Using data to improve quality of healthcare

在几乎每个主要行业中，质量改进都是基于测量和监控性能。有很好的证据表明，当对过程和结果数据进行衡量和比较时，就会有所改进。医疗保健在整个行业引入系统测量的速度一直很慢，但在国际上，这种情况正在迅速改变。在澳大利亚，我们有衡量质量和安全的数据；但各州和地区之间没有系统或一致地收集这些数据。我们的目标是开发一套全国性的高质量、可信的指标，利用现有的数据来源来衡量临床实践中的安全和质量方面的医疗表现。我们提出了一种“布丁中的探测器”方法，其中每个探测器测量医疗保健系统的一个方面的性能。主要出于财务原因收集的编码行政医院数据有可能被用作质量指标。然而，为了提供准确的图像，指标必须考虑人们的风险因素，医生无法控制这些因素，但这些因素可能影响结果。例如，在一个以老年人口为主的地区，一家医院的死亡率可能比一家接纳年轻患者的医院更高。实际上，指标必须将苹果与苹果进行比较。我们将进一步测试连接临床患者信息数据库以开发质量指标的方法。例如，通过将药房数据与医院数据联系起来，我们可以监控患者是否正在接受适当的药物治疗。用于测量质量的最可靠数据存在于登记处，在澳大利亚已经为许多程序和条件建立了登记处。我们将研究如何在国家层面上更好地实现登记标准化。我们将比较登记数据和医院行政数据，以调查如何制定质量指标，以提供卫生系统的准确情况。这一计划将获得成功，因为它可以使用大多数澳大利亚卫生机构中随时可用的数据以低成本实施。进行这项研究的小组由消费者权益倡导者和在使用行政数据和登记、卫生政策、法律、研究设计和统计方面有知识的人组成。

In virtually every major industry, quality improvement has been based on measuring and monitoring performance. There is good evidence to show that when processes and outcome data is measured and compared, improvements are made. Healthcare has been slow to introduce systematic measurement across the sector, however internationally this is rapidly changing. In Australia, we have data available to measure quality and safety; however it is not collected systematically or consistently between states and territories. Our aim is to develop a national set of high quality, credible indicators using existing data sources to measure healthcare performance in regard to safety and quality in clinical practice. We propose a “probes in the pudding” approach, where each probe measures the performance of one aspect of the healthcare system. Coded administrative hospital data collected mainly for financial reasons has potential to be used as quality indicators. However, to provide an accurate picture, indicators must take into account people’s risk factors over which the doctor has no control but which can influence outcome. For example, a hospital in an area surrounded by a principally elderly population would be more likely to have a higher death rate than a hospital admitting younger patients. In effect, indicators must compare apples with apples. We will further test ways in which clinical patient information databases can be linked to develop quality indicators. For example, by linking pharmacy data with hospital data we can monitor whether patients are receiving appropriate medication. The most reliable data for use in measuring quality exists in registries, which have been established for a number of procedures and conditions in Australia. We will investigate ways in which registries can be better standardised at a national level. We will compare registry data with administrative hospital data to investigate how quality indicators can be developed which provide an accurate picture of the health system. This program will succeed because it can be implemented at a low cost using data readily available in most Australian health settings. The group doing the research consists of consumer advocates and people with knowledge in the use of administrative data and registries, health policy, law, research design and statistics.