Wednesday, December 20, 2006
A few interesting articles from December's Journal of Pain and Symptom Management.
I'm going to be travelling for the next few days and likely won't post again until next week. For those of you who celebrate a winter holiday...Happy Holidays.
First is a randomized trial of oxygen for dyspnea in advanced cancer patients. It looked at 50 advanced cancer patients in an Australian cancer center (both in- and outpatients); all had dysnpea due to their cancer (although this was determined quite subjectively). Median age 65; most had ECOG performance statuses of 3. 17 were hypoxic. They received either oxygen or room air delivered at 4L/minute by nasal canula in a randomized, double-blinded fashion. They received this for 15 minutes, had a 30 minute wash-out, then crossed over to the other treatment. Basically both groups' dyspnea improved equally. The hypoxic patients (14 of whom became normoxic with oxygen) also improved equally with air or oxygen. The authors conclude that supplemental oxygen does not per se improve dyspnea in these patients. While I'm sympathetic to these findings, and appreciate all reminders of the complicated, pleomorphic nature of dyspnea, this conclusion needs to be taken with a grain of salt. This is not a mandate to rip off nasal prongs, as patients receiving air through them improved (just no less than with oxygen). In addition, while the study was powered adequately to determine a difference between oxygen and air for the entire 50, it was not powered adequately to determine a difference in just the hypoxic patients, so we can't conclude oxygen and air are equivalent in these patients, and giving patients a trial of normoxia seems very justified! This study however highlights how much more is going on with dyspnea than simply a lack of supplemental oxygen....
Next is a preliminary study of l-carnitine for cancer fatigue . This was a preliminary study that I'm not going to spend much time on other than to note it: cancer-related fatigue is widespread, poorly understood, and poorly-treated (insofar as we don't know how to treat it). This study looked at 27 advanced cancer patients with low serum l-carnitine levels and moderate to severe fatigue. L-carnitine supplementation wasn't toxic, some patients seemed to improve on it. One assumes we'll be seeing larger, placebo-controlled trials.
Finally is a another study on opioid use and mortality in hospice patients. It is a retrospective study looking at ~700 hospice patients, opioid use and dose changes, and death. Essentially, the authors looked at the amount of opioid use and the amount and timing of opioid dose changes, divided the subjects into different groups based on these factors, and tried to find any differences among them regarding time to death. Overall, they didn't. In their multivariate analysis, final opioid dose was associated with time to death (higher opioid dose = shorter time to death), but timing of final dose change and the percent amount of the final dose change were not. In addition the magnitude of the final opioid dose explained only 6% of the variance in survival (meaning that it didn't strongly predict shortened death). In their sample, the median time to death from the last opioid dose change was ~12 days, again suggesting that--in this population as a whole--opioid dose changes didn't lead to death (12 days is long enough that the effect of the last opioid dose change likely had little to do with the timing of death).
It should be noted that there was a lot of missing data in this study (the data wasn't initially collected for the study), limiting its conclusions. That being said, this adds to the slowly accumulating empiric evidence that opioid use near death just isn't a big factor in hastening death (all the studies have been poor quality, but all have had similar conclusions, meaning that if opioids are hastening death it is not on a magnitude large enough to be picked up by these studies). Because of this I officially call for a moratorium on the use of "morphine in the dying patient" as an example of the principle of double effect in the medical education literature. While I'm fine with the PoDE, using this as an example subtly promotes the idea that we are killing people (hastening death) by providing comfort for them. On the other hand, it must always be held clear that these are population based studies, and individually, there probably are some patients whose deaths are hastened directly due to the use of opioids, but these are rare enough not to be picked up. When I have family members ask me about this I usually reassure them that if opioids do hasten death it's likely on the order of hours or a day or two, but no longer.