Table 1.
Evidence table for use of selective serotonin reuptake inhibitors (SSRIs) with hormonal contraception
| Author, year, funding source | Design | Study population | Exposure psychotropic drug | Exposure contraceptive | Outcomes | Results | Strengths | Limitations | Quality |
|---|---|---|---|---|---|---|---|---|---|
| Clinical Studies for SSRIs | |||||||||
| Koke, S. 2002 Eli Lilly and Company |
Retrospective cohort; secondary analysis of 17 RCTs | 1698 females, ages 15–45, with major depression, OCD, or bulimia 353 OC users (232 fluoxetine/121 placebo) 1345 non-OC users (913 fluoxetine/432 placebo) |
Fluoxetine (5–80 mg) vs. placebo | OCs vs non-OCs, otherwise undefined, used at any point during the study period (range 5–16 weeks) | HRDS-17 scores for changes in depression symptoms Unplanned pregnancy Treatment-emergent adverse events (occurred for first time or worsened during treatment period) |
No statistical interaction between fluoxetine and OCs on changes in depression scores over the study period (p>.15 for all measures) Unintended pregnancy: Fluoxetine +OC: 2/232 Placebo +OC: 0/121 (p=.11) Adverse events equal except headache (p=.042), asthenia (p=.021), and pain (p=.031) greater in Fluoxetine group Increased headaches in OC users (OR 2.1,1.2–3.6) Fluoxetine was superior to placebo in all measures for improvement of depression equal among OC vs non-OC users |
Large sample size Only study to look at unplanned pregnancy outcomes |
Varied dosages of SSRI No information on follow-up rates across studies OC use may not have occurred at the time of unintended pregnancy; unclear how OC use was measured, unclear who was in the comparison group (may have included other hormonal contraceptive methods), range of doses and formulations No information on how unplanned pregnancy was measured No control for potential confounders other than study |
II-2, Fair |
| Kornstein, S. G. 2013 National Institute of Mental Health grant and medications provided by Bristol-Myers Squibb, Forest Laboratories,GlaxoSmithKline, King Pharmaceuticals, Organon, Pfizer, and Wyeth | Prospective cohort | Premenopausal women with Major Depressive Disorder, (HRSD-17>14) age<40 given citalopram for up to 14 weeks (n=896) 226 HC users 670 non-HC users |
Citalopram dose titrated to effect | Hormonal contraceptives (all types) | -Depression scales-HRSD17, QIDS-SR16 and the QIDS-C16 over 14 weeks -Frequency, Intensity, and Burden of Side Effects Rating Scale-Adverse events |
HC vs no HC, adjusted OR, 95% CI HRSD17: 1.2, 0.9–1.8 QIDS-SR remission:1.2, 0.9–1.9 QIDS-SR response:1.2, 0.8–1.6 No difference in side effect frequency (p=.62), intensity(p=.68) or burden (p=.21) and no difference in exiting the study (14.2% vs 17.8%; p=.21) or serious adverse events (3.5% vs. 3.6%; p=.98) |
Adjusted for baseline demographic and clinical features that differed between groups at baseline | No information on response rates or follow-up rates use was measured, unclear who was in the comparison group (may have included other hormonal contraceptive methods), range of doses and formulations. |
II-2, fair |
| PK studies for SSRIs | |||||||||
| Carlsson. 2001 and Reis, 2002 H. Lundbeck, the Lion research fund, and Swedish Medical Research Council |
PK Parallel | Adolescents females with major depressive disorder or dysthymia on citalopram (n=16) 10 COC users 6 non-OC users |
Citalopram (10–60 mg) for>2 months | COCs (any type) | Steady state plasm concentration of the S and R enantiomers of citalopram and metabolites (DCIT and DDCIT) Concentration/dose ratio correlations for citalopram and metabolites Single trough value (10–30 h) |
Sig. correlation of S/R ratio for citalopram/DCIT in women not taking COCs (r=0.963, pb.001) however no sig correlation in those taking COCs (r=0.403, p=.24) Concentration/dose correlations:citalopram: r2 0.33 (p=.08) COC users vs r2 0.75 (p=.02) DCIT: r2 0.33 (p=.08) COC users vs r2 0.71 (p=.03) DDCIT: r2 0.33 (p=.15) COC users vs r2 0.23 (p=.41) |
Unclear clinical relevance of primary outcomes Wide range of timing for blood samples (10–30 h after last dose) Wide range of citalopram dose Unclear how COC use was measured, unclear who was in the comparison group (may have included other hormonal contraceptive methods), range of doses and formulations |
poor | |
| Reis, M. 2007 H Lundbeck AB, Berzelius Clinical Research Center |
PK Parallel | Women taking escitalopram for various diagnoses on HC (n=11) vs age matched women not on any hormones (n=42) | Escitalopram (5–40 mg) | Hormonal contraceptives (all types) | S-DCIT/Scitalopram ratio Adverse events Two trough values (10–30 h) |
S-DCIT/S-citalopram ratio HC users vs non-users 0.46 vs 0.75 (p=.02) Authors concluded clinical impact likely low. |
Age matched controls | Unclear clinical relevance of primary outcome Wide range of citalopram dose (5–40 mg) Wide range of timing for blood samples (10–30 h after last dose) however when corrected for time then no differences seen. |
poor |
| Chen, G. 2013. Takeda Pharmaceutical company and H. Lunbeck A/S |
PK Single-blind, randomized, placebocontrolled, 2-sequence, 2-period, crossover | Healthy non pregnant women 18–45 years, BMI 18–30 kg/m2(n=28) | Placebo + COC for 21 days, Crossover 35 day washout Vortioxetine 10 mg + COC for 21 days, crossover | EE/LNG 30/150 mcg | AUC24, Cmax of EE/LNG (ratio of parameters vortioxetine/placebo) | EE: AUC ratio no change (test/reference) 99.1 (96.2–102.3) LNG: AUC ratio(test/reference) 93.9 (89.3–106.75) EE: Cmax ratio ↓ 6.1% (test/reference) 93.9 (89.3–98.7) LNG: Cmax ratio ↑ 7.1% (test/reference) 107.1 (100.9–113.65) |
Crossover design, adequate sample size, single COC used, meaningful outcome | Rare antidepressant used Not a traditional SSRI; works via serotonin reuptake inhibition and receptor activity. Healthy women, not women with depression |
good |
Cmax, Maximum concentration; OCD, obsessive–compulsive disorder; OR, odds ratio; QIDR-C16, quick inventory of depressive symptomatology–clinician rated; QIDS-SR16, quick inventory of depressive symptomatology–self-report; RCT, randomized control trial.