Data Ethics and Digital Privacy in Learning Health Systems for Palliative Medicine: Volume 23

This chapter introduces Learning Health Systems (LHS) and the impact of data science on such systems. It also examines the necessary properties of data used in LHS and identifies patients who may benefit from a transition to palliative care. Finally, it examines the way LHS can be used to identify racial and social disparities in access to palliative care.

A great concern regarding the use of data science in any field is privacy. Adequately protecting individuals from the negative effects of maliciously shared personal identifying information is essential. It is however, also important to understand the positive role that protected and shared information can play. This chapter provides a basic understanding of how the concept of privacy has developed in the United States (US) and suggests that continued development of that understanding and the protections provided will occur.

This chapter provides an overview of the role of technology and policy in shaping care plans for patients. Historically, healthcare has lagged behind other industry sectors in adopting and deploying useful technologies, and policy surrounding use is an important component of establishing a long-term strategy. This chapter evaluates the current state of technology in the clinical setting and extends the widely adopted policy-based approaches into the palliative care context.

This chapter will identify readily accessible existing sources of public data. Thechallenges of using that data are considerable and require extensive time to ensure validity for reporting purposes. Summaries of data field selection and data wrangling requirements are presented in conjunction with data aggregation strategies.

Characteristics that impact the levels of palliative care are introduced. Patients with the potential to be classified as palliative may be overlooked or simply so not seek medical attention. The population is much higher than those being treated on an annual basis. Data from the American Community Survey (ACS) and the Behavioral Risk Factor Surveillance System (BRFSS) are applied to the characteristics of palliative care and used to estimate the size of the palliative population in the United States (US).

This chapter more clearly identifies the distinction between Electronic Health Record (EHR) and Electronic Medical Record (EMR), and states their value in obtaining individual-level data. Synthetic medical records may be used as a surrogate for EHR data in order to ensure digital data privacy is maintained during the development of the LHS. Synthea is an open-source simulation tool available through GitHub.¹ Extensive descriptive analysis of synthesized data is provided as a foundation for the analysis in Chapter 7.

Synthetic patient data produced by Synthea was described in Chapter 6. That data is used to create a baseline for all patients, palliative patients, and deceased palliative patients. Distributions of comorbidities across the patient groups are examined and demographic characteristics. The factors used in palliative care groupings are presented with the synthesized data fields used. The size of the palliative population is again estimated to establish validity.

Data mapping from synthesized data to palliative care characteristics was the final step before the final analysis of survival. Background and foundation for Kaplan-Meier curves are provided before generating curves for the three Palliative Care Groups. Interpretations of the Kaplan-Meier curves are presented along with interpretation of the associated Hazard Curves. Three statistical hypothesis tests, completed on a pairwise basis, are used to verify that the survival curves differ by group. Patients mapped to specific groups may be further supported through advice, counseling, and other services to assist them in moving to a more advantageous care group.

This chapter discusses the steps taken to access and use the ACS five-year data. The format of the data is discussed pointing out the fact that there is no requirement that an ACS five-year variable holds data for the same field year to year. The development of a cross-reference table is discussed allowing the data to be accessed by a common label.

This chapter discusses the design of the LHS and the steps taken to ensure data privacy and security. Usage of the application programming interface (API) is discussed, paying attention to how an Electronic Health Record (EHR) provider would use the API. Finally, the clinician’s interaction with the system is discussed.

Having provided a feasible framework for the use of big data and a learning health system (LHS) in addressing disparities in access to palliative care, this chapter seeks to substantiate the ethical underpinning of that framework, drawing from well-regarded existing sources. The author will also address issues which will likely arise from a successful transition to LHSs such as the nature of informed consent, the impact it will have on medical decision-making in general, and the transformative effect big data and implementation of LHSs will have on existing data sources.