Sage Journals: Discover world-class research

Abstract

Background: The health insurance industry faces challenges like inefficiencies, fraud, lack of transparency, and limited customization. With its decentralized structure, smart contracts, and immutable records, blockchain technology offers a transformative solution by enhancing transparency and operational efficiency. Objective: This study proposes a blockchain-based health insurance model to improve transparency, fairness, and efficiency while addressing existing limitations. Methods: The proposed framework integrates smart contracts with quadratic voting (QV) and advanced validation techniques, creating a democratic, secure, and customizable insurance process. Results: The model tailors personalized insurance plans to collective preferences using QV-based decision-making and dynamic pricing. Blockchain enhances trust and system reliability, while the inclusion of QV fosters inclusivity and fairness. Conclusion: By combining blockchain’s decentralized architecture with QV, the proposed system overcomes the limitations of traditional insurance, offering a scalable, efficient, and equitable alternative that aligns individual preferences with societal health goals.

Keywords

blockchain health insurance quadratic voting smart contracts validation D72 I13 C52

Introduction

Health insurance and public health serve distinct roles within the healthcare ecosystem. Public health initiatives such as immunization programs and pollution control aim to enhance societal well-being by addressing collective health concerns. These initiatives are considered public goods—non-excludable and non-rivalrous—funded and implemented primarily by governments.¹ In contrast, health insurance is a market-based mechanism for mitigating individual health risks. It operates as a “club good,” providing financial protection and healthcare access to individuals who contribute through premiums or employment-linked memberships (i.e., it is excludable) while remaining non-rivalrous. This inherent exclusivity highlights significant differences in accessibility and equity between public health and health insurance systems.

Despite its critical role in providing financial security, traditional health insurance systems are fraught with inefficiencies, including fixed pricing structures, limited coverage, reliance on intermediaries, slow reimbursements, and opaque decision-making processes. These challenges contribute to inequities in coverage, erode stakeholder trust, and hinder the system’s ability to adapt to diverse and evolving needs.²

The distinction between public health and health insurance underscores the need for innovative solutions that address the latter’s limitations while leveraging its potential to enhance individual health outcomes. Unlike public health, which relies on centralized government funding, health insurance has the flexibility to integrate market-based mechanisms and emerging technologies to improve efficiency, inclusivity, and transparency. This study proposes a blockchain-based health insurance framework as a transformative approach to bridge these gaps.

Blockchain technology, a decentralized digital ledger system, has transformed finance, supply chain management, healthcare, and insurance industries by ensuring transparency, security, and immutability without relying on a central authority. Transactions are recorded across multiple computers, with each new transaction linked to the previous one, forming a chain. Initially introduced in 2008 with Bitcoin,³ blockchain has evolved through several generations. The second generation, marked by the introduction of smart contracts in 2014 through platforms like Ethereum, expanded blockchain’s application beyond digital currencies.⁴ The third generation, emerging around 2017, introduced significant scalability, interoperability, and energy efficiency improvements, addressing major limitations of earlier blockchain models. These enhancements enable real-time data processing, lower transaction costs, and improved integration with advanced technologies such as machine learning (ML) and privacy-preserving mechanisms.⁵ The current fourth generation integrates artificial intelligence to drive further innovation.⁶

Blockchain and smart contracts offer significant potential to address inefficiencies in health insurance systems. Blockchain can improve efficiency and transparency by streamlining processes, reducing fraud, and eliminating intermediaries, while smart contracts allow for customizable insurance plans with faster settlements, enhancing system performance.⁷

This study leverages third-generation blockchain technology for its enhanced transaction speed, cost efficiency, and integration capabilities with ML models and smart contracts. Unlike previous generations, third-generation blockchains introduce solutions like sharding and directed acyclic graphs to improve scalability, enabling faster and more efficient processing of health insurance transactions. Additionally, they facilitate interoperability between different blockchain networks, ensuring seamless communication with various healthcare platforms. Moreover, third-generation blockchains prioritize energy efficiency by adopting Proof of Stake (PoS) consensus mechanisms over the energy-intensive Proof of Work (PoW). These advancements are essential for developing a blockchain-based health insurance model that ensures transparency, fraud prevention, customization, and real-time decision-making, surpassing the limitations of earlier blockchain generations.

Literature review

Advancements in healthcare technology highlight the potential of integrating blockchain with emerging frameworks. The existing literature can be categorized into three main areas, as outlined below.

Blockchain in healthcare

Ghadi et al. emphasize the transformative role of the Internet of Medical Things and 5G-enabled Mobile Edge Computing systems in addressing latency, privacy, and computational inefficiencies within healthcare applications.⁸ These technologies improve data handling and real-time analytics, while blockchain enables secure management of decentralized health records and transactions, facilitating seamless patient data integration. Similarly, Corte-Real et al. highlight blockchain’s potential for health data sharing, emphasizing its alignment with regulations like Health Insurance Portability and Accountability Act (HIPAA) and General Data Protection Regulation (GDRP).⁹

Health data security and privacy

Ensuring patient data privacy is a major challenge in healthcare. Studies propose various blockchain-based solutions, such as hybrid architectures that balance on-chain transparency with off-chain data storage.⁹ Nunes et al. explore the use of non-fungible tokens (NFTs) in medical records and pathology exams, demonstrating how decentralized trust models enhance security and efficiency.¹⁰ Additionally, Agbo et al. analyze blockchain’s potential in drug supply chains and remote monitoring while identifying scalability and integration challenges.¹¹

Blockchain for health insurance

Blockchain has also been applied to health insurance to streamline claims processing and fraud detection. Bae and Yi discuss the eClaim system in South Korea, which automates claim verification using standardized healthcare data.¹² Ismail and Zeadally introduce the Block-HI framework, a blockchain-based fraud detection system that identifies fraudulent claims across multiple scenarios.¹³

Beyond automation, emerging research integrates ML techniques and optimization algorithms into blockchain-based healthcare models. Khan et al. demonstrate how advanced neural networks enhance disease prediction,¹⁴ while Shah et al. highlight the role of decentralized models in improving data security.¹⁵

Together, these studies lay the foundation for blockchain’s role in modern healthcare. However, challenges such as scalability, regulatory compliance, and real-world implementation remain, which this study aims to address through a blockchain-based insurance model integrating QV and fraud prevention mechanisms.

