Are X-rays overused in Irish hospitals?

A study in Dublin’s Mater hospital found that of 1,124 radiographs, 784 were unjustified, based on the clinical information provided

On November 8th, 1895, Wilhelm Röntgen (1845-1923) discovered an unusual type of radiation.

He called it "X-ray", grasped its significance and news spread fast. For example, 1896 saw the first operation based on results of diagnostic X-ray images, when a woman with a needle embedded in her hand had the object removed at Queen's Hospital, Birmingham, by a surgeon who'd viewed a radiograph showing its exact position.

X-rays are a routine diagnostic aid . . . but are they overused?

A recent retrospective review of 1,124 radiographs taken at Dublin's Mater Misericordiae University Hospital concluded that "[A] significant amount of inappropriate radiographs continue to be requested and performed, exposing patients to needless ionizing radiation and wasting staff members' time at a financial cost".

READ MORE

The study, in the Irish Journal of Medical Science, was conducted by Dr James Ryan – specialist registrar in radiology – and colleagues.

Concern about overuse of X-rays was raised in 1971 when a doctor asked in the British Medical Journal why colleagues should "bow cowardly to direction by 'public opinion' and submit children to numerous unnecessary X-ray exposures?" But another responded that "many patients consider an X-ray examination to be a therapeutic right . . . and the pressures on an inexperienced casualty officer to order X-rays indiscriminately are difficult to resist".

What are X-rays?

Dr Ryan says that they are a type of electromagnetic radiation (EMR): “An example of low-energy EMR is sunlight, and an example of high-energy EMR is gamma rays that can be generated from radioactive material. X-rays are in between these two. Excessive radiation adversely affects living tissue. For instance, sunburn from ultra-violet radiation, or mutations from exposure to high-energy gamma rays.”

X-rays, Ryan adds, are used safely every day worldwide, allowing radiologists to diagnose diverse conditions by generating conventional radiographs and/or three-dimensional computed tomography (CT) images.

Ryan and colleagues’ aims were: first, to assess the appropriateness of ordinary X-ray referrals before and after the introduction of evidence-based iRefer guidelines from the UK’s Royal College of Radiologists in 2015; and second, to analyse both the cost and X-ray dosage associated with inappropriate referrals.

But introducing iRefer guidelines failed to improve the percentage of inappropriate referrals – 42 per cent pre-introduction and 43 per cent post-introduction – and “the time spent performing inappropriate abdominal and spinal X-rays in 2017 yielded an estimated cost of €8,036.40”.

Commenting on their results, Dr Ryan says: “Out of 1,124 X-rays we identified, 784 were unjustified, based on the clinical information provided. These imparted a median cumulative radiation dose of 65.1 milli Sieverts from all 784 inappropriate X-rays.”

Ryan explains that the individual dose received by each patient was a fraction of this cumulative dose and akin to a day or two of background radiation that we are all exposed to in daily life: “The dose imparted by a single X-ray,” he says, “is small and harmless. Someone having many CT and X-ray investigations over a long period incurs a theoretical risk of adverse effects. But patients being referred for X-rays, inappropriate or not, shouldn’t be concerned about the associated radiation dose.”

A HSE spokesperson says: “The HSE and the National Clinical Programme for Radiology [NCPR] support the use of iRefer guidelines to guide clinicians to request appropriate imaging. The NCPR . . . welcomes the findings of this study,” adding, “the HSE ensures members of the public are not exposed to ‘needless ionizing radiation’ by undertaking a range of measures.”

These measures include ensuring radiologists and radiographers receive radiation protection as part of their training, with the ALARA (as low as reasonably achievable) principle a core component; the establishment of the National Radiological Protection Committee to guide service providers through the new legislation underpinning Radiation Protection in Ireland; and subjecting individual hospitals to audit by the regulatory authority to ensure that processes involving ionizing radiation are safe and comply with the new legislation.

Constantly increasing demands

Ryan notes that diagnostic radiology is a busy discipline, with constantly increasing demands unmatched by a proportionate increase in staff. “In recent years,” he observes, “we have experienced a dramatic increase in the number of imaging referrals to the Mater Radiology Department. We performed roughly 180,000 imaging tests in 2012 and roughly 206,000 in 2016.”

This upward trajectory, Ryan explains, has continued in 2017 and 2018, coinciding with an ever-increasing waiting list for outpatient imaging: “The central message of this paper is the importance of evidence-based justification for imaging. We’ve demonstrated that a large proportion of the X-ray referrals in this study were unjustified when compared to evidence-based guidelines.”

While acknowledging that their study did not address CT and magnetic resonance imaging, which are more time-consuming and expensive modalities, he emphasises the need for interventions to stem inappropriate referrals, and to ensure that public imaging resources are used appropriately, is clear.

How might the deficiencies highlighted by this study be addressed?

Ryan sees a role for clinical decision support (CDS) software, which can be incorporated into the scan-ordering process and provides feedback to physicians as to whether their scan request is evidence-based: "This type of software has been utilised in the United States for over 20 years and in some cases has been associated with fewer inappropriate referrals. I believe it's worth trialling CDS software in public hospitals in Ireland to try and promote efficient use of resources."

This seems a suggestion that the HSE might usefully consider . . . perhaps sooner than later?