Vaginal yeast infection in pregnancy: a small trial

One small Thai randomized trial compared two specific single-dose vaginal azoles in 96 pregnant adults with acute, microscopy-confirmed candidiasis. The two-week clinical and microscopic cure estimates were imprecise and pointed in opposite directions. The study did not test equivalence, compare oral with vaginal treatment, or establish rare pregnancy safety, so it does not support a new treatment action.

Most acute-treatment comparisons behind general guidance exclude pregnancy. This trial adds a narrow, direct comparison in pregnant participants, while its size and reporting limits leave substantial uncertainty.

This single-centre, double-blind randomized trial took place at Siriraj Hospital in Bangkok, Thailand. It screened 129 people, excluded 29, and randomized 100 pregnant participants (50 per group). Two participants per group were lost before the 14 ± 2-day assessment, leaving 96 in the primary analysis (48 per group). Participants were aged 18 or older and had symptoms plus pseudohyphae on a wet mount.

Trained staff inserted one 300 mg sertaconazole vaginal ovule or one 500 mg clotrimazole vaginal tablet. The primary outcome was self-reported resolution of all initial symptoms at 14 ± 2 days; microscopic cure meant no pseudohyphae on wet preparation. Participants also received lifestyle counselling. Those without microscopic cure at the first visit received further weekly sertaconazole regardless of their original assignment, so later outcomes do not preserve the initial randomized comparison.

At the primary assessment, clinical cure was 30/48 (62.5%) with sertaconazole and 24/48 (50.0%) with clotrimazole: a difference of 12.5 percentage points (95% CI −17.5 to 42.5; p=0.217). Microscopic cure was 23/48 (47.9%) with sertaconazole and 30/48 (62.5%) with clotrimazole: a difference of −14.6 points (95% CI −44.3 to 15.1; p=0.151). Neither result establishes equivalence or a better treatment.

Table 2 lists time to clinical cure among those cured as 3.1 ± 1.8 days for clotrimazole and 3.4 ± 2.7 days for sertaconazole (p=0.848). The abstract and Results prose reverse these group labels, and separate denominators are not reported, so this result cannot support a claim that either medicine worked faster. Table 2 lists substantial satisfaction in 32/48 (66.7%) clotrimazole participants and 29/48 (60.4%) sertaconazole participants (p=0.753); the abstract reverses this order, while the table and Results prose agree.

No local or systemic side effects were reported, but a sample of 96 cannot rule out rare harms. After initial microscopic cure, recurrence was reported in 2 clotrimazole and 4 sertaconazole participants; the denominators and arm-specific recurrence rates were not reported.

Pregnancy outcomes were reported for 60 participants who delivered at Siriraj Hospital (30 per group), not all 96 analyzed participants. The article reported no statistically significant differences. Table 3 lists gestational age at delivery as 37.9 ± 2.2 weeks with clotrimazole and 37.9 ± 1.5 weeks with sertaconazole (p=0.942); low birth weight as 2/30 (6.7%) versus 4/30 (13.3%) (p=0.389); and Apgar below 7 as 2/30 (6.7%) versus 0/30 (p=0.150). It lists 3 preterm events in the clotrimazole column, printed as 9.7% under n=30, versus 4/30 (13.3%) with sertaconazole (p=0.654); the denominator behind 9.7% is not stated.

The results describe two specific single doses in adults with confirmed acute candidiasis at one Thai hospital. They do not compare oral with vaginal treatment, establish a general treatment hierarchy, or determine what another product or regimen does. The study excluded recurrent infection and several higher-risk groups. Its findings do not replace current, exact-product pregnancy guidance in either locale.

The study was small, excluded four randomized participants from the primary analysis, and did not report an intention-to-treat result. The authors did not use fungal cultures and noted uncertainty about the meaning of microscopy-directed retreatment. Only 60 of 96 analyzed participants contributed hospital-delivery outcomes; the remaining pregnancy outcomes were unavailable or not reported. These counts cannot establish rare fetal or neonatal safety.

The arm labels for time to cure conflict between Table 2 and the abstract/Results prose; this summary reports the Table 2 values and preserves the unresolved discrepancy. The abstract also reverses the satisfaction arm order, so the summary uses the Table 2 and Results values. Table 3's printed 9.7% preterm rate does not match its n=30 column, and the article does not explain the denominator. Recruitment dates also conflict: the abstract says June 2020–May 2021, while Methods says June 2019–January 2021.

Contextual. The trial gives a limited view of two specific topical regimens in acute pregnancy, a population usually excluded from the broader comparisons. Its imprecise results and design do not establish equivalence, route superiority, or rare-event safety.

No practical change. The study does not change the nonpregnant claims summarized in the linked evidence map or the existing product-specific pregnancy action. It does not test an oral-versus-vaginal comparison or support a new treatment recommendation.

Mahidol University funded open access. Bayer Thailand and Pacific Health Care Thailand supplied the study drugs; Pacific Health Care also supported participant transport and administrative costs totaling 150,000 Baht. The two companies supported project expenses, and two authors received speaking honoraria from them. The authors reported that the companies had no role in data analysis or manuscript preparation; the other two authors reported no competing interests.

Read the original publication (published 4 April 2024).