Contribution to literature

The proposed blockchain framework builds on existing advancements by integrating blockchain technology with the QV technique and advanced validation mechanisms. By leveraging decentralized structures, smart contracts, and economic theories, the framework addresses key challenges in traditional health insurance, including inefficiencies, fraud, limited transparency, and lack of flexibility. This study contributes to the literature by proposing a blockchain-based model that combines economic principles with innovative technologies. Specifically, it:

(1) Develops a blockchain-based health insurance system with smart contracts to offer personalized insurance plans.

(2) Employs a decentralized architecture to enhance transparency, trust, and efficiency.

(3) Uses QV mechanisms to make decision-making more democratic and ensure fair outcomes.

(4) Incorporates KYC (Know Your Customer) processes and ML models to prevent fraud and maintain system security.

By bridging inherent gaps in traditional health insurance systems, the proposed framework delivers a scalable, equitable, and efficient alternative aligned with the health objectives of both individuals and society.

The rest of the paper is organized as follows: Section 2 describes the study’s methodology. Section 3 introduces the blockchain model developed for health insurance. Section 4 outlines a comprehensive evaluation plan for the proposed model. Section 5 discusses the findings, highlights the strengths and limitations of the proposed model, and offers recommendations for future research. Finally, Section 6 concludes the paper.

Methodology

This section outlines the theoretical framework and mathematical foundations of the proposed blockchain-based health insurance model.

Quadratic voting

Quadratic voting is a mechanism designed to enhance the efficiency of collective decision-making processes. Consider a scenario where a group of $N$ individuals must decide between two alternatives. Each individual has a private value of $u_{i}$ , representing their willingness to pay for one alternative over the other. Individuals are provided with a stock of voice credits, which they can use to vote for their preferred alternative. However, the cost of voting is calculated as the square of the votes cast.

In this decision-making process, the voter $i$ gains a value of $2 u_{i}$ if alternative $A$ is chosen over alternative $B$ , and a value of $- 2 u_{i}$ if $B$ is chosen over $A$ . The community of voters determines the alternative to be implemented. Alternative $A$ is implemented if the sum of all votes ( $\sum v_{i})$ is nonnegative. Each voter pays a fee of $C (v_{i})$ , where $C$ is a differentiable, convex, and strictly increasing function in $| v_{i} |$ .¹⁶

Private value and voting cost

Following the model proposed by Posner and Weyl,¹⁷ we consider a scenario involving independent symmetric private values with $N$ voters. Each voter’s willingness to pay is represented by $u_{i}$ . These independent values follow a probability density function $F$ . Each person knows only their value. However, the sampling distribution of $F$ is known to all. Let $σ^{2}$ , $μ$ , and $μ_{3}$ be the mean, variance, and third moment of $u$ under the $F$ distribution, respectively. Each voter $i$ buys $v_{i}$ votes at $v_{i}^{2}$ dollars.

Assuming the voting price ( $p$ ) across all voters, a rational voter chooses $v_{i}$ such that their utility function $U_{i} = 2 p v_{i} u_{i} - C (v_{i})$ is maximized. Assuming that voice credits are equally distributed, society tends to implement alternative $A$ when the summation of private values is nonnegative (i.e., $\sum u_{i} \geq 0$ ).¹⁸

Optimal voting pricing rule

Theorem: The vote pricing rule is optimal if it is quadratic.

Proof: The first-order condition for optimal voting is obtained by computing the derivative of the utility function:

{U_{i} = 2 v}_{i} u_{i} - C (v_{i})

(1)

Consider the vote pricing rules:

C (v_{i}) = {v_{i}}^{a}

(2)

Combining equations (1) and (2), we get:

U_{i} = 2 p v_{i} u_{i} - {v_{i}}^{a}

(3)

Maximizing the utility function gives:

\frac{d U_{i}}{d v_{i}} = 2 p u_{i} - {a v_{i}}^{a - 1} = 0 \to v_{i}^{*} = {(\frac{2 p u_{i}}{a})}^{\frac{1}{a - 1}}

(4)

If $a = 2$ , then $v_{i}^{*}$ is proportional to $u_{i}$ (i.e., $v_{i}^{*} = p u_{i}$ ), which is globally optimal. For other $a$ , $v_{i}^{*}$ is not proportional to $u_{i}$ , making the voting price rule sub-optimal. When $a$ approaches $1$ (nearly linear cost), $v_{i}^{*}$ approaches infinity, leading to a “dictatorship” of intense voters. When $a$ approaches infinity (highly convex cost), $v_{i}^{*}$ converges to 1, meaning each voter casts only one vote.¹⁹

Thus, QV balances dictatorship and majority rule. While acting with self-interest, voters unintentionally promote societal welfare.²⁰ The bottom line is that the quadratic pricing rule is the most optimal for voting.

