Pregnant women in China know their medicines but still need counseling, survey finds

When a woman becomes pregnant, every tablet, capsule, and herbal supplement she takes suddenly carries two sets of stakes: her own health and that of the developing baby. A new survey of nearly 500 pregnant and postpartum women at Taiyuan Taihang Hospital in northern China suggests that many of these women have absorbed years of public medication-safety education, yet still feel unprepared when a specific treatment decision arises. The study, published in Frontiers in Public Health, used a risk-oriented Knowledge-Attitude-Practice (KAP) questionnaire to map what women know, believe, and do about medicines during pregnancy and lactation, and the results reveal a striking gap between general preparedness and decision-specific confidence.

The distinction between knowing something in general and being able to apply it at a specific moment is central to interpreting the Taiyuan findings. The mean knowledge-risk score of 59.78 and behavior-risk score of 57.22 both fell within the prespecified good category, which suggests that the women surveyed had absorbed a reasonable amount of medication-safety information and reported generally sensible practices. Yet the attitude-risk score, averaging 37.34 and landing only in the pass category, tells a different story. Nearly six in ten participants were classified as pass on attitude risk, while one in six fell into the fail range and only about a quarter reached the good or excellent tiers. In practical terms, this means that a substantial share of women held beliefs or concerns about medication use that were less well aligned with safety guidance than their knowledge or reported behavior would suggest.

This pattern is not merely a statistical curiosity. Attitudes toward medication during pregnancy often carry an emotional dimension that factual knowledge alone does not resolve. A woman may know that untreated gestational hypertension poses serious risks, yet still feel deep reluctance to take an antihypertensive drug because of lingering fears about fetal harm. The weak positive correlation between attitude risk and behavior risk (r = 0.118) observed in the study hints that beliefs and actions were only loosely coupled in this sample, while the moderate inverse correlation between knowledge risk and attitude risk (r = −0.258) suggests that women with better factual knowledge tended to hold somewhat more favorable attitudes. The positive correlation between knowledge risk and behavior risk (r = 0.275) is more counterintuitive at first glance, since higher knowledge-risk scores mean less favorable knowledge, and these were associated with less favorable behavior. Taken together, the correlation structure implies that the three domains move in related but imperfect ways, and that improving one domain does not automatically improve the others.

The study’s measurement approach deserves some attention for readers unfamiliar with survey methodology. Risk-oriented scoring reverses the usual convention of KAP questionnaires: instead of higher scores indicating better outcomes, higher scores on the knowledge, attitude, and behavior domains indicate greater risk or less favorable profiles. This design choice keeps all unfavorable responses pointing in the same numerical direction, which simplifies interpretation of correlations across domains. The internal consistency of the questionnaire, assessed with Cronbach’s alpha, ranged from 0.808 to 0.957 across the scales, values generally considered acceptable to excellent for survey instruments. This indicates that the items within each domain tended to measure a coherent underlying construct rather than a grab bag of unrelated questions.

The statistical methods also reflect careful attention to the realities of survey data. Welch’s analysis of variance, which does not assume equal variances across groups, was followed by Games-Howell post hoc tests, a combination appropriate when group sizes and spreads differ. The multivariable regression models used ordinary least squares with HC3 robust standard errors, a technique that guards against misleading conclusions when the variability of residuals is not uniform across the range of predictor values. These choices do not change the cross-sectional nature of the design, which limits causal inference, but they do strengthen confidence in the associations that were reported.

Among the demographic findings, the elevated behavior-risk scores among women aged 35 to 49 years stand out. This age group encompasses women who may have more chronic health conditions, more prior medication experience, and potentially more entrenched habits around self-medication. It also includes women who may face higher-risk pregnancies by virtue of age alone. The finding that participants with other insurance types or self-payment showed higher behavior risk points to a possible structural dimension: access to consistent pharmaceutical counseling may vary with insurance status, and women paying out of pocket may face different incentives around seeking professional advice before using a medicine. The income analysis illustrates the hazards of overinterpreting isolated contrasts. Although the comparison between women earning 4,000 to 6,000 yuan per month and those earning below 2,000 yuan reached nominal statistical significance, the overall test of income across all groups did not (Wald p = 0.098). The researchers appropriately labeled the isolated contrast as exploratory, a decision that reflects sound statistical practice when a global test fails but a single pairwise comparison appears significant by chance or through a fragile pattern in the data.

The familiarity scores deserve particular emphasis because they reveal something that aggregate domain scores cannot. Across five pregnancy- and lactation-specific medication topics, self-rated familiarity ranged only from 2.46 to 2.60 on a five-point scale. These values sit just above the midpoint, indicating that women felt only moderately familiar with topics that are directly relevant to the decisions they face. The narrow range across topics suggests that no single area of medication knowledge stood out as well covered; instead, familiarity was uniformly modest. This is precisely the kind of finding that broad KAP totals can obscure. A woman might score well on general medication knowledge while feeling uncertain about, for example, which analgesics are compatible with breastfeeding or how to manage a chronic medication during the transition from pregnancy to lactation.

