[Table, COMBINED CONTRACEPTIVE PATCH (P) AND COMBINED CONTRACEPTIVE VAGINAL RING (CVR)]. - Medical Eligibility Criteria for Contraceptive Use

COMBINED HORMONAL CONTRACEPTIVES (CHCs)

CHCs do not protect against sexually transmitted infections (STIs), including HIV. If there is a risk of STI/HIV, the correct and consistent use of condoms is recommended. When used correctly and consistently, condoms offer one of the most effective methods of protection against STIs, including HIV. Female condoms are effective and safe, but are not used as widely by national programmes as male condoms.

CONDITION	CATEGORY I = initiation, C = continuation								CLARIFICATIONS/EVIDENCE

	COC		P		CVR		CIC

^† recommendations reviewed for the MEC 5^th edition, further details after this table * additional comments after this table	COC = combined oral contraceptive P = combined contraceptive patch CVR = combined contraceptive vaginal ring CIC = combined injectable contraceptive
PERSONAL CHARACTERISTICS AND REPRODUCTIVE HISTORY

PREGNANCY	NA		NA		NA		NA		NA = not applicable Clarification: Use of COCs, P, CVR or CICs is not required. There is no known harm to the woman, the course of her pregnancy, or the fetus if COCs, P, CVR or CICs are accidentally used during pregnancy.

AGE^†*									Evidence: Evidence about whether CHC use affects fracture risk is inconsistent (78–89), although 3 recent studies show no effect (90–92). CHC use may decrease bone mineral density (BMD) in adolescents, especially in those choosing very low dose formulations (< 30 μg ethinyl estradiol-containing COCs) (91, 93–105). CHC use has little to no effect on BMD in premenopausal women (90, 93–102, 106–109), and may preserve bone mass in those who are perimenopausal (103, 104, 110–117). BMD is a surrogate marker for fracture risk that may not be valid for premenopausal women, and which, therefore, may not accurately predict current or future (postmenopausal) fracture risk (118–120).
a) Menarche to < 40 years	1		1		1		1
b) ≥ 40 years	2		2		2		2

PARITY
a) Nulliparous	1		1		1		1
b) Parous	1		1		1		1

BREASTFEEDING ^†									Evidence: Clinical studies demonstrate conflicting results regarding effects on breastfeeding continuation or exclusivity in women exposed to COCs during lactation. No consistent effects on infant growth or illness have been reported (121–126). Adverse health outcomes or manifestations of exogenous estrogen in infants exposed to combined contraceptives through breast-milk have not been demonstrated; however, studies have been inadequately designed to determine whether a risk of either serious or subtle long-term effects exists.
a) < 6 weeks postpartum	4		4		4		4
b) ≥ 6 weeks to < 6 months postpartum (primarily breastfeeding)	3		3		3		3
c) ≥ 6 months postpartum	2		2		2		2

POSTPARTUM (IN NON-BREASTFEEDING WOMEN) ^† Although the risk of venous thromboembolism (VTE) is the same in breastfeeding and non-breastfeeding women, use of CHCs is generally not recommended prior to 6 months postpartum in women who are breastfeeding.

a) < 21 days									Clarification: For women up to 6 weeks postpartum with other risk factors for VTE, such as immobility, transfusion at delivery, BMI > 30 kg/m², postpartum haemorrhage, immediately post-caesarean delivery, pre-eclampsia or smoking, use of CHCs may pose an additional increased risk for VTE. Evidence: VTE risk is elevated during pregnancy and the postpartum period; this risk is most pronounced in the first 3 weeks after delivery, declining to near baseline levels by 42 days postpartum (127–131). Use of CHCs, which increases the risk of VTE in healthy reproductive-age women, may pose an additional risk during this time (132). Risk of pregnancy during the first 21 days postpartum is very low, but increases after that time in non-breastfeeding women; ovulation before first menses is common (133).
i) without other risk factors for VTE	3		3		3		3
ii) with other risk factors for VTE	4		4		4		4
b) ≥ 21 days to 42 days
i) without other risk factors for VTE	2		2		2		2
ii) with other risk factors for VTE	3		3		3		3
c) > 42 days	1		1		1		1