Validation of health insurance claims

Validating health insurance claims is crucial for ensuring accuracy, transparency, and reliability in any insurance model, particularly a blockchain-based system. Claims data is a vital source of real-world evidence, informing risk assessment, premium calculations, and evidence-based decision-making. Accurate validation ensures computational models reflect real-world phenomena, supporting fairness and sustainability in health insurance systems.²¹

In the insurance context, risk is often modeled as a random variable $X$ , representing potential claims. Using the Poisson distribution, where claims $X_{i}$ occur independently $N$ times, the total risk is expressed as:

X = \sum_{i = 1}^{N} X_{i} \sim P o i s s o n (λ)

(5)

Here, $λ > 0$ represents the expected number of claims, making it ideal for modeling count data like insurance claims.²² To link claims with individual characteristics, let $Z_{i}$ denote a vector of exogenous factors, including demographics like age, gender, and health status. The probability of $n_{i}$ claims for individual $i$ , given $Z_{i}$ , is:

P r o b (N_{i} = n_{i} | Z_{i}) = \frac{λ_{i}^{n_{i}} \cdot e^{λ_{i}}}{n_{i}!}

(6)

where

λ_{i} = \exp (Z_{i}^{'} β)

reflects expected claims based on individual traits.²³ This approach helps insurers tailor premiums fairly and sustainably. Premium (

P_{i})

is calculated using:

P_{i} = {\bar{P} + λ}_{i} C_{i} + R A C_{i}

(7)

where

\bar{P}

is the base premium,

C_{i}

is the average claim cost, and

R A C_{i}

(Risk Adjustment Coefficient) accounts for other individual risks not captured by

Z_{i}

. Advanced analytics, such as ML techniques, further enhance prediction accuracy and claims processing, optimizing premium pricing and improving customer satisfaction.

Blockchain framework for secure and transparent health insurance

This section presents a blockchain-based health insurance solution, structured into six layers and eight stages. We also incorporate attacks and solutions specific to each blockchain layer to address potential security vulnerabilities. The system’s effectiveness is further enhanced through the integration of the QV mechanism.

Blockchain architecture

The proposed blockchain architecture for health insurance is developed in six layers: Application, Identity and Data Management, Consensus, Smart Contract, Ledger and Storage, and Network. Each layer plays a crucial role in ensuring the security and efficiency of the health insurance system. Figure 1 displays the six-layer architecture of the proposed blockchain model, highlighting the processes managed by each layer and the associated potential threats.

Figure 1.

A six-layer blockchain model for health insurance with potential threats. Arrowed lines represent the logical flow and interactions between blockchain layers and stages. Dashed lines indicate the threats associated with each blockchain layer and stage.

To further elaborate on these security concerns, Table 1 summarizes the key threats across various system layers and stages, along with their corresponding countermeasures. The Threats column highlights common vulnerabilities. For example, data injections can occur in both the Application and Identity Management layers, disrupting operations by manipulating inputs or injecting malicious data. IP spoofing may occur in the Identity Management layer, where attackers falsify their IP address to gain unauthorized access or disrupt the verification process, such as tampering with KYC data or misrepresenting customer identity. In the Consensus layer, “51% attacks” could undermine the trust in the blockchain’s protocol, allowing a single entity to control a majority of mining power and alter transaction history.

Table 1.

Summary of threats, relevant layers & stages, and countermeasures.

Threat	Relevant layers & stages	Countermeasures
Data injection	Layer: Application, identity management Stage 1: Application submission, Stage 3: Data validation	Input validation, filtering, ML-based anomaly detection, robust validation
IP spoofing	Layer: Identity management Stage 2: Data verification	Decentralized identity solutions
Man-in-the-Middle	Layer: Identity management Stage 3: Data validation	TLS encryption
Flooding attacks	Layer: Consensus Stage 4: Insurance contract customization	PoW, rate-limiting, DDoS protection
Sybil attacks	Layer: Application Stage 5: Insurance plan selection	Identity verification, cryptographic proofs
Data tampering	Layer: Ledger & storage Stage 6: Information & contract review	Immutable blockchain ledger
Smart contract vulnerabilities	Layer: Smart contract Stage 7: Smart contract integration Stage 8: Smart contract implementation	Code audits, formal verification
51% attacks	Layer: Smart contract Stage 8: Smart contract implementation	PoW, PoS, sharding
Network threats	Layer: Network Peer-to-peer communication (supports all layers)	Firewalls, anomaly detection, secure protocols

The Countermeasures column outlines strategies to mitigate these risks, including input validation, ML-based anomaly detection, PoW, PoS, and sharding. These solutions are consistently applied across relevant layers and stages, enhancing the overall security and integrity of the system.

This framework highlights the dynamic interplay between the blockchain architecture layers, their functionalities, and the security mechanisms required to protect the health insurance system from potential threats.

Figure 2 provides a graphical representation of the designed blockchain for health insurance. It illustrates the interactions between key components, showing how blockchain and smart contracts automate processes, from application submission to claims processing, ensuring a streamlined and secure flow of information and transactions.

Figure 2.

A blockchain-based health insurance system: key components and process flow.

The following section delves into each stage of the blockchain development process, providing a detailed analysis of the identified threats and corresponding countermeasures.

Development stages of the blockchain-based health insurance system

The designed blockchain model for health insurance is structured in eight stages. The following sections provide a detailed overview of each stage.

Stage 1: Application submission

Customers submit an application requesting health insurance services. At this stage, the system may face some threats and require appropriate defenses.

• Potential Threat: Data injection attacks, where attackers may insert malicious data into the system.

• Recommended Solutions: Implement input validation and filtering techniques to detect and block malicious submissions. ML algorithms can also be employed to identify abnormal patterns in customer requests.²⁴

Stage 2: Data verification

The insurance company collects and verifies KYC information, including basic identity details such as age and gender, along with additional data such as phone number, address, and social security number. This forms a streamlined electronic KYC (eKYC) system, reducing costs and addressing anti-money laundering concerns.²⁵

To ensure accuracy, the insurance company verifies the submitted information by connecting to external databases. For example, the Social Security Administration and Department of Motor Vehicles are used to verify identity details, and the USPS Address Verification System ensures the accuracy of the insured’s address.

