icon-folder.gif   Conference Reports for NATAP  
  ID Week
Oct 8-12 2014
Back grey_arrow_rt.gif
Commercial and Medicaid Claims Databases Show No Evidence of Higher Suicidality With Efavirenz
  IDWeek 2014, October 8-12, 2014, Philadelphia
Mark Mascolini
Analysis of a commercial insurance database and a Medicaid database disclosed no evidence that people taking first-line efavirenz run a higher risk of suicidality than people taking nonefavirenz regimens [1]. The findings run counter to a recent multistudy analysis by the AIDS Clinical Trials Group (ACTG) [2].
ACTG investigators combined results of four clinical trials in which previously untreated people started either an efavirenz regimen or a regimen not containing efavirenz [2]. After a median follow-up of 96 weeks, suicidality (suicidal ideation or attempted or completed suicide) incidence was 8.08 per 1000 person-years with efavirenz versus 3.66 per 1000 with other regimens. Statistical analysis indicated more than a doubled risk of suicidality with efavirenz (hazard ratio 2.28, 95% confidence interval 1.27 to 4.10, P = 0.006). "When efavirenz is used as a component of antiretroviral therapy," the ACTG team suggested, "patients should be monitored carefully for exacerbation of depression or evidence of suicidal thoughts or behavior" [2].
But retrospective analysis of the FDA Adverse Event Reporting System by researchers from Bristol-Myers Squibb, maker of efavirenz, found no difference between efavirenz and other often-prescribed antiretrovirals in a predefined threshold for signals of suicidality [3].
Pursuing this line of research, Bristol-Myers Squibb investigators conducted this claims database analysis to estimate suicidality incidence in people starting a first-line efavirenz-containing regimen and people starting an efavirenz-free combination [1]. They scrutinized the MarketScan commercial claims database and the MarketScan Medicaid database to find people at least 12 years old who had at least one pharmacy claim for an antiretroviral from 2006 through 2013. Everyone had been in the database for at least 6 months before the date when they filed the antiretroviral claim. Efavirenz-free combinations could contain other nonnucleosides, protease inhibitors, integrase inhibitors, fusion inhibitors, or CCR5 antagonists.
The primary outcome was suicidality (suicidal ideation or a suicide attempt). Secondary outcomes were (1) suicide attempt and (2) an expanded definition of suicide attempt encompassing injuries consistent with a suicide attempt plus a psychiatric condition during the same healthcare encounter. The investigators used Cox regression to compute hazard ratios using propensity score stratification to adjust for baseline patient variables.
The commercial database included 11,187 people starting efavirenz and 8796 starting a nonefavirenz regimen. Age was slightly but significantly younger in the efavirenz group (mean 40.1 versus 40.8, P < 0.001), and a significantly higher proportion of efavirenz takers were men (86% versus 79%, P < 0.001). Significantly lower proportions of people taking efavirenz versus comparison regimens had depression, anxiety, bipolar disorder, schizophrenia, or drug dependence, and significantly lower proportions were taking antidepressants or second-generation antipsychotics.
Analysis of the commercial database found a low suicidality rate with efavirenz regimens (0.2%) and efavirenz-free regimens (0.4%) (P < 0.001). Rates of suicide attempts and the expanded definition of suicide attempts did not differ significantly between the two groups. Unadjusted suicidality incidence was 3.28 per 1000 person-years with efavirenz and 4.02 per 1000 with other regimens, and 95% confidence intervals (CI) broadly overlapped. Rates of suicide attempts and expanded-definition suicide attempts did not differ significantly between the two groups.
Analysis of the commercial database that adjusted for demographics, depression, and other relevant variables found no difference in the hazard ratio for suicidality with efavirenz versus efavirenz-free combinations (adjusted hazard ratio [aHR] 1.03, 95% CI 0.64 to 1.67). Suicide attempt risk was higher with efavirenz than nonefavirenz regimens, but this difference did not reach statistical significance (aHR 5.70, 95% CI 0.69 to 47.15). Expanded-definition suicide attempt risk was equivalent with efavirenz and efavirenz-free combinations (aHR 1.00, 95% CI 0.51 to 1.95).
