Using Serum Drug Levels

StudioLaMagica/AdobeStock

Are you interested in a diagnostic tool that provides significant objective information about your patient’s responsiveness to a psychiatric medication? Remarkably, and surprisingly to many prescribing clinicians, this tool has been in use since the 1970s. Serum drug levels, when obtained and interpreted with the well-established evidence base, provide a treasure trove of information to help you refine your treatment plan and improve patient outcomes. An appropriately drawn serum drug level can provide a wealth of information (Table 1).

TABLE 1. Information Gleaned From a Serum Drug Level

It is prudent to obtain a serum drug level when (1) a patient is on a stable dose of a medication for a clinically acceptable period and there is uncertainty about efficacy, (2) there is an absence of adverse effects, (3) there is an emergence of confusing symptoms, and/or (4) there is a presence of significant adverse effects on low dosages that are usually well tolerated. Additionally, if a patient has had an adequate clinical trial—regarding dose and duration—of a medication for a specific diagnosis with a suboptimal or no benefit, before cross-tapering or abruptly changing to a different medication, a serum drug level provides an opportunity to understand what’s happening before making a major change in the patient’s treatment. It may even explain why the drug was not effective or tolerated and may provide an opportunity to adjust the current dosage to achieve efficacy and tolerability. At the very least, the serum drug level provides a bookmark in the patient’s record of the consequences of that particular drug’s serum level at the prescribed dosage at that point in the patient’s treatment. Finally, it provides documentation to justify why the medication was stopped or increased to off-label dosages, or to explain the clinical appropriateness of the treatment plan to an insurance company or colleague.

A Little History

The importance of serum levels of psychiatric medications became apparent in the 1970s, when assays were developed and therapeutic serum levels were investigated for commonly prescribed drugs in use at that time: lithium,¹ nortriptyline,^2-4 carbamazepine,⁵ and valproate⁶ levels.

The late 1980s cemented the importance of serum drug levels and left the legacy of many psychopharmacological dogmas that remain today. For instance, Baldessarini et al⁷ published an article in 1988 that concluded:

In particular, assessment of immediate and later follow-up treatment of psychotic patients (1) indicates that moderate doses are adequate for most patients, (2) fails to support the utility of unusually high doses, and (3) even suggests the existence of a biphasic relationship of antipsychotic efficacy to dose of neuroleptics and possibly to plasma concentrations of the drugs as well.

A 1989 publication by psychiatry icons Gelenberg et al⁸ convincingly demonstrated that for maintenance treatment of patients with bipolar disorder with a serum lithium concentration of 0.4 to 0.6 mmol/L, there was a 2.6 times greater relapse risk than patients with a serum lithium concentration of 0.8 to 1.0 mmol/L.

Strict Adherence to Specifications

Before ordering a serum drug level, it is imperative to become familiar with the drug’s blood sample requirements for the particular laboratory you are ordering from. Each lab may calibrate its equipment based on specifications related to the drug’s unique properties, including its half-life, hours following the last dose taken by the patient, active or inactive metabolites present, and physical properties of the drug. Additionally, the lab’s reported serum drug level commonly provides ranges of a specific drug’s serum level across different dosages taken by a large number of individuals. This information can be quite helpful in interpreting the results in the various scenarios listed in Table 1.

A common scenario is to receive a mailed or faxed lab result, or even a phone call, from an emergency department or clinic where your patient was recently evaluated and a random serum drug level was obtained. It is imperative to put this random serum drug level in context with the time and circumstances that it was obtained. If this information is not available, it would be prudent to recheck the serum drug level with the clear parameters provided by the testing lab to allow for an accurate assessment of the meaning of the level in relation to the patient’s drug regimen. Some drugs have well-defined therapeutic serum levels that should be monitored as part of good clinical practice (eg, lithium, valproic acid, carbamazepine, and clozapine). Serum levels are available for virtually all medications, and although many drugs lack a clear therapeutic serum level, a tremendous amount of information can be obtained. (This topic will be explored further in future editorials.)

What About Pharmacogenomic Testing?