With the customer’s consent, the insurance company may access medical records from sources like the Medical Information Bureau to mitigate risks such as adverse selection. This comprehensive verification process enables a thorough and accurate evaluation of applicants.

• Potential Threat: IP spoofing attacks could occur during this stage, compromising the integrity and authenticity of KYC information.

• Recommended Solutions: Employ blockchain’s decentralized identity solutions to verify sources and ensure tamper-proof data. Moreover, security measures like firewalls and anomaly detection systems should be utilized.²⁶

Stage 3: Data validation

The data validation process involves several steps to ensure accurate health status prediction. First, the target and explanatory variables, as shown in Table 2, are identified. Then, the dataset is divided into training and testing sets in an 80/20 split, where the training set is used to build the model, and the testing set evaluates its accuracy. Subsequently, data normalization is applied to standardize variables like binary gender, numerical age, and monetary cost, enabling meaningful comparisons and effective model training. The multi-layer perceptron (MLP) algorithm is then implemented, defining characteristics and iterations, while weight training determines variable influence to enhance prediction accuracy. Finally, the model predicts health status (healthy or unhealthy) with minimal errors, and validation ensures precise classification to guide premium decisions.

• Potential Threat: Data injection attacks during the data preparation phase could compromise the training data. Also, Man-in-the-middle (MITM) attacks may occur when data is transmitted between systems.

• Recommended Solutions: Implement robust data validation techniques and use federated learning models to ensure data integrity. Digital signatures can also be utilized to authenticate data sources.²⁷ Moreover, encrypting all data communications with secure protocols like TLS (Transport Layer Security) can mitigate MITM attacks.²⁸

Table 2.

Main factors used for data validation.

Variable	Definition
Target: Health status	1 if healthy, 0 if unhealthy
Explanatory:
Gender	1 for males, 0 for females
Age	Age of individuals
Visits	Number of doctor/healthcare visits
Cost	Health insurance payout amount
Catastrophes	1 for severe illness, 0 otherwise

Notes. Catastrophes include severe conditions such as organ transplants, strokes, and cancer.

Stage 4: Insurance contract customization

After stage 3, the insurance company creates flexible and fixed-price contracts. The QV mechanism introduces $m$ coverage each with $j \geq 2$ price points, resulting in $m \times j$ coverage-price combinations. Let’s assume $j = 3$ . Then, there are three, for example, dental plans with maximum reimbursements of $1200, $1670, and $2330, priced at $36, $60, and $86 per month, respectively.

Table 3 outlines these plans. Columns (4)–(6) display the flexible-price plans, where $P_{m j}^{c}$ (for $j = 1, 2, 3$ ) is the premium of coverage $m$ under plan $j$ , adjusted for health status ( $c = 1$ for healthy, $c = 0$ for unhealthy), and $X_{m j}$ is the maximum reimbursement. For comparison, column (3) represents the maximum reimbursement ( ${\bar{X}}_{m}$ ) and premium ( ${\bar{P}}_{m}$ ) for coverage $m$ under a fixed-price model available for individuals opting out of the voting process.

• Potential Threat: Flooding attacks could impact this stage, where attackers overload the system to cause delays or crashes.

• Recommended Solutions: Introduce PoW or rate-limiting techniques to mitigate system overloading. In addition, distributed denial-of-service (DDoS) protection services are to be employed.²⁹

Table 3.

A flexible insurance contract.

(1)	(2)	(3)	(4)	(5)	(6)
Insurance coverage	Description	Maximum reimbursement [Premium]
		Fixed-price	Flexible-price
		Fixed-price	Plan 1	Plan 2	Plan 3
$m$	Detailed coverage description.	${\bar{X}}_{m}$ [ ${\bar{P}}_{m}]$	$X_{m 1}$ $[P_{m 1}^{c}]$	$X_{m 2}$ $[P_{m 2}^{c}]$	$X_{m 3}$ $[P_{m 3}^{c}]$

Notes. This table presents the insurance coverage details, including maximum reimbursements and premiums under fixed- and flexible-price models.

Stage 5: Insurance plan selection

At this stage, customers select their preferred coverage and insurance plan. Table 4, for instance, outlines 14 insurance coverages available for purchase—either at flexible prices for those who participate in voting or at a fixed price for others.

Table 4.

A sample of recommended insurance coverage and benefits.

Coverage	Description
Doctor visits & Medication	Covers routine doctor visits, prescription medications, and emergency care.
Diagnostic & Imaging services	Includes radiology, MRI, ultrasound, and select therapeutic procedures.
Outpatient surgeries	Covers minor surgeries, fracture treatments, and biopsies.
Refractive error surgery	Provides vision correction procedures, such as LASIK.
Maternity care	Covers labor and delivery, including natural and cesarean births.
Infertility treatments	Includes treatments like IVF and microinjections.
Dental services	General dental care.
Orthodontics for dependents	Provides orthodontic treatment for children under 20.
Vision care	Covers prescription glasses and contact lenses (every 2 years).
Organ transplants	Includes kidney and liver transplants.
Hospitalization & Surgery	Covers inpatient hospital stays, surgeries, and ICU care.
Durable medical equipment	Provides coverage for medical devices like prosthetics, wheelchairs, and canes.
Genetic & Diagnostic testing	Covers genetic testing and specialized diagnostic screenings.
Therapies & Home equipment	Includes physical therapy, rehabilitation, and home medical equipment rentals.

