Value
Constructing generic health value measures may not be possible,25 but assessing the value of health is more appropriate and having a framework like the triple aim provides a simple and comprehensive approach. With the complexity of measuring consequences of disease and the total value of health over various contexts, starting with pre-set measures to support decision-making is impossible. 25 This is complicated by substantial variability in reported healthcare data due to factors such as clinical inertia, patient expectations, and financial capacity, which differ greatly among different cultures and countries. It is therefore suggested to introduce value early in the decision-making process and work upwards from available data to develop decision options. Consequently, determining the best value relies on data gathering and adapting to local factors.
If data on effectiveness, harm, cost, and patient experience is deficient, then it may be challenging to clearly state the optimal decision. Hendrikx and colleagues (8) conducted an international comparative analysis to assess which triple aim measures are being used to evaluate population management (PM) initiatives. Of the 865 measures used by 20 PM initiatives, only 11 PM initiatives included all qualities of care domains. However, each triple aim domain has challenges for optimum judgment to be made at any of the meta-decision steps. First, improving the health of populations through healthcare has been measured by a limited number of studies. A well-known measure is the Center for Disease Control’s Summary Measures of Population Health,26 which combine information on mortality and non-fatal health outcomes to represent population health in a single number. Another example is the quality-adjusted life year (QALY) and the disability-adjusted life years, which were developed as measurement units to quantify the burden of disease and injury on human populations.27 Challenges with such measures are that they are as accurate as the data sources and context from which they were derived. The variation of healthcare systems, payment structures, and patients’ determinants of health is what makes data deficient.28 Therefore, it may not be suitable for generalizability, and may not have the statistical accuracy required to confidently estimate the desired outcomes. Additionally, the available metrics are never comprehensive enough to assist in all decision areas in any context’s unique details. Further, a major barrier for QALY is that it is assigns a weight between 0 (for death) and 1 (100% health) to each health state and then multiplies that value by how long the state lasts. 29 This method provides a crude idea for policy decision-making, but it is difficult to apply at the patient level.
Regarding cost, it is changing, unsustainable, and not unified for each encounter, patient, or setting. For example, the financial consequences after a myocardial infarction in an adult male varies with different influences on a patient’s life, family, work, and medical resources utilized. Less attention is paid to eliminating wasteful spending such as missed prevention, unnecessary service, inefficiently delivered care, high-priced services, excess admirative costs, and fraud—applicable to individual and population levels.
An even more challenging area in the value of health is patient/population values assessment. There are conflicting reports on the relationship between positive patient experience and patient outcomes, 30 which is beyond the scope of this review to investigate. Nevertheless, valuing health is only complete with patient perspectives, and their judgment of value is best when all relevant information is available, free of rational flaws like self-interest. Assessment of such judgment was suggested by using personal experience over personal preference due to preferences affecting judgment, or “being guilty of wanting something that could be detrimental.” 31 A more practical and simpler approach was then suggested by rationing health care “in terms of how severely they limit the range of valuable lives individuals can live in just two dimensions: activity limitations and health-related feelings.”32
The subjective nature of health and well-being ratings by patients may be biased. For example, the immediate emotional reactions could be misleading compared to the overall and long-term outcome. Therefore, deliberative focus groups rather than individual surveys should be used for such judgments. 32
Reflecting on population value is different from individual-level value. The population-level value is usually permanent, and the welfare of a country, context, and population diversity greatly contributes to it. Contribution to the judgment is the extent that the individual is involved in decision-making versus the government. For example, some countries see the introduction of colorectal screening programs by the government as a necessity, while others do not even if the average population preference is supportive.
Decisions regarding population needs are dependent on the principles that governments adopt when they prioritize alternative health programs. Should governments adopt minimal principles and leave decisions to individual self-motivation, or should it implement social goals and expectations for the population? Should the government use coercion to ensure participation in health programs, or should it use coercion and information? Answering any of these questions needs information33. Hausman described how we may reflect on the policy level of these principles: “the welfarist approach where one thinks of the government as everybody’s mother with advancing individual welfare might be requiring intrusions into individual life. Or liberals who regard the government as a protector, insurer, and arbitrator but not as an active partner in individual pursuits. The former may promote passivity in choosing personal priorities in health.”32 These preferences for alternative ethical principles are called meta-preferences. 33 The best approach is probably carefully demarcating the line between not harming and not being harmed in the bargaining in the evaluation step. It is also important to consider that public value cannot be sensitive to all relevant details, nor can they be accurately measured. Therefore, implementing meta-decision while reflecting on patients’ choices, will build data to inform better future decisions with adaptation to each country’s norms.
Finally, an essential component of meta-decisions based on the triple aim is the consideration of possible harm, which is essential when assessing population health. Harm cannot be reported without being clearly measured and weighed against effectiveness to assess net improvement. Potential benefits are not meaningful without the knowledge and quantification of harmful impacts. Interventions provided by healthcare services are administered with the best of intentions; nevertheless, most inevitably cause harm, ranging from minute to significant. Failure in transparently informing end users of health interventions’ potential impacts is inconsistent with the ideals in the triple aim. However, safety assessment is not easily separated from that of effectiveness.
A challenge of focusing on measures but not the overall value is evidenced from examples in healthcare where achieving value is worryingly not always the target but achieving individual measures is. For example, it has been reported that when improvements as a result of organizations’ strategies for quality improvement described by the domains of the triple aim affected revenue in for-profit organizations with a decrease in patient visits or orders, sustainability of these strategies were challenged. 34 Similarly, a payment system initiative used surrogate measures, such as hospital re-admission for heart failure instead of the target-improved outcome, resulting in negative patient outcomes due to decrease in needed readmissions, raising ethical concerns of implementing what is perceived as best value. 34–36 Thus, meeting the target outcome was through developing the wrong choice.