Opioid Limits: Means, Medians or Madness?

By Roger Chriss, Columnist

CVS recently announced it would impose a 7-day limit on opioid prescriptions for short-term acute pain for customers in its pharmacy benefit management program.  A pharmaceutical industry trade group also supports a 7-day limit and so does the U.S. Pain Foundation, a patient advocacy group.

Maine, New Jersey, Massachusetts and other states are also limiting prescriptions to a week or less, justifying the time frame by saying that’s what patients need on average.

But this represents a misunderstanding of how statistics work and ignores emerging research about opioid analgesia in the world of acute pain care.

In statistics, we have three values of fundamental importance: the mean, median, and variance.

The mean, also known as the “arithmetic mean,” is the sum of a collection of numbers divided by the number of numbers in the collection. The average height or weight of a group of people is the mean.

The median is the “middle value” of a data set that is ordered from lowest to highest. The mean is found in “median income” or “median price of a new home.” Importantly, the mean and median are not necessarily the same. In the set of numbers 2, 3, 3, 5, 7, 17 and 313, the median value is 5, but the mean value is 50.

The variance is the tendency of a set of numbers to cluster around the mean, or how spread out the numbers are. We know from experience that the height of adults is closely clustered around average height: Most people are over five feet tall and under seven feet tall. No one is 2 inches or 20 feet tall. But variance can also be significant, as is the case with annual income, home prices or family size.


The significance of these three values cannot be understated. In his essay “The Mean Isn’t the Message,” biologist Stephan J. Gould explains that a “median mortality of eight months” does not mean that a person will probably be dead in eight months. Some people, including Gould himself, live many times more than the median survival time for a disease. 

So when talking about opioid analgesia for acute pain, we cannot rely on just an “average” value. Physicians know this, but legislators, corporations and even some patient advocates do not seem to.

JAMA Surgery recently reported on the post-surgical acute pain needs of over 200,000 patients who had one of eight common surgical procedures. The results showed median values from 4 days for an appendectomy or gallbladder surgery to 7 days for a discectomy.  

The authors then used these values and the variance to calculate the range of time a patient would typically need opioids for acute pain after surgery:

  • 4 to 9 days for general surgery procedures
  • 4 to 13 days for women's health procedures
  • 6 to 15 days for musculoskeletal procedures

In other words, there is substantial variance, with the optimal length extending to as much as two weeks. And there is no way to know ahead of time where in this range an individual will fall.

To address this uncertainty, the Opioid Prescribing Recommendations for Surgery were developed at the University of Michigan. They list the recommended numbers of tablets of hydrocodone, codeine, tramadol, or oxycodone for a range of common surgical procedures, including laparoscopic cholecystectomy, open colectomy, and several types of biopsy.

This resource gives amounts that “represent the actual maximum opioid use reported by three-quarters of actual surgery patients.” Those amounts range from 10 to 40 pills, depending on the procedure, noting that “prescribers are encouraged to use their best judgment.”

The Opioid Prescribing Recommendations for Surgery also advise recovering unused pills to reduce the risk of diversion, which is a much more sensible policy than forcing people recovering from trauma or surgery to seek refills if they happen not to fit a mandated average.

In sum, the medical profession is offering evidence-based recommendations for pain management that include not just a simplistic mean, but the real-world variance found in individuals. This approach is likely to provide better results than blanket policies geared toward a statistical mean that does not capture vital features of medical care.

Roger Chriss.jpg

Roger Chriss lives with Ehlers Danlos syndrome and is a proud member of the Ehlers-Danlos Society.

Roger is a technical consultant in Washington state, where he specializes in mathematics and research.

The information in this column should not be considered as professional medical advice, diagnosis or treatment. It is for informational purposes only and represents the author’s opinions alone. It does not inherently express or reflect the views, opinions and/or positions of Pain News Network.