Primary Care Corner with Geoffrey Modest MD: opiates and benzos assoc with inc mortality

By Dr. Geoffrey Modest

Another observational study, this one using data-mining of large patient databases, found that concurrent use of prescription opiates and benzodiazepines was associated with increased ER visits and hospital admissions for opiate overdoses (see doi.org/10.1136/bmj.j760).

Details:

  • 315,428 privately insured patients aged 18-64, continuously enrolled in a health plan with medical and drug benefits from 2001 till 2013
  • About 40% men, comorbid conditions included: heart failure, peripheral vascular disease, hypertension, COPD, diabetes, CKD, cerebrovascular disease, dementia, MI, liver disease, alcohol “abuse” (??not clear how defined), drug “abuse” (also not defined), psychosis, and depression
  • The principle search was for patients on opiates who had at least one day of overlapping prescriptions for a benzo

Results:

  • 9% of opiate users also had a benzo in 2001, increasing to 17% in 2013
  • The increase was largest in those on intermittent opiates
  • Compared to those who did not use benzos, risk of ER visit or hospital admission for opioid overdose was:
    • 14x as often for all opiate users, adjusted OR 2.14 (2.05-2.24, p<0.001)
      • 42x as often for intermittent opiate users, adjusted OR 1.42 (1.33-1.51, p<0.001)
      • ​1.81x as often for chronic opiate users, adjusted OR 1.81 (1.67-1.96, p<0.001)
    • Assuming a causal link between the combo of meds and ER visits/hospitalizations for opioid overdose, eliminating this combo (i.e. not taking benzos) would decrease these services by 15%

Commentary:

  • This is yet another study linking opiates and benzos to overdoses. Other studies done in several different countries have confirmed this, overall finding that around 30% of fatal “opioid” overdoses involve the concommitant use of benzos.
  • There are clear limitations in this study, most of them articulated by the researchers, including that as a retrospective study there could be hidden biases (they controlled only for the ones they assessed), the study only looked at patients continuously enrolled in their health plans (though sensitivity analyses found not much difference if looking at those enrolled for at least 2 years), they did not include patients who died from an overdose, they do not have information on the doses of meds or how those changed over the time period, and they only looked at prescription drugs (i.e. not heroin use or prescription meds bought on the street). They also did a sensitivity analysis requiring both medications overlap for at least 25% of the days of opiate prescriptions, also finding similar results.
  • I would add to this list: it is important to note that the likelihood of hidden biases is increased given that the patients on concurrent opioids and benzos had a higher incidence of every medical condition (14 of them) they tried to control for. Perhaps very significantly in this sicker population, they did not have information on potential differences in the severity of any of these medical conditions in those on benzos vs not (not all diabetes is the same…). And there are some very important conditions that they did not include: for example, they did not look at patients with anxiety disorders, including panic, for which these patients might well be on benzos appropriately for therapy, and these underlying psych conditions themselves have considerable mortality associated with them. Or, bipolar disorders with predominant mania, perhaps misdiagnosed as anxiety and also treated with benzos, yet bipolar disorder itself also has significant attendant mortality. And I should note that of these 14 conditions, the one with the most profound difference between the non-benzo and benzo groups was “depression”, noted in 4.4% in the nonbenzo group and 17% in the benzo group. And, as mentioned, there is no differentiating by severity of depression or comorbid other psych issues (i.e. those on benzos may well have had more psych issues, both quantitatively and qualitatively, which put them at higher risk of overdosing).
  • But, the bottom line to me is that there is certainly a plausible mechanism for increased overdose in those on both meds (benzos increase the respiratory depression of opiates), there is an FDA “black box” caution about simultaneous use of these meds, prior studies have found up to a 10-fold increased mortality from overdose in those on the combo, and so I think the onus is on us as prescribers to minimize the concurrent use of these medications as much as possible. That being said, I certainly do see some patients who are on opiates and do have debilitating anxiety disorders refractory to other meds/therapies, who are also on benzos. Some prescribed by me, some by psychiatrists. I do try to minimize the doses of all of their opiates/benzo. And perhaps it makes sense to try to switch the opiates in those on both classes of medications to buprenorphine, since this is associated with less respiratory depression (though as far as I know this approach has not been validated through studies. I would welcome any information from blog readers). And, it makes sense to be especially certain that these patients have naloxone at home, with someone trained in its administration.