Over prescribing — high risk for the elderly

hugsnotdrugs

Elderly patients are taking too many drugs.  And, in most cases the drugs can be stopped.  This surprising idea was highlighted by Dr. Ezekiel J. Emanuel in his New York Times article 11/22/15.  He backed up his statements with a study from 2010 (JAMA).  The figures that follow are from that study.

The first figure is the logic diagram for stopping medications in elderly patients who have several chronic diseases.  The second figure is the list of medications that were stopped in the study.  Only 2% of the stopped medications eventually had to be restarted.  The cost savings alone is huge and the reduction in the risk of side effects is likewise tremendous.  “Less is more”.

Health care providers often prescribe drugs intended for younger healthier patients to older sicker patients.  Often the drug testing originally done to approve drugs excludes older sicker patients due to the risks of side effects.  A drug might add many years of life to a 50 year old but that’s not likely for an 80 year old with multiple problems.   In fact, over medicating elderly patients may hasten death — this has been proven in many studies.

Dr Ezekiel suggests asking the following questions to the health care providers who prescribe medications, tests or treatments to elderly patients (or perhaps any patients):

  1. What difference will it make?
  2. How much improvement is expected?
  3. How likely and severe are the side effects?

Evaluate the answers carefully, if the answers are: small difference, not much improvement, and fairly likely side effects then perhaps that drug, test or treatment is not needed.

Dr. Emanuel favors the advice of teaching hospital doctors like him, but it’s just not practical or necessary for patients to get a university consultation.  Stopping medications is not rocket science.  As the graphic says, it may be better for many elderly patients to get “hugs not drugs“.

 

Figure 1

drugstop

 

 

Figure 2

medsstopped

, , , , ,

%d bloggers like this: