1.
Imaging referral guidelines
The JRS and JCR jointly developed the imaging referral guidelines for diagnostic imaging in 2013 which covered 152 clinical questions (CQs),
9 areas of imaging techniques,
and 7 reviews (Jpn J Radiol.
2016 Jan;34(1):43-79). During the 2016 revision,
the CQs increased to 171.
The revised guideline was written by a total of 176 radiologists and was externally validated by experts in other academic societies.
The recommendation grades in the guideline were determined according to the Minds Guide for developing Clinical Practice Guidelines 2007 (Table 1).
The scientific evidence level was mainly considered but also the number of studies,
variation in conclusion,
magnitude of clinical efficacy,
clinical applicability (ability of physicians,
regional characteristics,
medical resources,
medical insurance system),
and harms and cost were incorporated.
Table 1 MINDS recommendation grades (from The Japanese
Imaging Guideline 2013)
Recommendation grade |
Description |
A |
There is firm scientific evidence strongly recommending the procedure |
B |
There is scientific evidence recommending the procedure |
C1 |
There is no scientific evidence,
but the procedure is recommended
|
C2 |
There is no scientific evidence,
and the procedure is disrecommended.
|
D |
There is scientific evidence suggesting ineffectiveness or harm,
and the procedure is disrecommended
|
Among a total of 171 CQs in the 2016 guideline,
37 exams were graded as C2 or D which means the procedure is disrecommended.
2.
Surveying the current situation and possible priority areas
After the publication of the Japanese imaging referral guideline 2016,
we conducted an online survey in January 2017 under the initiative of the JRS Appropriate Utilization of Imaging Committee.
The survey targeted the radiologists' perspective on how appropriately imaging is performed in Japan,
in terms of "indication".
We selected 189 hospitals that had been accredited as training institutions by the JRS/JCR,
and sent the URL of the survey to the radiologist in charge at each hospital.
All the hospitals had at least one board-certified radiologist.
The first portion of the questionnaire asked the radiologist what their subjective assessment was on the amount of imaging exams that are NOT recommended by the guidelines at their hospitals.
Also included were the questions about how the radiologists intervened in the imaging ordering process to achieve more appropriate imaging utilization.
The second portion of the questionnaire focused on the hospital's structural information (e.g.,
the number of full-time radiologists,
scanners,
CT/MR exams).
The response rate was 87.3% (165/189 hospitals),
and all questionnaires were completed by board-certified radiologists; for 85% (140/165) of the radiologists that answered the questionnaire,
more than 20 years had passed after medical school graduation.
The mean number of full-time radiologists,
CT exams/month,
MR exams/month at the responding hospitals were 8±5,
2569±909,
and 961±449,
respectively.
Among all the exams disrecommended by the guidelines,
the percentage of "Very frequently performed" and "Frequently performed" were higher in the head exams and breast exams (Fig. 2),
compared to other body parts.
In 21% of the hospitals (34/165),
radiologists rarely contacted the referring physicians when disrecommended exams are ordered.
The primary reason (32% of the all reasons) for the no/rare intervention by radiologists was that "it is difficult for radiologists to give advice on the imaging indication or appropriateness when referring physicians considered the exam to be necessary".
The second most frequent reason (30% of the all reasons) was "there was no time to intervene".
At these 34 hospitals,
median number of CT and MR exams that a full-time radiologist needed to interpret in a half day was significantly higher than that at other hospitals (median 98 vs 79.5 exams,
p=0.023).
There was no significant correlation between the hospital structural factors and the percentage of "Very frequent performed" and "Frequent performed" discrecommended exams.
3.
Institutional-level trial of the clinical decision support
Although the national-level CDS has not been developed yet,
an institution-level trial is being performed at some hospitals.Fig. 3 is an example of the CDS for venous thromboembolism (VTE) CT used at the University of Tokyo hospital,
an academic hospital in Tokyo.
VTE is common in Japan, as is the case with other countries (J Thromb Haemost.
2007;5: 692–9), and should be diagnosed appropriately because delayed diagnosis or misdiagnosis can potentially be lethal.
In Japan,
VTE is mainly diagnosed by VTE-CT (pulmonary CT angiography plus CT venography) because of limited accessibility to doppler ultrasonography.
Defensive medicine and lower CT cost also accelerate the utilization of VTE-CT.
At the University of Tokyo hospital,
VTE-CT is frequently ordered and performed when d-dimer testing is positive without careful assessment of other VTE signs (e.g.,
Well's score).
Positive d-dimer itself is not quite helpful to stratify the risk of VTE,
especially for patients with a low pretest probability of PE (Ann Intern Med. 2015;163(9):701-711), and thus the radiologists plus experts in vascular medicine at the hospital were concerned that utilization of VTE-CT is not optimized.
The University of Tokyo hospital then introduced the trial system that all referring physicians who order VTE-CT exam are required to fill the questionnaire embedded in the EMR (Fig. 3). The questionnaire includes questions about VTE risks.
Submission of the questionnaire before the VTE-CT acquisition is preferred,
but is not mandated in this trial; if the situation is urgent and the referring physicians needs VTE-CT as soon as possible,
they can order it anyway but are required to complete the questionnaire within 2 days of the VTE-CT acquisition.
Moreover,
VTE-CT is performed irrespective of their response to the questionnaire (e.g.,
even if the referring physician checks NO to all VTE risks,
CT is still performed).
In this trial system,
it is expected that the questionnaire leads to thorough physical exam before ordering VTE-CT,
and gives referring physicians a chance to reflect on their VTE-CT utilization.
More aggressive feedback,
e.g,
sending of the statistics of the responses to the referring physicians,
may be needed for more effective decision support.