That looks like alot of effort, you put in to this discussion.

I only look at studies when I have no personal experience in an area, or when an area is "new" or rare.

For 40 yrs I've dealt with patients who complain or worry about their medications. Who have negative experiences, or who are not getting what they expected, etc.

Prescribing data, and data off the drug insert is typically taken off short lived studies. Post operative pain, is rather unique and short lived for most patients. Also it can be subject to fraud as we learned about Pfizer and Celebrex, and Dr. Reuben:
http://www.theday.com/article/201001...100119833/1047
I tend to think these post operative studies are drug company funded to give them leverage with sales/marketing etc.

Insert reports do not often include post marketing data, unless it is extreme and dangerous, and getting that added may take a decade or more in some cases. Nausea? not like aplastic anemia or metabolic acidosis. The latter two might get added more quickly.

Basically when people get discharged home, with meds, and it is late at night and they cannot call their doctors, etc, they call me. I've had many many opiate side effects including Sphincter of Oddi attacks which are alarming in the extreme.

But nausea, is a biggie and more common with codeine, and morphine. Oxycodone and hydrocodone may cause nausea in high doses, and also severe itching spells at moderate doses.

But I never had an Ultram/tramadol call! These would be emergency calls. Over the long haul people may call their doctors the next day and ask for another medicine. But I did not see many discharged surgical patients on Ultram at all in fact. It is more of a chronic long term intervention. People coming to me with trauma and other painful things in the night, always had Vicodin at least. Maybe Percocet. Sometimes Darvocet, which is no longer available finally.
Many people with trauma are given Toradol (an NSAID) injection in the ER or hospital, to augment the narcotic given. Ultram? Seems like a foolish choice for acute or severe issues.

I forgot to list Hydrocodone +Tylenol = Lortab and Vicodin, Norco etc.
In this case, I believe the tylenol was added to prevent abuse of the hydrocodone portion which when used alone is a CII narcotic.

The Tylenol also functions in the other mixes as an abuse deterent but for determined abusers really isn't effective and just becomes a poison to the liver instead.

Let's look at Ultracet:
This drug is 37.5mg of tramadol and 325mg of Tylenol. The typical starting dose of tramadol plain is 50mg. But many people use 100.

Why 37.5mg tramadol? This is a trick used by drug companies so the doctor cannot give regular 50mg tramadol and tell the person to take a tylenol with it. It forces the dosing to use TWO tablets to give 75mg of tramadol to get a therapeutic dose + 650mg of Tylenol. TWO tablets for a response that maybe one tablet of plain tramadol 50mg would accomplish for many situations.

This is a frequent thing, drug companies do to sell more product.
In the case of tramadol, it also protects the patent, because any generic company would be forced to spend $$ to provide studies for the new 37.5mg dose.

Here is another example of manipulation of doses, with a well known brand name: Soma:
http://www.soma250.com/
This product really bombed because the insurance companies refused to pay for it. Generic Soma 350mg has been on the market almost for 20 yrs! I read about the launch of this bomb on Cafepharma.com and it was very amusing and enlightening!

So while I don't dispute the PubMed articles you found, empirically, they don't often reflect real life situations.
Today, unfortunately, if a drug company does something we need to look very carefully at the result, because often it is for THEIR edification, and not for the patient at all. I keep hoping we are turning a corner in this regard, but I have yet to see much improvement!