Making measurements simpler and more relevant

Making measurements simpler and more relevant

Some of the most important measures – for both internal and external use – are those that capture the overall outcomes of the care. To a diabetic patient, for example, the Hb1Ac (an intermediate outcome measure) is not a meaningful goal in itself; the true objective is to combine a full life with as few symptoms, exacerbations and long-term complications as possible. Similarly, for acute stroke or cancer patients, the core goals are survival, optimal recovery and, ultimately, quality of life.

Related content

“We need competition on outcomes, not on the metrics!” - Dr. Panigrahi, Head of Medical Operations, Fortis Healthcare, India

Once hospitals are able to reliably measure and report these outcomes, and demonstrate improvements over time, there is no need to publicly report a plethora of process and intermediate measures. Healthcare boards, patients, payers, governments and other stakeholders around the world will likely no longer be interested in processes for patient centeredness, timeliness and effectiveness; they will just want to know whether providers are delivering best possible outcomes.

This is a new approach and, not unexpectedly, the sector is still trying to define the key outcomes and find ways to measure these effectively.

“Ideally, internationally standardized measures would work best,” argues Ralf Kuhlen of Germany’s Helios Kliniken. “In that way, everyone would be doing the same; we could compare outcomes, and we could really bring medicine forward.” In the absence of such measures, Helios has developed its own set of measures based on hospital administrative data, with an emphasis upon key outcomes. This is published annually for every clinic that is part of the Helios group.

“Ideally, internationally standardized measures would work best. In that way, everyone would be doing the same; we could compare outcomes, and we could really bring medicine forward.” - Ralf Kuhlen, Chief Medical Officer, Helios Kliniken, Germany

Nevertheless, progress on outcome measurement is promising. In oncology and cardiovascular surgery, standardized outcome measures are becoming available through internationally coordinated clinical registries. For acute cardiovascular, chronic and elective care, the Dutch Health Insurers’ association has used its all-payer database to work with leading professionals to establish key outcome measures for conditions such as strokes, AMI and Parkinson’s disease. By combining this work with patient-reported outcome measures, it is possible to establish the key outcomes – including the appropriateness of interventions. In some of the leading US Accountable Care Organization (ACO) developments, payers and providers are combining administrative databases with clinical registries, to improve the validity of measured outcomes.

For providers, such measures should ensure attention on those outcomes that matter most, feeding directly to internal improvement efforts, and freeing time traditionally spent on reporting unwanted metrics. The new, limited set of outcome measures would more reliably demonstrate the organization’s level of control.

Does the increase in improving compliance with process measures translate into improved outcomes?

The NHS in England introduced a new policy relating specifically to venous thromboembolism (VTE) and pulmonary embolisms (PE). It required all hospitals in the country to complete a clinical risk assessment of 90 percent of their patients who were admitted to determine if they were at risk of developing a VTE or PE. This process measure was designed and enforced through policy, with financial penalties for non-compliance, in the hope that if patients were deemed to be at risk the appropriate steps would be taken, reducing clinical risk and improving outcome.

“Ask not what you need to do for clinical data entry and patient assessment, but what clinical data entry and assessment can do for you.” - Daniel Ray, Co-founder and Director of Quality and Outcomes Research Unit, University Hospitals Birmingham, UK

Adequate benchmarking, as made possible by the KPMG UHB benchmarking tool, demonstrates how improved process measures do not always yield the outcomes hoped for. Some hospitals who achieved the process measure target early actually had worse outcomes than in previous years. The ability to track long-term patient outcomes at disease level to refine clinical processes is paramount. The benchmark also allows hospitals to study whether the cost of implementing the process measure policy translated into saving lives. What would be better to report on, and more meaningful: the process measure or the outcomes?

Until recently, real outcome measures were rare

Processes are the most common measurements, because they are easier to measure and do not require detailed risk-adjustment. Measuring outcomes, on the other hand, means gathering data from a variety of different care providers that are often not connected. Providers usually do not have access to data from other systems; even clinical registries only follow the patient as far as the reach of the relevant profession goes. Payers will have data that transcend organizational boundaries, but until recently, these sources were rarely combined.External agencies have been focused primarily on the quality of care that a provider delivers, rather than on the quality of care a patient receives. The latter is a tougher challenge, as care is typically received from more than one provider.

Connect with us


Request for proposal



KPMG's new digital platform

KPMG's new digital platform