Indonesia faces a persistent crisis in the production and equitable distribution of medical specialists, a challenge that directly undermines the quality, accessibility, and equity of health services for its 287 million people [1]. As of January 2025, national data report only 51,443 specialists, equating to a density of 0.18 specialists per 1000 population [2], significantly below the Ministry of National Development Planning (Bappenas) target of 0.28 per 1000 [3]. This overall deficit is projected to widen to 70,000 specialists by 2032 [4]. Compounding this quantitative shortage is severe maldistribution: nearly 59% of specialists are concentrated in Java and major urban centers [5], leaving vast areas of the archipelago, particularly underserved regions (Daerah Tertinggal, Perbatasan, dan Kepulauan [underdeveloped, border, and island regions; DTPK]), critically underserved [4].

This specialist deficit translates directly into significant public health vulnerability and contributes to avoidable mortality and morbidity across national priorities [6,7]. Delays in accessing obstetric and pediatric specialists contribute to persistently high maternal and neonatal mortality, particularly in regions with the lowest specialist availability [8]. The inadequacy of specialist coverage critically undermines core initiatives, including Maternal and Child Health (Kesehatan Ibu dan Anak [KIA]) programs [9]. The growing burden of noncommunicable diseases, which account for up to 73% of all deaths in Indonesia [10], necessitates timely access to specialized cardiology, oncology, and neurology services—most of which remain out of reach for populations outside major cities [9]. Low specialist density is consistently associated with adverse public health outcomes, including higher hospitalization rates for ambulatory care–sensitive conditions, longer diagnostic delays, and reduced survival for time-sensitive emergencies [11,12]. Given that the current university-based system produces only about 2700 specialists per year, eliminating the existing deficit of roughly 30,000 specialists would take at least a decade, even before accounting for retirements or the rising demand driven by population aging [5].

To situate the present reform, we briefly describe the Indonesian medical workforce hierarchy. Generalists (dokter umum or general practitioners) complete a 5.5‐year to 6-year undergraduate medical program comprising preclinical and clinical training, followed by a 1-year mandatory national internship (Internship Dokter Indonesia) before obtaining independent licensure. As of January 2025, Indonesia had approximately 200,000 active generalists staffing community health centers (Puskesmas), district hospitals, and family practice settings [13]. Generalist training and the internship are governed by separate regulatory cascades and have been described elsewhere [14].

Specialists complete an additional residency program (Program Pendidikan Dokter Spesialis [specialist medical education program; PPDS]) lasting 2 to 6 years, in one of 39 specialties recognized by the Indonesian Health Council (Konsil Kesehatan Indonesia [KKI]) and governed by specialty collegia. Until 2023, all PPDS programs were exclusively university-based, with residents officially registered as tuition-paying postgraduate students. A consultant subspecialist tier requires a further 2 to 3 years. This paper concerns the specialist (PPDS) tier; the hospital-based RSPPU pathway introduced in 2024 sits within this tier and is the focus of analysis.

Indonesia’s existing residency framework presents a structural anomaly when benchmarked against global standards [15]. In the United States, the United Kingdom, the Netherlands, Japan, and Australia, residents are recognized as licensed members of the clinical workforce who receive salaries, benefits, and formal employment status [16-20]. This holds in Malaysia as well, which operates both hospital-based and university-based systems but pays all residents a government salary [21]. In Indonesia’s traditional university-based model, by contrast, residents are classified as tuition-paying students who pay substantial education fees while performing full clinical duties without remuneration [22]. This arrangement creates substantial equity barriers to entry, limiting the pool of aspiring physicians able to pursue specialization and constraining the size of the specialist workforce [23,24].

To urgently address the triple crisis of insufficient production, inequitable distribution, and the structural anomaly of unpaid residency, the Indonesian government mandated a transformative reform under the 2023 Health Law [25]. This landmark legislation introduced a dual-track specialist education system that formally integrated the Hospital-Based Specialist Program (Rumah Sakit Pendidikan Penyelenggara Utama [primary teaching hospital; RSPPU]) with existing university-based programs. The new model positions designated national referral hospitals as co-organizers of residency training, ensures that residents receive living stipends and pay no tuition, and explicitly links specialist training to equity goals by binding deployment to underserved regions upon graduation. This dual approach is a comprehensive, scalable, and competency-based strategy to accelerate specialist output and drastically improve equitable access to specialized care [26].