POST-ABORTION									Clarification: COCs, P, CVR or CICs may be started immediately post-abortion. Evidence: Women who started taking COCs immediately after first-trimester medical or surgical abortion did not experience more side-effects or adverse vaginal bleeding outcomes or clinically significant changes in coagulation parameters compared with women who used a placebo, an IUD, a non-hormonal contraceptive method, or delayed COC initiation (134–141). Limited evidence on women using the CVR immediately after first-trimester medical or surgical abortion indicated no serious adverse events and no infection related to CVR use during 3 cycles of follow-up post-abortion (77).
a) First trimester	1		1		1		1
b) Second trimester	1		1		1		1
c) Immediate post-septic abortion	1		1		1		1

PAST ECTOPIC PREGNANCY*	1		1		1		1

HISTORY OF PELVIC SURGERY	1		1		1		1

SMOKING									Evidence: COC users who smoked were at increased risk of cardiovascular diseases, especially myocardial infarction (MI), compared with those who did not smoke. Studies also showed an increased risk of MI with increasing number of cigarettes smoked per day (30, 31, 142–151).
a) Age < 35 years	2		2		2		2
b) Age ≥ 35 years
i) < 15 cigarettes/day	3		3		3		2
ii) ≥ 15 cigarettes/day	4		4		4		3

OBESITY									Evidence: Obese women who use COCs are more likely to experience VTE than obese women who do not use COCs. The absolute risk of VTE in healthy women of reproductive age is small. Limited evidence suggests that obese women who use COCs do not have a higher risk of acute MI or stroke than obese non-users (146, 147, 151–156). Limited evidence suggests obese women are no more likely to gain weight after 3 cycles of using CVR or COCs than overweight or normal-weight women. A similar weight gain during 3 months was noted in both the COC group and the CVR group across all BMI categories (76). Overall, evidence suggests that contraceptive effectiveness is maintained among obese CHC users (157–172); however, among women with very high BMI using COC, evidence is inconsistent (161, 167, 171). No association was found between pregnancy risk and BMI among P users (161, 167, 171). The effectiveness of the patch decreased among women who weighed > 90 kg in 1 study (172).
a) ≥ 30 kg/m² BMI	2		2		2		2
b) Menarche to < 18 years and ≥ 30 kg/m² BMI	2		2		2		2

BLOOD PRESSURE MEASUREMENT UNAVAILABLE	NA		NA		NA		NA		NA = not applicable Clarification: It is desirable to have blood pressure measurements taken before initiation of COC, P, CVR or CIC use. However, in some settings, blood pressure measurements are unavailable. In many of these settings, pregnancy-related morbidity and mortality risks are high, and COCs, P, CVR or CICs may be among the few methods widely available. In such settings, women should not be denied use of COCs, P, CVR or CICs simply because their blood pressure cannot be measured.

CARDIOVASCULAR DISEASE

MULTIPLE RISK FACTORS FOR ARTERIAL CARDIOVASCULAR DISEASE (e.g. older age, smoking, diabetes, hypertension and known dyslipidaemias)	3/4		3/4		3/4		3/4		Clarification: When a woman has multiple major risk factors, any of which alone would substantially increase the risk of cardiovascular disease, use of COCs, P, CVR or CICs may increase her risk to an unacceptable level. However, a simple addition of categories for multiple risk factors is not intended; for example, a combination of 2 risk factors assigned a Category 2 may not necessarily warrant a higher category.

HYPERTENSION For all categories of hypertension, classifications are based on the assumption that no other risk factors for cardiovascular disease exist. When multiple risk factors do exist, the risk of cardiovascular disease may increase substantially. A single reading of blood pressure level is not sufficient to classify a woman as hypertensive.

a) History of hypertension, where blood pressure CANNOT be evaluated (including hypertension in pregnancy)	3		3		3		3		Clarification: Evaluation of cause and level of hypertension is recommended, as soon as feasible. Evidence: Women who did not have a blood pressure check before initiation of COC use had an increased risk of acute MI and stroke (26, 32, 33, 173, 174).

b) Adequately controlled hypertension, where blood pressure CAN be evaluated	3		3		3		3		Clarification: Women adequately treated for hypertension are at reduced risk of acute MI and stroke as compared with untreated women. Although there are no data, COC, P, CVR or CIC users with adequately controlled and monitored hypertension should be at reduced risk of acute MI and stroke compared with untreated hypertensive COC, P, CVR or CIC users. Evidence: Among women with hypertension, COC users were at increased risk of stroke, acute MI, and peripheral arterial disease compared with non-users (14, 26, 31, 33, 142, 144, 150, 151, 173–185). Discontinuation of COCs in women with hypertension may improve blood pressure control (186).
c) Elevated blood pressure levels (properly taken measurements)
i) systolic 140–159 or diastolic 90–99 mm Hg	3		3		3		3
ii) systolic ≥ 160 or diastolic ≥ 100 mm Hg	4		4		4		4
d) Vascular disease	4		4		4		4

