Suicide Risk in Lithium vs. Divalproex Treated Bipolar Disorder
From MindLinc Wiki
SUICIDE RISK IN BIPOLAR DISORDER
Bottom Line: Among patients treated for bipolar disorder, risk of suicide attempt & suicide death is lower during treatment with lithium than during treatment with divalproex.
Clinical Question: TH is a 62 year old WM with a long-standing history of bipolar disorder. He is admitted during an acute manic phase, which was likely precipitated by his discontinuing treatment with Depakote after hearing that "Depakote causes suicide in patients with bipolar disorder" on National Public Radio. He expresses a high level of paranoia about psychiatrists & the healthcare system, and requests to be put on lithium instead of Depakote. In patients with bipolar disorder, does lithium reduce the risk of suicide attempt and suicide death compared to those treated with Depakote (divalproex)?
Methods:
Design - retrospective cohort study
Setting - two large integrated health plans in California (Kaiser Permanente) & Washington (Group Health Cooperative).
Patient population /Inclusion criteria - population-based sample of 20638 health plan members aged 14 years or older who had at least one outpatient diagnosis of bipolar disorder and at least one filled prescription for lithium, divalproex, or carbamezepine between January 1, 1994 and December 31, 2001. Follow-up for each individual began with first qualifying prescription & ended with death, dis enrollment from the health plan, or end of the study period.
Exclusion criteria - 1. schizophrenia; 2. schizoaffective disorder; 3. dementia or cognitive disorders.
Interventions - lithium, divalproex, carbamezepine, combination. Period of exposure began on Rx dispensing date & continued for the expected duration of the Rx plus a grace period of 14 days or 25% of the expected prescription duration. Each day during the study period was classified by exposure to lithium, divalproex, carbamezepine, a combination, or none.
Blinding - not discussed
Main outcome measures - (primary) suicide attempt, recorded as a hospital discharge diagnosis or an emergency department diagnosis; suicide death, recorded on a death certificate. -(secondary) 1. Could differences in suicide risk associated w/ diff m.s.'s reflect differences in preexisting illness severity or other factors affecting suicide risk (confounding by indication)? 2. Were risk differences between lithium & divalproex stable throughout study period? 3. Are patients w/ higher suicide risk more likely to be switched from one class of m.s. to another?
Analysis - exact binomial tests (to compare the unadjusted suicide rate during periods of lithium exposure with the suicide rate during periods of divalproex or Barb amezepine exposure), Cox proportional hazards regression models (to examine risk in relation to exposure to each type of mood stabilizer after adjustment for age, sex, health plan, year of diagnosis, comorbid medical & psychiatric conditions & concomitant use of other psychotropic drugs), and weighted Schoefeld residuals (to assess the proportional hazards assumption that the relative risks do not change over time). Periods of exposure to mood stabilizers & concomitant psychotropic drugs were analyzed as time-dependent variables. Thus, patients who switched from one m.s. to another contributed information to estimates of the suicide risks associated with each of the drugs. Risk during treatment with lithium alone was the reference category for comparisons among m.s.'s. Rates were reported per 1000 person-years of treatment exposure.
Follow-up - not specific as to frequency of checks on administrative databases used to identify treated patients, assess potential confounding factors & ascertain periods of exposure. Suicide attempts were identified using emergency department & hospital discharge diagnoses. Suicide deaths were identified using health plan mortality records & state death certificate records. Mean follow-up period of 2.9 yrs per individual (60060 person-years of observation).
Validity:
Randomization - no
Similarity at baseline of treatment groups - KP & GH groups were similar on age, sex, first mood stabilizer, exposure to various m.s.'s, rates of suicide attempts resulting in hospitalization & rates of suicide attempts resulting in death. Similarities & differences among individuals treated w/ the different types of m.s.'s were not discussed, but were adjusted for during analysis.
Blinding-not discussed
Intention to treat - not discussed; depending on pattern of med switches, an individual might contribute follow-up time to any or all of the exposure groups.
