Since 2004 the National Health System (NHS) in the UK has rewarded general practitioners for the provision of quality care through the ‘Quality and Outcomes Framework’ under which adherence to standardised protocols is assessed by process measures. “This comes with the assumption that adhering to these measures gives good quality outcomes. That is not always the case,” said Sally Lewis, Assistant Medical Director, NHS Wales.
In particular, the Quality and Outcomes Framework sets out standardised approaches to caring for patients with chronic diseases. But in diabetes for example, a varying patient population means a one-size-fits-all protocol is not appropriate, and in some cases could be harmful. That is not to say all process checklists should be abandoned however. “Look at the patient first, so you don’t subject them to a process that will not contribute to outcome,” Lewis said opening a session, ‘What are we measuring?’ at a Science|Business event on patient outcomes.
December 2016 saw the first publication of a national survey of patient satisfaction with hospitals in France, allowing the public to make comparisons between hospitals and highlighting where there is room for improvement. With the French healthcare system a mixture of public and private, and everyone having total freedom of choice in selecting physicians and hospitals, this is highly influential information. “We must be very careful in publishing outcomes, because the impact is high,” said Félix Faucon, Inspector General, L'Inspection générale des affaires sociales.
The publication of the data followed a two year pilot programme to develop the methodology. A patient’s view of their stay in hospital shifts over time, particularly in respect to pain, and for this reason, the patient questionnaires are all completed between two and ten weeks after treatment. When comparing results for different hospitals, it is important to understand the diversity of case mix, Faucon said. “The same medical intervention can produce a good outcome for one patient and a bad outcome for another, depending on their initial state of health.”
The data on patient satisfaction will be factored into league tables that are starting to be used as the basis of pay for performance, with 0.5 per cent of budgets assessed in this way currently. “This is the first step to make it possible to go to value-based payments,” said Faucon. It is important to note that payment for performance is made to the best institutions but also the most improved. “The aim is to make it possible for the quality of all to rise; hospitals are provided with an incentive,” Faucon said.
While league tales are to be published annually, each hospital gets reports on its ranking every 48 hours, so managers know how they are positioned in relation to their peers and where they may need to make improvements.
The pharmaceutical industry sees itself as key stakeholder in the discussion about patient outcomes, said Shahid Hanif, Head of Health Data and Outcomes at the Association of the British Pharmaceutical Industry. “Once you measure outcomes and value, you get onto price – this is a tense discussion. To get a proper debate, there needs to be early engagement with stakeholders,” he said. Measuring outcomes raises challenges for the industry, clinicians, politicians and patients, for example, in data and information access, collecting outcomes data and getting all stakeholders on board. The concept of learning health systems, in which performance information drives continuous improvement, puts questions about how data is effectively utilised and how changes in the way care is provided are implemented, under the spotlight.
Initiatives such as the European Medicines Agency’s Prime (priority medicines) scheme that aims to accelerate access to innovative and highly effective drugs, are challenging pharma companies to provide early demonstrations of safety and efficacy. “The industry has got to provide evidence, but it is not clear if evidence will be available,” said Hanif. “There is a lot to be done in the foothills to build an effective platform.”
Using outcomes data
There are many levels at which outcomes from patients treated in real world settings can be applied to improve healthcare, said Lewis suggested. For example, drugs for controlling tremor in Parkinson’s disease become ineffective over time, raising the question of when to start taking them. “It is very important for patients and physicians to look at this together, it is a powerful tool in consultations,” she said.
Another use is in evaluating new models of care, something that healthcare systems are “very bad at doing,” Lewis said. With outcomes data it becomes possible to demonstrate value, or to make a decision to stop doing something, allowing resources to be allocated in the best way. “There is a lot of opportunity cost to the things we do,” Lewis noted. In addition, outcomes data provides a means to compare between hospitals, regions and countries, to uncover disparities and drive improvements.
Outcomes and diversity
The value of looking at patient outcomes is that they provide a common denominator by which to assess the diverse range of European healthcare systems, with their different structures, motivations and values. A commonly used illustration is prostate cancer. Here, survival rates may be similar from one hospital, region or country, to another, but side effects including loss of sexual function or incontinence – which matter most to patients – are not. The disparities provide a huge opportunity to learn from clinically-relevant, patient-centric outcomes.
As Lewis noted, in many diseases, data on patient-relevant outcomes is not available. “Patients can’t believe it, but we don’t have the data,” she said. That shortage of information is being addressed in the Innovative Medicines Initiative’s ‘Big Data for Better Outcomes’ programme, which is examining how to develop patient relevant outcomes in a number of disease areas.
A barrier to collecting and putting patient outcomes data to use is that healthcare professionals are already collecting a lot of other data, much of which goes into a bureaucratic black hole, from where there is no feedback. Outcomes data has to be made available to data collectors so that people are engaged and see the value. “You need headroom to embed outcomes data collection and use into core practice,” said Lewis.
Comparing outcomes between surgeons is a way to speed up the adoption of innovation, believes Rishi Hazarika, VP Implementation at the International Consortium for Health Outcomes Measurement. Healthcare systems are notoriously resistant to innovation and it can take up to 17 years for full adoption. But when one surgeon at the Martini-Klinik in Hamburg, Germany’s biggest prostate cancer centre, adopted a new technique from Korea, his peers were able to see the positive impact on patient outcomes. “As a result, it was quickly adopted in the whole centre,” Hazarika said.
Therapy outcome measures
To date, much of the focus on patient outcomes concerns hospitals and primary care. There is a requirement to ensure the allied community services that play a key role in enabling people to get back to their own homes, thus reducing overall costs of care, are factored into patient-reported and clinical outcomes. Each patient should be considered in relation to their health and wellbeing in their life and not in the context of a hospital or a service.
Therapy outcome measures, in which health needs are assessed at the start of care and providers work with patients to meet those needs and assess subsequent success at the end of a cycle of treatment, are “hugely underestimated” in terms of their value in keeping people out of hospital, Lewis said. In addition, standard outcomes measures do not take account of different aims. For example, in mental health, if a patient is still depressed at the end of a course of treatment, but is now capable of doing her or his own shopping, that is a good outcome.
For patients, it is about how the system as whole delivers, not its individual parts. “The holy grail has to be whole system assessment,” said Lewis.
In summary, Hazarika said the benefits of process and outcomes measurements should be blended, as is the case in other service industries. When measuring outcomes be sure to reflect the whole patient journey; to be cautious in interpreting and using the data; demonstrate how transparency around patient outcomes can influence peers and promote innovation; and most importantly, include things that matter to patients, not just to hospitals or payers.