iRADiX

No doubt, most of our errors go unnoticed. It might be due to the fact that many clinicians do not know how to read films, or may not have access to the films. It is financially not feasible to do double blind reading of every film to make it as much fool proof as possible. Hence, we need to develop a method where we can identify the errors and learn from each others' mistakes.

1. Discrepancy meeting:

If the department has more than one radiologist, it is worth doing the exercise of discrepancy once a month. The key is keeping anonymity of the reporter, otherwise this may lead to blame culture. In a teaching hospital, the best person to take the lead of discrepancy meeting would be one of the senior residents or a final year post-graduate. All the cases should be collected by the designated person, and the cases should be shown to all the people in the meeting. 

This should not be treated like a quiz, or to test juniors or colleagues. The clinical information and all the films available at the time of reporting should be given to the audience.  Then discussion should be on the report and subsequent discrepancy on the reporting.

The most important bit is to learn from the mistake:
1. to identify why the mistake happened
2. to check if with further clinical information the findings would have picked up
3. to see if review of the previous images would have helped
4. to see if the previous report/ clinical information lead to the bias
(covered widely in my previous blog post)

Then we need to record the outcome by identifying the severity of the error and grade them. If the mistake/ error found in the meeting is going to change the management, then one should not hesitate to contact the concerned team and convey the findings.

Example:
A 45 year old man presented with increasing headache. CT scan was reported as bilateral old frontal lobe lacunar infarcts. The patient was referred to neurologist for TIA/ stroke evaluation as the patient was young for stroke/ TIA. Neurologist promptly asked for MRI of the brain. When the neuroradiologist was vetting (checking if the request is appropriate) the request form, he reviewed the CT images, and it was very much evident that the pattern of low attenuation in the both frontal lobes is typical of old frontal lobe contusions secondary to head injury. There was no need of MRI, review of the patient history suggested previous head injury. Hence need for starting life long aspirin was avoided. This case was discussed in discrepancy meeting, and the fellow general radiologists re-learnt the basics from a neuroradiologist.

2. Multidisciplinary team meeting/ clinical-radiological meeting:

These are great places for radiologists to learn the beats and pulses of referring clinicians. The radiologist with particular sub-specialty interest should review the images before meeting, and discuss them with the clinicians. The errors/ discrepancies should be conveyed back to the radiologists in the discrepancy meeting.

Multidisciplinary meetings typically occur for cancer patients, where the surgeons, oncologists, radiotherapists discuss the  diagnosis and management of the patient with radiologist and histopathologist. This is a great learning platform for the radiologists to pick up current trends in the management, and to throw more light in post-operative imaging. With further clinical output, the subtle findings which might have been over looked become important findings, or the same radiological abnormality may look like a different pathology. 

Clinical-radiological meetings are very educative, and at the same time help to pick the errors/ discrepancies and rectify them. Example: paediatrics-radiology meeting, neurology-radiology meeting, surgery-radiology meeting etc.

Example:
CT neck of 80 year old smoker was reported as residual tumour in the left laryngeal region. In multidisciplinary meeting, it was evident that the tumour was T1 (based on endoscopic findings and histopathology), and the surgeon has done laser quite recently. The appearances on the CT were not secondary to residual tumour, but from residual inflammation from recent laser. The reason why CT was asked was not to look for larynx, but to look for any nodal metastasis. The message was conveyed back in discrepancy meeting, and the lesson was learnt by all.

3. Subspecialty interest:

In a department of few/ many radiologists, it is worth developing one or two subspecialty interest, and start attending courses and conferences in that subspecialty. This will give a great depth to the department. It is not just about interventional radiology or neuroradiology. A department can have a general radiologist with sub-specialty interest in HRCT of the lungs.  The subspecialty radiologist will deal with the clinician's second referrals, and discuss the discrepancies.

Example:
75 year old smoker presented with breathlessness and weight-loss. CT thorax was reported a mass in the left lower lobe (possible T2/T3 tumor) with N0 M0. The radiologist with special interest in chest imaging was reviewing the images, and found that the mass was not a "mass" but an infarct, and pulmonary embolism in the left lower lobe artery was completely missed. The trauma of undergoing unnecessary biopsy was avoided. In discrepancy meeting, this was discussed, and again, basics are learnt again.

 (I hope you enjoyed the series of "to Err is Human", and i will come up another topic in my next blog. Till then, happy reporting!)

Reference:
http://www.rcr.ac.uk/publications.aspx?PageID=310&PublicationID=258