HISTORY OF HIGH BLOOD PRESSURE DURING PREGNANCY (where current blood pressure is measurable and normal)	2		2		2		2		Evidence: Women using COCs who had a history of high blood pressure in pregnancy had an increased risk of MI and VTE, compared with COC users who did not have a history of high blood pressure during pregnancy. The absolute risks of acute MI and VTE in this population remained small (32, 33, 151, 174, 176, 187–192).

DEEP VEIN THROMBOSIS (DVT)/PULMONARY EMBOLISM (PE)*
a) History of DVT/PE	4		4		4		4
b) Acute DVT/PE	4		4		4		4
c) DVT/PE and established on anticoagulant therapy	4		4		4		4
d) Family history (first-degree relatives)	2		2		2		2
e) Major surgery
i) with prolonged immobilization	4		4		4		4
ii) without prolonged immobilization	2		2		2		2
f) Minor surgery without immobilization	1		1		1		1

KNOWN THROMBOGENIC MUTATIONS (e.g. factor V Leiden; prothrombin mutation; protein S, protein C, and antithrombin deficiencies)	4		4		4		4		Clarification: Routine screening is not appropriate because of the rarity of the conditions and the high cost of screening. Evidence: Among women with thrombogenic mutations, COC users had a 2- to 20-fold higher risk of thrombosis than non-users (3, 155, 193–214).

SUPERFICIAL VENOUS DISORDERS ^†
a) Varicose veins	1		1		1		1		Evidence: One study suggested that among women with varicose veins, the rate of VTE and superficial venous thrombosis (SVT) was higher in oral contraceptive users compared with non-users; however, statistical significance was not reported and the number of events was small (215).
b) Superficial venous thrombosis (SVT)	2		2		2		2		Clarification: SVT may be associated with an increased risk of VTE. Evidence: One study demonstrated that among women with SVT, the risk of VTE was higher in oral contraceptive users compared with non-users (216).

CURRENT AND HISTORY OF ISCHAEMIC HEART DISEASE	4		4		4		4

STROKE (history of cerebrovascular accident)	4		4		4		4

KNOWN DYSLIPIDAEMIAS WITHOUT OTHER KNOWN CARDIOVASCULAR RISK FACTORS ^†	2		2		2		2		Clarification: Routine screening is not appropriate because of the rarity of the condition and the high cost of screening. Increased levels of total cholesterol, low-density lipoprotein (LDL) and triglycerides, as well as a decreased level of high-density lipoprotein (HDL), are known risk factors for cardiovascular disease. Women with known severe genetic lipid disorders are at much higher lifetime risk for cardiovascular disease and may warrant further clinical consideration. Evidence: Limited evidence on use of CHCs among women with dyslipidaemia and risk of cardiovascular outcomes provided inconsistent results. One study suggested an increased risk for MI among COC users with hypercholesterolaemia compared to non-users without hypercholesterolaemia (217); 1 study suggested an increased risk for VTE and for stroke among COC users with dyslipidaemia compared to COC users without dyslipidaemia (22); and 1 study suggested no worsening of lipid abnormalities among CHC users with dyslipidaemia compared to non-users with dyslipidaemia (218). No evidence of risk for pancreatitis was identified.

VALVULAR HEART DISEASE*
a) Uncomplicated	2		2		2		2
b) Complicated (pulmonary hypertension, risk of atrial fibrillation, history of subacute bacterial endocarditis)	4		4		4		4

RHEUMATIC DISEASES

SYSTEMIC LUPUS ERYTHEMATOSUS (SLE) People with SLE are at increased risk of ischaemic heart disease, stroke and venous thromboembolism (VTE). Categories assigned to such conditions in the Medical eligibility criteria for contraceptive use should be the same for women with SLE who present with these conditions. For all categories of SLE, classifications are based on the assumption that no other risk factors for cardiovascular disease are present; these classifications must be modified in the presence of such risk factors. Available evidence indicates that many women with SLE can be considered good candidates for most contraceptive methods, including hormonal contraceptives (219–236).

