ONE of the defining characteristics of modern medicine is the meteoric rise of diagnostic testing.

Diagnostic testing has become an indispensible tool in assisting with accurate diagnosis and supporting life-changing treatments and interventions. One only has to consider the role of computed tomography, angiogram and magnetic resonance imaging in the revolution in stroke management that has occurred over the past decade, or the new frontiers of medicine that genomic sequencing is opening up.

On the other hand, there lurks a dark side to this apparent abundance of data and knowledge. Many clinicians and health economists have pointed to the unsustainability of current health care models and alarm bells should be ringing about wasteful practice. We know from Runciman and colleagues that guidelines espousing best practice are not consistently followed by clinicians. Pathologists have warned that studies of clinical algorithms have found instances where 20–25% of frequently ordered tests were inappropriately requested. Doctors and consumers need to be an integral part of solving this problem so that precious resources can be directed to the therapies that actually make a difference to patients’ lives.

There is a groundswell of change afoot, with Choosing Wisely now a global movement engaging with peak medical societies, colleges and associations to create a powerful community of practice. Their mandate is to challenge the way we think about health care, questioning the notion that “more is always better”.

But talk is easy and creating change in behaviour is a difficult nut to crack.

Important as it is, the economic imperative is not the most important issue in this discussion; it is about what is right for patient care. Unnecessary tests harm patients in many ways: the needles, the radiation, time wasted, worry while awaiting a result, and further worry and investigation after a false positive result or incidental finding. For patients, these issues are rarely spoken of. For clinicians, unnecessary tests only serve to distract attention and contribute noise, chaos and cognitive overload in an already pressured working environment.

With executive support, we tackled the problem at Eastern Health, one of Melbourne’s large public health services, by creating No Unnecessary Tests (NUTs), a program designed to support clinicians’ test requesting such that patients undergo tests that are clinically indicated and do not receive tests that are unnecessary, where evidence does not support them or where they can be deferred to the ambulatory setting.

Although we believe that efficiencies are highly likely to be made, our patient-focused approach means that we are open to the possibility that one day we might examine an area where more testing is indicated, not less. The important question turns out to be: is the test needed in order to advance patient care? We have been very clear that this is not solely about revenue, but that if one applies good clinical reasoning, the rest will follow.

Starting 5 years ago, NUTs was born of emergency department (ED) coalface clinicians becoming increasingly frustrated with blatant wasteful practices they could see every day. Confronted with the vast numbers of normal imaging and microbiology results that required daily checking, it became apparent that a systematic way of examining the issue was needed. With 40% of the health service test requests originating in the ED, this seemed a logical place to begin.

To start this work, we selected a group of tests that we felt were often ordered with little clinical indication: several commonly ordered blood tests such as venous blood gases and coagulation studies were chosen alongside some tests where evidence-based guidelines existed that were not followed consistently: urinary microscopy and investigation of suspected pulmonary embolism. We then performed local audits, examined available literature and obtained best-practice advice from relevant senior clinicians regarding these tests.

Clinical guidelines cover only a few specific conditions. Most test decisions rely instead on clinical acumen, which is subject to knowledge, skill, experience and bias. Clinicians are prey to a multitude of influences and our next task was to discover what was causing this overordering. Multidisciplinary focus groups uncovered many drivers: fear related to missing a diagnosis, looking foolish on the ward round, having an adverse event, getting sued, clinical uncertainty; pressure related to inpatient unit requesting extra tests, time pressure and access targets; bias, influence from prior clinical rotations, previous teaching and experience; ignorance of cost, consequence, evidence base or guideline; and habit (identified by older clinicians), inertia in changing set patterns of thinking and behaviour.

Based on this focus group feedback, and our more recent use of a formal cognitive domains framework (Table 1), we selected interventions that best matched the identified causes. The interventions that proved to be most effective in our setting were the use of decision support at the point of care, a weekly program of punchy peer-to-peer teaching modules and a global incorporation of the NUTs program into key departmental meetings.

Table 1. Overordering issues and interventions

Cause of overordering issues Intervention
Knowledge §  Inclusion in orientation

§  Test of the month focus

§  Harness existing forums to get the messages across

§  Education modules developed: 5 minutes, designed to be delivered weekly in routine teaching slots

Professional identity §  Involve staff in NUTs projects: clinician as teacher/researcher

§  Nurture champions

§  End-user ownership of the program

§  Use of the buddy program to reinforce messages

§  Debiasing at the start of the ED rotation

§  ED inpatient interface meetings

Beliefs and capability §  Promote teamwork: joint patient assessments and discussion of test strategy

§  Promote awareness of factors affecting the patient journey – unnecessary tests slow the journey for everyone

§  Promote the pathology grid to HMOs and nurses

Motivation §  Provide feedback on results

§  Provide incentives

§  Use of peer-to-peer teaching

§  Fun, engaging and gamification (eg, meme competition)

Memory/attention/decisions §  Booklet of common test indications (expert consensus) and why

§  Pathology testing decision grid for junior staff

§  Pathways (eg, PE investigation, urine culture testing flowchart)

§  EMR senior doctor sign-off for designated tests (eg, VBG, D-dimer, Ca/Mg/P, coagulation)

§  Minimalist approach to care set development in EMR

ED = emergency department. EMR = electronic medical record. HMOs = hospital medical officers. NUTs = No Unnecessary Tests program. PE = pulmonary embolism. VBG = venous blood gas.

