshold. It was not until 2016 that all PLHIV, regardless of CD4 cell count, were able to obtain free ARVs under this policy (Liu et al., Reference Liu, Wang, Hu, Song, Liao, Feng, Li, Xing and Ruan2023). After 2017, several newly developed ARVs that would normally require full out-of-pocket payment began to be included in China’s medical insurance coverage and thus required only partial out-of-pocket payment. The incorporation of medical insurance marked the beginning of a period in which free ARVs, insurance-covered ARVs, and fully self-paid ARVs coexist (He et al., Reference He, Chen, Chen, Guo and Li2023).
Recent studies have evaluated the implementation outcomes of China’s free ARVs policy and highlighted the challenges it faces. For example, previous studies have found that greater health literacy among PLHIV may lead them to pay closer attention to drug side effects and become less willing to use free ARVs provided under the policy when these medications are perceived to have more severe adverse effects (Xie et al., Reference Xie, Ngok, Yang and Duan2026). Analyses of mortality data among Chinese adults with AIDS have shown that although the free ARVs policy significantly reduced mortality, its limited supply of second-line treatment drugs has severely affected treatment outcomes for PLHIV (Zhang et al., Reference Zhang, Dou, Ma, Zhao, Liu, Bulterys and Chen2009). Additional research has found that China’s free ARVs list lacks several first-line and second-line drugs recommended in WHO guidelines and does not provide free access to third-line treatment drugs (Zhang and Ma, Reference Zhang and Ma2019). These gaps hinder effective responses to the rising proportion of PLHIV with transmitted drug resistance (Ye et al., Reference Ye, Dong, Lan, Chen, Zhou, Liu, Yuan, Lu, Guo, Zheng, Yang, Song, Liu, Zhou, Zheng, Guo, Yang, Zhang, Ge, Liu, Yu, Han, Huang, Hao, Ruan, Wu, Li, Chen, Ning, Ling, Zhou, Liu, Bai, Gao, Tong, Zhou, Mei, Yang, Wang, Wei, Qiao, Luo, Huang, Wang, Shen, Hu, Zhang, Tan, Fan, Tu, Yu, Fang, He, Chen, Wu, Zhang, Xin, He, Ren, Xu, Sun, Li, Liu, Li, Duan, Huang, Shao, Feng, Pan, Su, Jiang, Zhao, Zhang, Chen, Hu, Wang, Zhao, Cai, Sun, Gao, Ning, Liang, Huo, Fu, Li, Lin, Xing and Lu2024; Zhu et al., Reference Zhu, Han, Zhang and Yang2025). Other review studies note that HIV monitoring and diagnosis in China are primarily the responsibility of CDC, while treatment and care are delivered in designated medical institutions. This arrangement creates physical separation between the locations where PLHIV receive diagnosis and where they access treatment and long-term care, which constitutes a structural barrier to integrated ART service delivery (Cao et al., Reference Cao, Hsieh and Li2020).
The literature reviewed above indicates that China’s current ARVs-related policy environment has entered a phase in which the ‘Four Frees and One Care’policy overlaps with the medical insurance policy. Existing empirical studies have mainly focused on treatment outcomes and have emphasised the challenges the free ARVs list faces in updating its contents. Review studies have primarily identified coordination challenges among institutions responsible for different components of ARVs service delivery. However, insufficient attention has been given to the interaction between different policy-implementing institutions. Empirical investigations are especially lacking regarding how different institutions understand and implement relevant policies in under policy overlap. Based on this gap, the present study takes CDC and medical institutions as research subjects. It examines their understanding of the free ARVs provisions under the ‘Four Frees and One Care’ policy in conditions of policy overlap, their interactions during implementation, differences in their implementation preferences and the reasons underlying these differences, and the implications for HIV prevention and control.