The aim of this paper is threefold: (1) to document the rationale, policy design, and early implementation of Indonesia’s dual hospital–university specialist medical education reform; (2) to interpret these processes through an 8-step change management framework developed by Kotter; and (3) to extract transferable policy lessons for low- and middle-income countries (LMICs) facing analogous workforce crises. We address three guiding research questions: (RQ1) What policy logic and political processes catalyzed the dual-track reform? (RQ2) How is the hospital-based pathway being implemented during the first 18 months and with what early outcomes? (RQ3) What systemic tensions and policy implications emerge for domestic scale-up and international adaptation?

We used an integrative qualitative policy process review design to examine the design and early operational dynamics of Indonesia’s dual-track specialist education reform. An 8-step change management framework developed by Kotter [27-29] served as both an a-priori interpretive lens and a deductive coding frame. Each step was mapped to observable indicators in the RSPPU rollout: (1) establishing urgency: quantification of the specialist deficit and high-stakes political articulation; (2) forming a guiding coalition: cross-ministerial governance arrangements involving the Ministry of Health (MoH), Ministry of Higher Education, KKI, Indonesian Accreditation Agency for Higher Education in Health (Lembaga Akreditasi Mandiri Pendidikan Tinggi Kesehatan Indonesia [LAM-PTKes]; IAAHEH), and the specialty collegia; (3) developing a change vision: Law 17/2023 and its successor regulations; (4) communicating the vision: subsequent public communication; (5) empowering broad-based action: Government Regulation 28/2024 [30], Permenkes 14/2024 [31], and Lembaga Pengelola Dana Pendidikan (Indonesia Endowment Fund for Education; LPDP) financing; (6) generating short-term wins: pilot cohort enrollment and the unified digital platform; (7) consolidating gains: joint monitoring, dual accreditation, and expansion to additional RSPPUs; and (8) anchoring change in culture: embedding mechanisms including the integrity pact, Pangkalan Data Pendidikan Tinggi (Higher Education Database; PDDikti) recognition, and the Badan Penyelenggara Jaminan Sosial (National Social Security Agency; BPJS)–linked financing pipeline.

Data collection ran from October to November 2025, covering the first 18 months of program operation. At that point, all 52 inaugural residents were in their fourth or fifth semester; none had yet graduated, as the shortest pilot specialty (ophthalmology) requires a minimum of 7 semesters before the national board examination. The study therefore examines design and early implementation fidelity, not graduate outcomes.

Verbal informed consent was obtained from all interviewees prior to recording. Data were anonymized, and any identifiable information was excluded to protect participant privacy. Formal ethical approval was not required for this study, as it involved a policy evaluation using nonclinical, nonidentifiable data [34]. No patient-level data were used.

The integrative policy review confirms that Indonesia’s reform establishes a dual-track specialist training system that is operationally feasible and addresses critical shortages and maldistribution. Across the 2 evidence streams, four cross-cutting themes structure the findings: (1) financial barrier removal as an equity lever; (2) dual governance pragmatism; (3) accreditation strain and capacity-building; and (4) equity-anchored deployment. Triangulated evidence supports strong fidelity to Kotter’s steps 1 to 6 and early evidence for steps 7 to 8 (Table 2). Key successes include eliminating financial barriers through stipends and waived tuition, ensuring equitable access via a centralized digital selection system, and strengthening quality assurance and resident well-being through digital tools (an e-logbook with anonymous reporting). Persistent operational tensions are detailed in the dedicated subsection below.

The inaugural cohort comprised 52 residents selected from 412 (overall acceptance rate 12.6%) eligible applicants nationally. All 52 announced positions were filled. Distribution across pilot specialties was as follows: ophthalmology (Cicendo Eye Hospital, n=8 slots filled, from 96 applicants); cardiology (National Cardiovascular Center Hospital Harapan Kita, n=10, from 87 applicants); pediatrics (National Women and Children Health Center Hospital Harapan Kita, n=8, from 79 applicants); orthopedics (Prof Dr Soeharso Orthopedic Hospital, n=10, from 64 applicants); neurology (National Brain Center Hospital, n=10, from 45 applicants); and oncology (Dharmais Cancer Hospital, n=6, from 41 applicants). As of December 2025, no resident had withdrawn; all 52 remained enrolled.