a) Positive (or unknown) antiphospholipid antibodies	4		4		4		4		Evidence: Antiphospholipid antibodies are associated with a higher risk for both arterial and venous thrombosis (237-239).
b) Severe thrombocytopenia	2		2		2		2
c) Immunosuppressive treatment	2		2		2		2
d) None of the above	2		2		2		2

NEUROLOGIC CONDITIONS

HEADACHES*	I	C	I	C	I	C	I	C	Clarification: Classification depends on accurate diagnosis of those severe headaches that are migrainous and those that are not. Any new headaches or marked changes in headaches should be evaluated. Classification is for women without any other risk factors for stroke. Risk of stroke increases with age, hypertension and smoking. Evidence: Among women with migraine, women who also had aura had a higher risk of stroke than those without aura (240–242). Women with a history of migraine who use COCs are about 2–4 times as likely to have an ischaemic stroke as non-users with a history of migraine (142, 154, 181, 182, 240–246).
HEADACHES*
a) Non-migrainous (mild or severe)	1	2	1	2	1	2	1	2
b) Migraine
i) without aura
age < 35 years	2	3	2	3	2	3	2	3
age ≥ 35 years	3	4	3	4	3	4	3	4
ii) with aura, at any age	4	4	4	4	4	4	4	4

EPILEPSY	1		1		1		1		Clarification: If a woman is taking anticonvulsants, refer to the last section of this table, on drug interactions. Certain anticonvulsants lower COC effectiveness. The extent to which P, CVR or CIC use is similar to COC use in this regard remains unclear.

DEPRESSIVE DISORDERS

DEPRESSIVE DISORDERS	1		1		1		1		Clarification: The classification is based on data for women with selected depressive disorders. No data on bipolar disorder or postpartum depression were available. There is a potential for drug interactions between certain antidepressant medications and hormonal contraceptives. Evidence: COC use did not increase depressive symptoms in women with depression compared to baseline or to non-users with depression (247–256).

REPRODUCTIVE TRACT INFECTIONS AND DISORDERS

VAGINAL BLEEDING PATTERNS*
a) Irregular pattern without heavy bleeding	1		1		1		1
b) Heavy or prolonged bleeding (includes regular and irregular patterns)	1		1		1		1		Clarification: Unusually heavy bleeding should raise the suspicion of a serious underlying condition. Evidence: A Cochrane Collaboration review identified 1 randomized controlled trial evaluating the effectiveness of COC use compared with naproxen and danazol in treating menorrhagic women. Women with menorrhagia did not report worsening of the condition or any adverse events related to COC use (257).

UNEXPLAINED VAGINAL BLEEDING* (suspicious for serious condition)
a) Before evaluation	2		2		2		2		Clarification: If pregnancy or an underlying pathological condition (such as pelvic malignancy) is suspected, it must be evaluated and the category adjusted after evaluation.

ENDOMETRIOSIS	1		1		1		1		Evidence: A Cochrane review identified 1 randomized controlled trial evaluating the effectiveness of COC use compared with a gonadotropin-releasing hormone (GnRH) analogue in treating the symptoms of endometriosis. Women with endometriosis did not report worsening of the condition or any adverse events related to COC use (258).

BENIGN OVARIAN TUMOURS (INCLUDING CYSTS)	1		1		1		1

SEVERE DYSMENORRHOEA	1		1		1		1		Evidence: There was no increased risk of side-effects with COC use among women with dysmenorrhoea compared with women not using COCs. Some COC users had a reduction in pain and bleeding (259, 260).

GESTATIONAL TROPHOBLASTIC DISEASE									Evidence: Following molar pregnancy evacuation, the balance of evidence found COC use did not increase the risk of post-molar trophoblastic disease, and some COC users experienced a more rapid regression in human chorionic gonadotropin (hCG) levels, compared with non-users (261–268). Limited evidence suggests that use of COCs during chemotherapeutic treatment does not significantly affect the regression or treatment of post-molar trophoblastic disease compared with women who used a non-hormonal contraceptive method or depot medroxyprogesterone acetate (DMPA) during chemotherapeutic treatment (269).
a) Decreasing or undetectable β-hCG levels	1		1		1		1
b) Persistently elevated β-hCG levels or malignant disease	1		1		1		1

CERVICAL ECTROPION*	1		1		1		1

CERVICAL INTRAEPITHELIAL NEOPLASIA (CIN)	2		2		2		2		Evidence: Among women with persistent human papillomavirus (HPV) infection, long-term COC use (≥ 5 years) may increase the risk of carcinoma in situ and invasive carcinoma (64, 270). Limited evidence on women with low-grade squamous intraepithelial lesions found use of the vaginal ring did not worsen the condition (64).

CERVICAL CANCER* (AWAITING TREATMENT)	2		2		2		2

BREAST DISEASE*
a) Undiagnosed mass	2		2		2		2		Clarification: Evaluation should be pursued as early as possible.
b) Benign breast disease	1		1		1		1
c) Family history of cancer	1		1		1		1		Evidence: Women with breast cancer susceptibility genes (such as BRCA1 and BRCA2) have a higher baseline risk of breast cancer than women without these genes. The baseline risk of breast cancer is also higher among women with a family history of breast cancer than among those who do not have such a history. Current evidence, however, does not suggest that the increased risk of breast cancer among women with either a family history of breast cancer or breast cancer susceptibility genes is modified by the use of combined oral contraceptives (175, 271–293).
d) Breast cancer
i) current	4		4		4		4
ii) past and no evidence of current disease for 5 years	3		3		3		3

ENDOMETRIAL CANCER*	1		1		1		1

OVARIAN CANCER*	1		1		1		1

UTERINE FIBROIDS*
a) Without distortion of the uterine cavity	1		1		1		1
b) With distortion of the uterine cavity	1		1		1		1

PELVIC INFLAMMATORY DISEASE (PID)*
a) Past PID (assuming no current risk factors for STIs)
i) with subsequent pregnancy	1		1		1		1
ii) without subsequent pregnancy	1		1		1		1
b) PID – current	1		1		1		1

STIs
a) Current purulent cervicitis or chlamydial infection or gonorrhoea	1		1		1		1
b) Other STIs (excluding HIV and hepatitis)	1		1		1		1
c) Vaginitis (including Trichomonas vaginalis and bacterial vaginosis)	1		1		1		1
d) Increased risk of STIs	1		1		1		1		Evidence: Evidence suggests that there may be an increased risk of chlamydial cervicitis among COC users at high risk of STIs. For other STIs, there is either evidence of no association between COC use and STI acquisition or too limited evidence to draw any conclusions (289–369).

HIV/AIDS ^†

High risk of HIV	1		1		1		1		Evidence: Eight studies assessed the use of COCs and were considered to be “informative but with important limitations” (370). Seven of these studies found no statistically significant association between use of COCs and HIV acquisition (371–378), although 1 study among sex workers in Kenya did (379).

Asymptomatic or mild HIV clinical disease (WHO stage 1 or 2)	1		1		1		1		Clarification for asymptomatic or mild HIV disease (WHO stage 1 or 2) and severe or advanced HIV disease (WHO stage 3 or 4): Because there may be drug interactions between hormonal contraceptives and ARV therapy, refer to the last section of this table, on drug interactions. Evidence for asymptomatic or mild HIV disease (WHO stage 1 or 2) and severe or advanced HIV disease (WHO stage 3 or 4): Out of 8 available studies, 7 suggested no association between use of COCs and progression of HIV, as measured by CD4 count < 200 cells/mm3, initiation of antiretroviral therapy (ART), or mortality (380–386). One randomized controlled trial found an increased risk of a composite outcome of declining CD4 count or death among COC users when compared with users of copper-bearing IUDs (387, 388). Two prospective observational studies directly assessed the effects of different hormonal contraceptive methods on female-to-male HIV transmission by measuring seroconversions in male partners of women known to be using hormonal contraceptives. One of these studies reported an elevated, but not statistically significant, point estimate for COCs (378). The other study also did not find a statistically significant association for COCs (389). Studies indirectly assessing the effect of various hormonal contraceptive methods on female-to-male HIV transmission by measuring genital viral shedding as a proxy for infectivity have had mixed results. The majority of indirect studies measuring whether various hormonal contraceptive methods affect plasma HIV viral load have found no effect (381, 390–404).
Severe or advanced HIV clinical disease (WHO stage 3 or 4)	1		1		1		1

OTHER INFECTIONS

SCHISTOSOMIASIS
a) Uncomplicated	1		1		1		1		Evidence: Among women with uncomplicated schistosomiasis, COC use had no adverse effects on liver function (405–411).
b) Fibrosis of the liver (if severe, see cirrhosis)	1		1		1		1