2.2. Theoretical framework
To explore the interactions between CDC staff and hospital physicians in the provision of free ARVs under China’s ‘Four Frees and One Care’ policy, this study adopts the Institutional Analysis and Development (IAD) framework as its theoretical framework. The IAD framework is used to analyse how institutions, defined as the rules-in-use, shape the incentives faced by individuals and influence the outcomes they produce through interactions in action situations (Ostrom, Reference Ostrom2009a). This analytical framework has been widely applied in policy studies across various practical contexts (Grossman, Reference Grossman2019; Nigussie et al., Reference Nigussie, Tsunekawa, Haregeweyn, Adgo, Cochrane, Floquet and Abele2018). The IAD framework consists of exogenous variables, the action arena, patterns of interaction and outcomes, and evaluation criteria and feedback (Ostrom, Reference Ostrom2011). The action arena refers to the social space in which participants interact, exchange goods and services, solve problems, exercise authority, or engage in conflict (Ostrom, Reference Ostrom2019). It comprises action situations and the participants involved in these situations (Ostrom, Reference Ostrom2014). In simple terms, if the action arena is understood as an institutionalised setting in which interactions take place, then an action situation refers to the specific setting in which participants interact and make decisions under a particular configuration of rules (Ostrom, Reference Ostrom2009b). In this study, the ARV supply system clearly constitutes a typical action arena. However, the policy overlap between the ‘Four Frees and One Care’ policy and medical insurance policy gives rise to multiple interconnected action situations. Given that the core component of the IAD framework is the action situation, this study focuses on the provision of free ARVs under the ‘Four Frees and One Care’ policy as the focal action situation for analysis. The internal characteristics of an action situation can be described through seven sets of variables: (1) participants; (2) positions; (3) potential outcomes; (4) action–outcome linkages; (5) control over actions; (6) information; and (7) payoffs or net benefits (Ostrom, Reference Ostrom2019). Exogenous variables are external factors that influence the action arena, and they include biophysical/material conditions, attributes of the community, and rules-in-use. Among these, rules-in-use play a critical role, as they shape participants’ behavioural choices, help maintain order within the action situation, and ensure the predictability of outcomes resulting from participants’ actions (Ostrom, Reference Ostrom2011). Rules-in-use can be further classified into seven types, each corresponding to one of the seven components of the action situation: boundary rules, position rules, choice rules, scope rules, aggregation rules, information rules, and payoff rules (Ostrom et al., Reference Ostrom, Gardner and Walker1994). Specifically, in this study, the ‘Four Frees and One Care’ policy establishes boundary rules, position rules, and choice rules for two groups of participants at the implementation level, namely the CDC and medical institutions. In addition, the degree of information asymmetry generated under information rules further shapes their potential behavioural choices. As a result, the coordination of these potentially divergent choices between the two institutions relies on aggregation rules in order to produce outcomes that are more consistent with policy objectives. However, to fully understand the implications of these differences, it is necessary to examine scope rules and payoff rules within the broader goals of HIV prevention and control and to further assess them in conjunction with other action situations, particularly those represented by health insurance policies.
In short, the IAD framework functions as a tool for analysing how participants, under the influence of exogenous variables, produce patterns of interaction and outcomes within a specific action situation, and for subsequently evaluating these outcomes. The framework is not a fixed model but a flexible analytical tool that can be extended or simplified according to research needs at different levels or variables (Cole, Reference Cole2017). In line with the objectives of this study, which investigates the challenges faced in implementing China’s ‘Four Frees and One Care’ policy, the IAD framework is applied with a focus on the core elements of exogenous variables, action arena, and patterns of interaction and outcomes to ensure the analysis is targeted and operational. Specifically, advancements in HIV medical technologies, the evolution of ARVs, and rules-in-use constitute the exogenous variables in this study. Within the action arena of ARV supply system, the interactions between CDC and medical institutions in implementing the ‘Four Frees and One Care’ policy reflect patterns of interaction, while their implementation preferences in the context of policy overlap with medical insurance constitute corresponding outcomes.
3. Methods
3.1. Research design and sampling strategy
This study employed a qualitative design, collecting data through semi-structured interviews. The IAD framework was used as an analytical lens for interpreting interactions between CDC staff and hospital physicians in implementing the provision of free ARVs under China’s ‘Four Frees and One Care’ policy, as well as the implementation preferences arising under overlapping medical insurance policy. The research team consisted of four members, all with academic backgrounds in public policy and health policy research and training in qualitative research methods. The team jointly conducted the interviews and data analysis, maintaining reflexive notes throughout the process to enhance transparency and mitigate interpretive bias. The researchers’ academic backgrounds facilitated communication with policy implementers and helped establish rapport during interviews.
Because free ARVs in China are provided under the nationwide ‘Four Frees and One Care’ policy, this study conducted interviews with frontline bureaucrats responsible for implementing the policy in CDC and medical institutions located across China’s seven geographic regions: Northeast, North China, Central China, South China, East China, Northwest, and Southwest. Each geographic region comprises multiple cities. The study used cross-site purposive sampling to reduce the likelihood that interview findings reflect only local phenomena. Within each geographic region, multiple cities were sampled while only one implementing institution in each city was selected for interviews. The inclusion criteria for this study required participants to have at least five years of experience in HIV-related work within CDC or medical institutions, as well as direct contact with PLHIV, given the importance of implementers’ familiarity with the policy. Individuals without direct involvement in ART provision or policy implementation were excluded. In total, 42 respondents were interviewed, including 21 from CDC and 21 from medical institutions.
