---
title: Multi-Disease Health System Models
url: https://www.emergentmind.com/topics/multi-disease-health-system-models-hsms
type: topic
---

# Multi-Disease Health System Models

Multi-disease health system models (HSMs) are computational models that jointly represent multiple diseases and, in some formulations, “capture how healthcare delivery systems respond to population health needs” [2508.11730]. In the recent literature, the term encompasses whole-population individual-based simulation, compartmental disease trajectory models, multi-state survival models, foundation models for disease trajectories, multimodal graph predictors, reinforcement-learning treatment policies, and modular health digital twins [2508.11730] [2403.14296] [2605.14227] [2606.11264]. Across these variants, the unifying objective is to model multimorbidity, longitudinal disease evolution, cross-organ dependence, and system-level constraints more faithfully than single-disease or single-organ approaches.

## 1. Conceptual scope and formal objects

A central organizing concept is **multimorbidity**: chronic diseases “frequently co-occur in patterns that are unlikely to arise by chance,” and these patterns can be represented as longitudinal health states rather than isolated diagnoses [2510.09680]. In the compartmental disease trajectory model (CDTM), compartments correspond to multimorbidity patterns, and the model describes chronic disease trajectories across **132 distinct multimorbidity patterns (compartments)** derived from **131 ICD-10 diagnostic groups (A00–N99)** [2403.14296]. Each individual occupies one compartment per year, and transitions are governed by empirically estimated age- and sex-specific probabilities \( q_{g,a,k,j} \), yielding a multilayer directed network of disease-state evolution [2403.14296].

A related formalism is the **multi-state model**, in which disease histories are modeled as transitions between discrete states such as onset, progression, competing events, and death. Multi-state models are specifically described as suitable for “transitions between different disease stages in presence of competing risks,” but their estimation is complicated by dependent left-truncation, multiple time scales, index event bias, and interval-censoring [2509.19956]. In that framework, the transition hazard for transition \(k\) is written as
\[
h_k(t \mid \mathbf{x}_{i,k}) := \lim_{\Delta t \to 0} \frac{1}{\Delta t} P\left(t \leq T < t + \Delta t, o \mid T \geq t, \ell, \mathbf{x}_{i,k}\right),
\]
and piecewise exponential additive models (PAMs) are extended to estimate such hazards under the stated observational complexities [2509.19956].

The same conceptual space also includes **whole-population HSMs**. The Thanzi la Onse (TLO) model of Malawi is described as “the first of its kind” among multi-disease HSMs that build on individual-level epidemiological models of multiple diseases while also modeling healthcare delivery [2508.11730]. By contrast, foundation models such as DT-Transformer treat a patient’s EHR as an ordered event sequence and learn a single model for broad next-event prediction across hundreds of disease categories [2605.14227]. These variants differ in granularity and purpose, but all treat disease histories as coupled trajectories rather than independent endpoints.

## 2. Data substrates and state representations

Recent HSMs are distinguished by the breadth of their data substrates. DT-Transformer was trained on **57.1M structured EHR entries over 1.7M patients from Mass General Brigham (MGB), spanning 11 hospitals and a broad network of outpatient clinics** [2605.14227]. Its inputs include diagnoses coded as ICD-9/10 and truncated to 3-character codes, age in days, masked static covariates such as sex, smoking status, and alcohol status, intermittent “no-event” tokens, and death events [2605.14227]. This representation turns disease trajectories into a temporally ordered token stream.

Other HSMs are explicitly **multimodal**. DiffDT integrates **multi-organ sensor data** from brain, heart, liver, and kidney with **tokenized healthcare events** and **digitalized SDoH proxies**, specifically ICD-10 Chapters **Z and V–Y** [2605.09771]. The UK Biobank data used in that study include **44,834 brain**, **23,987 heart**, **28,722 liver**, and **32,155 kidney** imaging samples, together with **nearly 500k medical history sequences** spanning ages **25–89 years** [2605.09771]. HGDC-Fuse uses EHR and chest X-ray data from **MIMIC-IV** and **MIMIC-CXR**, constructing a patient-centric heterogeneous graph with EHR nodes, one CXR node per image, cross-modal edges encoding relative acquisition time, and inter-patient EHR–EHR edges linking similar patients [2509.15852].

Population-level HSMs rely on registry-scale aggregation. CDTM uses Austria-wide data from approximately **45 million hospital stays spanning 17 years**, while the comparative Austria–Denmark study defines yearly health states as binary diagnosis vectors over **131 disease blocks** and derives **132 distinct, interpretable clusters per country** via divisive hierarchical clustering [2403.14296] [2510.09680]. TLO, by contrast, simulates individuals in a **synthetic Malawi population** and represents demand through disease and symptom status, health-seeking behavior, and health system interaction events that consume workforce time, consumables, equipment, and beds [2508.11730].

