The quality of healthcare is regulated in a variety of ways internationally. Routinely collected data often provides a foundation to that regulation. In England, quality is regulated by the Care Quality Commission (CQC). Attention is focused on the inspections carried out by the CQC, however, routine data is used extensively before and during inspections (CQC 2015).
Over 200 routinely collected provider level indicators are tracked by the CQC, in relation to acute and mental health hospital providers and GP services. These indicators are aggregated, for each sector, to give a risk banding for each provider. They cover areas as diverse as waiting times, mortality rates and staff survey results. This process is known as Intelligent Monitoring.
This information is published (Department of Health 2012) and is used to prioritise providers for inspection. In the inspection planning phase, more bespoke indicators are requested from providers, sometimes down to the service level. All of these indicators are combined in a “data pack” relating to a provider. The inspection team use the data pack to guide their inspection and to triangulate findings.
The indicators used for Intelligent Monitoring are sourced from a variety health service databases and represent indicators that are available, rather than the indicators that might ideally be selected. Many providers, such as private mental health providers, do not report enough indicators to produce a sensible interpretation. Even with the maximum available set of indicators, Intelligent Monitoring can only provide ‘smoke signals’ and comprehensive inspection is required to make a judgement on the safety, effectiveness, responsiveness, compassion and leadership of each organisation.
In the UK, early Learning Healthcare Systems, including analysis of HES data (Morrow 2015) and patient feedback websites (Munro 2015) are already finding their way into the quality regulation process. Broader implementation of comprehensive outcome measurement would give regulators and patients a much better impression of the quality of providers and even individual services than is currently available. This could allow much more cost effective, timely and targeted inspection. Predictive modelling might take this a step further by actually allowing regulators and providers to identify emerging risks before harm is caused to patients.
Evidence