3.2. Data collection
The study was conducted following ethical approval obtained from the institutional ethics committee of the corresponding author institution. Initial access to participants was facilitated through national-level HIV prevention and control civil society organisations, which helped establish preliminary contact with potential respondents. Participants were recruited purposively to ensure coverage of CDC and medical institutions across China’s seven geographic regions. Semi-structured interviews were used to collect data. Interviews were conducted from July 2025 to November 2025. This study forms part of a large-scale nationwide research project on the implementation of HIV ART policies in China. All interviews were conducted face-to-face by trained members of the research team during fieldwork. Prior to each interview, researchers fully explained the study purpose, content, and relevant considerations to participants and obtained their informed consent. Interviews were conducted face-to-face at locations chosen by participants that were quiet and private to ensure comfort and authenticity. Each interview lasted approximately fifty to eighty minutes and was audio-recorded with the participant’s permission. Audio recordings were transcribed within 24 hours after each interview to ensure completeness and accuracy. The interview guide was developed based on the research questions and relevant literature and covered the following topics: participants’ work history and tenure; their understanding of and views on the provision of free ARVs under the ‘Four Frees and One Care’ policy; experiences of interaction with the other institution during policy implementation; and implementation preferences and decision considerations in the context of policy overlap with the medical insurance policy.
3.3. Data analysis
Reflexive thematic analysis was applied to the semi-structured interview data (Braun and Clarke, Reference Braun and Clarke2019), following Braun and Clarke’s six-stage procedure for coding and theme development (Braun and Clarke, Reference Braun and Clarke2006). After initial coding and the development of themes, the IAD framework was used as an interpretive lens to examine how the identified themes related to core analytical dimensions of the IAD framework, including exogenous variables, the action arena, and patterns of interaction and outcomes. Prior to analysis, all interview records were de-identified to protect participant privacy, and respondents were assigned random identifiers: CDC participants were labelled CDC-01 to CDC-21 and medical institution participants were labelled Physician-01 to Physician-21. Coding proceeded iteratively. The researcher first produced initial codes from the interview transcripts and then examined the relationships among codes to develop themes that captured key features of inter-institutional interaction in policy implementation. The themes were subsequently interpreted in relation to relevant IAD dimensions, which helped situate the findings within a broader institutional perspective. To enhance the credibility and reflexivity of the analysis, multiple strategies were employed. Initial coding was completed by one author of the study and then subjected to peer discussion with other members of the research team, who provided feedback and critical scrutiny of preliminary themes to assess conceptual clarity, empirical support, and potential bias. The entire analytic process was documented transparently, recording each step of coding, theme generation, and revision to create an auditable trail. Iterative returns to the raw data were conducted to refine and verify themes, ensuring that the final themes captured both commonalities and differences across institutions. Ultimately, three themes capturing interactions between CDC and medical institutions in policy implementation were identified for analysis.
4. Results
As noted above, the core components of the IAD Framework are the action arena and its subcomponent, the action situation. In this study, the provision of free ARVs under the ‘Four Frees and One Care’ policy is taken as the focal action situation for analysis. Within this setting, the CDC and medical institutions jointly constitute the participants at the implementation level. Recognising the heterogeneity of these actors not only serves as a key entry point for analysing the action situation, but also provides an analytical lens for interpreting the themes emerging from the interview data. Figure 1 presents the thematic map of the study’s findings, while Figure 2 illustrates the connections between the IAD framework and the study’s findings.
Thematic map of study findings.
Connections between the IAD framework and study findings.
4.1. Expansion and controversy of the target population for free ARVs
Within the ARV supply system, the ‘Four Frees and One Care’ policy provides an important and observable action situation. Specifically, this policy defines boundary rules that determine eligibility for free ARVs, establishes position rules that allocate implementation responsibilities to the CDC and medical institutions, and sets out choice rules that guide decision-making by implementing actors in the course of policy execution. In this study, differences in institutional goals and professional perspectives between these two types of implementing actors lead to divergent interpretations of the policy, which in turn result in differences in their implementation tendencies. Such divergence is particularly evident in debates over the appropriate target population for free ARVs.
According to the provisions of China’s ‘Four Frees and One Care’ policy, free ARVs are intended for ‘PLHIV who are rural residents or urban residents without basic medical insurance and who face economic hardship’. Although local governments across China initially required PLHIV to meet certain CD4 thresholds before becoming eligible to receive ARVs at no cost, the requirement that beneficiaries must demonstrate economic hardship has never been enforced in practice. Faced with this clear discrepancy between policy text and policy implementation, CDC staff, as one group of policy implementers, tended to express support for expanding the target population from a broader public health perspective.
In the early days, all the drugs were imported and really expensive, so at first, free ARVs were only given to people with a CD4 count below 200. Later it went up to 350, then 500. Now there is no limit; anyone who is confirmed positive can get free drugs. I think the approach made sense at the time. The government had budget constraints, so it made sense to start with those with lower CD4 counts. But basing it purely on the economic situation of PLHIV really was not very suitable. (CDC-09)
I think it is really necessary to provide free ARVs to everyone because with HIV, treatment is also prevention. Some middle-aged and older people, even if they have medical insurance, are reluctant to spend money on the drugs. Offering free drugs actually helps protect others who are not infected to the greatest extent. (CDC-01)
In contrast to the CDC staff’s generally positive stance, physicians in medical institutions displayed more complex and at times negative attitudes toward the universal provision of free ARVs. On one hand, physicians drew on their practical experience in implementing the policy to describe unintended negative behaviours that emerged among PLHIV in response to this universal benefit. On the other hand, although China’s public hospitals are nonprofit institutions, their clinical activities often involve direct considerations of economic costs and benefits. As a result, when physicians could clearly compare in practice the differences in financial burden between HIV and other diseases, their negative attitudes toward the universal provision of free ARVs became even more pronounced.
