Therapeutic drug monitoring is an integral part of a care plan for any individuals on medications with narrow therapeutic range. Some common examples include digoxin, phenytoin, lithium and vancomycin. Similar to other laboratory monitoring, one needs to understand why the test is needed, when it is needed and how the results may prompt us to adjust the dosage of the medication.
I don’t intend to give a detailed talk on pharmacokinetics or discuss about different compartmental models on drug distribution. But below are some important questions every clinician should ask before ordering any drug levels:
Sometimes drug level monitoring helps to assess the efficacy of a drug, other times the adverse effects and sometimes both. We need to know why.
For many anti-epileptic drugs, ensuring the drug level is within the therapeutic range will help to prevent seizure occurrence and to minimize side effects. This is particularly true for phenytoin where the drug is highly protein bound, has many drug interactions that can alter its plasma level as well as having a non-linear relationship between its dose and the corresponding level.
However, some anti-epileptics are used as a mood stabilizer, such as carbamazepine or valproic acid. In these cases, drug levels don’t offer us helpful information on the mood stability. But if we suspect adverse effects, then an elevated drug level may confirm our suspicion.
Also not all anti-epileptics need routine drug monitoring. For instance, lamotrigine levels have not been shown to consistently correlate with outcome, as such routine monitoring is not indicated
Digoxin is excreted renally and as such, there is a higher risk of digoxin toxicity in renal impairment. But does the digoxin level correlate with outcome or efficacy? Given digoxin can be used for heart failure or as a rate control agent, the outcome measures for efficacy can differ depending on use. Based on the DIG trial, digoxin level has not been shown to correlate with efficacy in the setting of heart failure but higher levels may put the patient at risk for digoxin toxicity. And if used as a rate control agent, one can easily monitor the heart rate to assess the response to medication. Hence, digoxin monitoring should be focused on ensuring level is not elevated as a means to prevent toxicity. If the level is considered “low” even to the point of “undetectable”, that doesn’t mean we have to increase the dose automatically. One needs to assess the stability of the heart failure or the control of ventricular rate – if there is no concern, then there isn’t a need to adjust the digoxin dose. But if the digoxin level is elevated, then I would jump at the chance to lower the dose immediately.
Know WHEN …(Stat level? Steady State? Trough level? Peak level?)
When a drug level is ordered, it is important to know when the result should be taken. This will depend on the clinical situation and the pharmacokinetic parameters of the medication.
For instance when toxicity is suspected, a stat level is ordered. In many cases, we want to ensure the dose is adequate and as such, we would like to order the drug level at the steady state concentration. Ordering the drug level too soon will result in under-estimation of the drug level. This in turn can mislead the clinician to increase the dose prematurely, leading to the risk of toxicity.
The time to steady state concentration depends on the half life of the drug. But in general, it takes 5 half lives to achieve steady state concentration. Hence keep this mind when checking for any levels when a drug is started or dosage has changed recently.
It is also important to know if we are looking for the peak level or the trough level and this depends why we are checking the drug level as well as the pharmacodynamic properties of the medications. In many situations, we aim to check the trough level which is the drug level immediately before the next dose. This is the level to check for antiepileptic, digoxin, vancomycin. For aminoglycosides such as gentamicin or tobramycin where the bactericidal activity is concentration dependent, we may want to check the peak level and wants to ensure it is reaching a high enough concentration for optimal activity.
Treat the patient, not the number
Often we focus on the drug level results, we forget to assess the patient. The drug levels are only markers to help us evaluate the drug therapy. There are other clinical markers that are just as important. For instance, the temperature, WBC and other symptoms of an infection are important markers to evaluate the antibiotic in question. For someone who has a history of seizure but has been seizure-free for the last 5 years with consistent sub-therapeutic levels of phenytoin, I will think twice before increasing the dose. Also if the patient is stable with no worsening symptoms of heart failure, I will leave the digoxin dose alone, even if the lab result indicates it is subtherapeutic.
How to adjust dose and recheck drug level
If we need to adjust the dose, how and when to check drug level again? This is a very important question that sometimes doesn’t get enough attention. If we have all the information such as the half life, volume of distribution, exact time of administration etc, then we probably can calculate the dose based on the drug level we want to achieve.
However, life is never that easy and simple and we never have all the information in front of us. So we have to make an educated guess in most cases, especially in the community setting. The basic rule is change the dose by 10-15% at a time, or round to the nearest tablet or capsule size available and recheck the drug level again when the steady state concentration is reached ( = 5 half lives of the drug). Sometimes, I would suggest checking the level sooner, especially for drugs with long half lives and there is concern that we may be overshooting. In these cases, I will check sooner but qualify the result that it may not be at steady state concentration and the drug level will continue to rise and the need to recheck at a later time.
Finally, therapeutic drug monitoring is only one tool we have to help assess and evaluate drug therapies. It is not the only tool. We should also look at other reliable markers or clinical outcome measures. Just because we can order a drug level does not mean it needs to be ordered – it depends if we have a valid reason to check the drug level. Also making sure ordering the drug level at the right time is important. Otherwise, the results cannot be interpreted and the resources that have been invested into ordering and checking the drug levels would have gone to waste.