Critical Care Medicine - List
http://www.pitt.edu/~crippen/

Valium (diazepam) as a sedative for the ICU


Judith Hwang:

Diazepam infusions are strongly encouraged here, especially if the patient will need to be sedated for several days. Before I started, there was at least one reported incident in which the patient received the maximum dose of flumazanil without effect. Further workup included a head CT, neurology consult and EEG. A couple of weeks later, the patient was more awake and interactive. When we use the diazepam infusion, I encourage the residents and the nurses, daily, to decrease the dose whenever possible. Over the last few months, diazepam at scheduled intervals has been more popular--costs less and is easier to "titrate" while minimizing the risk of overdosing. (Usually the diazepam is not started until the patient has already been on a midazolam drip for greater than 24 hours, and we anticipate the need for prolonged use.)

David Crippen:

Long term sedation with Valium is the worst possible usage (IMHO). Diazepam is a drug that accumulated very quickly although the initial effect is brief. Once it builds up, very long lasting intermediaries build up too. The drug is not titratable in any sense, and VERY difficult to make and use in continuous infusion. There is pitiful little reason to use a drug with so many problems when infinitely better drugs are freely available. If cheap is an issue, why not Ativan (lorazepam). It is an infinitely better drug than Valium and it's cheap too.

My experience with "encouragement" of cheap drugs is that someone somewhere has an incentive to save money, pure and simple. I cannot help but wonder..."encouraged by whom". In some cases, support from groups of physicians assigned to committees is manipulated (on the basis of information provided by those who desire Valium to be used). Then..."Our committee on sedation usage fully supports the use of (fashionable cheap drug)". Right...most committee members are so bored they would rubber stamp propositions making everyone in the hospital wear propeller beanies.

The ICU is a place where highly titrated care is afforded to very sick patients. Valium is not titratable at all after it's initial effects, it is difficult to use and has many more untoward side effects than newer drugs. There should be a very flexible and titrated approach to the sedation and analgesia problems of ICU patients. One drug does not fit all, especially a drug like Valium. So why is there any interest in Valium? Because it's REALLY cheap. If they all cost the same, the entire world supply of Valium would be buried with David Crosby's liver 500 feet under the desert in Los Alamos in a titanium box.

But physicians get pissey when they are told they have to do something, especially to save money so the current "encouragement" of drugs like Valium must be packaged differently. It is less fulfilling to come up with a hard line like "we need to save money so we're going to force you to use the cheapest drug on the market. There's nothing you can do about it so don't bother to complain". A much more palatable line is: "Our committee of physicians after perusing the data we gave them, has concluded that all sedation drugs in the ICU are the same, and that Valium is one of the best...isn't that nice? So we know you wouldn't want to use anything else so we'll kind of get rid of the rest". The proof is in the fact that Valium is mandated. There are no other options. If there were, few would use it. So I am not impressed that you use Valium because it is the best drug for ICU use. If that were the case there would be need to make the rest unavailable.

Usually the diazepam is not started until the patient has already been on a midazolam drip for greater than 24 hours, and we anticipate the need for prolonged use.

I would ask you to consider, on the basis of a lot of experience with this kind of thing, that lorazepam is an infinitely better drug after the midazolam runs its course (about 48-72 hours). It's cleaner at least.

Lori Schoonover PharmD:

As Dr. Crippen pointed out in a previous posting, " Diazepam is cheap, but so is lorazepam." I would argue that lorazepam is cheaper than diazepam. One must consider that the cost of drug therapy is not just the cost of the drug, but all of the effects induced by that agent. If a diazepam infusion is used because it is cheap, yet the patient requires a neurology work-up, CT scan and loads of flumazenil to wake up, then the diazepam infusion just became about five times more expensive than even a midazolam drip.

I agree with Dr. Crippen, diazepam has little if any role in the setting of ICU sedation. Committee agreement on drug usage does decrease cost because the pharmacy doesn't have to keep on inventory several drugs. However, the decision regarding which agent to choose should be based, like all drug therapy decisions, on efficacy first, safety second and cost third. If efficacy and safety are not equal then likely the cost of the less efficacious or less safe drug will outweight the potential cost savings when only looking at drug acquisition cost.

Andrew D. Barnes PharmD:

I agree wholeheartedly. Diazepam is best used as an intermittent bolus drug in the short term. I particularly find it appropriate for EtOH withdrawl prophylaxis for patients who don't require the titratability of something like propofol or midazolam. The long "effective" half life of diazepam and its metabolites gives a kind of "self taper". But as for diazepam, the savings in drug costs are miniscule compared to the cost of an extra week in the ICU on a vent because they're too gorked on Valium.

Aviel Roy-Shapira:

The reason that I have stayed away from diazepam is the presence of active metabolites that in normals have half lives of 96 hours or more, certainly longer in an ICU population. We usually use Ativan (lorazepam) drips for patient we expect to need more than three days of sedation, propofol for less than three days. We have had a number of patients take a week or more to wake up from 24-48 hr of Ativan, thus the switch to propofol. Despite the cost advantages of Ativan, it has a big cost disadvantage with length of stay.

Yes, I agree. For a long time we have used midazolam drip, but even that tends to accumulate after a while. We now use either propofol or a cocktail with ketamine (low dose) midazolam and fentanyl.

Pauline L. Wong: David Crippen on diazepam infusion:

However, during my travels this week I found one hospital claiming to use it in a continuous infusion. I didn't know that was possible. One of the ICU nurses admited it took a patient a week to wake up after a typical sojourn. Has anyone ever used diazepam in continuous infusion? I always thought didn't work decause of it's progressively long length of action and the improbability of mixing it.

Here are some refs I found:

I had these in my files, but haven't looked at the lit on this topic for over 10 yr. In the days b/f midazolam, folks used diazepam by continuous infusion, but I don't know about its efficacy. It does seem logical that recovery would be longer than that seen after the newer drugs such as midazolam or propofol.

Joe Dasta:

Several yrs ago, we used diazepam by infusion in a patient with tetanus and published our experience in: Southern Medical Journal 1981;74:278. It was difficult to get into solution, but it finally dissolved: we used, i think, 50 mg/250 ml. Please note that this was before we had drugs like midazolam.

I too am aware of several institutions using diazepam infusion. If you have a patient with, for example, severe ARDS and you anticipate a long ICU stay and are willing to deal with prolonged effects after stopping it, it can be used. For long-term therapy like the above, I'd go with lorazepam....

Louis Brusco:

The reason that I have stayed away from diazepam is the presence of active metabolites that in normals have half lives of 96 hours or more, certainly longer in an ICU population. We usually use Ativan (lorazepam) drips for patient we expect to need more than three days of sedation, propofol for less than three days. We have had a number of patients take a week or more to wake up from 24-48 hr of Ativan, thus the switch to propofol. Despite the cost advantages of Ativan, it has a big cost disadvantage with length of stay.

Gloria Fortune:

Do you use ativan drips in pts suffering alcohol withdrawal? In our MICU, we see many pts with their medical diagnoses complicated with ETOH withdrawal, and they are quite ill. Some of our docs are reluctant to allow us to sedate these folks for fear of respiratory depression, etc. The nurses find that ativan works wonders in these people but I have never used it in drip form. Could you tell me what concentration you use and some idea of you protocol for determining dosing?

Louis Brusco:

According to the manufacturer, lorazepam is stable in concentrations up to 0.2 mg/cc. We use 20 mg /250 cc of d5w or NS for our baseline concentration, and increase it to 40mg/250 cc if we are giving a lot and need to cut down on IV fluid. That is our maximum concentration. We usually start at 1-2 mg/hr and have gone as high as 20 mg/hr in an unintubated ETOH withdrawal patient. In that patient, I switched to a syringe pump to save on IV volume, but, since we don't use them in our ICU, it was a bit of a problem. Just hook up a syringe pump at the hub of whatever line you are using and be careful of what goes in that line. There also is something about a certain type of IV Bag not being compatible with the lorazepam, but, since we don't use that type, I don't remember what it was. Wyeth-Ayerst, while they have lost the patent, was very helpful in setting up our protocol.

Joe Dasta:

It seems to me that the diazepam discussion emphasizes the range of opinions on a given drug therapy. Several sedative agents have simultaneously been tauted as both "great and terrible" "cheap and expensive","never use it and always use it"!

Three points ring true:

  1. There are few absolutes in critical care. One approach may fail miserably in one unit and be the salvation in another.

  2. We need good, comparative, randomized trials of sedative use in the ICU to address the issues of efficacy, cost of care, side effects. Finding a funding source would be tough...

  3. It will be interesting to see the SCCM guidelines on ICU sedation/paralysis whenever they are published! Anyone know when they will appear????

Louis Brusco:

We usually start lorazepam at 1-2 mg/hr and have gone as high as 20 mg/hr in an unintubated ETOH withdrawal patient.

David Crippen:

That's not surprising. The two drugs that are particularly cross tolerant to alcohol are barbiturates and benzodiazepines. For the chronic alcoholic, very doses of benzos are the norm. In fact, the doses can get so high that their side effects begin to predominate. It is not unusual to have to intubate patients for benzo respiratory insufficiency and they are still agitated!

Lorazepam is a good drug for uncomplicated DT or ETOH withdrawal because it's clean. However, it ain't very potent. Propofol seems to have a narrower cross tolerance profile than the benzos..therefore, you can use less of it to get the same effect. And it is so accurately titratable that you can ride out the fluctuations of these patients very nicely by having the nurse simply go up and down as needed with the dose. 48 hours later they come around and you throw them back in the sea to fight again another day. That is not to say that propofol is all that safe. For patients with any degree of heart failure or intravascular depletion, it will quicklycause hypotension. That is why we rarely bolus the drug. We start with 1 mg/kg/hr and titrate up to effect. Most DT patients are well controlled somewhere in the 2 to 2.5mg/kg/hr range.

Louis Brusco:

I still can't see why anyone would want to use a midazolam drip.