The Medicaid database included 2224 people starting efavirenz and 2930 starting an efavirenz-free combination. Age averaged 41.7 in the efavirenz group and 39.7 in the comparison group (P < 0.001), and a higher proportion starting efavirenz were men (56.7% versus 50.2%, P < 0.001). About two thirds in each group were black. Prevalence of depression, anxiety, bipolar disorder, schizophrenia, and drug dependence were significantly lower in the efavirenz group than in the efavirenz-free group. A significantly lower proportion of people taking efavirenz also took second-generation antipsychotics (5.8% versus 9.4%, P < 0.001).
Analysis of the Medicaid database indicated a lower suicidality rate with efavirenz than with efavirenz-free regimens (1.3% versus 2.9%, P < 0.001) and a lower rate of expanded-definition suicide attempts (0.3% versus 0.8%, P = 0.028). In unadjusted analyses, suicidality incidence and expanded-definition suicide attempt incidence were nonsignificantly lower in efavirenz takers. Suicide attempt incidence was significantly lower in people taking efavirenz: 0.56 per 1000 person-years (95% CI 0.01 to 3.13) versus 6.42 per 1000 person-years (95% CI 3.32 to 11.22).
Adjusted analysis of the Medicaid database disclosed no difference in suicidality risk with efavirenz versus efavirenz-free regimens (aHR 0.90, 95% CI 0.62 to 1.32). The expanded-definition suicide attempt risk did not differ significantly between the two groups (aHR 0.71, 95% CI 0.33 to 1.51). But suicide attempt risk was significantly lower with efavirenz than with nonefavirenz regimens (aHR 0.11, 95% CI 0.20 to 0.89).
The researchers noted that their analysis is limited by (1) potential outcome misclassification, especially with suicidal ideation, (2) difficulty in capturing completed suicides, and (3) inability to assess other risk factors such as family history, socioeconomic status, and psychiatric evaluations. With these limitations in mind, they concluded that, "using propensity-score adjusted Cox regression, we observed no evidence of an increased risk of suicidality among patients initiating efavirenz-containing regimens when compared with efavirenz-free regimens."
Why these results [1] differ from the 4-trial ACTG analysis [2] is hard to pin down. The ACTG analysis included about as many people (3241 on efavirenz, 2091 on other regimens) as the Medicaid database but many fewer than the commercial database. Everyone included in the insurance claims analysis lived in the United States, compared with 74% in the ACTG trials. Median age in the ACTG population (37 years) was a few years younger than in the commercial or Medicaid populations. Similar proportions in the ACTG group and the insurance groups were taking antidepressants. One third in the ACTG trials had a documented psychiatric history, slightly more than had depression or anxiety in the commercial database but about the same as had depression or anxiety in the Medicaid database. Yet suicidality incidence (per 1000 person-years) was much higher in the ACTG efavirenz group (8.08) than in the ACTG nonefavirenz group (3.66), the commercial database efavirenz group (3.28), or the commercial database nonefavirenz group (4.02). Compared with these rates, suicidality incidence was much greater in the Medicaid efavirenz group (25.71) or the Medicaid nonefavirenz group (40.56).
1. Nkhoma E, Coumbis J, Farr A, et al. Using real world data to assess the risk of suicidality among patients initiating an efavirenz-containing regimen versus an efavirenz-free antiretroviral regimen. IDWeek 2014. October 8-12, 2014, Philadelphia. Abstract 818
2. Mollan KR, Smurzynski M, Eron JJ, et al. Association between efavirenz as initial therapy for HIV-1 infection and increased risk for suicidal ideation or attempted or completed suicide: an analysis of trial data. Ann Intern Med. 2014;161:1-10. http://www.natap.org/2014/HIV/071414_10.htm
3. Napoli AA, Wood JJ, Coumbis JJ, Soitkar AM, Seekins DW, Tilson HH. No evident association between efavirenz use and suicidality was identified from a disproportionality analysis using the FAERS database. J Int AIDS Soc. 2014;17:19214. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4156595/