How can we support evidence-based health policy?

Health policy can take many forms and cover a range of topics. The development of health policy is a complex and non-linear process. Theories of policy development – such as group theory, systems theory, or a streams-and-windows model – acknowledge that a multitude of factors interact and influence which topics are given priority as well as how these topics are shaped, discussed and converted into policy. The factors highlighted in such policy theories include:

• values, morals and preferences of policymakers, society and advocacy groups (patient groups, civic organisations, lobbyists etc.)
• internal environment: topographic, geographic, demographic and climatic changes, economic restrictions, scientific evidence
• external pressures: geopolitics or wider motivation from other countries, international movements.

Evidence-based policy promises greater efficacy

Evidence-based health policy prioritises an accumulation of relevant and appropriate data over opinions, values or political sway. The concept gained traction in the UK in the late 1990s as politicians used evidence as a gold standard, giving them leverage to overcome political biases and gridlock.

Evidence can be defined as ‘the available body of facts or information indicating whether a belief or proposition is true or valid’. The extent to which evidence is prioritised over other influences will vary, but it is unlikely that any policy will be built solely on evidence. Some argue that this means we can only talk about ‘evidence-informed policy’. Others suggest the term ‘data-driven policy’ is more appropriate. While all terms have their benefits, ‘evidence-based policy’ is the most commonly used term and encapsulates the widest range of policies.

Evidence-based health policies are meant to reflect real-world scenarios, and therefore promise efficacy. Building policies based on evidence is meant to ensure that the real needs and issues in a system are identified, understood and addressed, with solutions developed in the absence of biased perspectives, malicious intent or outliers. With problems and solutions grounded in reality, such policies are more likely to be relevant and appropriate, increasing the likelihood that they will have the desired impact.

Not all evidence is created equal

The success of evidence-based policy depends on the quality of the evidence. Of course, there are many areas where we simply don’t have the data to generate evidence-based policy. Women’s health has historically been a prime example. But rare diseases, mental health and other areas have also been under-recognised. There are also large gaps in data collection from certain under-represented groups. In rare instances, data collection may not be feasible. However, often the decision to collect data in certain areas but not others is based on values, opinions and preferences.

Where evidence does exist, quality issues may be broadly attributed to issues either with data collection, or with its selection and interpretation.

Data collection:

• The collection of data in an unbiased and rigorous manner is the subject of extensive study. Data entry systems, from surveys to financial reports, may influence respondents’ answers or researchers’ interpretation of the results – either by design or unintentionally. Moreover, data may not always represent the relevant population or situation.
• Data collection must balance comprehensiveness with usability and cost. Detailed data collection systems with extensive questions can support more nuanced and accurate understanding of a situation. However, these systems are often more costly; they may be too time-consuming; and the results may be inaccessible – even to researchers. While data collection systems with fewer fields might be completed more frequently, they may not capture all information required to appropriately contextualise and interpret data.

Data selection and interpretation:

• It is possible to ‘cherry pick’ data that support a desired hypothesis – missing, ignoring or, at worst, hiding data to the contrary.
• Data can be presented without appropriate context or consideration for quality, leading to acceptance of findings that may not represent reality.

Health policies created with incomplete, inaccurate or biased data are not likely to have the desired impact – and they may actually be detrimental to public health.

How can we ensure we present high-quality evidence?

Evidence should be generated in a systematic way, with priority given to higher-quality data. A number of tools can help structure evidence generation and ensure that all relevant data are captured. For example, rigorously developed models, scorecards and frameworks can form the foundation of research. These tools should support the collection of diverse perspectives, accounting for both scientific evidence and people’s lived experiences. They can also make evidence generation more efficient and cost-effective.

Evidence should be presented to relevant decision-makers in an appropriate way. This means that scientific literature and academic publications must be synthesised and adapted to account for the fact that policymakers may not be experts in the topic of interest. Moreover, it is essential that evidence presented to policymakers includes recognition of the potential shortcomings in data collection, selection and interpretation.

High-quality data should be at the heart of evidence-based policy. Taking the time for appropriate data collection, selection and distribution supports the longer-term sustainability and efficiency of the health system – so it is worth doing it right.