[EXTERNAL][IMPROVEDX] Balancing misclassification costs & numerators\denominators

Robert Bell rmsbell at ESEDONA.NET
Wed Dec 18 15:17:38 UTC 2013


FOR POSTING TO THE LIST

Some truth, but the probabilities are related to the cost of a work up. With limited resources one needs to conserve.

So indirectly, probabilities are firmly bound to diagnoses.

Rob Bell

Sent from my iPad

On Dec 18, 2013, at 7:06 AM, Bimal Jain <bjain at PARTNERS.ORG> wrote:

> I find the discussion about role of probabilities in diagnosis, for example, of acute cardiac ischemia fascinating. I believe, probabilities have little or no role in clinical diagnosis, as our goal is to diagnose a disease correctly in a particular, individual patient. In such a patient, a probability, whether high or low is not evidence for presence or absence of disease. A probability only informs us about distribution of disease in a series of patients but tells us nothing about disease in a particular patient. A clinical presentation should be employed, I suggest, only to suspect a disease, regardless of whether the prior probability is low or high. The suspected disease should then be evaluated by tests which yield highly informative results. For example, acute Q wave and ST elevation changes in EkG alone should lead to definitive diagnosis of acute myocardial infarction, regardless of whether its prior probability is high or low. I believe, basing clinical diagnosis on probabilities is a major source of diagnostic error which can be avoided by recognising their non-role in diagnosis.
> 
> Bimal P Jain, MD
> Pulmonary-Critical Care
> North Shore Medical Center (Union)
> Lynn, MA 01904
> 
> 
> 
> 
> 
> 
> Moderator: Lorri Zipperer Lorri at ZPM1.com, Communication co-chair, Society for Improving Diagnosis in Medicine
> 
> To unsubscribe from the IMPROVEDX list, click the following link:<br>
> <a href="http://list.improvediagnosis.org/scripts/wa-IMPDIAG.exe?SUBED1=IMPROVEDX&A=1" target="_blank">http://list.improvediagnosis.org/scripts/wa-IMPDIAG.exe?SUBED1=IMPROVEDX&A=1</a>
> </p>
> 










More information about the Test mailing list