In the first year of the NUTs program, we saw a reduction in all targeted ED tests of 30–50%, with no adverse event related to test omission or a recordable increase in downstream inpatient ordering.

Since then, we have added additional tests to our repertoire, while also focusing heavily on sustainability strategies to ensure these changes were not eroded by the constant flow of new ED clinical staff.

With Better Care Victoria funding in 2017, the NUTS program was expanded to the other EDs of Eastern Health, along with general medicine and intensive care. In each area, clinicians identified where the likely hot spots of unnecessary testing were and we adapted our interventions to fit. In general medicine, an important focus was on the repeat intervals of frequent “routine” tests, while in intensive care units a clinical pathway was introduced to standardise the use of chest x-rays.

Our next challenge is to expand NUTs to the other clinical programs within Eastern Health and to increase our collaborative work with other health services. One of our major sources of inspiration and optimism comes from working with medical students, whose “millennial” approach to the issue of wise stewardship gives us hope for continued innovation into the future.

Deb Leach is clinical director of the ED at Box Hill Hospital, an acute campus of Eastern Health, one of the largest health services in Melbourne. Eastern Health is affiliated with Monash and Deakin Universities and has a clinical school for undergraduate medical training.

Dr Paul Buntine is a full-time emergency physician working at Box Hill Hospital in Melbourne. In 2015, he completed a 6-month sabbatical period focusing on rational investigation ordering within the hospital setting, and he is currently the Eastern Health NUTs project lead. In addition to his clinical work, he also has a strong interest in evidence-based research and ED workflow practices.


To find a doctor, or a job, to use GP Desktop and Doctors Health, book and track your CPD, and buy textbooks and guidelines, visit doctorportal.



4 thoughts on “Drowning in data: the argument for no unnecessary tests

  1. Linda Mann says:

    NO, the real challenge is to spread this across the red brick silo divide, and include primary care in the results sharing. If you really want to facilitate the patient’s jouiney, remember they are still travelling when they leave your institution, Let me , the GP , see your test results, like the NZ GPs in Christchurch can. Allow my results , which you can access via My Health record, to be part of your diagnostic thinking.

  2. Anonymous says:

    I think the prime driver for ordering unnecessary tests, is the fear of being sued for ‘medical negligence’. Clinicians are more often covering that remote possibility of a missed diagnosis. This cost is the price we unfortunately have to pay in this age of Litigany. Once a clinician has been ordered to present to the coroner, you soon learn that no test ordered is too expensive, if fact , it is the opposite. We have to cover for all possible outcomes, no matter how unlikely.

  3. Brett Lidbury says:

    Thanks for the terrific commentary on this very modern, digital-age problem. Please know that some of us in the medical research community are very keen consolidate patterns within mass health data to ultimately simplify the information overload, as well as improve predictive power. The Quality Use of Pathology Programme (QUPP – Commonwealth Department of Health) fortunately provides funding to assist such investigations in the diagnostic pathology context.

  4. Sue Ieraci says:

    Thanks to the authors for this important piece of work. They are spot-on about what is driving over-testing, particularly in hospitals: “fear related to missing a diagnosis, looking foolish on the ward round, having an adverse event, getting sued, clinical uncertainty; pressure related to inpatient unit requesting extra tests, time pressure and access targets; bias, influence from prior clinical rotations, previous teaching and experience; ignorance of cost, consequence, evidence base or guideline; and habit (identified by older clinicians), inertia in changing set patterns of thinking and behaviour.” To this, I would add the current push to accelerate throughout, and the mistaken assumption that “doing all the tests” on arrival, prior to medical assessment, “saves time”. For many patients, it takes less time to do a detailed history, focused examination and detailed explanation to the patient than to wait for test results.

    The real challenge is to gain a wise understanding of what patients require from acute medical consultations, and how to influence expectations. Physiologically well people do not need an immediate diagnosis – especially considering the fact that clinical conditions develop over time.

    Better communication between the ED and both inpatient units and community providers works both ways. EDs must not just “rule out” rare but catastrophic diagnoses, but must also help the patient understand the realistic risks, and develop a plan forward, should the symptoms persist. At the same time, inpatient units and community providers must not expect an ED assessment to exclude all patient risks, no matter how small, and think about a “deferred” diagnosis rather than a “missed” one.

    A diagnosis that is made some time after ED consultation, according to an agreed plan for follow-up, is not a “missed” diagnosis. Waiting a day or a week to see how the condition develops, in a physiologically well patient, is good medicine.

Leave a Reply

Your email address will not be published.