TUBERCULOSIS
a) Non-pelvic	1		1		1		1		Clarification: If a woman is taking rifampicin, refer to the last section of this table, on drug interactions. Rifampicin is likely to decrease COC effectiveness. The extent to which P or CVR use is similar to COC use in this regard remains unclear.
b) Pelvic	1		1		1		1

MALARIA	1		1		1		1

ENDOCRINE CONDITIONS

DIABETES
a) History of gestational disease	1		1		1		1		Evidence: The development of non-insulin-dependent diabetes in women with a history of gestational diabetes is not increased by the use of COCs (412–419). Likewise, lipid levels appear to be unaffected by COC use (420–422).
b) Non-vascular disease									Evidence: Among women with insulin- or non-insulin-dependent diabetes, COC use had limited effect on daily insulin requirements and no effect on long-term diabetes control (e.g. haemoglobin A1c levels) or progression to retinopathy. Changes in lipid profile and haemostatic markers were limited, and most changes remained within normal values (419, 422–430).
i) non-insulin dependent	2		2		2		2
ii) insulin dependent	2		2		2		2
c) Nephropathy/retinopathy/neuropathy	3/4		3/4		3/4		3/4		Clarification: The category should be assessed according to the severity of the condition.
d) Other vascular disease or diabetes of > 20 years' duration	3/4		3/4		3/4		3/4		Clarification: The category should be assessed according to the severity of the condition.

THYROID DISORDERS
a) Simple goitre	1		1		1		1
b) Hyperthyroid	1		1		1		1
c) Hypothyroid	1		1		1		1

GASTROINTESTINAL CONDITIONS

GALL BLADDER DISEASE*
a) Symptomatic
i) treated by cholecystectomy	2		2		2		2
ii) medically treated	3		3		3		2
iii) current	3		3		3		2
b) Asymptomatic	2		2		2		2

HISTORY OF CHOLESTASIS*
a) Pregnancy related	2		2		2		2
b) Past-COC related	3		3		3		2

VIRAL HEPATITIS	I	C	I	C	I	C	I	C
VIRAL HEPATITIS
a) Acute or flare	3/4	2	3/4	2	3/4	2	3	2	Clarification: The category should be assessed according to the severity of the condition.
b) Carrier	1	1	1	1	1	1	1	1	Evidence: Data suggest that in women with chronic hepatitis, COC use does not increase the rate or severity of cirrhotic fibrosis, nor does it increase the risk of hepatocellular carcinoma (431, 432). For women who are carriers, COC use does not appear to trigger liver failure or severe dysfunction (408, 433, 434). Evidence is limited for COC use during active hepatitis (435, 436).
c) Chronic	1	1	1	1	1	1	1	1

CIRRHOSIS
a) Mild (compensated)	1		1		1		1
b) Severe (decompensated)	4		4		4		3

LIVER TUMOURS*
a) Benign
i) focal nodular hyperplasia	2		2		2		2		Evidence: There is limited, direct evidence that hormonal contraceptive use does not influence either progression or regression of liver lesions among women with focal nodular hyperplasia (437–439).
ii) hepatocellular adenoma	4		4		4		3
b) Malignant (hepatoma)	4		4		4		3/4

ANAEMIAS

THALASSAEMIA*	1		1		1		1

SICKLE CELL DISEASE	2		2		2		2

IRON-DEFICIENCY ANAEMIA*	1		1		1		1

DRUG INTERACTIONS

ANTIRETROVIRAL THERAPY (ART) ^†
a) Nucleoside reverse transcriptase inhibitors (NRTIs)									Evidence: NRTIs do not appear to have significant risk of interactions with hormonal contraceptive methods (440, 441).
Abacavir (ABC)	1		1		1		1
Tenofovir (TDF)	1		1		1		1
Zidovudine (AZT)	1		1		1		1
Lamivudine (3TC)	1		1		1		1
Didanosine (DDI)	1		1		1		1
Emtricitabine (FTC)	1		1		1		1
Stavudine (D4T)	1		1		1		1

