Lithium dose estimation / level prediction methods: What is your preferred method in an inpatient se

[Lauraaa]

---

Members don't see this ad.

I have typically done empirically dosing for individual patients. I’m reading a review paper from 2013 J of Affective disorders about 38 different methods for predictive models: a prior using patient characteristics and test dose methods based on level after a set test dose. From the article, the majority of predictive models had inconsistent or poor results in the studies evaluating them. But, I’m curious if others out there have a preferred method other than empirical dosing (such as Cooper, Perry, etc) for inpatient settings – in acute mania – with healthy adults and rationale / clinical experience with that method.

 

[shan564]

I like to uptitrate as quickly as tolerated, check the trough level after 3 days, and use that to calculate what the steady state level should be. That's always given me an accurate estimate of the eventual steady state level, but my experience is somewhat limited.

 

[slappy]

Why 3 days? Half-life of lithium is ~1 day, so it would be enough to check the levels after 5 days.

Titrate based on clinical response until then, and of course monitor for signs and symptoms of toxicity.

 

[Nasrudin]

Why 3 days? Half-life of lithium is ~1 day, so it would be enough to check the levels after 5 days.

Titrate based on clinical response until then, and of course monitor for signs and symptoms of toxicity.

Wouldn't you want to avoid symptoms of toxicity though? Hence all the predictive methods?

 

[MacDonaldTriad]

I cannot claim to have reviewed dozens of dosing strategies, but I’m not fond of loading doses or test doses with calculations for lithium. Maybe, if you are lucky, this could speed things up by a day or so.
LiCO3 has a half-life of 24 hours. Concentration =Ke to the t/24 hours. So after a day you are half way to steady state, 2 days 75% there, 3 days 87.5% there, 4 days 93.75%... So if you get a level after only 3 days remember it will be going 12.5% higher without changing the dose. Four days or more and you don’t need to worry. Be sure levels are 8 – 12 hours post dose. If you screw this up, the level will be artificially too high so not dangerous, just slowing things down. On inpatient services with a Mon – Friday rounds, this means get levels or change doses Monday Thursday, or Tuesday Friday. With any given GFR, the stead state to dose relationship is linear so it is simple cross multiplication Dose-1/Dose-2 = level-1/level-2. Since you are shooting for a level of 1.0 generally: dose patient is on/dose you need = level you measured/1.0
Watch out for things that change GFR, dehydration, diuretics, NSAIDS... Going on a low sodium diet also can raise Li+ because Li+ will be co-transported with Na+ as your kidneys try to reabsorb Na+ in the distal tubules.
Don’t forget common sense. If you start someone on 900 mg and the level comes back 0.33, probably wise not to jump to 2.7 grams. Doubling is probably OK, tripling is a little scary. Maybe you wrote the order a few hours before a once a day dosing, but it wasn’t take off or given until the next day, or maybe some doses were refused and this was poorly communicated. I’m in the habit of looking at the RNs log myself. I’m sure most people know all of this, I just can’t help myself.

 

[slappy]

Wouldn't you want to avoid symptoms of toxicity though? Hence all the predictive methods?

Of course. Hence why we exercise caution by starting/restarting the drug at responsible doses (900mg od) , not doing loading doses, making sure the patient isn't taking drugs like NSAIDs or thiazides, drinking enough water, not restricting sodium, and watching for signs and symptoms of toxicity every day. Most become symptomatic at levels over 1.5 mmol/l so considering the fatal dose is generally over 2.5 mmol/l, it'd be extremely unlikely for someone to reach such a high level within 5 days especially without becoming symptomatic. And even if that were to happen because of something going terribly wrong somewhere, it could've happened in the first three days as well since the theoretical difference between the drug levels on day 3 and day 5 is only 1/2*1/2*1/2 - 1/2*1/2*1/2*1/2*1/2 = 0.09375 x 100%. Those are the reasons why I don't see the point in doing a "half-hearted" plasma level testing 3 days after starting lithium.

 

[Nasrudin]

Of course. Hence why we exercise caution by starting/restarting the drug at responsible doses (900mg od) , not doing loading doses, making sure the patient isn't taking drugs like NSAIDs or thiazides, drinking enough water, not restricting sodium, and watching for signs and symptoms of toxicity every day. Most become symptomatic at levels over 1.5 mmol/l so considering the fatal dose is generally over 2.5 mmol/l, it'd be extremely unlikely for someone to reach such a high level within 5 days especially without becoming symptomatic. And even if that were to happen because of something going terribly wrong somewhere, it could've happened in the first three days as well since the theoretical difference between the drug levels on day 3 and day 5 is only 1/2*1/2*1/2 - 1/2*1/2*1/2*1/2*1/2 = 0.09375 x 100%. Those are the reasons why I don't see the point in doing a "half-hearted" plasma level testing 3 days after starting lithium.

I see.

 

[shan564]

Why 3 days? Half-life of lithium is ~1 day, so it would be enough to check the levels after 5 days.

Because you can use the 3-day level to accurately predict the 5-day level. After three half-lives, you'll have 87.5% of the steady state level, so it's a pretty simple calculation. You can theoretically predict the 5-day level even after a 1-day level, but I haven't tried that because there are too many other factors that can affect the level after one day.

It gets a bit more complicated if you're titrating the dose during those first few days, but you can still calculate it. Here's the math:
http://www.summitpk.com/equations/equations.htm#P3

 

[psychattending]

Because you can use the 3-day level to accurately predict the 5-day level. After three half-lives, you'll have 87.5% of the steady state level, so it's a pretty simple calculation. You can theoretically predict the 5-day level even after a 1-day level, but I haven't tried that because there are too many other factors that can affect the level after one day.

It gets a bit more complicated if you're titrating the dose during those first few days, but you can still calculate it. Here's the math:
http://www.summitpk.com/equations/equations.htm#P3

The thing to remember is that while we all say 24 hours is Li's half-life, it really is different in each person. Thus, while one can exactly calculate the steady state level given the person's known level one has to know the half-life of Li in that particular person. The earlier one checks the level, the larger the estimate will be off. My approach is similar to shan564 when I am doing inpatient work since it allows me to speed up the discharge. In the output setting when I do not have an acute crisis, I wait at least 5 days so that my estimate will be closer to reality. In patients that will have a longer half-life (e.g. renal insufficiency), I would wait even longer.