David Crippen:

Several reasons. It is much more rapid acting than lorazepam, gets your sedation going quicker and with more potency. For the first 48 hours or so it is also titratable. That means you can go up and down with the natural fluctuations of your patient's processes. For dressing changes and procedures and any other tasks that require discomfort, you can titrate quickly, then bring them back to where you want them. If the patient needs to go down to X-Ray and get all their tubes pulled out, you can bring them up. If they are sleeping nicely at 0300, you can bring it down. If the surgeon or consultant is coming around at 0600, you can lighten them for an exam. You can keep them nicely in the "comfort zone".

Louis Brusco:

For the long term, lorazepam is much cheaper, and has the side effects that you mentioned.

David Crippen:

Lorazepam has a MUCH longer onset of action and it is VERY weaker sedative. I could show you cerebral function tracings that show the onset of all these drugs. However, it is a very clean drug and it is nice when you simply want to push the "cruise mode" switch and send the patient down the road snowed. Also when the patient simply needs light sedation with no reason to titrate.

We start most everyone on Midazolam for quick onset, potency and titratability, then switch them to lorazepam after 48 hours for the long haul if needed. We also use a lot of midazolam/fentanyl combos to bring down the dose of midazolam (and get the same effect as the high does) and when pain complicates discomfort.

For the short term, propofol has a better offset time, is cheaper and has most, if not all of the advantages of midazolam without any additional down side, except for the lipid administration and infectious possibilities.

Propofol is very much like midazolam. We reserve Propofol for life threatening agitation syndromes where we want to get absolute control and we want it quickly. It is God's gift to the treatment of Delirium Tremens. It is exceptionally titratable as well and there is no evidence of tolerance over even very long time periods. Big drawback is the cost of a week`s supply of the stuff would support several South American drug lords in the style they have become accustomed. Valium sucks large quantities of canal water.

Judith Hwang:

Occasionally, it is possible to get a lorazepam infusion--but its cost is still significantly greater than diazepam and therefore access is restricted.

David Crippen:

Restricted by who? Who is running this ICU anyway, the ICU Director or the pharmacy? Someone needs to storm into the P-and-Whatever Committee and start raising hell, kicking asses and taking names.

  1. FORCIBLY ELIMINATING drugs that are effective and necessary for the treatment of ICU patients so that the cheapest possible substitute MUST be used is not acceptible to the mission of a critical care unit. If the hospital is so broke they cannot fund the difference between diazepam and lorazepam, it should be closed as a menace to the public. They can't afford to take even adequate care of patients.

  2. Specifically who by name is responsible for FORCIBLY REMOVING all other drugs but the cheapest available in the hold of a Polish Freighter in Taiwan Harbor? I want names...then I think the local and national media might be interested in the kind of care these patients are receiving and might be interested in parking their truck outside with the Night Beat anchorman pointing their finger at the well lit entrance for the 11 o-clock news.

The ICU Director must fight for the right to care for sick patients properly, and sometimes play hardball. If that doesn't happen, the next thing you will see is large boxes of diazepam all alone on the shelf.

Lori Schoonover:

But as for diazepam, the savings in drug costs are miniscule compared to the cost of an extra week in the ICU on a vent because they're too gorked on Valium.

Judith Hwang:

True, but the cost does not come out of Pharmacy's budget!

David Crippen:

The ICU is the LAST place budget should be the primary concern and the LAST place where patients should be stiffed with cheap substitutes of genuine articles. The ICU is a place where the sickest, most unstable patients require, and should get, the most effective and titrated care available in the hospital. We had a few overtures from the budget bean counters toward this end. It was a noble battle. In the end, We beat them. They came, they failed , and we gnawed on their skulls. They screeched like hyenas for a while, then ran like rats.

Some will march on a road of bones, others will be nailed up on telephone poles. That's the way it works.

Louis Brusco:

However, when we did our cost analysis we came out with about $200 per day for propofol versus about $500 per day with midazolam. I would be interested in your information on cost. Not that I would advocate the choice to be purely on cost, but when the drugs are fairly equal (or even propofol has a slight edge) it was the deciding factor in our decision.

David Crippen:

This just goes to show you that the cost of these things varies wildly depending on area and usage habits. I can hardly believe that your cost for Midazolam is more than twice that of propofol!!!! How is that happening.

The average dose of midazolam for the average patient in our unit is 3 - 5 mg per hour, titrated to effect, and we switch to lorazepam after 48 hours when the titratability runs out. Midazolam runs about $1.30 per milligram. Propofol costs about 5 cents per milligram and the average dose for the average patient in DT is around 2 mg / kg /hour for about 48 hours (unless Hansen is in charge, then it's 20 mg/kg/hr, but that's another story.)

If you used an average of 4 mg per hour of midazolam for 48 hours, you would pay $250.00.

Assuming an 80 kg patient and using 2 mg/kg/hr of propofol for 48 hours you would be in for $384.00

In order for you to spend that kind of money on midazolam, you must be using enough to put Godzilla into La-La land for a week.

When we get over 5 mg / hr of midazolam, we start cutting it with fentanyl (which costs peanuts- off patent). This allows you to dramatically cut back on the amount of the expensive drug (midazolam) and still get the same effect as the larger dose, and also the concurrent use of both drugs preserves titratability for 48 hours as well.