Data ethics in action: transforming Swiss healthcare with policyholder insights
Gaël Saillen, Nicola Diviani, Sara Rubinelli
Healthcare costs and data-driven solutions
The Swiss parliament is deliberating on the second phase of the Federal Council’s strategy to manage escalating healthcare expenses. One aspect of this involves a proposal, supported by the National Council’s Health Committee, which aims to allow health insurers to use policyholder billing data. While this initiative seeks to uncover new pathways for enhancing care and allowing policyholders to make informed choices, it raises several ethical concerns.
At the core of this debate is the challenge of balancing the potential benefits of data-driven advancements in healthcare with the need to maintain strict ethical standards. The discussion emphasises the crucial role of transparent communication for health insurers. Using patient data to improve healthcare while strictly adhering to privacy norms and ethical principles is a complex task.
In this comprehensive analysis, we explore the ethical complexities associated with health insurers accessing patient data. We examine the moral dilemmas, assess the implications, and consider various available strategies. We propose alternative approaches to address these ethical challenges, aiming to strike a balance between technological innovation and ethical responsibility in healthcare.
Costs, data, and policyholder expectations
Rising healthcare costs are a reality in many countries, including Switzerland. The total cost of public healthcare is nearing 12% of the national GDP. It is crucial to address this issue while maintaining high-quality care. According to an Accenture study, most (96%) Swiss residents feel responsible for their health. While the interpretation of this percentage could be debated, it implies that policyholders expect to be well-informed regarding their health. Through their access to extensive data, insurers could play a role in this transmission of information. However, they are currently prohibited from using this data to make recommendations to policyholders. A circular issued by the Federal Office of Public Health, written in collaboration with the Data Protection Commissioner in 2022, explicitly forbids this practice, stating a “targeted recommendation to policyholders selected on the basis of a specific state of health or illness, […] as it has no legal basis, […] is unlawful and must be prohibited” (own translation).
The law would need to be amended or clarified to enable insurers to use the data they hold and communicate the results to policyholders. This is the objective of the proposal currently under consideration by the National Council (art. 56, P-LAMal): “In order to guarantee the measures provided for in para. 1, insurers may inform insured persons individually of the benefits they receive, measures to prevent illness and offers to improve the economy and coordination of the medical services provided; with their consent, insurers may also make information on the benefits received by insured persons available to service providers” (own translation).
Another aspect that requires closer scrutiny is the concept of risk compensation. Embedded in the legal framework by the Swiss Parliament in 2007, risk compensation was established with the intention of mitigating risk selection and promoting equitable competition among health insurers. This legislative action also granted the Federal Council the authority to refine this mechanism. In 2020, 13 years after its enactment, a pivotal advancement in risk compensation significantly shifted the dynamics within the health insurance sector. This reform repositioned health insurers, incentivising them to optimize service management. This shift aimed to enhance the quality of information provided to policyholders about their healthcare journey; however, it carried significant economic implications. Improved risk compensation strategies have a dual impact: they limit risk selection by insurers and increase the potential for savings, benefitting policyholders.
Using insurers’ data, it is possible to identify gaps in policyholders’ care pathways. A study conducted by a Swiss health insurance company reported several shortcomings in public health. It highlighted that over the last 10 years, only 50–60% of the Swiss population aged between 50–75 had undergone preventive medical examinations for bowel cancer detection, whereas the recommendation would be for 80% to have undergone this.
In the field of compulsory health insurance, as with other social insurance, there exists the principle of an “obligation to collaborate”. This duty to cooperate accompanies the principle of solidarity of compulsory basic insurance in Switzerland. Cooperation means that the insured must provide information and documents or undergo examinations relevant to the assessment of benefits [10]. While they are obliged to cooperate under numerous provisions of the Federal Act on the General Part of Social Insurance Law (LPGA), this obligation is not unlimited. This raises a question about whether the principle of solidarity, which implies a duty to cooperate, also includes a duty to be informed.
Ethical dilemmas
Leveraging health insurance data for health promotion and disease prevention can result in substantial public health advantages. It can lead to improvements in healthcare quality while simultaneously reducing the associated costs. However, there are ethical considerations to take into account. A significant moral conflict associated with the use of patient data by health insurers involves privacy and data security, given the personal and sensitive nature of this data. Health insurers must ensure this data is adequately protected against unauthorised access, breaches, and misuse.
Furthermore, the use of data analysis and algorithmic applications across various sectors has the potential to exacerbate existing inequalities. This necessitates heightened responsibility, particularly when using billing data to inform recommendations for insurers. In this context, insurers bear the responsibility to ensure that the algorithms they use are technically proficient and equitable. The ethical imperative is clear: insurers must vigilantly maintain fairness in their algorithmic decision-making processes, ensuring that technological advancements contribute positively to healthcare equity.
While policyholders provide data to insurers for the reimbursement or payment of medical bills, they may not fully understand how their data is used, by whom, and with what potential consequences. This creates a problematic challenge of information asymmetry.
Finally, certain rights and duties of the policyholder may conflict. The right to collaborate should be considered in relation to the right to be informed and the right not to be informed. In Switzerland, patients can waive their right to information without having to forego treatment, and we can assume that the same applies to policyholders. A question then arises around to whom the right to know applies. While it applies primarily to the policyholder, depending on the situation, one could ask whether it extends to the spouse or parents. The right not to be informed presents another ethical dilemma regarding a potential conflict with the obligation to cooperate in social insurance. The alternatives outlined below in the potential application of the new article 56 P-LAMal provide possible answers.