Wearable and movement sensing extends HSMs beyond EHR and imaging. ADH-MTL uses **wearable sensor data** and patient profile information to jointly assess **diabetes, cardiovascular disease, high cholesterol, and depression** in a multi-task setting [2511.16398]. The gait foundation model uses **3D skeletal motion** from a **single depth camera** during **five standardized motor/postural tasks**, generating embeddings from **26 3D joints** over **900 frames (30 s)** and testing them against **3,210 phenotypic targets spanning 18 body systems** [2603.25283]. This broadening of input space suggests that HSM state representations are increasingly multi-scale, combining clinical events, sensor-derived phenotypes, and contextual covariates.

| Paradigm | State representation | Data scope |
|---|---|---|
| CDTM [2403.14296] | Annual multimorbidity compartments and transition network | Austria-wide hospital stays |
| DT-Transformer [2605.14227] | Temporal token sequence of diagnoses, ages, static covariates, no-event and death tokens | MGB health system EHR |
| DiffDT [2605.09771] | Tokenized event history plus multi-organ digital twins and SDoH proxies | UK Biobank |
| HGDC-Fuse [2509.15852] | Patient-centric multimodal heterogeneous graph | MIMIC-IV and MIMIC-CXR |
| TLO [2508.11730] | Individual life-course simulation with health system interaction events | Synthetic Malawi population |
| OmniBioTwin [2606.11264] | Modular digital twins coupled in a multi-layer network | Multiscale HDT framework |

## 3. Principal computational paradigms

A dominant line of work models disease histories as **sequence prediction**. DT-Transformer adapts the GPT architecture for disease trajectory prediction, using **12 transformer layers**, **12 attention heads**, an **embedding dimension of 120**, and approximately **2.2M parameters** [2605.14227]. It predicts both the next event and time-to-event, combining cross-entropy for event prediction with negative log-likelihood for an exponential waiting-time model [2605.14227]. DiffDT also uses an autoregressive transformer for tokenized ICD sequences, with an adaptive tokenizer and embeddings
\[
y_t = \mathrm{Embed}_\text{ICD}(s_t) + \mathrm{Embed}_\text{age}(\tau_t),
\]
and an autoregressive loss
\[
\mathcal{L}_{\mathrm{AR}} = -\sum_t \log p(s_{t+1} \mid s_0, \ldots, s_t; \phi)
\]
to encode history before conditioning a generative sensor twin [2605.09771].

A second line uses **generative mediation**. DiffDT formalizes disease reasoning through
\[
P(\text{future disease} \mid \text{history}) = \int P(\text{future disease} \mid \text{biomarker}) \; P(\text{biomarker} \mid \text{history}) \; d\,\text{biomarker},
\]
thereby coupling event histories to multi-organ biomarkers [2605.09771]. For tabular organ traits it uses conditional DDPMs with classifier-free guidance, and for topological brain connectivity it introduces **SPD-VQVAE**, which encodes symmetric positive definite matrices through the Cholesky factorization \( M = LL^\top \) so that generated connectomes remain valid SPD objects while avoiding \( O(N^3) \) spectral operations [2605.09771]. This is a distinct shift from purely event-level disease predictors toward explicit physiological mediation.

A third family uses **heterogeneous graph learning and disease-aware fusion**. HGDC-Fuse represents each patient as a graph \(\mathcal{G}_s = (\mathcal{V}_s, \mathcal{E}_s)\) with modality-specific node types and time-aware edge attributes [2509.15852]. It aggregates messages separately from same-type neighbors and learns a disease correlation matrix
\[
A_{ij} = \frac{\text{co-occur}(i, j)}{\text{count}(i)}, \quad i \neq j,
\]
which is thresholded and processed by a GCN to produce disease prototypes that guide disease-specific attention over EHR features, similar-patient messages, and temporally weighted CXR messages [2509.15852]. The defining feature is that fusion weights are disease-specific rather than globally shared.

A fourth line uses **multi-task heterogeneity modeling**. ADH-MTL treats each disease assessment as a task and models both disease heterogeneity and patient heterogeneity through group-level models \(M_{d,k}\), a decomposition
\[
R_{(d,k),(d',k')} = \alpha R^d_{d,d'} + (1-\alpha)R^g_{k,k'},
\]
and a Bayesian network with variational inference over relationship parameters and model parameters [2511.16398]. This reduces the parameter count from \(D^2K^2\) to \(D^2 + K^2 + 1\) while supporting new-patient predictions via cluster assignment [2511.16398].

Finally, some HSMs are organized as **modular digital-twin systems** rather than a single predictor. OmniBioTwin proposes a **System-of-Twinned-Systems (SoTS)** architecture in which each twin \(T_i^t = (S_i^t, f_i, U_i^t)\) is an autonomous computational subsystem, and heterogeneous twins are coupled through explicit interaction operators in a **seven-layer** architecture spanning data integration, twin modeling, cross-scale coupling, temporal synchronization, and human-in-the-loop decision support [2606.11264]. This framework addresses what the paper identifies as structural fragmentation in current health digital twins.

## 4. Intervention, control, and health-system response

A distinguishing property of HSMs is that they can represent not only disease evolution but also **intervention**. In CDTM, preventive interventions are implemented by modifying transition probabilities to reduce acquisition of a diagnosis \(d\):
\[
q'_{g,a,k,j} = (1-p_{j,d})q_{g,a,k,j} + p_{j,d}p_{k,d}q_{g,a,k,j} + (1-p_{k,d})p_{j,d}q_{g,a,k,j} \cdot \gamma,
\]
where \(\gamma = 0.95\) corresponds to a **5% reduction** in new cases [2403.14296]. The model then propagates the intervention through the transition network to estimate long-term effects on multimorbidity and mortality [2403.14296].

DiffDT implements **counterfactual simulation** at the patient level. At inference, a subject’s event history up to age \(t\) is encoded, a conditional diffusion model generates age-\(t\) multi-organ digital twins given history and SDoH, and a predictive model estimates the next disease event from the generated twin [2605.09771]. The paper describes “do(healthy)” experiments in which an exposure event in history is replaced with healthy, and the generated twin becomes significantly closer to ground-truth healthy subjects than to diseased subjects by FID and WD, with \(p < 1e{-5}\) [2605.09771]. This is a specific implementation of intervention reasoning through generated mediators.

Clinical treatment policy is addressed by **hierarchical multi-agent reinforcement learning**. HMARL decomposes multi-organ sepsis management into a **Root Agent** \(M_{Rt}\), **organ-specific agents** \(M_{Neu}, M_{Car}, M_{Ren}\), a **Mixture Agent** \(M_{OMix}\), and leaf-level dosage agents [2409.04224]. Inter-agent communication is explicit: organ mixture agents consult treatment-specific sub-agents within a system, and \(M_{OMix}\) recursively passes recommendations across organ systems so that, for example, neuro decisions can depend on communicated cardio and renal dosages [2409.04224]. The state representation is dual-layer, with a unified root representation built from **48 physiological variables**, dense embeddings with \(k=8\), higher-order interaction terms, exponentially decayed temporal context, and targeted organ-level refinements [2409.04224]. The learning formulation combines hierarchical Q-learning, options, and QMix-style monotonic mixing [2409.04224].

HSMs that explicitly model **health system production** move from treatment recommendation to delivery feasibility. TLO’s health system module models healthcare services as outputs of joint inputs—worker time, consumables, and facility assets—using a **Leontief production function** such as
\[
Q = \min \left(\frac{L_1}{a_1}, \frac{L_2}{a_2}, \dots, \frac{C_1}{b_1}, \frac{C_2}{b_2}, \dots \right).
\]
It also uses a **Modes framework**: **Mode 0** imposes no requirement for specific workers, **Mode 1** requires relevant worker presence, and **Mode 2** enforces both presence and time-budget constraints [2508.11730]. Workforce absenteeism, motivation, ownership category, management practice, and consumable stock-outs are treated as determinants of effective service output [2508.11730]. In this formulation, an HSM does not merely forecast disease; it simulates whether interventions can actually be delivered under resource constraints.

## 5. Empirical performance and validation regimes

The empirical literature reports strong but heterogeneous performance, with metrics aligned to task structure rather than a single benchmark.

| System | Task and metric | Reported result |
|---|---|---|
| DT-Transformer [2605.14227] | Next-event prediction, median age- and sex-stratified AUC | **0.871** across **896 disease categories**, with all categories exceeding **AUC 0.5** |
| DiffDT-Qwen3 [2605.09771] | Next disease prediction, macro-average AUC / F1 macro-average | **0.9171** / **20.92** vs Delphi-Qwen3 **0.8931** / **18.17** |
| HGDC-Fuse [2509.15852] | Macro-PRAUC across 25 diseases | **0.434** on full dataset; **0.470** on matched subset |
| HMARL [2409.04224] | Estimated mortality rate in sepsis management | **8.81 ± 0.24** vs clinician **16.27** |
| CDTM [2403.14296] | Model fit and intervention effect | Validation fit **\( r = 0.99 \)**; 5% fewer new hypertensive cases yields **0.57% (0.06)** lower all-cause mortality over 15 years |
| Gait foundation model [2603.25283] | Phenotype prediction from gait embeddings | Age **\(r=0.69\)**, BMI **\(r=0.90\)**, VAT **\(r=0.82\)**; **1,980 of 3,210** targets significant |