For a nonparticipant individual $i$ , the final premium is the sum of the fixed prices for the coverages they choose ( $\sum {\bar{P}}_{m}) .$ In contrast, for those who participate in voting, the final premium ( $P_{i}$ ) is determined as follows:

P_{i} = \sum_{m = 1}^{14} \sum_{j = 1}^{3} P_{i m j}^{c}

(8)

Each vote incurs a cost, with the total voting cost ( $C_{i}$ ) calculated as the square of the votes cast ( $v_{i}$ ):

C_{i} = v_{i}^{2}

(9)

The total fees collected from all voters are then redistributed equally, giving each voter a share ( $I$ ) calculated as:

I = \frac{\sum_{i = 1}^{N} C_{i}}{N}

(10)

where

N

is the number of voters. This approach fosters informed decision-making while ensuring fair cost-sharing.

It is important to acknowledge that the plan selection stage of the blockchain is susceptible to certain risks:

• Potential Threat: Sybil attacks, where malicious entities create fake identities to influence voting outcomes, could disrupt the QV process.

• Recommended Solutions: Use blockchain-based identity verification and ensure only authenticated customers participate. Leverage cryptographic proof mechanisms to prevent false identities.²⁷

Stage 6: Customer information and contract review

The insurance company reviews the customers’ eKYC information, medical history (obtained in stage 2), and their selected draft contract (specified in stage 5).

• Potential Threat: Data tampering attacks could alter contract data during this review process.

• Recommended Solutions: Use the blockchain’s immutable ledger to log contract data and track any unauthorized changes.³⁰

Stage 7: Smart contract integration

The insurance company defines smart contract policies and encodes customer data obtained in stage 6.

• Potential Threat: Smart contract vulnerabilities, such as bugs in coding, can lead to exploitation.

• Recommended Solutions: Conduct rigorous audits of smart contract code using automated tools and manual reviews. Adopt formal verification methods to ensure code correctness.²⁹

Stage 8: Smart contract implementation

Smart contracts containing finalized data from stages 6 and 7 are reviewed, approved, and deployed on the Ethereum platform. Ethereum is offered for its role as a leading platform for smart contracts, offering a customizable blockchain architecture that supports decentralized applications. Known for its extensive developer support, well-established ecosystem, and robust security infrastructure, Ethereum provides the necessary functionality to automate health insurance processes. These smart contracts enable secure, seamless, and automated interactions between customers and service providers, enhancing operational efficiency and fostering trust.¹

• Potential Threat: A significant vulnerability is the risk of “51% attacks,” where attackers gain control over the network’s computational power.

• Recommended Solutions: Transitioning to PoS consensus mechanisms can significantly reduce the risk of “51% attacks.” Sharding techniques further enhance security and scalability.²⁴

Evaluation plan

To assess the feasibility and effectiveness of the blockchain-based health insurance model, we propose a multi-stage evaluation strategy. This approach includes a simulation-based evaluation, a real-world feasibility study, and a comparative analysis.

Simulation-based evaluation

The performance of the blockchain system can be evaluated through simulations using synthetic and real-world insurance datasets, where smart contracts could be deployed on Ethereum test networks (e.g., Rinkeby or Goerli) to assess transaction efficiency and scalability. Simulated participants with predefined risk profiles and claim histories will help analyze system behavior under various scenarios.

Moreover, the ML models like XGBoost (Extreme Gradient Boosting)³¹ can be integrated to improve fraud detection and cost efficiency (e.g., see Srivastava et al.).³² Recently, a novel ensemble technique combining Random Forest (RF), XGBoost, and Support Vector Machine (SVM) was introduced to further improve fraud detection accuracy in the Ethereum blockchain, achieving a score of over 98% across key classification metrics, including accuracy, precision, recall, and F1-score.³³

Table 5 summarizes the key performance metrics to be evaluated in the simulation.

Table 5.

Key performance metrics of the health insurance blockchain model.

Metric	Description	Evaluation method
Transaction speed	Time required to process claims via blockchain	Compare transaction speed under the blockchain model versus the traditional method.
Cost reduction	Savings in administrative costs and fraud prevention	Conduct a cost-benefit analysis using claims data for a control group.
Quadratic voting impact	Effect of quadratic voting on policy customization	Measure fairness and diversity of selected coverages.
Fraud detection rate	Percentage of fraudulent claims identified	Apply ML models to detect anomalies in claims.
Scalability	Performance under increasing transaction volume	Stress testing with varying numbers of simulated policyholders.

Notes. The control group includes customers under the traditional health insurance system.

Real-world feasibility study design

A pilot study could test the model in a real-world setting using data from a mid- or large-sized company. Smart contracts would automate policy selection and claims processing, while quadratic voting would allow employees to prioritize coverage options.

ML models can be used for fraud detection, customer validation, and premium adjustments (i.e., step 3 of the blockchain model). For example, Shouri used health insurance claims data from 29,785 employees of the East Oil Company of Iran, analyzing 148,925 records in 2023.³⁴ The study used the XGBoost technique to classify individuals as healthy or unhealthy and found that 1.78% of policyholders were classified as unhealthy but accounted for 17.48% of total claims. The model achieved an accuracy of 99.98%, highlighting its effectiveness in identifying high-risk individuals and enabling risk-based premium adjustments.

The pilot study design consists of four main steps: Planning, Data Collection, Evaluation and Analysis, and Finalization and Reporting. Figure 3 represents a hierarchical flowchart of the study design, following a logical grouping of steps and their subsets.

Figure 3.

Study design for evaluating the efficiency of the blockchain model for health insurance. Solid arrowed lines represent the logical flow between study phases and their interactions with key tasks. Dashed lines indicate study elements, including data points, processes, and actions.

Below, we provide an overview of the main phases of the study design.

Planning

To ensure a robust study, the pilot should begin with well-defined objectives, such as evaluating operational feasibility, user adoption, efficiency improvements, cost-effectiveness, and regulatory compliance. Employees from diverse departments, age groups, and health conditions should be included to ensure broad representation. A clear participant recruitment plan must be established, selection criteria must be defined, enrollment procedures must be established, and informed consent must be obtained.

