Medical Matters: Doctors are human and will always make diagnostic errors

A number of large studies carried out worldwide have concluded that some 10 per cent of hospital admissions are associated with an adverse outcome. Most are not life-threatening, but the figure reflects the complexity of secondary medical care.

The research points to a “Swiss cheese model” when things go badly wrong and a series of individually minor omissions or errors line up together: if any one of those had been nipped in the bud, it could have prevented death or serious injury.

Improving the system of care, rather than demonising individual healthcare professionals, is what will ultimately reduce the rate of healthcare adverse outcomes.

Much has been written and researched about errors in treatment, so I was interested to read a new report from the Institute of Medicine (IOM) in the US, Improving Diagnosis in Healthcare.

It concluded that most people will experience at least one diagnostic error, defined as an inaccurate or delayed diagnosis, in their lifetime.

In terms of numbers it’s estimated that some 5 per cent of US adults who seek outpatient care each year experience a diagnostic error; about 10 per cent of patient deaths have diagnostic error as a contributory factor; and diagnostic errors account for between 6 and 17 per cent of hospital adverse events.

Wide range

According to the authors, diagnostic errors have a wide range of causes including poor collaboration among clinicians, patients, and their families; a healthcare work system that is not well designed to support the diagnostic process; limited feedback to clinicians about their diagnostic performance; and a culture that discourages disclosure of errors.

The authors cite the benefit of the US OpenNotes initiative, which invites patients to view the notes physicians made about them during their visit. Not alone does this promote patient engagement, it can add an extra check to help avert diagnostic errors.

“This is not just about really smart doctors sitting alone in a room trying to figure out a tough problem,” Dr Christine Cassel, a member of the report committee says.

“In order to be open to questioning your thinking, you need to be welcoming of patients questioning what you’re thinking.”

I was disappointed not to see any reference in the report to the need for longer consultations to facilitate this enhanced doctor-patient interaction. Much of US medicine is fee-per-service based, but rarely does the fee structure allow payment for time spent interacting with patients. Rather it encourages a procedure-based, production-line mode of practice.

And at the risk of sounding nihilistic, a diagnostic error rate of 5 per cent may be difficult to improve on to any substantial degree.

Doctors are human and therefore will always make some mistakes. An X-ray or scan diagnosis may require the human eye to pick up on a subtle anomaly. Brian Toft, professor of risk management at Coventry University, has identified a phenomenon called involuntary automaticity (IA), which might help to explain human error and why it persists.

IA is a cognitive mechanism that causes people to miss cues that are right before their eyes, despite double-checking protocols. Significantly, Toft says high workloads, stringent time pressures and staff shortages are all causes of IA.

Despite advances in medical technology, there is currently no medical test with 100 per cent accuracy. False negative results will continue to contribute to diagnostic inaccuracy. And patients are not mechanised vehicles with the same number of parts of the same length and width put together with metronomic efficiency. So diagnosis can never be a linear process based on scientific certainty.

The IOM report calls for work systems and a culture that supports the diagnostic process and one that encourages learning from error and near misses. But to expect to reduce diagnostic error to zero is unreasonable. To paraphrase Donald Rumsfeld, former US secretary of State, there will always be known unknowns in the diagnostic process.

mhouston@irishtimes.com muirishouston.com