How to Run Analytics for More Actionable, Timely Insights: A Healthcare Data Quality Framework

COVID-19 response复苏需要数据来推动及时、可操作的见解，达到前所未有的水平。As a result, health systems increasingly recognizedata质量是临床、财务和运营分析的先决条件。为了量化数据质量，医疗保健数据团队可以使用可测量的数据属性来证明它是否适合于特定用途。良好和透明的数据质量会让人们对所提供的见解充满信心，从而加快合理的决策。相反，糟糕的数据质量会降低信心，最终推迟或导致错误的决策。

Organizations tend to understand the value of data quality, but the fundamentals of a system that generates quality data and analytics are complex. To meet the COVID-19 urgency for quality data and ongoing data quality challenges, health systems need an actionable structure to navigate the essential phases of a comprehensive and proactive data quality strategy. A framework for healthcare data quality provides a systematic way to measure, monitor, and determine if data is “fit for purpose” (i.e., it can serve its intended purpose).

The Four Levels of Healthcare Data Quality

Defining data quality levels helps an organization understand the current state of its data quality and whether its data is improving. Data users can follow the Four Levels of Data Quality (Figure 1) to determine quality checkpoints, including whether data quality depends on the context of the data or purpose for its use and whether defining data quality requires subject matter expertise.

A Framework to Measure Quality Throughout the Data Pipeline

遵循卫生保健数据质量框架(图2)的卫生系统将从头开始建立数据质量文化，并积累必要的信息，以推动有意义的改进，对危机做出反应，并为未来的突发事件做好准备。

The Healthcare Data Quality Framework guides quality assurance throughout the data lifespan:

Think of Data as a Product

在数据质量的背景下，将数据视为产品意味着数据来自于一个过程或系统，该过程或系统对其质量进行评估和处理——类似于汽车从原材料到装配线、经销商到专家杂志评论的过程。为了成功地通过汽车制造、销售和评估过程，汽车制造商需要优质的原材料(如车身和发动机部件)来将他们的汽车从概念车推向消费者。

To create a fit-for-purpose healthcare data product, analytics professionals need to prioritize quality at the beginning of the data pipeline and shepherd that quality as data traverses the system. In healthcare, it is important to confirm that data is an accurate representation of its source (e.g., EMR, payer/claims, costing, human resources, etc.).

To ensure they are thinking of data as a product, data engineers can ask themselves the following questions:

• Have I defined user personas or data users representing people who will use the data now and in the future?
• 我是否定义了描述用户完成特定于其目标的任务的用户故事或数据用例?
• 我是否定义了单元测试或数据质量评估来评估流程是否按预期运行，以及数据是否适合使用?
• Have I deployed data quality assessments at the earliest appropriate point in the data pipeline?
• Have I identified and deployed user personas, user stories, and data quality assessments for the components of the data production process that are upstream of my data product?
• Have I documented each assessment in a transparent, accessible, and centralized place—allowing new, existing, and upstream/downstream data users to understand what users and use cases the data supports and how the organization defines and ensures data quality?

Address Structural Data Quality First

Health systems struggle to move to higher data quality levels if the data is not first structurally sound. The levels described above build on each other, and while content and utility assessments will expose structural issues, understanding the root cause is more efficient when leveraging specific structural assessments. For instance, determining whether an encounter identifier is unique across encounters and not NULL promotes referential integrity. When a data user then leverages that identifier as a foreign key to link the flowsheet and encounter data together, the user can focus on assessing the quality of the content across these two subject areas, like whether the flow sheet recorded date is during the encounter.

Define Content Level Data Quality with Subject Matter Experts

Understanding data use cases is extremely important for defining and ensuring the quality of the data content because it requires subject matter expertise and can be context dependent (for multiple-subject areas). Potentially different from a data user, organizations must identify a data subject matter expert (SME) for defining content level data quality because that expert will understand the content (e.g., the temperature makes sense given the unit of measure). The SME tailor definitions based on the context (e.g., the heart rate is appropriate given the patient’s age) to assess whether data quality is sufficient for the intended use cases.

Create a Coalition

Typically, organizations take a grassroots approach to data quality by addressing it within individual projects or department silos. However, creating a data quality coalition brings together organizational leaders, managers, subject matter experts, and analytics professionals—all with a vested and shared interest in ensuring data quality because it facilitates better decisions. The team agrees on a standard approach to advance proven processes and avoid spending resources reinventing the wheel. The coalition must have support from leadership at the highest level for organizational alignment in terms of objectives and resources focused on the work.

Building Data Quality from the Ground Up

在卫生系统继续应对COVID-19并为应对紧急情况做好准备的未来之际，符合目的的高质量数据已成为一项必要的战略。确保医疗保健组织领导人、管理人员和提供商拥有适合关键决策(如COVID-19应对)的数据的唯一方法是在数据生命周期的一开始就建立质量，并在所有流程中维护数据。结构化的质量流程(如医疗保健数据质量框架)利用技术和主题方面的专业知识来定义、评估和监控整个流程中的数据质量。因此，卫生系统不仅做出有数据依据的决定，而且做出有数据依据的高质量决定。

Additional Reading

Would you like to learn more about this topic? Here are some articles we suggest:

1. Why Health Systems Must Use Data Science to Improve Outcomes
2. Smartsourcing Clinical Data Abstraction Improves Quality, Reduces Costs, and Optimizes Team Member Engagement
3. Quality Data Is Essential for Doctors Concerned with Patient Engagement
4. Self-Service Data Tools Unlock Healthcare’s Most Valuable Asset
5. Achieve Data-Informed Healthcare in Eight Steps

---

PowerPoint Slides

Would you like to use or share these concepts? Download the presentation highlighting the key main points.

Click Here to Download the Slides

How to Run Analytics for More Actionable, Timely Insights: A Healthcare Data Quality Frameworkfrom世界杯葡萄牙vs加纳即时走地