Psychiatric pharmacogenomic testing is still in its early stages for clinical utility, with a few exceptions. Although tens of thousands of publications exist, the current evidence-based clinical applications in psychiatry include only 6 genes, as vetted and reported by the gold-standard organization the Clinical Pharmacogenetics Implementation Consortium (CPIC). CPIC maintains an online open-access database that is constantly updated as new gene/drug information emerges in each medical specialty.⁹

The current 6 evidence-based genes in psychiatry include 4 cytochrome P450 genes (CYP2D6, CYP2B6, CYP2C19, and CYP2C9) and 2 HLA genes (HLA-B*15:02 and HLA-A*31:01).¹⁰ Of these 6 genes, CYP2D6 has the most clinical applicability for psychiatric medications, and this gene can be ordered as a single test by most hospital labs. Depending upon a patient’s medications, or new candidate drugs, testing for the other appropriate gene(s) can also be ordered independently.

The Rx Journey and Serum Drug Levels

TABLE 2. Factors That Determine Serum Drug Levels

What is so special about a serum drug level? Numerous factors will impact a drug’s journey from the time it enters your body until it is excreted (Table 2), and each factor can impact a drug’s serum level. Ultimately, what is clinically important in assessing a drug’s risk/benefit for a patient is the serum level using an established reference range with prespecified criteria to create a meaningful and reproducible result. The 2 variables that are out of our control are the patient’s cooperation with the parameters of the blood draw and adherence with the drug, eg, nonadherent, poorly adherent, adherent, taking extra doses, or not following the prescribed dosage schedule. These scenarios and the dangers each represents can best be addressed with psychoeducation for the patient. Hence, with the patient doing their part in preparing for the lab’s blood draw, all of the other variables listed in Table 2 will be factored in automatically, resulting in a serum drug level that provides us with a tremendous amount of clinically useful information.

Concluding Thoughts

If you had a magical and powerful tool to make life easier, you would use it, wouldn’t you? In psychiatry, we are fortunate to have serum drug levels to guide us in making the best treatment decisions for our patients. Serum drug levels have been available since the 1970s as the field transitioned from psychoanalytically based care to psychopharmacological care. Despite this, this testing continues to be underutilized in clinical practice for unclear reasons.¹¹

Dr Miller is Medical Director, Brain Health, Exeter, New Hampshire; Editor in Chief, Psychiatric Times; Staff Psychiatrist, Seacoast Mental Health Center, Exeter; Consulting Psychiatrist, Insight Meditation Society, Barre, Massachusetts.

References

1. Amdisen A. Serum lithium determinations for clinical use. Scand J Clin Lab Invest. 1967;20(2):104-108.

2. Asberg M, Cronholm B, Sjöqvist F, Tuck D. Correlation of subjective side effects with plasma concentrations of nortriptyline. Br Med J. 1970;4(5726):18-21.

3. Asberg M, Crönholm B, Sjöqvist F, Tuck D. Relationship between plasma level and therapeutic effect of nortriptyline. Br Med J. 1971;3(5770):331-334.

4. Ziegler VE, Clayton PJ, Taylor JR, et al. Nortriptyline plasma levels and therapeutic response. Clin Pharmacol Ther. 1976;20(4):458-463.

5. Bertilsson, L. Clinical pharmacokinetics of carbamazepine. Clin Pharmacokinet. 1978;3(2):128–143.

6. Gram L, Flachs H, Würtz-Jørgensen A, et al. Sodium valproate, serum level and clinical effect in epilepsy: a controlled study. Epilepsia. 1979;20(3):303-311.

7. Baldessarini RJ, Cohen BM, Teicher MH. Significance of neuroleptic dose and plasma level in the pharmacological treatment of psychoses. Arch Gen Psychiatry. 1988;45(1):79–91.

8. Gelenberg AJ, Kane JM, Keller MB, et al. Comparison of standard and low serum levels of lithium for maintenance treatment of bipolar disorder. N Engl J Med. 1989;321(22):1489-1493.

9. Guidelines. CPIC. Accessed June 6, 2025. https://cpicpgx.org/guidelines/

10. Miller JJ. Psychiatric pharmacogenomics: the evidence base. Psychiatric Times. 2022:39(7):1,6-8.

11. Miller JJ. Serum levels of psychiatric drugs. Psychiatric Times. 2014;31(11):64-67.