I feel that free ARVs can make PLHIV not value them as much. I once had a case where a person had just picked up their free drugs, and shortly after came back telling me they had lost them. You can imagine what I did. So I think they should bear at least some of the cost themselves. (Physician-21)
Right now, China does not invest as much in any other disease as it does in HIV. You can see that if you compare it with something like tuberculosis. Plus, free ARVs require extra manpower and resources to manage. (Physician-16)
Despite the differing views among institutions on the increasingly widespread provision of free ARVs, all respondents emphasised that the relationship between the ‘Four Frees and One Care’ policy and its practical implementation should not be rigid. Instead, they argued that the policy should remain flexible and subject to dynamic adjustment in response to changing historical contexts and the evolving HIV epidemic in China.
I think whether the policy of giving out free drugs needs to continue in the future mainly depends on how the epidemic develops. For example, around here, the number of newly reported cases has gradually gone down in recent years, but I don’t think we have reached a turning point yet. So for now, we still need to use free ARVs to push HIV treatment, though it’s possible that at some point in the future it might not be necessary anymore. (CDC-06)
We have many PLHIV here who were among the first batch infected through blood in China. Back then, the ‘Four Frees and One Care’ policy was not only about controlling the epidemic but also about protecting people’s rights. But things are different now. Most infections are through sexual transmission, so whether we still need a universal free ARVs policy needs to be reconsidered based on both the epidemic situation and the national budget. (Physician-01)
4.2. Tensions in the responsibilities of institutions providing free ARVs
In addition to the influence of the various rules discussed in the previous theme, interactions between the CDC and medical institutions are also constrained by information rules. In this study, because the two groups of participants responsible for policy implementation belong to distinct administrative and professional systems, information asymmetry often arises with regard to PLHIV follow-up, clinical care, and drug management. The persistence of role misalignment and information asymmetry limits the effectiveness of aggregation rules in coordinating different participants. As a result, tensions emerge between the two implementing institutions regarding the appropriate allocation of responsibilities in the provision of free ARVs.
Since the implementation of the ‘Four Frees and One Care’ policy, the division of responsibilities among institutions involved in the provision of free ARVs has evolved differently across regions. Overall, three major models of responsibility allocation have emerged: (1) the CDC is fully responsible for all tasks; (2) the CDC stores ARVs centrally and delivers them periodically to officially designated hospitals responsible for HIV diagnosis and treatment, and these hospitals then dispense the drugs to PLHIV; (3) designated hospitals are responsible for storing, transporting, and dispensing ARVs, and the CDC provides support as needed. In summary, although CDC typically managed the direct provision of free ARVs during the early period, responsibility was later transferred to designated hospitals because CDC staff do not possess prescribing authority. However, considerable regional variation remains regarding the specific role that CDC play. Beyond the question of whether CDC continue to store and transport free ARVs, inconsistencies also exist across regions concerning the authority to approve a switch from the standard free ARVs regimen, tenofovir alafenamide, lamivudine, and efavirenz (TAF+3TC+EFV), to alternative regimens such as lopinavir/ritonavir (LPV/r) or dolutegravir (DTG) when drug resistance develops. Although the inclusion of these alternative regimens in the free ARVs list is not fully uniform nationwide, most regions have incorporated them. Regardless of the model used, even in regions where designated hospitals undertake all tasks related to storage, transportation, and dispensing, CDC continue to participate in work related to free ARVs to some degree in a collaborative manner.
In the early days, because medical knowledge about HIV/AIDS was pretty limited, hospitals didn’t really want to take on the treatment work, so back then the government had us at the CDC handle it. Later on, it was gradually handed over to designated hospitals. Now, our main job is just to assist the hospitals with the distribution of free ARVs. (CDC-03)
Ever since I joined the CDC Department of STD and HIV Prevention and Control in 2013, I haven’t handled ARVs at all. The province just sends the drugs straight to the designated hospitals. But we still coordinate a lot with the local hospitals. For example, HIV confirmation is still done here, and PLHIV first get their positive result from us before being referred to the designated hospital for ART treatment. (CDC-05)
Within this context of intersecting institutional responsibilities, both CDC staff and hospital physicians expressed ambivalent views about the allocation of duties related to the provision of free ARVs. On the one hand, respondents from each type of institution believed that the other institution was better suited, either professionally or in terms of resources, to carry out these tasks. On the other hand, they were concerned that a complete withdrawal from relevant responsibilities would negatively affect coordination and workflow efficiency within their own institution. As a result, their interpretations of responsibility boundaries reflected both a tendency to shift responsibilities and a simultaneous reliance on the other institution. Specifically, CDC staff are oriented toward disease surveillance and prevention rather than clinical treatment. Consequently, CDC respondents generally viewed it as more appropriate for medical institutions to assume responsibility for treatment and drug management. At the same time, they were concerned that such a shift would create challenges for subsequent follow-up responsibilities for PLHIV. Physicians in medical institutions emphasised considerations related to workload and clinical effectiveness. They noted that transferring the management of free ARVs to hospitals would add administrative burdens, whereas leaving these tasks to the CDC would create additional coordination work. This tension in responsibility between policy implementers from different institutions further contributes to coordination challenges during policy implementation.