Several studies emphasize not only discrimination but also **mechanistic or structural validation**. DiffDT reports that the best F1 for each disease category is always obtained using the matching organ’s digital twin as mediator, and that the generated mediators halve the mean absolute average treatment effect error relative to Delphi (**0.004 vs 0.007**) [2605.09771]. HGDC-Fuse reports ablations showing that removing similar-patient connections, temporal CXR aggregation, or disease correlation-guided attention degrades performance, supporting the relevance of its graph construction and fusion design [2509.15852]. In HMARL, disabling either hierarchical state representations or cross-agent communication leads to substantial deterioration, with mortality increases reported as high as **33% higher** than the full model in ablation experiments [2409.04224].

Other validation regimes target generalizability, explainability, or transferability. The Austria–Denmark multimorbidity study reports **Adjusted Rand Index 0.998** and **Normalized Mutual Information 0.88**, indicating strong alignment between independently learned clusterings despite different healthcare systems [2510.09680]. The explainable surveillance system for eight chronic diseases reports **F1 scores >0.75** and **AUROC >0.80** for most diseases and time points, using routine EHR variables rather than laboratory tests and integrating SHAP, surrogate models, and a rule-engineering framework into an EMR deployment pathway [2501.15969]. ADH-MTL reports **15–17% F1** improvement for depression over single-task deep learning baselines and **10% F1** gain for diabetes plus **14%** for depression over the best multi-task baseline, with smaller performance gaps across age, income, race, and gender [2511.16398]. These results indicate that evaluation in multi-disease HSMs is multi-criterion: predictive accuracy, counterfactual validity, policy relevance, robustness under missingness, and system transferability are all treated as first-order concerns.

## 6. Limitations, open problems, and research directions

The literature repeatedly identifies limits in both data and model structure. Existing generative disease models “largely depend on event-level representations from hospital and registry data,” and the absence of explicit modeling of social determinants of health limits personalized disease modeling and clinical decision support; DiffDT addresses this only through **ICD-coded proxies** rather than direct SDoH measurements [2605.09771]. The same study notes **UKB’s healthy volunteer bias** and that its imaging data are “mostly cross-sectional rather than longitudinal,” although the architecture is described as modular for future incorporation of direct SDoH, environmental data, and higher-resolution multimodal inputs [2605.09771].

Real-world multimodal prediction remains constrained by **modality missingness, noise, temporal asynchrony, and evidentiary inconsistency across modalities for different diseases**, which HGDC-Fuse treats as central obstacles to clinical deployment [2509.15852]. Statistical estimation for disease histories remains difficult even when the state space is well defined: PAM-based multi-state models can handle dependent left-truncation and multiple time scales, but multiple-time-scale models were found less robust to the data-generating process than stratified single-time-scale models, and baseline hazards were not well recovered under interval-censoring [2509.19956]. These are methodological limits, not merely engineering details, because they affect the interpretation of estimated transition hazards and progression risks.

At the systems level, current health digital twins are characterized as **structurally fragmented**: monolithic models addressing a single organ or task lack cross-scale fidelity, while system-level twins lack a generalizable architectural framework [2606.11264]. OmniBioTwin responds with a seven-layer SoTS architecture, but the paper presents a demonstration focused on **GLP-1 signaling pathways in Alzheimer’s disease**, so broader validation across multi-disease clinical settings remains to be established [2606.11264]. Similarly, the TLO line of work is still extending its production-side modeling to account for ownership forms, management practices, and worker absence in greater detail [2508.11730].

A common misconception is that multi-disease HSMs are a single model class. The cited work instead shows a heterogeneous field: some models forecast next diagnoses in health-system EHRs, some estimate transition hazards under censoring and left-truncation, some simulate preventive counterfactuals through digital mediators, some optimize multi-organ treatment through hierarchical RL, and some model whether constrained health systems can deliver care at all [2605.14227] [2509.19956] [2605.09771] [2409.04224] [2508.11730]. A plausible implication is that future HSMs will increasingly combine these strands: multimodal state representation, explicit causal or mechanistic mediation, intervention and resource simulation, and modular cross-scale composition. The current literature already points toward that synthesis, even though each component remains under active development.

Source: https://www.emergentmind.com/topics/multi-disease-health-system-models-hsms