The joint monitoring exercise was documented in 2 official interministerial site-visit reports that were analyzed as part of the documentary evidence stream (see Methods). A joint monitoring team composed of representatives from the MoH (Directorate of Health Human Resources Provision), the Ministry of Higher Education, IAAHEH/LAM-PTKes, and the relevant specialty collegia conducted 2 structured site visits: the first in December 2024 (approximately 7 mo after the May 2024 presidential launch) and the second in June 2025 (approximately 13 mo post-launch). At each visit, a standardized 23-item instrument adapted from the ACGME-International Self-Study framework assessed five governance domains: (1) governance and DIO leadership; (2) curriculum implementation and milestone tracking; (3) faculty development and supervision ratios; (4) infrastructure (simulation, library, and e-logbook uptime); and (5) learner support (well-being, anonymous reporting use, duty-hour compliance). Each domain was scored on a 4-point ordinal scale (excellent, satisfactory, needs improvement, or unsatisfactory). Across both visits, all 6 pilot sites were rated satisfactory or better on all 5 domains, with no domain at any site scored needs improvement or unsatisfactory (Table 2, Kotter step 7).

Triangulated evidence mapped to an 8-step framework developed by Kotter showed strong fidelity for steps 1 to 6 and early, partial evidence for steps 7 to 8 (Table 2). Step 1 (urgency) was established through the 70,000-specialist deficit projection and high-level political articulation. Step 2 (guiding coalition) is operationalized through the joint interministerial monitoring team. Step 3 (vision) is codified in Law 17/2023. Step 4 (communication) was executed through the nationally broadcast May 2024 presidential launch and sustained public communication by the MoH. Step 5 (empowerment) is reflected in successor regulations (Government Regulation 28/2024, Permenkes 14/2024) and LPDP financing. Step 6 (short-term wins) is evidenced by the enrollment of 52 residents and the operational unified digital platform. Step 7 (consolidating gains) is partially evidenced through dual accreditation and announced expansion to additional RSPPUs in 2025 to 2026. Step 8 (anchoring in culture) remains incomplete: the integrity pact, PDDikti integration, and BPJS-linked financing each represent unfinished embedding mechanisms.

This integrative policy process review of Indonesia’s dual-track specialist training reform yields 3 principal findings. First, the RSPPU model is operationally feasible at 18 months: all 6 pilot hospitals established the requisite educational infrastructure, all 52 selected residents remain enrolled, and a multiagency joint monitoring framework documented satisfactory or better performance across governance, curriculum, faculty, infrastructure, and learner support. Second, the reform’s defining innovation is structural rather than pedagogical. By abolishing resident tuition, reclassifying residents as supervised health professionals, and binding selection to an equity-anchored integrity pact, the model recasts specialist training as a public investment in spatial health equity rather than a private financial transaction. Third, the persistent implementation tensions, including dual-accreditation adaptation, the absence of a national clinical-educator compensation framework, PDDikti integration delays, and Java-centric examination logistics, are operational and tractable rather than conceptual and define a clear policy agenda for scale-up.

The RSPPU model corrects a longstanding misalignment: specialist training has always taken place predominantly in hospitals, yet formal authority resided with universities. The 2023 Health Law resolves this not through institutional substitution but through functional realignment. Hospitals now lead curriculum delivery, mentorship, and day-to-day program management; universities retain cosignatory authority over degrees; and professional collegia ensure that both pathways adhere to the National Specialist Education Standard [39]. Internationally, Indonesia’s design sits between fully hospital-employer models such as the United Kingdom NHS and the United States ACGME, and university-mandated systems found in parts of Continental Europe and Japan. Malaysia’s parallel dual system [21] is the closest structural comparator, but Indonesia’s model is distinctive in 3 respects: the explicit binding of admission to a return-of-service integrity pact; LPDP-financed full coverage of resident fees; and a dual ACGME-I/IAAHEH accreditation requirement applied to every pilot site. Unlike the US model, Indonesia’s pathway emerged from negotiated tripartite compromise rather than a top-down mandate. Unlike the UK’s centrally employed model, it preserves academic coauthority. This hybridity may be what makes the model relevant to other LMICs whose health and higher-education ministries operate in parallel rather than in aligned hierarchies.