If the local CDC gets the drugs from the province and then delivers them to the hospitals, it’s really just adding an extra step without much real benefit. Plus, our CDC isn’t a clinical unit. For example, once PLHIV start ART and show abnormal lab results or symptoms, we don’t really know the cause, so hospitals are definitely better suited for that work. But ever since PLHIV started picking up their drugs directly at the hospitals, our follow-up work has definitely become more challenging. (CDC-20)
These specially managed drugs under vertical administration basically added extra work for us. We need space and dedicated staff, even though the drugs are free. However, having the CDC manage the drugs is not necessarily better. They need to collect treatment data, so we still have to coordinate with them and regularly cross-check the information. (Physician-10)
In addition to these issues related to institutional roles and HIV clinical care, external institutional factors, including funding arrangements for drug management, also influence how responsibilities for managing free ARVs are allocated between CDC and medical institutions.
We receive a dedicated fund for drug management every year. A portion of the annual government allocation is set aside specifically for this purpose. I’m not sure whether the local Health Commission’s current decision to have the CDC manage the free ARVs is related to this. (CDC-15)
4.3. Differences in recommendations for PLHIV choice of ARVs
Beyond differences in institutional responsibilities, variations in institutional preferences for recommending different ARVs are also shaped by scope rules and payoff rules. Within the ARV supply system, scope rules define the range of ARVs provided under the ‘Four Frees and One Care’ policy, while payoff rules determine the costs and benefits associated with selecting different ARVs. As the range of ARVs covered by health insurance policies continues to expand, the interaction between these rules becomes increasingly pronounced under conditions of policy overlap, giving rise to multiple interconnected action situations related to ARV recommendations. In this context, the CDC and medical institutions incorporate their respective institutional priorities into their evaluations of ARVs, leading to differing recommendation preferences for PLHIV.
With continued advances in HIV diagnosis and treatment, the range of available ARVs has become increasingly diverse. In addition to the free ARVs provided under China’s ‘Four Frees and One Care’ policy, some newly developed ARVs that previously required full out-of-pocket payment have gradually been included within medical insurance coverage. This has created a situation of policy overlap between the ‘Four Frees and One Care’ policy and the medical insurance system in terms of access to ARVs. For instance, DTG, which is typically purchased through partial reimbursement under medical insurance, is provided free of charge in some regions when PLHIV develop resistance to LPV/r. Overall, PLHIV may obtain ARVs either through the ‘Four Frees and One Care’ policy or through medical insurance, which covers part of the drug cost. As CDC staff and hospital physicians are the primary healthcare providers PLHIV encounter after diagnosis, both groups play a guiding role in influencing how PLHIV select their ARVs. Our interviews revealed that both CDC staff and physicians consider multiple factors when recommending ARVs. Because obtaining ARVs through medical insurance requires PLHIV to pay part of the cost out-of-pocket, the economic capacity of PLHIV is unsurprisingly the most significant consideration influencing recommendations from both CDC personnel and physicians.
A few years ago, this phenomenon of choosing between free ARVs and ARVs partially covered by medical insurance wasn’t that obvious, because there weren’t that many options included in the insurance. But in recent years, it has become much more noticeable, especially in economically developed areas. Still, we still highly respect PLHIV’ own choices. Even when we make recommendations, we usually take their financial situation into account first. (CDC-18)
When I make a recommendation, I still have to consider what kind of PLHIV they are. If it’s someone who clearly isn’t well off, I think it’s better to let them take the free drugs. Otherwise, even if I try to explain the pharmacological differences between the free drugs and the partially self-paid drugs, it’s hard for them to really understand. Then later, they might even complain why someone else is getting the drugs for free while they are being recommended drugs that cost money. (Physician-04)
Despite the shared emphasis on PLHIV’s financial situation, our interviews also revealed clear institution-driven differences in the underlying goals shaping recommendations. CDC staff generally prioritise improving treatment coverage and enhancing adherence. In contrast, hospital physicians place greater emphasis on clinical efficacy and individualised treatment matching, noting that ARVs reimbursed through medical insurance often provide advantages such as better treatment effects and fewer side effects.
