Post-op surgery, opiates and subsequent misuse

​​A recent large database mining study found that future “opioid misuse” was more related to the length of opioid prescriptions after surgery and not so much by the dose prescribed (See doi: 10.1136/bmj.j5790).

Details:

--surgical claims from 1,015,116 opioid naïve patients undergoing surgery from 2008-2016 (opioid naïve=total opioid use in 60 days prior to surgery was <7days), from the Aetna administrative database

--568,612 (56%) received postoperative opioids (90% of scripts were filled within 3 days)

--patients followed median of 2.67 years

Results:

--5906 developed opioid misuse (0.6%, rate 183/100K person-yrs).  Opioid misuse= ICD diagnostic codes for opioid dependence, abuse or overdose

--though surgeries were more common in older people, opioid misuse was more common in younger ones

--surgeries:

   -- musculoskeletal problems (367,317 surgeries, 2448 misuse events at rate of 206/100K person-yrs)

   -- digestive problems (293,905 surgeries, 1825 misuse events at rate of 198/100K person-yrs)

   -- integumentary problems (106,914 surgeries, 533 misuse events at rate of 161/100K person-yrs)

--though opioid scripts filled plateaued from 2010 to 2013 and then fell off in 2014, opioid misuse increased: 183/100K person-yrs in 2009, up to 269/100K person-yrs in 2014

--median duration and dose remained stable over the study (5 days, and 50 MME/d = morphine milligram equivalent/d), BUT this actually reflected fewer short courses and more longer duration scripts, and a trend toward lower doses/refill

--no difference in these trends by type of surgery

--the number of post-discharge opioid scripts was the best predictor of eventual misuse: each additional opioid refill was associated with an adjusted 44% increased rate of misuse (40.8-47.2%, p<0.001), from 145/100K person-yrs to 293/100K person-yrs

    --and, each additional week of opioids was associated with an adjusted 19.9% increase (18.5-2.4%, p<0.001)

    --BUT, the dosage of opiates prescribed was a much weaker predictor of subsequent misuse: 0.8% increase per additional 10 MME/d of opiates; even those on >150 MME/d had only mild increases in misuse rates, eg:

        --if duration of opiates was <2 weeks, there were similar rates of misuse if on 40-50 MME/d compared to 100-150 MME/d

        --if duration >9 weeks, higher doses dramatically increased misuse (if <20 MME,  476/100K person-yrs vs if 50-60 MME/d, 2398/100K person-yrs vs if >150 MME/d,  5689/100K person-yrs)

Commentary:

--some statistics:

    --though we are (rightfully) very concerned about opiate-related deaths, 7-11 times more patients have non-fatal overdoses

    --2/3 of opiate misuse is from opiates prescribed from a single source (ie, most are not from patients shopping around to multiple providers/ERs)

    ​--post-surgical opiate prescribing: 3-10% of opiate-naive patients become chronic users; up to 80% of prescribed opiate pills are unused in the remaining group [which speaks somewhat to how hard it is to gauge the appropriate dose, for example an article in Lancet on pain found that only <20% of patients who develop persistent post-op pain could be predicted by the severity of the immediate post-op pain (other risk factors seem important: older age, female sex, obesity, depression), see Wu CK. Lancet 2011; 377: 2215]

--overall, and likely related to increased awareness of adverse medical/social effects of prescribed opiates, the number of post-surgical prescriptions for opiates dropped pretty dramatically from 2013 to 2014 (around 65% to about 50%), though the misuse rate per 100K person-yrs increased from about 200 to about 260.  There have been accompanying small changes in opiate doses prescribed, in the 3-18 MME/d range, though not much change in mean duration of use

--what could explain this finding that giving higher doses but for shorter periods after surgery lead to less misuse? one suggestion floated is that inadequate opiate doses directly after surgery may inadequately saturate mu-receptors and may lead to requests for more opiates for longer term and risk of future dependence, overdoses, etc.

--there are several caveats to this study. This was a large database study from a private insurer, raising several potential issues: the data were not granular (no specifics on the individual patients, how extensive the surgery was/extent of post-op pain/surgical techniques used), the analysis relied on billing codes which might not have been complete or accurate, the patients involved were privately insured (overall, VA patients and those on Medicaid/Medicare have higher rates of opioid misuse, perhaps related to their being a sicker population/more comorbidities/different supports/different socioeconomic conditions), and there may well have been more “overcoding” for opioid misuse over the time period involved as clinicians have become more attuned to this as a problem. Also, patients who had a higher likelihood of opioid misuse might have requested more opiates (differences in pain thresholds, mu receptor activity, etc). also, these data reflect only prescription opioids (the increase in opioid misuse over time, despite decreasing prescriptions, could reflect opiates bought on the streets)

So,

--interesting study. for reasons above, it is not definitive, though notable that:

    --prescriptions for opioids seem to be decreasing, but there is an increasing amount of opioid "misuse", per their definition above; and opioid deaths still seem to be increasing....

    --in the same cohort, they found that the intensity of opioid medications is much less of a factor in misuse than the duration of the medications. to me, that seems likely to be a valid conclusion, since the constraints on the validity of the study would be the same for each of these prescribing patterns. in particular, they found that even short-term high intensity opiates led to much less misuse than long-term (though the opiate intensity was a factor in those on longer-term opiates)

--so, it might well make sense that we clinicians treat post-op (and perhaps acute non-postsurgical) pain more aggressively initially (ie, adequate pain control, as per the assessment of the trauma/pain and the patient's response to it), but advise the patient that if we prescribe opiates,  this is a short-term pain remedy for only a few weeks, to be followed by non-opiates/other adjunctive therapies/etc, and that it is really important not to develop opiate dependence....   

--one option to consider is topical lidocaine/steroid injections. i have found that many non-surgical patients with acute pain which is localizable in their necks, backs, etc, do get really dramatic and long-lasting relief from an injection (i often do find a clear muscle spasm in that area of maximal pain and inject that).  Also, as noted in prior blogs, there are studies finding that some pain (eg kidney stones) are pretty equally effectively treated by NSAIDs as opiates. or the recent article of of patients going to the ED with acute pain, finding no real benefit of opiates over NSAIDs (see http://gmodestmedblogs.blogspot.com/2017/11/opioids-not-better-than-nsaids-for.html​ )

--so, to me, the bottom line is that we should actively engage the patient in discussion about post-op and acute non-surgical pain, we should encourage use of NSAIDs and nonmedical management of pain as a first line, we should explain the risks and benefits of opioids if we think these are an appropriate option for pain management, and make clear our concerns about prolonged opiate-taking (perhaps more than 2 weeks or so) in developing dependence/tolerance/etc.

for prior blogs of interest:

-- http://gmodestmedblogs.blogspot.com/2017/07/decreasing-opiate-prescriptions.html notes decreasing opiate prescriptions, but increasing opioid-related deaths

-- http://gmodestmedblogs.blogspot.com/2017/02/opiate-prescribing-in-elderly-and.html /​ has an interesting/pretty powerful retrospective study in older patients and naive to opiates (mean age 69) finding that those seeing high-intensity opiate prescribing ED physicians vs low-intensity ones for similar clinical conditions had a 30% increased likelihood of long-term use of opiates

and http://gmodestmedblogs.blogspot.com/search/label/pain%2Fopiates for more recent ones

also, there was a good review article in New England Journal which reviews the physiology of both opiate tolerance and addiction as well as suggestions to mitigate the risks(see Volkow ND. N Engl J Med 2016; 374: 1253)

to be added to the mailing list, please contact me at gmodest@uphams.org