Data collection

Data collection will involve both quantitative and qualitative methods. Surveys and questionnaires will measure key metrics such as adoption rates, user satisfaction, trust in the system, and efficiency gains, while in-depth interviews and focus groups will provide richer insights into user perceptions, challenges, and preferences. Both baseline and follow-up measurements will track the impact of blockchain adoption over time. Additionally, a control group using traditional insurance methods will allow direct performance comparisons by analyzing differences in claim processing time, administrative costs, and fraud detection rates.

For example, data can be collected electronically via questionnaires to analyze the effectiveness of the QV model. Similarly, surveys can assess whether policyholders feel that quadratic voting results in fair and customized coverage options. Additionally, historical data could be used to validate health insurance customers. Finally, ML-based risk assessment models can also be integrated to analyze premium calculations and enhance fraud detection by identifying patterns in fraudulent claims.

Evaluation and analysis

The evaluation will also incorporate expanded metrics such as adoption rates, efficiency improvements, user experience scores, cost savings, and regulatory compliance. To refine the system before full-scale deployment, feedback should be collected at multiple stages. A risk management plan will identify potential challenges, including technical issues, data privacy concerns, user resistance, and mitigation strategies.

Table 6 represents the key evaluation factors and the corresponding methods for data collection. This table outlines how the real-world feasibility study will measure adoption rate, efficiency gains, user experience, cost savings, regulatory compliance,³⁵ and fraud detection effectiveness.

Table 6.

Evaluation factors for assessing the blockchain-based health insurance model.

Factor	Description	Data collection method
Adoption rate	Percentage of employees opting for blockchain-based plans	Surveys and system usage data
Efficiency gains	Reduction in claims processing time	Time tracking before and after blockchain adoption
User experience	Ease of use, trust in the system	User satisfaction surveys and feedback
Cost savings	Reduction in claims, fraud, and administrative costs	Financial analysis comparing traditional versus blockchain systems
Regulatory compliance	Alignment with health data regulations (e.g., HIPAA and GDRP)	Legal review and compliance assessment
Fraud detection effectiveness	Accuracy of fraud identification via ML models	Comparison of flagged fraudulent claims versus actual cases

Finalization, reporting, and documentation

Comprehensive reporting and documentation will be essential, covering study design, data collection instruments, analysis methods, and ethical considerations such as participant confidentiality and compliance with healthcare regulations. Additionally, scalability and integration testing will assess the system’s ability to handle increased users and its compatibility with existing health insurance providers and hospital systems.

Comparative analysis

Finally, a comparative analysis could be conducted to compare the blockchain-based model with traditional health insurance systems, focusing on key aspects such as transparency, claim processing speed, fraud detection, user control, cost efficiency, and risk-based premiums (as in Table 7). Overall, this comparison illustrates the improvements in all these aspects when adopting blockchain technology. However, to formally evaluate the improvement of some aspects (such as claim processing speed, cost efficiency, and premium), one must first implement the blockchain model and collect data under both models. For example, an empirical study like Bae and Yi can be conducted to validate the findings.¹² User surveys and expert reviews will further assess trust, satisfaction, and perceived benefits of the blockchain model.

Table 7.

Comparative analysis of blockchain-based and traditional health insurance models.

Aspect	Traditional insurance	Blockchain-based model
Transparency	Opaque claims processing	Fully auditable on blockchain ledger
Claims processing speed	Weeks (manual verification)	Near-instant (smart contracts)
Fraud detection	Manual, prone to errors	Automated ML-based fraud detection
User control	Predefined policies	Customization through quadratic voting
Cost efficiency	High due to intermediaries	Lower due to automation and decentralization
Risk-based premiums	Standardized for all users	Adjusted based on ML-validated risk profiles

Discussion

Blockchain technology has transformative potential in health insurance, where traditional fixed-coverage contracts limit customer choices. By incorporating QV techniques, this study introduces a democratic approach that offers flexible contracts with customizable coverage and pricing options.

An eight-stage blockchain system was developed to enhance the health insurance system. The MLP neural network validates customer information in stage 3, while stages 4 and 5 implement the QV model. These stages empower customers to vote on preferred coverage-price combinations, optimizing medical reimbursement and introducing a dynamic pricing model. Additionally, voting is crucial in identifying optimal contracts and enhancing reimbursements. Finally, blockchain strengthens fraud prevention through KYC validation, eliminates intermediaries, and enables instant settlement of treatment costs, enhancing trust and efficiency.

To assess the effectiveness of the designed blockchain, we offered a comprehensive evaluation plan, including simulation-based assessments for transaction efficiency and scalability, a real-world pilot study to test user adoption, and a comparative analysis with traditional models to measure improvements in fraud detection, cost efficiency, and user control.

Strengths of the proposed model

The blockchain-based health insurance model presents several advantages over traditional systems. Blockchain technology enhances transparency and trust, as all transactions are recorded on an immutable ledger, ensuring auditability and fraud resistance. The integration of QV empowers customers to actively participate in the plan selection process, creating a more democratic and inclusive system. This model also strengthens fraud prevention by utilizing KYC processes and ML models to detect and mitigate fraudulent activities. Additionally, smart contracts automate administrative tasks, eliminating intermediaries, reducing operational costs, and ultimately improving efficiency.

Comparison with recent studies

The study’s proposed blockchain-based health insurance model advances beyond recent studies by seamlessly integrating multiple innovative features into a unified and comprehensive framework. Prior research primarily focuses on automating claims, enhancing transparency, or addressing fraud.^12,13,26 However, our model combines these aspects with a user-centric approach. Specifically, it incorporates ML models with decentralized KYC validation for robust fraud prevention and introduces QV to enable customer-driven customization of insurance plans. This holistic integration addresses system inefficiencies and fosters inclusivity, delivering broader applicability and functionality compared to earlier models.