In practice, it’s understandable that many doctors mainly recommend drugs covered by medical insurance to increase revenue. But when we deliver HIV-positive results, if we see that a PLHIV financial situation isn’t great, we’ll emphasize to them that if the doctors at the designated hospital recommend insurance-covered drugs, it’s better for them to insist on taking the free drugs. Our main goal is to help them maintain better adherence to their medication. (CDC-10)
Of course, taking drugs through medical insurance is generally better since with free drugs you still need to be concerned about drug resistance and greater side effects. If a PLHIV has employee health insurance and decent financial means, I would even recommend a slightly more expensive drug like Biktarvy. (Physician-05)
These differences again reflect the previously discussed divergences between institutions regarding whether the scope of the provision of free ARVs should be expanded. The core logic behind CDC staff recommending specific ARVs is largely shaped by broader public health goals focused on maximising ART uptake among PLHIV. Physicians, however, concentrate on clinical outcomes and potential implications for healthcare service costs and benefits. When asked about PLHIV who obtain generic ARVs from other countries through informal channels, these differences in underlying logic became even more pronounced.
I think this situation is quite risky because usually, in cases like this, our CDC system won’t show the PLHIV ART status, and we can’t keep track of whether they are continuing to take their ARVs. (CDC-21)
If we don’t take the medical insurance-covered drugs into account, I actually think the overseas generic drugs are still better than the free ones, setting aside the intellectual property issues. They save PLHIV money too. The only real concern is making sure they don’t end up buying fake drugs that don’t work. (Physician-07)
5. Discussion
This study focuses on the implementation of the provision of free ARVs under China’s ‘Four Frees and One Care’ policy. Using the IAD framework, it revealed differentiation between CDC and medical institutions in policy cognition, responsibility allocation, and drug recommendations under conditions of policy overlap with medical insurance. These differences constitute structural issues arising from differences in institutional positions and policy overlap. Their consequences significantly affect the overall effectiveness of ART at the societal level, as well as the continuity and equity of ART treatment for PLHIV. The following discussion integrates the core components of the IAD framework with the specific institutional contexts to analyse the study’s main findings.
First, the study found that CDC support expanding the target population for free ARVs, whereas medical institutions take a more cautious approach. From the perspective of the IAD framework, within the specific action situation of implementing the ‘Four Frees and One Care’ policy, boundary rules determine the inclusion of the CDC and medical institutions as joint implementing actors. The different choice tendencies exhibited by participants in different positions, based on their interpretations of the policy, are directly shaped by the combined effects of position rules and choice rules. Although both are under the administration of China’s National Health Commission, CDC emphasise disease prevention and control, whereas medical institutions focus on treatment. In the context of HIV prevention and treatment, as many studies have noted, the ‘U=U’ principle underscores the interdependence of prevention and treatment; however, in practice, there remains an inherent tension between these two functions (Mody et al., Reference Mody, Sohn, Iwuji, Tan, Venter and Geng2024; Xie and Duan, Reference Xie and Duan2024d). From a policy design perspective, the logic of the free ARVs policy is rooted in a public health-centred approach that emphasises expanding ART coverage by removing economic barriers for PLHIV (Doutre et al., Reference Doutre, Godin, Dmitriev, Pena-Gralle, Bergeron, Blais and Lemire2025). In contrast, the medical insurance policy introduces a different resource allocation logic: it reduces PLHIV’s financial burden without fully eliminating costs, rendering ARVs obtained through insurance more aligned with market-oriented healthcare service logic. The divergence in institutional logic between these two policies corresponds closely with the respective mandates of CDC and medical institutions (Ostrom, Reference Ostrom2009b). When institutional responsibilities become misaligned with policy objectives, medical institutions, which are embedded in the logic of health insurance policies, are likely to experience role ambiguity upon entering the action situation of implementing the ‘Four Frees and One Care’ policy. From a broader perspective, at the level of institutional design, the ‘Four Frees and One Care’ policy and health insurance policies, which operate within the same action arena, are intended to complement each other and jointly contribute to the effective achievement of HIV prevention and control goals. However, when situated within specific action situations, differences in participants’ responsibilities lead to tensions and strategic interactions based on rules-in-use. It should be emphasised that neither policy logic departs from the overarching objective of HIV prevention and control. Empirical research indicates that economic thresholds may reduce ART coverage and consequently lower population-level viral suppression (Jaafari et al., Reference Jaafari, McFarland, Eybpoosh, Tabatabaei, Bafti, Ranjbar and Sharifi2022), while other studies suggest that the choice of ART regimen also affects PLHIV adherence and overall viral suppression (Nachega et al., Reference Nachega, Scarsi, Gandhi, Scott, Mofenson, Archary, Nachman, Decloedt, Geng, Wilson, Rawat and Mellors2023). In the context of almost universal medical insurance coverage in China (Fang et al., Reference Fang, Eggleston, Hanson and Wu2019), partial reimbursement through the medical insurance policy can enable PLHIV to select ARVs most suited to their individual circumstances. At the same time, however, the inter-institutional negotiation arising from policy overlap may redirect potential beneficiaries of the free ARVs policy toward obtaining ARVs through insurance, potentially delaying treatment for economically disadvantaged PLHIV, undermining the high coverage rates that underpin successful HIV control, and exacerbating disparities in ART accessibility. This distinction highlights the value of conceptualising the implementation of the two policies as separate action situations within the same action arena of the ARV supply system. Doing so allows for a more nuanced analysis of their respective rules and interaction patterns, thereby informing more targeted policy reforms. Therefore, given that China’s ‘Four Frees and One Care’ policy functions as a guarantee policy, the future direction of reform should focus on how to expand the applicability of the medical insurance policy in the provision of ARVs, taking into account variations in individuals’ economic conditions.