The reform’s equity imperative is operationalized through 3 reinforcing mechanisms. First, a unified national selection platform removes geographic barriers, allowing physicians from Aceh to Papua to compete on equal terms. Second, a binding integrity pact at the point of entry obliges graduates to serve in their regions of origin, a commitment built into admission rather than imposed after graduation. Third, residents are now salaried professionals rather than fee-paying students. With tuition fully covered by LPDP and monthly stipends provided, the model removes the largest financial barrier that historically excluded talented but underresourced candidates, bringing Indonesia into line with the global norm of recognizing residents as members of the clinical workforce. The equity mechanism operates at entry, not exit. Prior postgraduation deployment schemes such as the WKDS were judicially invalidated [40], and exit-based coercion has failed in multiple LMICs. By embedding the return-of-service obligation in the admission contract and financing the residency with public funds, the reform converts deployment from a contested instrument of coercion into a consensual condition of participation.

Early implementation across 6 national referral hospitals confirms operational feasibility. The joint monitoring framework, itself a Kotter step 7 consolidation mechanism, produced satisfactory or better ratings in both the 2024 and 2025 site visits. Several tensions identified in the Results nonetheless require interpretation. The variability in clinical-educator compensation across pilot sites is not a temporary inconvenience but a fundamental design gap. Without a national framework, scale-up risks producing a 2-tier system in which residents at well-resourced flagship hospitals receive structured mentorship while those at expanded RSPPUs encounter underincentivized faculty. The PDDikti integration delay signals a related institutional question: are RSPPU residents learners, employees, or both? Resolving this status ambiguity will require coordinated action by the MoH and the Ministry of Higher Education, ideally through a dual-status entry in PDDikti that preserves academic benefits while reflecting the residents’ employment-grade compensation. From a Kotter perspective, these tensions are characteristic of steps 7 and 8. The reform has generated visible wins and structural commitments, but cultural and bureaucratic embedding will require sustained leadership attention through at least the first graduating cohort in 2028.

The reform’s domestic durability hinges on five concrete actions: (1) issuance of a national clinical-educator compensation framework anchored in BPJS Health case-mix reimbursement, with a minimum hourly teaching credit codified in a Permenkes successor regulation; (2) full PDDikti integration with an explicit dual learner-employee status to resolve resident benefit ambiguity; (3) regional decentralization of selection examination hubs to neutralize Java-centric travel costs; (4) statutory consolidation of the integrity-pact return-of-service obligation into the new specialist license cycle so that it is enforceable through licensure renewal; and (5) a national tracking platform for RSPPU graduates’ geographic deployment, retention beyond the initial bond, and downstream clinical outcomes in destination districts.

Four design lessons are transferable to LMICs facing the same triple workforce challenge of undersupply, maldistribution, and structurally unpaid residency, though direct replication is not implied. (1) Anchor reform in primary legislation, not subordinate regulation alone. Indonesia’s Law 17/2023 created the political foundation under which negotiated compromises could be operationalized through successor regulations. Lower legal instruments are reversible by a single minister. (2) Resolve the resident-status anomaly first. Removing financial barriers is the largest single equity lever and the most politically defensible against reversal. (3) Build dual accreditation deliberately. Treating ACGME-I and a strong national accreditor as complementary rather than competing reduces foreign-perception risk while preserving sovereignty. (4) Engineer equity at entry, not at graduation. A binding integrity pact at admission is administratively simpler and politically more durable than postgraduation deployment schemes, which have failed in multiple LMICs including Indonesia’s own invalidated WKDS [40] and earlier Malaysian iterations [21].

Indonesia’s dual hospital-university residency model addresses the nation’s specialist shortage and geographic maldistribution through structural rather than incremental reform. By designating national referral hospitals as co-organizers of training under collegium-led standards, the RSPPU pathway eliminates tuition barriers, provides living stipends, and directs deployment to underserved regions through a binding integrity pact. Early implementation across 6 pilot sites demonstrates operational feasibility: residents are salaried professionals trained in high-volume clinical settings, supported by a digital platform for transparent selection, competency tracking, workload monitoring, and anonymous well-being reporting. Remaining challenges, including accreditation adaptation, educator incentives, learner-database integration, and centralized examination logistics, are operational and tractable rather than conceptual.

Pending longitudinal evaluation of graduate output, geographic retention, and clinical outcomes, this model offers a transferable design template for LMICs confronting parallel workforce crises, though it is not yet a definitively replicable blueprint. Whether the reform delivers on its equity promise will be answered not by its design but by whether the first 52 residents and the cohorts that follow practice in the districts where Indonesia’s health system most needs them.