Making population health measurement simpler for clinicians, in Ontario Health Teams and daily practice

This article was originally posted on OntarioMD.blog

With fairly radical transformation happening in health care in Ontario as we speak, a question looms large: how do we measure success?

How do we know that the change we get is what we need? Which lens do we look through as success will appear differently for patients, physicians and the health system itself? Is there a benchmark we can compare to that goes beyond “at least it’s better than it used to be”?

We have experience with measurement.

Understanding that not all that can be measured matters, and not all that matters can be measured, we still need to ante up useful performance indicators that will allow course correction or celebrations of success.

OntarioMD has been measuring practice change for nearly a decade already. We have been measuring maturity of EMR use by over 5,000 doctors for 7 years. The Insights4Care Dashboard is installed in 500 offices now, with at least another 500 more to come this year. We implemented indicators that matter (valid, meaningful, measurable, open for improvement), although this was no easy task. The dashboard uses real world data from practices with the robust data set found in primary care EMRs. It pulls results in real time. It visualizes the data to be easily consumed. And it allows for direct action from the dashboard itself.

The first batch of indicators was strong in four major domains: roster management, chronic disease, screening and prevention and opioid management. Opioid indicators were added as a direct response to a partnership with over a dozen provincial organizations who want to bend the curve on opioid prescribing in the Ontario. They are pithy and impactful, taking into account current guidelines on risk levels. Clinicians can now quickly see all patients using over 50 and 90 morphine mg equivalents per day, those on dangerous opioid / benzodiazepine combinations, and those who have been on the drugs for a prolonged period of time signaling risk of addiction. This information can help clinics and their staff develop a plan of action to handle complex and complicated patient problems, and proactively move to change behaviours and reduce risk.

Advanced measures will be required for showing the progress of OHTs, and for the comparators needed for accountable care.

For OHTs, the same basic principles apply.

• Indicators chosen must matter not just to policy makers, but also to patients and providers, taking Quadruple Aim concepts into full consideration. We must ask ourselves “do we know what those we care for really need?” Do we know? Have we asked?
• Evidence, standards and guidelines do not always translate into actionable measures of the practice or data source level. For example, when looking at Quality Standards, some of these are aspirational. There are often few that can be boiled down to indicators at the coal face.
• Basic principles of indicators being measurable, valid, subject to improvement and actionable apply. If and indicator does not lead to change, then it should stop. As much attention should be paid to removing indicators that are not adding to improvement as to adding new ones that might.
• Too many indicators are a bad thing. Where possible, and where insights are not lost in doing so, similar indicators should be combined for a larger world view.
• Balancing measures are important to ensure that focusing on one specific area does not have an unintended consequence in another.
• Real-time access to data is important. Insights should derive from the data at its source, wherever possible, with little delay.
• Accountabilities for any recommended action need to be shared, especially in integrated care delivery systems. Outcomes are not solely controlled by clinicians. The health system must support their work. Citizens need to be included and empowered to make changes in their health. This will likely require exposing them to their own data and even the higher lever metrics. We will have to grapple with the daunting issue of public reporting sooner rather than later.
• Wherever possible, measurement should be automated. Manual extraction and reporting must not be accepted as the norm as this net-new work distracts from the main task of delivering excellent care.

Success takes many forms when it comes to measurement and reporting for Ontario Health Teams. We have created a framework for success with the OntarioMD Insights4Care Dashboard. A series of metrics have been carefully chosen from provincial and national measurement frameworks. They have been analyzed tested and vetted. They are automatic and in real time. They can be acted upon in a meaningful way. Combined with effective change processes and practice advice and coaching, OHTs being built with primary care at the core have a tool that showcases their great work, and which will ultimately produce better population outcomes. When this primary care data is combined with administrative data from government sources, utilization information, statistical data on social determinants of health, prevention, and health promotion, and knowledge, even data from patients’ own devices, we have generated wisdom and clarity.

We have seventy more clinical indicators in development for the dashboard. Let’s scale up this important tool. Doing so will help create a culture of knowing and lead to lasting, sticky change in the health of the communities we care for.