Second, the analysis continues to focus on the specific action situation constituted by the implementation of the ‘Four Frees and One Care’ policy. The study found that CDC and medical institutions operate in a state of both responsibility-shifting and practical interdependence regarding the provision of free ARVs. This actually reflects the conflicts in cross-institutional coordination obligations and the mismatch of policy implementers that have arisen in the execution of the ‘Four Frees and One Care’ policy. It should be noted that aggregation rules can only play a constructive role when the roles of different participants are clearly differentiated during policy implementation, thereby enabling the allocation of decision-making authority through technical means (Schlager and Villamayor-Tomas, Reference Schlager and Villamayor-Tomas2023). However, in the evolution of this system, policymakers transferred the responsibility for dispensing free ARVs from CDC to medical institutions solely based on the criterion of prescription authority, without reforming the system to reconcile the differing objectives of the institutions. The result is that when a policy must be implemented by an agency whose objectives are not fully aligned with the policy, and that agency is also responsible for collaborative duties with another institution, the conflicts arising from implementer misalignment and insufficient collaborative mechanisms are progressively exacerbated. In other words, when position rules are ambiguous, aggregation rules are unable to function as intended. In addition, actors are simultaneously affected by information asymmetry between the administrative and medical systems arising from information rules. Taken together, these factors may further intensify tensions and conflicts among institutional actors during policy implementation. This is consistent with our finding that micro-level factors, such as government allocations for the management of free ARVs, generate competing interests between the two institutions, and that the inability of free ARVs to create additional benefits for medical institutions also influence the behavioural tendencies of policy implementers. In short, the division of HIV prevention and treatment responsibilities between two different institutions, combined with the absence of effective coordination mechanisms, further intensifies inter-institutional conflicts. Such a fragmented ART service system not only increases the frequency of PLHIV moving between institutions but also raises the risk of treatment interruption and ultimately hinders the integration of the HIV prevention and treatment system (Sierra-Madero et al., Reference Sierra-Madero, Belaunzaran-Zamudio, Crabtree-Ramírez and Magis-Rodriguez2019).
The study also identified differences in ARVs recommendation preferences. It is important to note that within the same action arena, different action situations may be analytically separated but cannot exist in isolation. The potential outcomes of one action situation often constitute the exogenous variables of another action situation (Ostrom, Reference Ostrom2019). In the context of this study, although the ‘Four Frees and One Care’ policy and health insurance policies can be analysed separately at a conceptual level, in practice PLHIV must ultimately make medication choices within a structure where both policies coexist. This underscores the necessity of examining the interactions among different action situations. In this study, it was found that CDC tend to recommend free ARVs to PLHIV, whereas medical institutions are more inclined to recommend ARVs obtained through the medical insurance channel. While both institutions consider the PLHIV’s economic situation as the primary factor in making recommendations, once this economic factor is excluded, the divergence in institutional preferences becomes more apparent. Prior research on physician prescribing preferences for HIV treatment indicates that, when resources are sufficient, physicians tend to select newer ARVs with lower side effects and more stable efficacy (Gelhorn et al., Reference Gelhorn, Garris, Arthurs, Spinelli, Cutts, Chua, Collacott, Lebouché, Lowman, Rice and Heidenreich2022), which was also observed in this study. However, this difference is not merely a professional judgment based on treatment outcomes; it also reflects differentiated incentive structures created by overlapping policies. Compared with free ARVs, insurance-covered ARVs are governed by complex rules related to reimbursement, hospital performance, and drug cost ratios, constituting a new institutional incentive system that subtly encourages medical institutions to treat insurance-covered ARVs as the standard option and free ARVs as alternative pathways. CDC, responsible for follow-up and adherence monitoring, prefer to recommend free ARVs to reduce the risk of treatment interruption due to financial barriers, thereby enhancing overall ART coverage and adherence. Studies have shown that inconsistent treatment recommendations can undermine PLHIV’s trust in the ART system and negatively affect adherence (Sanchez et al., Reference Sanchez, Mills, Coley, Hoskins, Momplaisir, Gross, Brady and Beidas2025). In practice, these potential conflicts between implementing institutions can lead to PLHIV receiving contradictory guidance, which adversely impacts individual and population-level ART outcomes.
