Radiological Errors/ Discrepancies -- Taking stock

Salient  points across the  countries ,  literature, studies, modalities and others: By Dr MGK Murthy

1. Approximately 1 billion  radiological examinations are performed annually across the world 

2. An average real time errors  in daily practice  is  estimated at 3-5% (i.e.40 million  discrepancies in an year) 

3. 75%of Radiology malpractice claims  pertain to diagnostic errors.( Diagnostic errors in US hospitals contribute 40-80 000  deaths  per year, apart from many more non lethal incidents).

4. In view of the  prevalent proactive societal actions,  most radiologists prefer   false +ve reading from safety point of view . 

5. Diagnostic errors = missed/ wrong/ delayed  as detected by  some subsequent  definitive test/ finding by the same test .  

Radiology errors  =   failure to detect/ interpret/ communicate the results / suggest an appropriate followup test 

6. Opinion  =   A view held  about a particular subject or point ; A judgement formed  OR  belief ( Terminology of  "Error" is being actively  replaced with" Discrepancy", a more  appropriate term )

7. Some classify  discrepancies in to 2 varieties  (a)  Cognitive (failure to notice) (74%) (b) System failure (processes/ Equipment/Team failure etc)(60%) , or at  times both 

 where as some others group them  as 

(a) cognitive (failure to detect/ notice ) (20-40%) (b) Perceptive (60-80%) (picked up retrospectively and debated  in hind sight bias etc) 

8. Studies involving modalities  suggest   X-rays leading  in the  discrepancies  (54%) / CTs (30.5%)/MRIs (11.4%). Ironically not much difference in discrepancies  exist  between with (77%)  and without clinical data (80%)

9.   MGH  study involving CT Abdomen gave 26%  Inter observer discrepancies   Vs 32% intra observer discrepancies ( Blinding up the scans , prior reported by few of the participants) .  Oncological CT  studies  suggested discrepancies of 21-30%.

10. Structured formatting  reporting as suggested remedy by few  , surprisingly has not been preferred by clinicians  in the studies (who  are comfortable with  free style conventional dictation method  of organ based paragraphs )

12. Suggested remedies to reduce discrepancies  are pretty standard  like  clinico- Radiology or clinic- pathology meets/ Peer review/ information technology  enabled tools/ Computer  Aided Diagnosis / limiting workload (more than 20 cases of CT per day resulted in more discrepancies)  / Education etc 

13. It is  a misnomer to think Experts(in a field)  do not make  mistakes. "An expert is someone who is 50 miles from home, and has no responsibility for implementing the advice  he gives and shows slides "= Ed Meese , US Attorney General 1985-88

14. "Finally Errors of Judgement must occur  in the  practice of an art , which  consists largely in balancing probabilities "  and  all effort will continue towards this  hypothetical Zero terminus 

.(presuming it  exists)