In a geriatric Canadian community sample, lithium level monitoring did not meet the frequency and consistency of international standards, according to research results published in The Journal of Clinical Psychiatry.

Investigators conducted a prospective cohort study of older adults (age ≥66) residing in Ontario, Canada, who initiated treatment with either lithium or valproate between April 1, 2002 and March 31, 2014 (N=11,006). Healthcare data were abstracted from multiple databases housed at the Institute for Clinical Evaluative Services, including the Ontario Health Insurance Plan and the Ontario Drug Benefit Program. Lithium users (n=5503) were propensity score matched 1:1 to valproate users (n=5503). Over a 1-year follow-up period, investigators captured the frequency with which serum lithium levels were monitored and renal and endocrine laboratory testing was performed. Outpatient laboratory tests were identified using the Ontario Health Insurance Plan database.

In the total cohort, the mean age (SD) was 70.6 (5.9), and 59.0% were women. After propensity score matching, baseline characteristics did not differ between the lithium and valproate user groups. Lithium level monitoring was infrequent; at 90, 180, and 365 days of follow-up, just 24.1%, 42.4%, and 66.8% of lithium users had undergone lithium level monitoring, respectively. Monitoring rates for serum creatinine were statistically higher in lithium compared with valproate users at 90 (29.6% vs 26.2%), 180 (50.4% vs 45.5%), and 365 days (75.4% vs 70.3%), with a hazard ratio of 1.19 (95% CI, 1.12-1.27; P <.001). In a similar fashion, at 90, 180, and 365 days, thyroid-stimulating hormone (TSH) and calcium annual testing rates were statistically higher in lithium users compared with valproate users: 22.6%, 40.1%, and 64.1%, respectively, vs 16.8%, 30.8%, and 52.2% for TSH (HR, 1.47; 95% CI, 1.37-1.58; P <.001) and 3.5%, 6.9%, and 13.3%, respectively, vs 2.9%, 5.8%, and 11.2% for calcium levels (HR, 1.15; 95% CI, 1.02-1.29; P =.018).

Related Articles

Per these data, only 24.1% of lithium users met the standards from the International Society for Bipolar Disorders to screen for lithium levels and renal function every 3 months in older adults. Although monitoring rates for serum creatinine, TSH, and calcium levels were statistically higher in lithium users compared with valproate users, these differences “may not have…been clinically important,” the researchers wrote. Prior research has suggested an association between suboptimal lithium level monitoring and lithium-related adverse events, underscoring the importance of these data. Clinician education strategies and “treatment algorithms” for improving monitoring are potential to surmount these obstacles, the investigators wrote, and should be explored in future studies.


Rej S, Herrmann N, Gruneir A, et al. Blood lithium monitoring practices in a population-based sample of older adults. J Clin Psychiatry. 2018;79(6):17m12095.