Safe Health Information Technology

•   Rattner, J. reported Intel Developer Forum, San Francisco. Sourced at 29 August 2008 from http://www.intel.com/pressroom/archive/releases/20080821comp.htm

Drug dosages and weight calculations are a well understood source of serious errors and all manner of adverse outcomes including death. This is particularly noted in inexperienced, fatigued, panicky and high workload situations.

Mathematical skills can vary considerably among clinical staff and while a base level of numeracy is usual present various causes of problems can occur including the use of an incorrect formulae, miscalculation of a formula, legibility of the original order or subsequent directions and lack of checking. 

None of these errors are necessarily eradicated in moving to a computerised solution be it handheld or spreadsheet process.

All the issues of poor software design live synergistically within new clinical information systems along with the pre-existing causative factors.

Poor software design may not lock in constants such as weight or dosage rates allowing them to be over-written by mistake, poor mouse or pen focus control and particularly on the latest micro and touch screens simply inaccurate hand eye coordination.  Software and hardware focus control is essential as following completing one entry, movement to another cell by mouse may leave the focus on the preceding entry and unanticipated changes may result and may not be picked up by the inexperienced or casual user.  Having typed a couple of key strokes and noting no change they click on the new entry point not realising the few key entries have gone into the preceeding entry.

 ”Wrong data in wrong cell” – a new catch cry or defense?

Windows is particularly prone to the focus-scrolling error where you can place your cursor on a cell and scroll through alternatives and then rather than using your mouse to move down to the next cell and click where you intend to type you forget to and start typing, meanwhile off screen an other cell is changing unbeknown to the user and even if known may not be able to be detected if the sheets or templates are complex.

Saskia N de Wildt, Ron Verzijden, John N van den Anker and  Matthijs de Hoog in “Information technology cannot guarantee patient safety” ( BMJ, 2007:334:851-2) looked at some of these issues.

The e-health environment is complex. Unlike aviation or space flight it is not a single goal directed activity and while safety is paramount it is tempered by enormous numbers of patients and pathologies, the ever present likelihood of adverse outcomes from any intervention, the impact of individual human variation, intricate biological and chemical processes and on it goes.  Add to this computing and its steady advance into more complex areas of human intellectual engineering and you have a recipe for error and adverse outcomes. This site examines analogies with non-clinical design process errors, anecdotal observations and reviews recent human factor research to suggest future health informatics research directions in applying human factors and systems designs through encouraging standardised clinical interface development based on evidence and analysis of errors as well as to suggest the need for greater uses of standardised interfaces across vendors.

All e-health developments are a web of incredible complexity – the interweaving of people,  systems and knowledge. The complexity we enter into in managing sources of error impacts substantially in many different areas of human endeavour:

• Computer Sciences – graphics, operating systems, programming languages.
• Systems design and Development
• Communication Theory
• Graphic and Industrial design
• Linguistics
• The Social Sciences and Cognitive Psychology
• Human factors including Physiology and Anatomy
• Engineering

There are a multitude of points of interaction in all these areas where design may be critical including graphical presentation,  icon design and placement,  standardisation of terminologies and screen actions,  physical environment, interface tool design, warnings and alerts.

We will explore all this and more on IT Human Factors. Enjoy my ramblings.

There are a vast range of causitive factors in clincal adverse outcomes and the introduction of information technology and new advances in systems design, languages and algorithms exposes patients to changed and often new sources of adverse outcomes.

In order to understand these outcomes it is necessary to come to grips with the type of potential causative factors.

The sites current task is to develop a suitable hierarchical classification and term set for IT related adverse outcomes. I welcome your input.

• I thought you said you were gonna cut the red wire!
• Well I did, didn’t I?
• No! You cut the blue wire!
• Well I meant the red wire.

 Lethal Weapon 3 (1992)

The notion of making a mistake is so engrained in the human psyche that the study and practice of preventing a mistake is an industry in its own right. An industry which paradoxically also makes mistakes. Mistakes are quite simply unintentional errors or misunderstandings usually (but not always) associated with a potential or real adverse or negative outcome for someone or something. History is littered with large demonstrable mistakes with little or no impact on humankind but equally there are events where small errors of judgement that have led on to catastrophic outcomes.

The rush to embrace Clinical Information Systems in the heath care environment has as one of its drivers a reduction in adverse clinical events. The design of clinical systems is often more dependent on underlying operating software, vendor preferences, legacy system economies and less with ensuring that clinicians, often in highly stressful situations, do not make errors when interacting with systems.

Such errors leading on to new adverse outcomes that the implementations were intended to reduce in the first place. There are a multitude of points of interaction where design may be critical including graphical presentation, icon design and placement, standardisation of terminologies and screen actions, physical environment, interface tool design, warnings and alerts.

There are also less obvious sources of error including the increasing push for deeply embedded decision support algorithms that may as they become hidden from the clinician increase the opportunities for misinterpretation and poor outcomes.

This site explores analogies with non-clinical design process errors, anecdotal observations and reviews recent human factor research to suggest future health informatics research directions in applying human factors and systems designs through encouraging standardised clinical interface development based on evidence and analysis of errors as well as to suggest the need for greater uses of standardised interfaces across vendors.