Editorials

Communication among health professionals

Human factors engineering can help make sense of the chaosInformation in practice p 673

Last year, my father was told by his family doctor that the cardiologist had found aortic stenosis during a diagnostic evaluation for hypertension. Some time later it transpired that the specialist's diagnosis had been wrongly transmitted. Instead of a major valve defect, my father actually had atherosclerosis, a much more benign diagnosis. The kind of culture that makes this sort of unfortunate miscommunication possible is examined in a paper in this week's BMJ and a recently published government report. ^{1 2} Their conclusions will come as no surprise to many BMJ readersthat communication between health professionals is a mess.

Both sets of authors offer a series of insightful recommendations on what might be done to improve things. However, there is also a pressing need to define the role of applied research in this area and to accept that other disciplines have a lot to teach health professionals on how to design, evaluate, select, and set up efficient communication systems. Without this dialogue between disciplines, useful concepts and theories will simply languish in journals instead of being used by doctors and managers to improve efficiency and reduce mishaps in medical practice.

Coiera and Tombs' observational study confirms that face to face, telephone, or pager based communication is common in hospitals and often driven by events.¹ They found that hospital communications commonly interrupt tasks, including patient consultations, and are inefficient. They suggest that we evaluate and consider investing in asynchronous methods of communication, such as electronic mail or message boards, which are potentially less disruptive to professionals' work and patients' welfare.

The Clinical Systems Group, set up in 1996 to advise the NHS on information management, used questionnaires to study patients' and doctors' views on how health professionals talk to each other and what they say.² Despite finding that both groups wanted most types of patient information shared freely, doctors estimated that most of the time important patient details were missing. Similar to Coiera and Tombs, the authors recommend procedural and educational measures to improve communication and urge the NHS to pursue research in this area. A further study in the same report also concludes that documentation in several healthcare delivery systems, and communication between the health professionals in those delivery systems, is chaotic. The authors' recommendations to doctors include more training in information technology, more structured data collection, and adoption of new technology.

These authors should be congratulated for trying to inform and improve policy, education, and deployment of technology. The inefficiencies they uncover may even be enough to prompt some action in the most deficient areas. Poor communication is not only a waste of time, it can threaten patient care and is the chief culprit behind avoidable errors in clinical practice, which can lead to injury and even death. ^{3 4} We should therefore push for more and better research into clinical communication and, of course, more funding. We should also heed the Clinical Systems Group's advice for education to fill the gaps in doctors' knowledge about collecting, sharing, and analysing clinical information.

The authors of the study and report agree that their methods were limited (small unrepresentative samples) or potentially misleading (reporting anecdotes and self reported survey data), but once again other disciplines can help.⁵ We must be more open to the theories and methods used in subjects like cognitive psychology and linguistics.⁶ Methods that go beyond questionnaires and interviews, like applied ethnography, are often unfamiliar to medical informatics researchers. ^{7 8} Human factors engineering, also known as cognitive ergonomics or usability engineering, is another discipline that applies knowledge of human capabilities and limitations to the design of devices and software. ^{9 10} Such methods of research and development have been useful in high risk domains (aerospace), complex systems (nuclear power), and consumer products.¹¹

You, as purchasers and users of information systems as well as the guardians of patients' interests, hold the key to changing this situation. Your influence will encourage researchers, administrators, and developers to base their projects on your information needs and use human factors engineering methods that result in usable and useful systems. My father seems to think that your influence surpasses his.