Welcome to my blog, a place to explore and learn about the experience of running a psychiatric practice. I post about things that I find useful to know or think about. So, enjoy, and let me know what you think.

Saturday, July 6, 2013

Az, The Great and Powerful

I remember studying Pharmacology in medical school, and some of the mnemonics. If it ended with, "alol", it was a beta blocker (propranalol, metopralol). And if it had an "az" in it, it was a benzodiazepine (diazepam, lorazepam, chlordiazepoxide).  Now I just call them valium, ativan, librium, etc., so my mnemonic isn't much use.

I've noticed that there's an excessive and ironic amount of anxiety associated with prescribing benzos for anxiety. My residency had a particularly strong addiction psychiatry program, and it was great to work on the dual diagnosis unit, believe it or not. But the ethos there was, absolutely no benzos. Benzos are bad. If patients ask for them, it's because they're drug seeking. And if doctors suggest their use, it's because the doctors are either too foolish to recognize that they're being duped by drug seekers, or because the doctors are coddling their patients.

As a PGY-3, in the outpatient clinic, I had to force myself to unlearn some of this attitude. But even then, it was with a fair degree of trepidation that I wrote my first independent script for a benzo (I think klonopin).

So with the I-STOP program looming large, gearing up to check on all the patients I write controlled substance scripts for, in real time, I need to ask myself, what's the truth about benzos?

Here are some specific questions, in no particular order:

* Do benzos help with anxiety, in the long run?
* Do benzos help with sleep?
* How likely is it that a patient without a significant substance
    history will abuse or divert benzos?
* How likely is it that a patient with a significant substance
    history will abuse or divert benzos?
* Does the particular substance of abuse matter?
* How dangerous is benzo use?
* How helpful can a benzo be to someone with a significant substance history?

In an extensive review of the literature since 1966, Posternak and Mueller (2001) found that:

Although most benzodiazepine abusers concurrently abuse other substances, there is little evidence to indicate that a history of substance abuse is a major risk factor for future benzodiazepine abuse or dependence. Furthermore, benzodiazepines do not appear to induce relapse of substance abuse in these patients. The authors conclude that the position that benzodiazepines are contraindicated in former substance abusers appears to lack empirical justification. Benzodiazepines may be indicated in certain patients with anxiety disorders and a history of substance abuse or dependence.

Conversely, there's a really nice summary entitled, Addiction: Part I. Benzodiazepines—Side Effects, Abuse Risk and Alternatives on the American Family Physician site. It reviews things I hadn't thought about in a while, like the neurochemistry of BZDs. According to this article, benzos are pretty safe in overdose, when used alone, but are frequently used with other substances, which can increase toxicity and danger. The authors claim that people who take benzos become tolerant to the hypnotic effects pretty quickly, which is why they're not useful for insomnia, in the long run. Tolerance to anxiolytic effects develops more slowly, but it does develop, so that the authors discourage long-term benzo use for anxiety. Finally, the article claims that:

Benzodiazepines are rarely the preferred or sole drug of abuse. An estimated 80 percent of benzodiazepine abuse is part of polydrug abuse, most commonly with opioids...Studies indicate that 3 to 41 percent of alcoholic persons report that they abused benzodiazepines at some time, often to modulate intoxication or withdrawal effects...Most addiction medicine specialists believe that benzodiazepines are relatively contraindicated in patients with current alcohol or drug abuse problems and in patients who are in recovery.

This article predates the Posternak and Mueller paper, so it isn't clear what the fact that they disagree implies.

This lack of clarity is reflected in a 2012 study of prescribing practices, Psychiatrist Decision-Making Towards Prescribing Benzodiazepines: The Dilemma with Substance Abusers. The study surveyed outpatient psychiatrists, and found that:

Sixty-six percent of (respondents) experienced requests for behaviors suspicious for abuse...Patient characteristics such as ‘history of abuse’, ‘unknown patient’, and ‘patient use of illicit substances’ were occasional or common reasons for NOT prescribing BZDs (75 %). (And) The most common contexts in which the majority of (the) sample was uncomfortable prescribing BZDs involved a patient history of substance abuse, fear of initiation of dependence, diversion, and feeling manipulated by the patient.

According to the CDC's Emergency Department Visits Involving Nonmedical Use of Selected Prescription Drugs --- United States, 2004--2008:

The estimated number of ED visits involving nonmedical use of benzodiazepines increased from 143,500 in 2004 to 271,700 in 2008 (89%, p=0.01), and rates increased from 49.0 to 89.4 per 100,000, an increase of 82% (p<0.05). The increases in numbers of ED visits during 2004--2008 for individual benzodiazepines were significant: alprazolam (125%, p=0.01), clonazepam (72%, p<0.001), diazepam (70%, p=0.02), and lorazepam (107%, p=0.006)...Among opioid analgesic--related visits, 38% did not involve any other drug (including alcohol); the corresponding figure was 21% for benzodiazepine-related visits. Benzodiazepines were involved in 26% of opioid analgesic--related visits. Alcohol was involved in 15% and 25% of visits for opioids and benzodiazepines, respectively.

The overall opinion, aside from the Posternak article, at least that I could find in a semi-brief search that was more extensive than what I've recorded, seems to be that BZD treatment of anxiety is contraindicated in substance abusers, at least while they're actively using. Once the substance use is in remission, it appears that benzos can be used, with caution, to treat anxiety, but are not recommended as first line.
In more general populations, benzos are useful for the short term treatment of anxiety, especially as PRNs, and the even shorter term treatment of insomnia.
Benzos are generally safe in overdose if used alone, but are frequently used with other substances which can make them much more toxic in overdose.
Benzos are also making increasing appearances in emergency rooms.

Somehow, I'm not satisfied with this answer. Maybe because it doesn't jive with my clinical experience. I've known benzos to be helpful to patients with a history of substance abuse, and even to be helpful in curtailing the substance abuse, with no subsequent escalation in dose. I've worked with patients whose insomnia improved significantly with benzo use QHS, even long term, and even after failing with ambien or trazadone.

So I'm not quite sure what to do with this information. I recognize that my sample size is not large, but since I've read mostly abstracts of the above articles, I can't speak to the validity of their conclusions, either. Any thoughts?


  1. I see a different population in the med-surg units and, in general, I avoid benzodiazepines because they're typically associated with delirium.

    But I agree that clinical context is important. Patients in the ICU or general medical units can suffer from catatonia and the treatment for that, at least initially, is a benzodiazepine.

    Jim Amos, MD
    The Practical Psychosomaticist

    1. This is a good point. Thanks for noting it.
      I was thinking in terms of my own patient population, which is outpatient, working, and self-pay. But when I worked in an inpatient medical hospital setting, I tended to avoid benzos, too, because of the risk of delirium. Then there was always the question of what to use for anxiety or agitation in those patients.
      On inpatient psychiatry, it really depended. The non-MICA units were more liberal than the MICA unit. And the outpatient clinic population was yet another context.