Exposure to Macrolides During Pregnancy and the Risk for Spontaneous Abortions: A Population-Based Retrospective Cohort Study.

Previous studies evaluating the risk of spontaneous abortions following exposure to macrolides reported controversial results. The goal of the current study was to examine the risk for spontaneous abortions following exposure to macrolides during pregnancy. We conducted a population-based retrospective cohort study by linking three computerized databases: Clalit Health Services drug dispensation database, Soroka Medical Center (SMC) birth database, and SMC hospitalizations database. Multivariate time-varying Cox regressions were performed and adjusted for suspected confounders and known risk factors for spontaneous abortions. Hazard ratios (HR) and 95% confidence intervals (CI) were calculated. A secondary analysis was performed to assess the association between exposure to macrolides in terms of the defined daily dose dispensed and spontaneous abortions. The study cohort included 65,457 pregnancies that ended at Soroka Medical Center between 2004 and 2009, of which 6508 (9.9%) resulted in a spontaneous abortion. A total of 825 (1.26%) pregnancies were exposed to macrolides during the exposure period. Exposure to macrolides was not associated with spontaneous abortions as a group (adjusted HR 1.00 95% CI 0.77-1.31) or as specific medications. There was no evidence of a dose-response relationship between exposure to macrolides and spontaneous abortions. In conclusion, this population-based retrospective cohort study did not detect an increased risk for spontaneous abortion following exposure to macrolides during the first trimester of pregnancy.

The Association between Ovarian Hyperstimulation Syndrome and Pregnancy Complications following Fertility Treatments.

OBJECTIVE: This study was aimed to assess the association between ovarian hyperstimulation syndrome (OHSS) and pregnancy complications among women who conceived following fertility treatment. STUDY DESIGN: A retrospective population-based cohort study, including all singleton deliveries of patients conceived following ovulation induction (OI) or in vitro fertilization (IVF) between 1988 and 2016, was conducted. All births occurred in a single tertiary medical center. A comparison was performed between deliveries of women who had experienced OHSS at early gestation and subsequently had a pregnancy and women without OHSS. Women lacking prenatal care, multiple gestations, and stillbirths were excluded from the analyses. A multivariable logistic regression model was used to control for confounders. RESULTS: During the study period, 351,373 deliveries met the inclusion criteria, of which 6,748 were deliveries of infants who were conceived by either IVF or OI. Of this study population, 105 cases (1.6%) composed the exposed group, that is, women who had experienced OHSS with a subsequent live birth. In the multivariate analyses, after controlling for confounders, OHSS was not found as an independent risk factor for preeclampsia, gestational diabetes mellitus (GDM), intrauterine growth restriction (IUGR), preterm delivery (both <37 and <34 weeks), low birth weight (LBW), very LBW (VLBW), small for gestational age (SGA), and caesarean delivery. In a subanalysis conducted solely on the IVF population, similar results were found, aside from the association between OHSS and preterm delivery before 34 weeks of gestation which was statistically significant (adjusted odds ratio [AOR] = 2.3 95% confidence interval [CI]: 1.0-5.3, p = 0.049). CONCLUSION: In our population, OHSS was not found as a risk factor for adverse pregnancy and perinatal outcome. In IVF patients, OHSS is a risk factor for preterm delivery before 34 weeks of gestation. KEY POINTS: · OHSS is not a risk factor for pregnancy complications.. · Complications investigated were preeclampsia, GDM, prematurity, and others.. · In IVF patients, OHSS is a risk factor for preterm delivery..

Cesarean versus vaginal delivery for breech presentation is an independent risk factor for long-term pediatric respiratory hospitalization of the offspring.

OBJECTIVES: To compare the long-term respiratory morbidity of offspring born by cesarean delivery for breech presentation with that of those delivered vaginally. METHODS: A population-based cohort analysis including all singleton breech deliveries between the years 1991 and 2014, comparing long-term respiratory morbidity of offspring born in breech presentation, according to mode of delivery. Offspring with congenital malformations, perinatal deaths, and instrumental deliveries were excluded. Respiratory morbidity included hospitalizations (up to age 18 years), as recorded in hospital records. A Kaplan-Meier survival curve compared cumulative respiratory morbidity. A Weibull parametric survival model controlled for confounders and repeat deliveries. RESULTS: A total of 7337 breech deliveries were included; 6376 (86.9%) cesarean deliveries and 961 (13.1%) vaginal breech deliveries. The Kaplan-Meier survival curve demonstrated higher cumulative incidence of respiratory morbidity in the cesarean delivery group compared with vaginal delivery (log rank test P = 0.006). Using a Weibull parametric survival model to control for confounders, cesarean delivery was found to be an independent risk factor for long-term respiratory morbidity of the offspring (adjusted hazard ratio 1.87, 95% confidence interval 1.32-2.65, P < 0.001). CONCLUSIONS: Cesarean versus vaginal delivery for breech presentation is an independent risk factor for long-term pediatric respiratory morbidity of the offspring.

Risk factors for hospitalization at the pediatric intensive care unit among infants and children younger than 5 years of age diagnosed with infectious diseases.

