JAMA Health Forum. 2023 Jul 7;4(7):e231974. doi: 10.1001/jamahealthforum.2023.1974.
ABSTRACT
IMPORTANCE: Cervical cancer screening is associated with reduced cervical cancer mortality; however, clinical trials have also shown an association between excisional procedures for cervical neoplasia and an increased risk of preterm delivery (PTD). National screening guidelines must weigh adverse effects on birth outcomes against benefits of cancer prevention.
OBJECTIVE: To ascertain the population-level association between the number of guideline-recommended cervical cancer screenings and downstream PTD risk among females aged 18 to 24 years.
DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study used a difference-in-differences approach based on variation in the recommended number of cervical cancer screenings (over time and across individuals giving birth at different ages) to estimate the association between the cumulative recommended number of screenings (by the time of childbirth) and PTD risk. National Vital Statistics System data from females aged 18 to 24 years who had a singleton, nulliparous birth in the US between 1996 and 2018 were used. Regression models were adjusted for maternal educational level, race and ethnicity, comorbidities, marital status, and prenatal care visits. Data were analyzed between June 2020 and March 2023.
EXPOSURE: A constructed variable capturing the cumulative number of guideline-recommended cervical cancer screenings for an individual based on their age and year of childbirth.
MAIN OUTCOMES AND MEASURES: Binary indicators for PTD and very preterm delivery (VPTD), defined as delivery before 37 and 34 weeks’ gestational age, respectively, and gestational age was measured in weeks from the first day of the last menstrual period.
RESULTS: Among 11 333 151 females aged 18 to 24 years who gave birth between 1996 and 2018, 2 069 713 were Black (18.3%), 2 601 225 were Hispanic (23.0%), 6 068 498 were White (53.5%) individuals, and 593 715 (5.2%) were individuals of other race or ethnicity (Alaska Native; American Indian; Asian; Pacific Islander; multiracial; or unknown or missing race or ethnicity). Mean (SD) age was 20.9 (1.9) years, and 766 001 individuals (6.8%) had hypertension or diabetes. The mean (SD) number of guideline-recommended screenings by time of childbirth was 2.4 (2.2). Overall, PTD and VPTD occurred in 1 140 490 individuals (10.1%) and 333 040 (2.9%) of births, respectively. One additional recommended screening was associated with a 0.073 (95% CI, 0.026-0.120) percentage-point increase in PTD risk but no statistically significant change in VPTD risk. Females with hypertension or diabetes had a higher increase in PTD risk than those without these comorbidities (0.26 [95% CI, 0.11-0.4] vs 0.06 [95% CI, 0.01-0.10] percentage points; Wald test P < .001).
CONCLUSIONS AND RELEVANCE: Findings of this cross-sectional study suggest that additional recommended cervical cancer screenings before birth were associated with an increased risk of PTD. These results can be used in future simulation models integrating oncological trade-offs to help ascertain optimal screening strategies.
PMID:37477927 | DOI:10.1001/jamahealthforum.2023.1974