b) Non-nucleoside reverse transcriptase inhibitors (NNRTIs)									Clarification: Antiretroviral drugs have the potential to either decrease or increase the levels of steroid hormones in women using hormonal contraceptives. Pharmacokinetic data suggest potential drug interactions between some antiretroviral drugs (particularly some NNRTIs and ritonavir-boosted PIs) and some hormonal contraceptives. These interactions may reduce the effectiveness of the hormonal contraceptive. Evidence: Three clinical studies, including 1 large study, found use of nevirapine-containing ART did not increase ovulation or pregnancy rates in women taking COCs (442–445). For efavirenz-containing ART, a pharmacokinetic study showed consistent significant decreases in contraceptive hormone levels in women taking COCs, and a small clinical study showed higher ovulation rates in women taking efavirenz-containing ART and COCs (445–447). Etravirine and rilpivirine do not interact with COCs (448, 449).
Efavirenz (EFV)	2		2		2		2
Etravirine (ETR)	1		1		1		1
Nevirapine (NVP)	2		2		2		2
Rilpivirine (RPV)	1		1		1		1
c) Protease inhibitors (PIs)
Ritonavir-boosted atazanavir (ATV/r)	2		2		2		2
Ritonavir-boosted lopinavir (LPV/r)	2		2		2		2
Ritonavir-boosted darunavir (DRV/r)	2		2		2		2
Ritonavir (RTV)	2		2		2		2
d) Integrase inhibitors									Evidence: The integrase inhibitor raltegravir does not appear to interact with COCs (440, 441, 454, 455).
Raltegravir (RAL)	1		1		1		1

ANTICONVULSANT THERAPY
a) Certain anticonvulsants (phenytoin, carbamazepine, barbiturates, primidone, topiramate, oxcarbazepine)	3		3		3		2		Clarification: Although the interaction of certain anticonvulsants with COCs, P or CVR is not harmful to women, it is likely to reduce the effectiveness of COCs, P or CVR. Use of other contraceptives should be encouraged for women who are long-term users of any of these drugs. When a COC is chosen, a preparation containing a minimum of 30 μg of ethinyl estradiol (EE) should be used. Evidence: Use of certain anticonvulsants may decrease the effectiveness of COCs (456-459).
b) Lamotrigine	3		3		3		3		Clarification: The recommendation for lamotrigine does not apply when lamotrigine is already being taken with other drugs that strongly inhibit (such as sodium valproate) or induce (such as carbamazepine) its metabolism, since, in these cases, the moderate effect of the combined contraceptive is unlikely to be apparent. Evidence: Pharmacokinetic studies show levels of lamotrigine decrease significantly during COC use and increase significantly during the pill-free interval (460-464). Some women who used both COCs and lamotrigine experienced increased seizure activity in 1 trial (464).

ANTIMICROBIAL THERAPY
a) Broad-spectrum antibiotics	1		1		1		1		Evidence: Most broad-spectrum antibiotics do not affect the contraceptive effectiveness of COCs (465–501), P (502), or CVR (503).
b) Antifungals	1		1		1		1		Evidence: Studies of antifungal agents have shown no clinically significant pharmacokinetic interactions with COCs (504–513) or CVR (514).
c) Antiparasitics	1		1		1		1		Evidence: Studies of antiparasitic agents have shown no clinically significant pharmacokinetic interactions with COCs (411, 515–519).
d) Rifampicin or rifabutin therapy	3		3		3		2		Clarification: Although the interaction of rifampicin or rifabutin therapy with COCs, P, CVR or CICs is not harmful to women, it is likely to reduce the effectiveness of COCs, P, CVR or CICs. Use of other contraceptives should be encouraged for women who are long-term users of either of these drugs. When a COC is chosen, a preparation containing a minimum of 30 μg EE should be used. Evidence: The balance of the evidence suggests that rifampicin reduces the effectiveness of COCs (520–535). Data on rifabutin are limited, but effects on metabolism of COCs are less than with rifampicin, and small studies have not shown evidence of ovulation (363, 522, 535).

From: II, Using the recommendations

Cover of Medical Eligibility Criteria for Contraceptive Use

Medical Eligibility Criteria for Contraceptive Use. 5th edition.

Geneva: World Health Organization; 2015.

All rights reserved. Publications of the World Health Organization are available on the WHO website (www.who.int) or can be purchased from WHO Press, World Health Organization, 20 Avenue Appia, 1211 Geneva 27, Switzerland (tel.: +41 22 791 3264; fax: +41 22 791 4857; e-mail: tni.ohw@sredrokoob).

Requests for permission to reproduce or translate WHO publications –whether for sale or for non-commercial distribution– should be addressed to WHO Press through the WHO website (www.who.int/about/licensing/copyright_form/en/index.html).

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.