Limitations of the proposed model

Despite its advantages, the model faces several challenges to implementation. Integrating blockchain into healthcare systems requires technical expertise, resources, and organizational changes, which could slow adoption. High implementation costs may be a barrier for smaller entities, and regional regulatory compliance introduces complexity. While blockchain provides a secure framework, continuous monitoring is necessary to protect data privacy, particularly during the KYC process. The lack of real-world testing further underscores the need for pilot programs to assess its practical application. Lastly, the implementation of QV and blockchain may not be readily generalizable, necessitating further research to evaluate its broader applicability.

Recommendations for future research

Future research should focus on evaluating the framework using real-world data. Conducting pilot programs across diverse healthcare systems and expanding the model to national, public, and private settings will provide insights into its performance and help assess its broader applicability. Integrating the framework with existing insurance platforms is crucial to address interoperability challenges. Additionally, exploring cryptographic methods such as zero-knowledge proofs will be key to ensuring data protection and regulatory compliance.

Conclusion

This study demonstrates the transformative potential of blockchain technology in health insurance by addressing long-standing inefficiencies and limitations of traditional systems. The proposed model incorporates an eight-stage blockchain framework and integrates QV techniques, enabling customers to access customizable contracts and dynamic pricing. This design fosters a more inclusive and democratic insurance process.

Additionally, integrating ML techniques for fraud prevention and decentralized KYC validation strengthens trust and improves operational efficiency. Smart contracts reduce administrative overhead and expedite reimbursements, marking a significant advancement over traditional models.

As blockchain technology evolves, this framework underscores the value of combining technological innovation with economic principles to create scalable, equitable, and efficient systems. Although challenges in adoption and implementation remain, this work provides a foundation for future interdisciplinary research and real-world applications, with the potential to modernize health insurance and inspire similar transformations across various industries.

Footnotes

Acknowledgments

The authors extend their gratitude to the two anonymous reviewers for their insightful comments and constructive feedback, which significantly enhanced the quality of this paper. We are also grateful to the editors for their guidance and support throughout the review process. Special thanks to Glen Cooper for his assistance in proofreading and improving the clarity of the manuscript.

ORCID iD

Rasoul Ramezani

Ethical statement

Author contributions

Saeed Shouri led the development of the blockchain framework and authored the discussion and conclusion sections. Rasoul Ramezani contributed to the development of the introduction and methodology sections and played a key role in designing the blockchain architecture. He also prepared both the original and revised versions of the manuscript. Both authors collaborated on designing the evaluation plan and thoroughly reviewed and approved the final manuscript.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Note

References

Galea

. Healthier: fifty thoughts on the foundations of population health. 1 ed. Oxford University Press, 2017.

Chen

C-L

Deng

Y-Y

Tsaur

W-J

, et al. A traceable online insurance claims system based on blockchain and smart contract technology. Sustainability. 2021; 13: 9386. https://doi.org/10.3390/su13169386

Tanwar

. Blockchain technology. Springer, 2022.

Amponsah

Adekoya

Weyori

. Blockchain in insurance: exploratory analysis of prospects and threats. Int J Adv Comput Sci Appl. 2021; 12: 445–466. https://doi.org/10.14569/IJACSA.2021.0120153

Efanov

Roschin

. The all-pervasiveness of the blockchain technology. Procedia Comput Sci. 2018; 123: 116–121. https://doi.org/10.1016/j.procs.2018.01.019

Tripathi

Ahad

Casalino

. A comprehensive review of blockchain technology: underlying principles and historical background with future challenges. Decision Analytics Journal. 2023; 9: 100344. https://doi.org/10.1016/j.dajour.2023.100344

Ullah

Al-Rahmi

Alblehai

, et al. Blockchain-powered grids: paving the way for a sustainable and efficient future. Heliyon. 2024; 10: e31592. https://doi.org/10.1016/j.heliyon.2024.e31592

Ghadi

Shah

SFA

Mazhar

, et al. Enhancing patient healthcare with mobile edge computing and 5G: challenges and solutions for secure online health tools. J Cloud Comput. 2024; 13: 93. https://doi.org/10.1186/s13677-024-00654-4

Corte-Real

Nunes

Da Cunha

. Reflections about blockchain in health data sharing: navigating a disruptive technology. Int J Environ Res Public Health. 2024; 21: 230. https://doi.org/10.3390/ijerph21020230

10.

Nunes

Da Cunha

de Abreu

, et al. Non-fungible tokens (NFTs) in healthcare: a systematic review. Int J Environ Res Public Health. 2024; 21: 965. https://doi.org/10.3390/ijerph21080965

11.

Agbo

Mahmoud

Eklund

. Blockchain technology in healthcare: a systematic review. Healthcare. 2019; 7: 56. https://doi.org/10.3390/healthcare7020056

12.

Bae

B-K

. Development of eClaim system for private indemnity health insurance in South Korea: compatibility and interoperability. Health Informatics J. 2022; 28. https://doi.org/10.1177/14604582211071019

13.

Ismail

Zeadally

. Healthcare insurance frauds: taxonomy and blockchain-based detection framework (Block-HI). IT Prof. 2021; 23: 36–43. https://doi.org/10.1109/MITP.2021.3071534

14.

Khan

Mazhar

Mateen Yaqoob

, et al. Optimal feature selection for heart disease prediction using modified Artificial Bee colony (M-ABC) and K-nearest neighbors (KNN). Sci Rep. 2024; 14: 26241. https://doi.org/10.1038/s41598-024-78021-1

15.