The pharmacological backdrop helps explain why such uncertainty matters. Pregnancy alters gastric emptying, plasma volume, hepatic metabolism, and renal clearance, all of which can change how much of a drug reaches maternal circulation and how quickly it is eliminated. These changes evolve across gestation, meaning that a dose that was appropriate in the second trimester may require adjustment later or after delivery. Medications taken while breastfeeding add another layer of complexity, since infant exposure depends on drug properties, milk composition, and feeding patterns. The study’s authors note that avoiding medication altogether can itself cause harm when necessary treatment is delayed or inadequate, a point that deserves emphasis because excessive caution is a common and underrecognized medication-safety problem in pregnancy. Untreated asthma, epilepsy, depression, hypertension, and infections can carry risks to both mother and fetus that exceed the risks of appropriately chosen therapy.

The treatment decisions women face also shift across the maternity timeline. During pregnancy, questions may center on managing nausea, heartburn, anemia, or chronic disease. Around delivery, analgesia and anesthesia choices come to the fore. After discharge, women may confront lactation suppression, postpartum pain management, wound care, mental health treatment, and the resumption or adjustment of pre-pregnancy medications. A single educational session delivered at one point in this sequence cannot address all of these questions. The study’s recommendation to integrate medication-safety counseling broadly into antenatal and postpartum care reflects this reality, suggesting repeated touchpoints rather than one-time instruction. The additional suggestion of proactive medication review for women aged 35 to 49, those with other insurance or self-payment, and those with complex medication needs represents a targeted approach that concentrates resources where the adjusted models indicated higher risk.

It is worth situating these findings within the broader Chinese policy context. The Healthy China 2030 strategy frames maternal care within a life-course approach to health services, and national policy calls for integrated care spanning pregnancy and the postpartum period. Medication safety fits naturally within this framework, since pharmacies, community health services, antenatal clinics, and maternity wards all encounter women at moments when medicines can be reviewed. The Taiyuan study provides local evidence that can inform how such integration is designed, though the authors are careful to note that a single-hospital convenience sample cannot represent all women in Taiyuan or elsewhere in China. Convenience sampling, by definition, recruits participants who are accessible rather than randomly selected, which can introduce selection effects that are difficult to quantify.

Comparisons with other settings reinforce the value of examining knowledge, attitudes, and practice separately. Research in western Uganda found that 42.6 percent of pregnant women had adequate knowledge and 34.5 percent demonstrated rational medication practice despite generally positive attitudes, illustrating that these domains can diverge in different populations and different directions. National health-insurance database studies in mainland China have documented which medications are actually used during pregnancy, including some potentially higher-risk exposures, but such data do not reveal how women understood the risks or reached their decisions. The KAP approach addresses this complementary question, probing the reasoning and beliefs that underlie observed behavior.

For clinicians and public health planners, several practical implications emerge from this work. First, counseling should not assume that favorable knowledge translates into confident decision-making; the attitude-risk findings suggest that addressing fears and beliefs directly may be as important as conveying facts. Second, topic-specific familiarity assessments could become a routine part of antenatal visits, helping clinicians identify which specific decisions a woman feels unprepared to make. Third, the demographic correlates point to groups who might benefit from proactive outreach, particularly older pregnant and postpartum women and those navigating the healthcare system without standard insurance coverage. Finally, the authors’ call for multicenter prospective studies is a reminder that this cross-sectional survey describes a snapshot at one hospital and one moment in time. Prospective designs could follow women across pregnancy and the postpartum period, tracking how knowledge, attitudes, and behavior evolve as treatment decisions change, and could evaluate whether integrated counseling services actually improve medication safety outcomes for mothers and infants alike.

Subject of Research: Medication-related knowledge, attitudes, and practices among pregnant and postpartum women in Taiyuan, China

Article Title: Medication-related knowledge and behavior profiles are generally favorable, but counseling needs remain among pregnant and postpartum women in Taiyuan, China: a cross-sectional study

Article References: Zhang, C., He, Y., Yang, Y., Zhang, L., Zhang, W., Yang, L., & Ma, M. (2026). Medication-related knowledge and behavior profiles are generally favorable, but counseling needs remain among pregnant and postpartum women in Taiyuan, China: a cross-sectional study. Frontiers in Public Health, 14, Article 1969907. https://doi.org/10.3389/fpubh.2026.1969907

Image Credits: AI Generated

DOI: 10.3389/fpubh.2026.1969907

Keywords: medication safety, pregnancy, postpartum, lactation, Knowledge-Attitude-Practice, counseling, pharmacists, Taiyuan, China, antenatal care, self-medication, cross-sectional study