Ethical strategies for data utilisation
Addressing the ethical complexities outlined requires exploring various strategic scenarios. The first scenario, allowing insurers unrestricted access to data, falls short of addressing the ethical complexities involved. This approach overlooks the nuanced considerations necessary for responsible data use. The second, adhering to the current regulatory landscape, risks foregoing the potential public health benefits that controlled data usage could offer.
There are viable alternatives beyond the binary choice of unrestricted access or maintaining the status quo. Central to these is the development of a robust, well-defined regulatory framework. This framework would tailor the permissions granted to insurers, ensuring a balanced approach that respects ethical boundaries while unlocking the potential benefits of data utilisation in healthcare. If the parliament decides to create this framework, it may answer several of the moral issues raised.
Regarding data security, insurers should demonstrate the capability to secure the data and reduce the risk of bias. As insurers are already subject to the Federal Law of Data Protection (FADP), the article P-LAMal 56 should not alter this. However, the issue of potential bias remains. To address this, insurers’ communication must be limited to recommendations, which could be accompanied by advice, such as to consult health professionals.
This authorisation granted to insurers, in connection with their shift from being a payer to a player, invokes the concept of trust. Acting as a player by communicating health information to policyholders adds a new dimension to the role of trust in policyholder-insurer relationships. The importance of trust in the patient-doctor relationships is well recognised and may be gaining significance in policyholder-insurer relationships.
Reducing information asymmetry is crucial to consider when addressing trust-related issues within the context of health insurance. The relationship between trust and information asymmetry is inversely proportional: higher levels of trust typically correlate with lower degrees of uncertainty stemming from information gaps. To effectively reduce this asymmetry, insurers must adopt a transparent approach in terms of their data processing and recommendation derivation methods. This requires clear communication and the development and pre-implementation testing of ethically designed algorithms.
Insurers could extend their impact beyond direct healthcare services. Investing in community health initiatives, particularly those addressing social determinants of health and supporting underserved populations, could result in substantial public health benefits. Such engagements contribute to the overall well-being of communities and solidify the insurers’ role as proactive participants in the healthcare ecosystem. Through these multifaceted actions, health insurers can reinforce their commitment to ethical practices and, consequently, strengthen trust within the healthcare system.
To ensure the right not to know, an opt-in or opt-out procedure that allows policyholders to state their preference, could be recommended. However, significantly fewer policyholders are expected to select the opt-in option, diminishing the measure’s value. To increase participation, the opt-in option could be accompanied by incentives, such as a discount on insurance premiums, which should reflect a real economic advantage. However, this approach potentially raises two problems. Firstly, using discounts could be seen as a form of data “monetisation”, which could worsen social inequalities by introducing an economic incentive that policyholders, particularly those in challenging economic circumstances, might find difficult to refuse. Secondly, this approach overlooks a key aspect of social insurance dynamics, which is the obligation of the insured to cooperate within the established framework. As Davies highlighted, there are no absolutes regarding the “obligation to know”. For these reasons, an opt-out option should be preferred and is probably justified in terms of public health as it enhances quality, potentially reduces cost, and guarantees the rights of the policyholder.
Balancing patient rights with healthcare innovation
The ethical use of patient data by health insurers to inform patients about potential care optimisation is a complex issue requiring careful consideration. While data-driven insights could improve healthcare outcomes, they also raise significant ethical concerns related to privacy, consent, bias, and fairness. To navigate this ethical landscape, legislators and health insurers have several alternatives, such as including an opt-out option, ensuring transparent and ethical algorithm development, and addressing disparities. Effective communication is pivotal in this process as it builds trust and understanding among all stakeholders. Clear, consistent, and empathetic communication strategies are crucial for maintaining insurers’ reputations and ensuring that policyholders feel informed and respected. For these reasons, the opt-out option, accompanied by transparent communication, seems the most beneficial for policyholders, insurers, and public health considerations, without harming the rights of policyholders.
A direction taken on November 8, 2023, by the Council of States’ Committee on Social Security and Public Health proposes a framework “stipulating that insurers may only use this data to provide insured persons with targeted information on the following three elements: more advantageous benefits such as generics, specific forms of appropriate insurance, such as networks of regional general practitioners, or preventive measures for chronic illnesses”. This framework supports the opt-out procedure and highlights the need for ongoing dialogue and feedback mechanisms to adapt to the evolving needs and concerns of policyholders. By prioritizing open and honest communication, health insurers can reinforce their commitment to ethical practices, thereby safeguarding their reputation and fostering a more transparent and accountable healthcare environment.
Potential conflicts of interest
The manuscript is part of a doctoral project conducted without external funding. No funding has been received for the manuscript. However, one of the authors, Gaël Saillen, has a professional activity within Helsana Versicherungen AG
Reference
1 Miksch A, Andres E, Stammann C, Szecsenyi J. QISA Band E2: Krebsfrüherkennung. [Qualitätsindikaotren zu Früherkennung von Krebserkrankungen].
Gaël Saillen, Faculty of Health Sciences and Medicine, University of Lucerne, Lucerne, Switzerland / Helsana Insurance Group, Zurich, Switzerland
Nicola Diviani, Faculty of Health Sciences and Medicine, University of Lucerne, Lucerne, Switzerland / Swiss Paraplegic Research, Nottwil, Switzerland
Sara Rubinelli, Faculty of Health Sciences and Medicine, University of Lucerne, Lucerne, Switzerland / Swiss Paraplegic Research, Nottwil, Switzerland