“We need to come up with more things that work”: Prof Mary Dixon-Woods on patient safety

26 Feb

by Gavin Daker-White, Research Fellow for Multimorbidity theme


Today, Sally Giles and I were present at the inaugural meeting of the Yorkshire Medical Sociology Group, where Mary Dixon-Woods gave the keynote address: “Knowing about patient safety: how hard can it be?”

Mary, began with a historical survey of the measurement of adverse events in hospital settings, beginning with William Petty (1623-1687). In the UK, such efforts are currently encapsulated in the NHS Atlas, which shows variations in quality and outcomes for different regions. She then considered the various problems with “intelligence” and “measurement” in patient safety research and initiatives.  For example, hospitals found to be ‘unsafe’ previously ranked highly on various quality indicators. The ways in which information is collected, coded and classified is subject to variation. For example, current incident reporting systems around medication errors mean that such information is unlikely to reflect the “true” rate of such incidents in clinical practice.

Intelligence about patient safety has a “fugitive” character; difficult to define or grasp. For Mary Dixon-Woods, such information can be as much about “the business of judgement” as it is about harder, objective, scientific matters. For example, her own research with the Health Foundation on catheter-associated bloodstream infections found that clinical workers were excluding patients judged to be at either low or high risk of such infections. In a version of Goodhart’s Law, when data are used with regulation or control purposes, they may cease to be effective as a measure of the phenomenon of interest.

There are problems in collecting, recording and counting patient safety-relevant statistics. From the point of view of health boards and those commissioning services, what is needed is “intelligence.” On the face of it, “problem-sensing behaviour” – perhaps associated with an avowed honesty around clinical outcomes would be preferable. However, “comfort seeking behaviour” – such as routinely ticking off a patient checklist once an hour during a hospital stay – can lead to patient safety “blind spots.”

The technical and measurement tasks essential to patient safety are tied to normal social practices. For example, the primary literatures sources used as “scientific facts” for patient safety statistics were written based on historical data collected by imperfect methods (e.g. reviews of patients’ hospital records). Other studies have used “trigger tools” to measure adverse events, which are strongly biased according to the skills of the individual using them in each case.

Mary, ended her presentation with a call for translational research to “come up with more things that work” in relation to improving patient safety. A current dearth of such examples means it is not always easy to engage clinicians and others in patient safety initiatives. In discussion afterwards, one audience member pointed to the importance of the “ownership” of systems of reporting and control residing with the health workers who collect and handle patient safety information. I was left wondering whether the main issues in hospital safety: “clinical monitoring, diagnosis, drugs and fluids” would also apply to primary care.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: