Clinical decision-making at the heart of eHealth
As the federal election approaches like an express train, it would seem worthwhile to take a look at where eHealth now fits on the national agenda. It would appear that the government has put eHealth on the unpalatable ‘burnt toast’ ledger as it is not perceived as a vote-winner.
Reading what is available from the federal opposition, it would seem that it does not have eHealth on its agenda either, possibly because it does not deem it as very significant.
Having tried to establish a functioning PCEHR and underachieved in that area, it can be said that the government of whatever persuasion after September 2013 will need some significant reform to health policy. Why?
The answer to this question is very clear. Health is unaffordable. This is not only an Australian phenomenon but a worldwide one. It means that the established ‘systems’ of healthcare are wrong and health expenditures have been focused upon the wrong mechanisms of healthcare delivery.
OECD progressive evaluations show Australia sits within the top one-third of nations in terms of health costs as a percentage of GDP, but all of these nations are seeing this ratio rising in parallel. Australia has health costs running at approximately 9.1 per cent of GDP (i.e. ~$US3670 per person compared with per capita GDP of $US40,300) and on current linear models this will reach 16 to 20 per cent by 2045.
In an attempt to present a different image of healthcare functionality and funding models, it is worthwhile addressing healthcare delivery from the ‘clinical’ perspective i.e. direct patient care and the complex decision-making processes that are involved.
Clinical information management
The dynamics of clinical decision-making (CDM) and information management (IM) are regularly not addressed when eHealth projects are undertaken, yet they are critical to any eHealth implementation.
The critical nature to healthcare improvement of these two processes can be seen within the definitions of healthcare delivery by expert organisations. For example, the World Health Organisation (WHO) defines healthcare in its charter as follows: “There is no health without management, and there is no management without information”.
So what does this mean for day to day patient care? Here is a description of what clinicians – physicians, nurses, pharmacists and physiotherapists, and in the current era, patients – do as information managers. We also have documentation of the outcomes of this information management with inadequate clinical decision support tools. These rules apply to the stages of acute, intermediate and chronic care.
Even though clinicians are seen as providing services, their primary role is the management of information. But what does this actually mean? In routine care, clinicians collect data such as patient history, physical examination, create reports, access laboratory data, read x-rays results then record this data (through the production of notes, operative reports, prescriptions and diagnostic test results).
This data is transmitted through various means such as telephone, paper documents, electronic charts and email. Finally, they process this information to arrive at a diagnosis or deduce a hierarchy of possible diagnoses and initiate treatment(s). This process becomes an iterative cycle of data and information management so that care can be monitored, adjusted and measured.
This understanding of clinical information management, based upon physician clinical decision-making, was shown by Bruce Blum in his book Clinical Information Systems, first published more than four decades ago, and was reaffirmed more recently by Australia’s Enrico Coiera in the third edition of his text, Guide to Health Informatics.
We also know that modern health technologies have created a situation of intolerable information and knowledge overload in the absence of an increase in human cognitive capacity to manage it. Therefore we need tools to assist the unaided human mind to affect the poor outcomes of care such as underuse, overuse and inappropriate use of health resources and the perpetuation of inadequate quality of care. (See Lucian Leape’s report, Five Years after To Err Is Human. What have we Learned?)
Across health systems globally, we have actual measures of poor health outcomes that have resulted from clinical decision-making overload.
In Canada, it was shown in 2005 that five per cent of chronic kidney disease patients unnecessarily occupied 19 per cent of in-hospital bed days and consumed 25 per cent of duplicate unnecessary tests that each cost $C4.50. This represented $C4.55m in a 12-month period.
Similar results were found for CT scans, MRIs and prescriptions in the Canadian health system between 2005 and 2009.
In the UK, in an evaluation of resource utilisation based on clinical decision-making during after-hours in a UK emergency department, it was found that 87 per cent of the tests were unnecessary.
The reasons for this were diagnostic uncertainty, medico-legal protection, avoiding leaving work for colleagues, prevention of criticism from staff (especially consultants), to lessen anxiety and reduce stress levels. As shown as far back as 1975, these are the processes that lead to further information overload and impaired clinical decision-making.
Additional measures of impaired clinical decision-making include clinicians’ inability to detect and reduce preventable adverse drug events, with this being closely correlated with medical indemnity premiums.
The health system’s dependency on poorly designed communication tools like our PCEHR between hospitals, primary care and patients exacerbates poor care outcomes.
The list of adverse health care measures is extensive, so can any eHealth processes – including clinical decision support and information management tools – improve care delivery? The answer is yes.
From the 1980s, we know well-designed, clinically appropriate, e-summary records improve care communication and quality of care.
The use of standardised alerts and reminders leads not only to health cost reductions, but to more appropriate resource utilisation and better bed utilisation in hospitals. The costs saving alone can be measured in billions of dollars.
For example, using computerised antibiotic guidelines in Intermountain Health Care hospitals in the US, major cost, quality and access outcomes have been achieved.
In addition, the use of these large clinical databases and direct clinical data capture has allowed researchers to measure how current funding models of more beds, more doctors and nurses do not produce better healthcare.
The emphatic messages we need to understand and to address within our unaffordable healthcare systems involve politics, economic and social change.
We must remove ourselves from unscientific, non-data-driven ‘personal recommendations’ for care and move the care delivery system away from the costly and inefficient ‘widgets’ model of care based on a pre-determined ‘appointment system’ involving healthcare encounters.
Care needs to be directed at patient self-management by giving them the HIT tools to manage their own health.
Dr Terry Hannan is a Clinical Associate Professor at the School of Human Health Sciences in the Department of Medicine, University of Tasmania. He is a consultant physician at Launceston General Hospital and has a deep knowledge of eHealth and health IT. He blogs at www.austemrs.com.au.
Posted in Australian eHealth