Analysis in group assigned - does not apply
Groups treated similarly outside of intervention - not discussed; no mention made of amount/type of/access to mental health & other follow-up, # of hospitalizations, etc.
Reference to clinical case - patient population is restricted to members of the above-mentioned health plans in California & Washington, was predominantly female & younger than 45 years. My patient was a 62 year old white North Carolinian male veteran.
Results:
- In both health plans, unadjusted rates were greater during treatment with divalproex than during treatment with lithium for ED suicide attempt (31.3 vs 10.8 per 1000 person-years; P<.001), suicide attempt resulting in hospitalization (10.5 vs 4.2 per 1000 person-years, P<.001), and suicide death (1.7 vs 0.7 per 1000 person-years, P<.002).
- After adjustment for age, sex, health plan, year of diagnosis, comorbid medical & psychiatric conditions, and concomitant use of other psychotropic drugs, risk of suicide death was 2.7 x higher (95% CI, 1.1-6.3; P=.03) during treatment with divalproex than during treatment with lithium.
- Corresponding hazard ratios for nonfatal attempts were 1.7 (95% CI, 1.2-2.3; P=.002) for attempts resulting in hospitalization and 1.8 (95% CI, 1.4-2.2; P<.001) for attempts diagnosed in the ED.
- Results for other comparisons (lithium vs. carb, li vs combo tx, li vs no ms.) were not statistically significant (see Table 3). They were qualitatively similar to those for divalproex, but were less precise.
- Differences in selection of m.s.'s were determined by temporal trends in prescribing rather than by characteristics of individual patients.
- For suicide attempts, rates during lithium exposure were consistently lower throughout the study, but for suicide deaths, risk difference was less consistent over time,& appeared smaller during the first half of the study period.
- A history of any medication switch was associated w/ a higher risk of suicide attempt, but this effect did not differ by direction of the switch (from lithium to divalproex vs the reverse).
Adverse Effects - not discussed
Compliance w/therapies - not addressed, just assumed based on Rx fills & refills
Comments / Application:
Were the study populations similar to my patient? Not really, aside from the diagnosis of bipolar disorder. This was, however, the article that precipitated his medication non-compliance & inpatient admission.
Are the likely benefits worth the potential harms & costs? It appears so. As long as the patient has no direct contraindications to the use of lithium (such as previous adverse reaction), its use in patients with bipolar disorder should be considered. Each clinician must balance the risks (thyroid & renal dysfunction, low therapeutic index) & benefits (decreased risk of suicide attempt or death, good efficacy) for each patient.
Weaknesses:
- Use of administrative databases: Reliance on clinician's discharge or encounter diagnoses rather than structured research evaluations to i.d patients treated for bipolar disorder. Reliance on diagnosis & prescription data rather than clinical assessments to adjust for differences in illness severity or underlying suicide risk (due to other medical or psychiatric conditions, concomitant use of other psychotropic drugs). Visit or discharge diagnosis data may miss a significant proportion of true suicide attempts.
- No observational study can completely exclude the possibility of confounding by indication or bias due to unmeasured differences in illness severity or suicide risk.
Study in context of other available literature and/or current standard of care? This study complements previous studies comparing lithium & carbamezepine, and data regarding the anti-suicide effects of lithium. The current prescribing practices, however, favor divalproex 2:1 over lithium. Few studies up to this point addressed the effect of anticonvulsant m.s.'s on suicide risk.
How will this study affect your management of the putative patient? Patient was not switched to lithium, despite repeat requests. Multiple medical problems, substance abuse, thyroid dysfunction, hx of med non-compliance & poor follow up were taken into consideration. These factors, & accurate study results were explained to him.
Next steps for further study of this problem. Evaluate whether the apparently lower suicide risk during treatment with lithium also occurs during treatment with lithium-anticonvulsant combinations. Further evaluate the effects of treatment with carbamezepine on suicide risk.
Presented by Ana Carla P. Smith, MD
