RESUMO

PURPOSE: A strength of SuperLearner is that it may accommodate key interactions between model variables without a priori specification. In prior research, protective associations between fruit intake and preeclampsia were stronger when estimated using SuperLearner with targeted maximum likelihood estimation (TMLE) compared with multivariable logistic regression without any interaction terms. We explored whether heterogeneity (i.e., differences in the effect estimate due to interactions between fruit intake and covariates) may partly explain differences in estimates from these two models. METHODS: Using a U.S. prospective pregnancy cohort (2010-2013, n = 7781), we estimated preeclampsia risk differences (RDs) for higher versus lower fruit density using multivariable logistic regression and included two-way statistical interactions between fruit density and each of the 25 model covariates. We compared the RDs with those from SuperLearner with TMLE (gold standard) and logistic regression with no interaction. RESULTS: From the logistic regression models with two-way statistical interactions, 48% of the preeclampsia RDs were ≤-0.02 (closer to SuperLearner with TMLE estimate); 40% equaled -0.01 (same as logistic regression with no interaction estimate); the minority of RDs were at or crossed the null. CONCLUSIONS: Our exploratory analysis provided preliminary evidence that heterogeneity may partly explain differences in estimates from logistic regression versus SuperLearner with TMLE.

Assuntos

RESUMO

In trials with noncompliance to assigned treatment, researchers might be interested in estimating a per-protocol effect-a comparison of two counterfactual outcomes defined by treatment assignment and (often time-varying) compliance with a well-defined treatment protocol. Here, we provide a general counterfactual definition of a per-protocol effect and discuss examples of per-protocol effects that are of either substantive or methodologic interest. In doing so, we seek to make more concrete what per-protocol effects are and highlight that one can estimate per-protocol effects that are more than just a comparison of always taking treatment in two distinct treatment arms. We then discuss one set of identifiability conditions that allow for identification of a causal per-protocol effect, highlighting some potential violations of those conditions that might arise when estimating per-protocol effects.

Assuntos

Assuntos

Assuntos

RESUMO

BACKGROUND: Short interpregnancy intervals (IPI) are associated with poor birth outcomes. Often, only livebirths are considered to estimate IPI. The objective of our work is to explore whether the associations between demographic, behavioural, and pregnancy variables and IPI change when events other than livebirth are included. METHODS: We used data from the 2006-10 and 2011-13 period of the National Survey of Family Growth (NSFG). We defined IPI using the conception date of the index pregnancy and the event date of the previous one ending in (i) livebirth; (ii) stillbirth; (iii) miscarriage; (iv) abortion; or (v) any of these events. Risk ratios (RR) were estimated for short IPI (<18 months), and demographic, pregnancy, and behavioural variables using log-linear models. RESULTS: When intervening events are included, the association between short IPI and its predictors vary by definition, especially for unintended versus intended pregnancies (only livebirth risk ratio [RR] 1.34, 95% confidence interval [CI] 1.2, 1.5) versus livebirth and miscarriage RR 1.14, 95% CI 1.0, 1.3) and women older than 30 vs. younger than 20 at resolution of the previous pregnancy (only livebirth RR 1.22, 95% CI 1.0, 1.5 versus livebirth and miscarriage RR 1.36, 95% CI 1.2, 1.6). CONCLUSIONS: Including miscarriage as an intervening event in the calculation of IPI changes the association between several risk factors and short IPI. However, the association between short IPI and preterm birth does not vary when different IPI calculations are used.

Assuntos

RESUMO

BACKGROUND: Pregnant women treated with methadone as opioid maintenance therapy are more likely than women treated with buprenorphine to deliver preterm. Preterm birth is associated with less risk of neonatal abstinence syndrome (NAS). We sought to assess the role of preterm birth as a mediator of the relationship between in utero exposure to methadone and NAS compared with buprenorphine. METHODS: We studied 716 women receiving methadone or buprenorphine and delivering liveborn infants at Magee-Womens Hospital, Pittsburgh, Pennsylvania (2013-15). We implemented inverse probability weighted marginal structural models to isolate the role of preterm birth (<37 weeks' gestation). Weights accounted for confounding by maternal age, race, insurance, parity, delivery year, marital, employment, hepatitis C, and smoking status. RESULTS: Approximately 57% of the cohort were treated with methadone. Preterm birth was more common in methadone-exposed pregnancies (25% versus 14%). The incidence of NAS treatment was higher in methadone compared with buprenorphine-exposed infants (65% vs 49%), and term compared with preterm births (64% vs 36%). For every 100 infants liveborn to mothers treated for opioid dependence, there were 13 excess cases of NAS among infants exposed to methadone compared with buprenorphine (adjusted risk difference [RD] 13.3, 95% confidence interval [CI] 5.7, 20.9). Among term births, this increased to 17 excess cases of NAS in methadone- compared with buprenorphine-exposed (RD 16.7, 95% CI 9.3, 24.0). CONCLUSION: The further increased risk of NAS associated with methadone use vs buprenorphine in term deliveries emphasises the utility of buprenorphine in clinical settings aimed at decreasing NAS.

Assuntos

Assuntos

RESUMO

OBJECTIVE: To compare rates of stillbirth among Haitians and non-Haitians in Canada. METHODS: A retrospective cohort study was performed using data on all stillborn and live-born singletons weighing at least 500 g in the province of Quebec, Canada, from 1981 to 2010. Stillbirth rates were computed, and hazard ratios (HRs) and 95% confidence intervals (CIs) were estimated for Haitians relative to non-Haitians. The main outcome measure was stillbirth by cause of death. RESULTS: Data for 9657 stillbirths (124 Haitian) and 2 414 751 live births (17 165 Haitian) were included. Stillbirth rates were higher for Haitians than non-Haitians (7.17 [95% CI 5.91-8.43] vs 3.96 [95% CI 3.88-4.04] per 1000 births), particularly for cord prolapse (adjusted HR 1.87, 95% CI 1.10-3.18) and placental abruption (adjusted HR 2.84, 95% CI 1.95-4.15). Haitians had higher risks of stillbirth due to cord prolapse and abruption at every week of pregnancy. Risks were not elevated for stillbirth due to congenital anomaly, a cause less responsive to urgent intervention. CONCLUSION: Stillbirth rates among Haitians are disproportionately high in Canada, particularly fetal death due to cord prolapse and placental abruption. The potential to reduce stillbirth rates through optimal emergency care in vulnerable minorities requires further investigation.

Assuntos

RESUMO

BACKGROUND: Data on cultural groups at risk of stillbirth in high-income countries are scarce. We sought to determine disparities in stillbirth by gestational age for Haitian vs. non-Haitian Canadians. METHODS: We used data on 10,287 stillbirths and 2,482,364 livebirths from 1981-2010 in the province of Quebec, Canada. Stillbirth rates for Haitians were compared with non-Haitians using fetuses at risk denominators, and Cox proportional hazards regression models with gestational age as the time scale. RESULTS: Stillbirth rates were much higher for Haitians than non-Haitians during the study period (7.2 vs. 3.9 per 1000 total births). Disparities between Haitians and non-Haitians were largest at 32-36 weeks of gestation [hazard ratio 2.22, 95% confidence interval 1.61, 3.07]. CONCLUSIONS: Stillbirth rates in Haitian Canadians giving birth in Quebec are exceptionally high. Disparities were greatest during the late preterm period.

Assuntos