BACKGROUND: Children hospitalized with infectious diseases may develop severe, life-threatening conditions, often requiring admission to pediatric intensive care unit (PICU). The objectives of this study were to identify independent risk factors for PICU hospitalization with an infectious disease in children <5 years of age. METHODS: In southern Israel, two populations live side by side: the middle-high income Jewish population and the low-income Bedouin population, both receiving equal and free medical care at the only tertiary medical center in the area. The study population included all children born in southern Israel and hospitalized at PICU with an infectious disease during 1991-2012. Risk factors for PICU hospitalizations were retrospectively studied by Kaplan-Meier and Cox proportional hazard survival analyses. RESULTS: 9951 Jewish children and 18,002 Bedouin children were enrolled; overall, 1135 episodes of PICU hospitalizations with an infectious disease were recorded (879, 77.4% Bedouin and 256, 22.6% Jewish patients). Bedouin children had a higher risk for PICU hospitalization with an infectious disease compared with Jewish children (adjusted Hazard Ratio [adj. HR] 1.7, 95% CI 1.5-2.0); maternal multiparity and low-birth weight (<2500 g) were additional risk factors for PICU hospitalization with an infectious disease compared to firstborns (adj. HR = 1.2, 95% CI 1.0-1.5) or to children with a birth weight ≥2500 g (adj. HR = 1.5, 95% 1.2-1.9). Older age was a protective factor for PICU hospitalization (adj. HR = 0.98, 95% CI 0.97-0.99). Children hospitalized with a central nervous system infection had the highest risk of PICU hospitalization (adj. HR 6.8, 95% CI 5.5-8.4), followed by those with urinary tract infections (UTI, adj. HR 3.1, 95% CI 2.5-3.8) and those with lower respiratory tract infections (LRTI, adj. HR 2.9, 95% CI 2.4-3.4). CONCLUSION: Bedouin ethnicity, low birth weight, maternal multiparity and younger age were significant risk factors for PICU hospitalizations with an infectious disease. Among the infectious diseases analyzed, CNS infection had the highest risk for PICU hospitalization, followed by UTI and LRTI.

Prenatal exposure to heavy metal mixtures and anthropometric birth outcomes: a cross-sectional study.

BACKGROUND: Numerous studies have suggested significant associations between prenatal exposure to heavy metals and newborn anthropometric measures. However, little is known about the effect of various heavy metal mixtures at relatively low concentrations. Hence, this study aimed to investigate associations between prenatal exposures to a wide range of individual heavy metals and heavy metal mixtures with anthropometric measures of newborns. METHODS: We recruited 975 mother-term infant pairs from two major hospitals in Israel. Associations between eight heavy metals (arsenic, cadmium, chromium, mercury, nickel, lead, selenium, and thallium) detected in maternal urine samples on the day of delivery with weight, length, and head circumference at birth were estimated using linear and Bayesian kernel machine regression (BKMR) models. RESULTS: Most heavy metals examined in our study were observed in lower concentrations than in other studies, except for selenium. In the linear as well as the BKMR models, birth weight and length were negatively associated with levels of chromium. Birth weight was found to be negatively associated with thallium and positively associated with nickel. CONCLUSION: By using a large sample size and advanced statistical models, we could examine the association between prenatal exposure to metals in relatively low concentrations and anthropometric measures of newborns. Chromium was suggested to be the most influential metal in the mixture, and its associations with birth weight and length were found negative. Head circumference was neither associated with any of the metals, yet the levels of metals detected in our sample were relatively low. The suggested associations should be further investigated and could shed light on complex biochemical processes involved in intrauterine fetal development.

Association Between Mode of Delivery of the Breech Fetus and Hospitalizations Due to Inflammatory Bowel Disease During Childhood.

BACKGROUND AND GOALS: While evidence suggests short-term benefits in neonatal morbidity and mortality from cesarean delivery of the fetus in breech presentation, the long-term implications for the offspring are less clear. To assess the implications of the mode of delivery on offspring's health, we have evaluated the long-term gastrointestinal (GI) morbidity of offspring with a breech presentation delivered in either way. MATERIALS AND METHODS: A population-based retrospective cohort study including singleton deliveries in breech presentation occurring between 1991 and 2014 at a tertiary referral hospital. Incidence of hospitalizations of the offspring up to the age of 18 years involving GI morbidity was compared between those delivered via cesarean section or vaginally. A Kaplan-Meier survival curve compared cumulative GI morbidity. A Weibull parametric survival model controlled for confounders while accounting for repeated occurrence of mothers and dependence among siblings. RESULTS: Overall, 86.9% (n=6376) of the 7337 fetuses in breech presentation, were delivered abdominally. Hospitalizations involving GI morbidity were higher in offspring delivered by cesarean section, specifically due to inflammatory bowel disease (IBD). Kaplan-Meier survival curve revealed the higher cumulative incidence of total GI morbidity and IBD specifically in the cesarean delivery group (P<0.001 and P=0.004, respectively). Using a Weibull parametric while controlling for relevant confounders, cesarean delivery emerged as an independent risk factor for long-term IBD-related morbidity of the offspring delivered in breech presentation (adjusted hazard ratio=3.18, 95% confidence interval: 1.47-6.87, P=0.003). CONCLUSION: Cesarean delivery is associated with higher rates of hospitalizations due to IBD and total GI morbidity during childhood in term singleton in breech presentation.

