Clinical informatics can make a major contribution to improving the quality and safety of patient care. Some promising technologies which are potentially applicable throughout medicine are active decision support, care planning and flexible workflow management.. However, no-one knows how to integrate and practically deploy such services in the complex, unpredictable and ever-changing context of healthcare delivery, or how to do it in a sound, scalable and safe way. Our proposal is to carry out a comprehensive study of this challenge, starting with the medical need through to technical requirements and options. We will demonstrate the feasibility of at least one technical solution and deliver a report of our results, including a road map for further development at the end of the project.
Rationale
Medical errors and deviations from best practice are known to have a major impact on the quality and safety of patient care. Sir Muir Gray has frequently observed that the “application of what is known already will have a greater impact on health and disease than any single drug or technology likely to be introduced in the next decade”. This is supported by evidence published by partners in this proposal, which shows that (a) over 10% of admissions to NHS acute hospitals result in adverse events, and (b) many kinds of clinical decision can be improved by a factor between 10% and 50% using decision support and other ICT services. Improving outcome and increasing safety while also reducing costs is a massive challenge facing all healthcare services worldwide (see http://www.openclinical.org/publicreportsSafetyDetails.html). It cannot be achieved by simply providing more resources, demanding that healthcare professionals work better or harder, or threats of litigation. However research strongly suggests that ICT can significantly reduce, and in some cases eradicate, many problems.
The causes of adverse events can be separated into the immediate causes, which are the result of actions, or omissions, by people at the scene. However, other factors further back in the causal chain can also play a part in the genesis of an accident. These ‘latent conditions’, as they are often termed, lay the foundations for accidents in the sense that they create the conditions in which errors and failures can occur (Reason 1997)(2). These concern issues such as poor training, problems with scheduling, conflicts between safety and profit, communication failures, inadequate procedures, failure to address known safety problems and general sloppiness of management and procedures. Human beings also have the opportunity to contribute to adverse events in the design, testing, and implementation of a new system; in healthcare however, regrettably, we do not usually consider the design of the basic systems, many of which are highly vulnerable to error.
Bates and Gawande point out that the sheer quantity of medical information, even within a single specialty, is often beyond the power of one person to comprehend. People, that is the human brain, simply cannot cope with the amount of information that they need to function safely and effectively. For instance, more than 600 drugs require adjustment of doses for multiple levels of renal dysfunction; an easy task for a computer, but one which will inevitably be performed poorly by a person1. The disparity between human capabilities and the results that it should, given our knowledge, be possible to achieve has led to the situation in which patients receive varying levels of care, with the likelihood of recovery often dependent on which medical centre the patient visits. Adverse events which seriously impact the health and even the lives of patients are all too common. The challenge is to develop the technology needed to integrate the vast pool of existing information and knowledge relevant to the care of any specific patient and deliver it in an effective and coordinated manner at the point of care.
We must improve performance
As remarked above over 10% of patients admitted to NHS hospitals experienced an adverse event; around half of these events were judged preventable with ordinary standards of care. A third of adverse events led to moderate disability or death. According to a US Institute of Medicine report in 2001 “Between the health care we have and the care we could have lies not just a gap, but a chasm.” Safety is the most emotive problem. An earlier report from the IoM says “at least 44,000 and perhaps as many as 98,000 Americans die in hospitals each year as a result of medical errors. … Deaths due to preventable adverse events exceed the deaths attributable to motor vehicle accidents (43,458), breast cancer (42,297) or AIDS (16,516)."
Evidence is accumulating that failing to provide standard treatment is a problem of epidemic proportions which itself poses a serious threat to health. McGlynn and colleagues carried out a study of 6712 adults in the United States examining their medical records and conducting telephone interviews . Incredibly, even in the United States with its legendarily high health costs, albeit mainly spent on 80% of the population, patients received only 55% of recommended care overall. McGlynn and her colleagues argue that these findings have important implications for the general health of the population implying, in a sense, avoidable harm. For example only 24% of the diabetics in the study had regular blood tests, a requirement for close blood glucose control and the avoidance of complications. People with hypertension received 65% of the recommended care; uncontrolled hypertension increases risk for heart disease, stroke and death. Studies of this kind do not directly assess harm, but provide important information about the conditions in which patients can be harmed.
How can ICT help?
A significant way of addressing these problems is to ensure timely delivery of patient information and up-to-date medical knowledge at the point of need, the clinical decision. Systematic reviews of published research on decision support show that even simple aids, such as alerts and reminders at the point of care, measurably improve outcome in about 70% of studies. If a handful of key design principles are all followed benefits are found in 90+% of studies. More sophisticated technologies developed by partners for decision-making and workflow (Oxford) and care planning (Edinburgh) have also been rigorously evaluated in seven published trials covering genetic risk assessment, referral decisions, drug prescribing, image interpretation and treatment planning. These studies uniformly show substantial benefits, up to 50% improvements in decision making on a wide range of measures. In one case a 90% reduction in safety-critical deviations from clinical guidelines for early diagnosis of breast cancer was obtained (reduced from 10% to 1%) and in another a complete eradication of deviations from the protocol for treating leukaemia in children (from 23%).
After more than 20 years of research, decision support technologies which provide patient-specific advice are becoming mainstream. Decision support and related knowledge based services can help to improve performance in most areas of medicine. It seems evident that it is time for a substantial effort to show how such technologies can be implemented in a sound, usable and scalable way, providing an integrated set of services for supporting care from initial presentation and risk assessment to diagnosis and the planning of treatment. A goal of this proposal is to develop the case for this in detail, comprehensively setting out requirements and constraints, and proving the feasibility of the concept with a practical demonstration.
Our proposals can help to improve quality of service for all stakeholders. Busy clinicians are offered decision support and other services, with rapid access to current research and evidence, helping them to maintain consistently high quality, safe care even when clinical demands are heavy. They will also be better informed about the decisions and actions of other colleagues involved in a patient's care and the rationales for their decisions. The importance of patient-centred care identified by the IoM (and many others) can also be addressed. Patients will be able to review their treatment options, and see the pros and cons of those options as they relate to their individual circumstances. This will foster informed consent and empower those patients who wish to influence their care based on their personal values and preferences, resulting in greater patient satisfaction. (Concrete evidence of many of these benefits can be provided.) From the healthcare organisation’s perspective, better dissemination of knowledge and more consistent compliance with clinical and service guidelines can increase efficiency as well as improving outcome and safety of the healthcare process. If such services are implemented using ICT then we can routinely capture decisions and outcomes along with better records of safety-related and other incidents for subsequent clinical audit. These will make a major contribution to establishing the organisational memory called for by the Chief Medical Officer of the NHS , helping to reduce repeated adverse events and the risk of litigation.