Returning to the core issue, the tension between CDC and medical institutions in the provision of free ARVs under China’s ‘Four Frees and One Care’ policy also reflects the ongoing societal debate over whether HIV should be treated as a special disease or integrated into general healthcare. It is undeniable that, since the implementation of the ‘Four Frees and One Care’ policy in 2003, the provision of free ARVs to all PLHIV beyond the original policy text has played an irreplaceable role in advancing HIV prevention and treatment in China (Xu et al., Reference Xu, Han, Jiang, Ding, Li, Han, Lv, Chen, Zhang, Cui, Geng, Zhang, Wang, Kang, Li, Sun, Fu, An, Hu, Chu, Liu and Shang2021). At the same time, with rapid medical advances and the continuous development and updating of new ARVs, older drugs with significant side effects are gradually being phased out, while an increasing number of new drugs are being included in the medical insurance reimbursement system, thereby expanding treatment options for PLHIV (Lee et al., Reference Lee, Lee, Tang, Tsai, Yang, Lin, Wang and Lu2023). Both CDC and medical institutions are aware of this evolving context, yet differing assessments persist regarding whether China’s HIV epidemic has reached a historic turning point. At the policy level, de-specialising HIV certainly helps reduce existing stigma at the societal level, but at the same time, it is essential to consider how to better address the challenge posed by regional disparities in medical insurance reimbursement rates due to differences in local economic development (Yip et al., Reference Yip, Fu, Chen, Zhai, Jian, Xu, Pan, Hu, Zhou, Chen, Mao, Sun and Chen2019), which is a prerequisite for further optimising China’s ART system based on the free ARVs provision under the ‘Four Frees and One Care’ policy. In this regard, ensuring effective coordination between the two institutional systems while reconciling their differing objectives will be key to improving the implementation of China’s ‘Four Frees and One Care’ policy.
6. Conclusion
This study, based on the IAD framework, conducted in-depth interviews with policy implementers to examine the interactions between CDC and medical institutions responsible for the provision of free ARVs under China’s ‘Four Frees and One Care’ policy in the context of overlapping policies. The study found that the scope of the target population for free ARVs under the policy, the allocation of responsibilities for drug provision, and the tendencies in ARVs recommendations are the main sources of tension between CDC and medical institutions during policy implementation. These tensions stem from the fact that current policies have not reconciled the differences between the two types of institutions in terms of institutional objectives, resource structures, and performance incentives, which may adversely affect PLHIV’s access to ART services, ART adherence, and ultimately public health outcomes of viral suppression.
In terms of academic contributions, this study situates the implementation of China’s free ARVs policy within the IAD framework, revealing the interaction mechanisms between public health institutions and medical institutions under conditions of policy overlap, and providing empirical evidence for understanding why policy execution can deviate from intended outcomes. In terms of practical contributions, the study analyzes the implementation challenges of the ‘Four Frees and One Care’ policy regarding the provision of free ARVs and its negative impacts on PLHIV and ART service delivery. In light of this, it is recommended that policymakers in China more clearly delineate the boundaries of applicability between the Four Frees and One Care policy and health insurance policies, and advance policy optimisation in a differentiated manner. One possible approach is to adjust the current positioning of the ‘Four Frees and One Care’ policy as a universal policy and instead realign its implementation with its original intent as a safety-net policy for economically disadvantaged PLHIV. At the same time, efforts should be made to expand the coverage and accessibility of different ARVs within health insurance policies, and to systematically explore the nationwide inclusion of HIV under outpatient chronic and special disease coverage schemes. This would increase reimbursement rates for more effective ARVs obtained through health insurance and help strengthen the complementary roles of the two policies, thereby advancing the overall goals of HIV prevention and control.
It is important to acknowledge the study’s limitations. As the focus was on the tensions between different implementing institutions and their underlying causes, the study relied primarily on qualitative interview data and lacked quantitative evidence to assess the specific impacts of institutional tensions on treatment and prevention outcomes. Future research could combine quantitative data with longitudinal tracking of individuals to examine how policy implementation affects overall viral suppression rates and ART adherence among PLHIV, as well as broader public health outcomes.
Data availability statement
The datasets generated and analysed during the current study are not publicly available due to privacy concerns but are available from the corresponding author on reasonable request.
Acknowledgements
The authors gratefully acknowledge the support of the National Social Science Fund of China: Research on the Practical Challenges and Optimization Pathways of HIV/AIDS ART Medical Security Policies (25BSH151), of which this work represents a stage outcome. The authors also thank all CDC and medical institution staff who participated in this study for their valuable insights.
Author contributions
ZX and ZD designed the study, collected and analysed the data, and wrote the manuscript. KN and ZY contributed to revising the manuscript. All authors read and approved the final manuscript.
Financial support
This research was supported by the National Social Science Fund of China: Research on the Practical Challenges and Optimization Pathways of HIV/AIDS ART Medical Security Policies (25BSH151).
Competing interests
The authors declare that they have no competing interests.
Ethical approval and consent to participate
This research complied with the Helsinki Declaration and was approved by the Institutional Review Board at School of Law, Xingzhi College Zhejiang Normal University. The approval number is XZFX2503. Informed consent was obtained from each participant.
Consent for publication
Not applicable.
Open access