Faisal Abbas Shah

Mazhar

Shloul

, et al. Applications, challenges, and solutions of unmanned aerial vehicles in smart city using blockchain. PeerJ Comput Sci. 2024; 10: e1776. https://doi.org/10.7717/peerj-cs.1776

16.

Lalley

Weyl

. Quadratic voting: how mechanism design can radicalize democracy. In: AEA Papers and Proceedings. American Economic Association, 2018, pp. 33–37.

17.

Posner

Weyl

. Quadratic voting and the public good: introduction. Public Choice. 2017; 172: 1–22. https://doi.org/10.1007/s11127-017-0404-5

18.

Casella

. Storable votes. Game Econ Behav. 2005; 51: 391–419. https://doi.org/10.1016/j.geb.2004.09.009

19.

Weyl

. The robustness of quadratic voting. Public Choice. 2017; 172: 75–107. https://doi.org/10.1007/s11127-017-0405-4

20.

Lalley

Weyl

. Nash equilbria for quadratic voting. arXiv. 2014; 1: arXiv:1409.0264. https://doi.org/10.48550/arXiv.1409.0264

21.

Schad

Thronicke

. Real-world evidence-current developments and perspectives. Int J Environ Res Public Health. 2022; 19: 10159. https://doi.org/10.3390/ijerph191610159

22.

Lima Ramos

. Premium calculation in insurance activity. J Stat Manag Syst. 2017; 20: 39–65. https://doi.org/10.1080/09720510.2016.1187927

23.

Boucher

J-P

Denuit

Guillén

. Models of insurance claim counts with time dependence based on generalization of Poisson and negative binomial distributions. Variance 2008; 2: 135–162.

24.

Mazhar

Irfan

Khan

, et al. Analysis of cyber security attacks and its solutions for the smart grid using machine learning and blockchain methods. Future Internet. 2023; 15: 83. https://doi.org/10.3390/fi15020083

25.

Arner

Zetzsche

Buckley

, et al. The identity challenge in finance: from analogue identity to digitized identification to digital KYC utilities. Eur Bus Org Law Rev. 2019; 20: 55–80. https://doi.org/10.1007/s40804-019-00135-1

26.

Mohammed

Boujelben

Abid

. A novel approach for fraud detection in blockchain-based healthcare networks using machine learning. Future Internet. 2023; 15: 250. https://doi.org/10.3390/fi15080250

27.

Al Amin

Shah

Tummala

, et al. Utilizing blockchain and smart contracts for enhanced fraud prevention and minimization in health insurance through multi-signature claim processing. In: 2024 international conference on Emerging Trends in Networks and Computer Communications (ETNCC), 23-25 July 2024, Windhoek, Namibia, pp. 1–9. IEEE. https://doi.org/10.1109/ETNCC63262.2024.10767491

28.

Riadi

Umar

Busthomi

, et al. Block-hash of blockchain framework against man-in-the-middle attacks. Regist J Ilm Teknol Sist Inf. 2022; 8: 1–9. https://doi.org/10.26594/register.v8i1.2190

29.

Anwar

Tariq

Ashraf

, et al. A blockchain-based attack detection mechanism in healthcare cyber physical systems against jamming attacks. In: Ullah

Anwar

Calandra

, et al. (eds). Proceedings of International Conference on Information Technology and Applications. Springer Nature, 2022, pp. 147–157.

30.

Romero

Gerstoft

. Spoofing attack detection in the physical layer with robustness to user movement. In: 2024 IEEE Wireless Communications and Networking Conference (WCNC), 21–24 April 2024, Dubai, United Arab Emirates, pp. 1–6. IEEE. https://doi.org/10.1109/WCNC57260.2024.10570909

31.

Chen

Guestrin

. Xgboost: a scalable tree boosting system. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2016, pp. 785–794.

32.

Srivastava

Singh

Dwivedi

. Blockchain based prophecy of cardiovascular disease using modified XGBoost. Artificial Intelligence, Blockchain, Computing and Security 2023; 1: 250–258, CRC Press.

33.

Dib

. Enhancing fraud detection in the Ethereum blockchain using ensemble learning. PeerJ Comput Sci. 2025; 11: e2716. https://doi.org/10.7717/peerj-cs.2716

34.

Shouri

. Validation of health insurance customers using XGBoost modeling. Journal of International Crisis and Risk Communication Research 2024; 7: 1016–1026.

35.

Hasselgren

Wan

Horn

, et al. GDPR compliance for blockchain applications in healthcare. arXiv 2020. arXiv preprint arXiv:200912913. https://doi.org/10.48550/arXiv.2009.12913

A blockchain-based health insurance model enhanced with quadratic voting

Abstract

Keywords

Introduction

Literature review

Blockchain in healthcare

Health data security and privacy

Blockchain for health insurance

Contribution to literature

Methodology

Quadratic voting

Private value and voting cost

Optimal voting pricing rule

Validation of health insurance claims

Blockchain framework for secure and transparent health insurance

Blockchain architecture

Development stages of the blockchain-based health insurance system

Stage 1: Application submission

Stage 2: Data verification

Stage 3: Data validation

Stage 4: Insurance contract customization

Stage 5: Insurance plan selection

Stage 6: Customer information and contract review

Stage 7: Smart contract integration

Stage 8: Smart contract implementation

Evaluation plan

Simulation-based evaluation

Real-world feasibility study design

Planning

Data collection

Evaluation and analysis

Finalization, reporting, and documentation

Comparative analysis

Discussion

Strengths of the proposed model

Comparison with recent studies

Limitations of the proposed model

Recommendations for future research

Conclusion

Footnotes

Acknowledgments

ORCID iD

Ethical statement

Author contributions

Funding

Declaration of conflicting interests

Note

References