Risk for preeclampsia following exposure to PM2.5 during pregnancy.

BACKGROUND: Previous findings concerning the risk for preeclampsia following exposure to particulate matter are inconclusive. METHODS: We used data from all singleton pregnancies of women insured by the "Clalit health services" (CHS) maintenance organization in southern Israel that resulted in delivery or perinatal mortality at Soroka Medical Center (SMC). Daily PM2.5 concentrations were estimated by a hybrid satellite-based model at one-squared kilometer spatial resolution. We used Cox proportional hazard models coupled with distributed lag models to examine the association between the mean exposure to PM2.5 in every gestational week and the diagnosis of preeclampsia, adjusting for maternal age, parity, year of birth, season of birth and socio-economic status. Hazard Ratios (HR) and 95% Confidence Intervals (CI) were calculated for individual gestational weeks and for cumulative exposure until the 25th gestational week. RESULTS: A total of 133,197 pregnancies ended at SMC during the study period, of which 68,126 (51.1%) were Jewish and 65,071 (48.9%) were Bedouin. For pregnancies of Jewish women, exposure to PM2.5 from the 7th until the 14st gestational week was significantly associated with preeclampsia (maximal HR = 1.06; 95%CI: 1.01 - 1.11 during the 10th gestational week per 10 µg/m3 increase in PM2.5). Cumulative exposure to PM2.5 during the first 25th gestational weeks was also significantly associated with preeclampsia (HR = 2.08; 95%CI: 1.10 - 3.94 per 10 µg/m3 increase in PM2.5). We observed no association for pregnancies of Bedouin women. CONCLUSIONS: Exposure to PM2.5 between the 7th and the 14st gestational weeks was associated with preeclampsia among Jewish women but not among Bedouin women.

Increasing adverse drug reaction reporting-How can we do better?

Adverse drug reactions (ADRs) are associated with morbidity and mortality worldwide. Although national systems for reporting ADRs exist there is a low reporting rate. The aim of the current study was to evaluate an intervention plan for improving ADRs reporting among medical professionals (physicians and nurses). A multicentre intervention study was conducted, in which one medical centre was randomly assigned to the intervention group and two medical centres to the control group. The study consisted of 3 phases: baseline data collection, intervention and follow-up of the reporting rate. The questionnaire that was filled in at base line and at the end of study, contained questions about personal/professional demographic variables, and statements regarding knowledge of and behaviour toward ADRs reporting. The intervention program consisted of posters, lectures, distant electronic learning and reminders. An increase in the number of ADRs reports was noted in the intervention group (74 times higher than in the control group) during the intervention period, which was gradually decreased with as the study progressed (adjusted O.R = 74.1, 95% CI = 21.11-260.1, p<0.001). The changes in the "knowledge related to behaviour" (p = 0.01) and in the "behaviour related to reporting" (p<0.001) score was significantly higher in the intervention group. Specialist physicians and nurses (p<0.001), fulfilling additional positions (p<0.001) and those working in other places (p = 0.05) demonstrated a high rate of report. Lectures were preferable as a method to encourage ADRs reporting. The most convenient reporting tools were telephone and online reporting. Thus, implementation and maintenance of a continuous intervention program, by a pharmacovigilance specialist staff member, will improve ADRs reporting rates.

Associations between intrauterine exposure to polychlorinated biphenyls on neonatal ano-genital distance.

Polychlorinated Biphenyls (PCBs) are widespread environmental contaminants. PCBs have endocrine disrupting properties which raises concerns regarding their effect on the developing fetus. This study aimed to examine the association between prenatal exposure to PCBs and anogenital distance (AGD) in newborns. Serum concentrations of PCB congeners -118, -138, -153 and -180 were measured in 175 pregnant women presenting to the delivery room. AGD was measured in their newborns. Regression models were used to estimate associations between maternal PCB exposure and infant anogenital measurements, controlling for possible confounding variables. Mean maternal serum concentrations were 2.95 ± 2.18 ng/g, 4.62 ± 3.54 ng/g, 7.67 ± 6.42 ng/g and 5.10 ± 3.91 ng/g for congeners -118, -138, -153 and -180, respectively. Higher maternal concentrations of PCBs were associated with reduced AGD measures in male infants. Higher maternal concentrations of PCB-138 and PCB-153 were associated with reduced ano-scrotal distances and higher maternal concentrations of all four PCB congeners were associated with reduced ano-penile distances. No significant associations were found between any PCB congener and any AGD measure in female newborns. This study demonstrates that intrauterine exposure to PCBs may be associated with reduced AGD in male newborns. More research is needed to reveal the implications for adult reproductive health.

The effect of an intervention program on the knowledge and attitudes among medical staff regarding adverse drug reaction reporting.

PURPOSE: Adverse drug reactions (ADRs) are a growing important public health problem; however, underreporting of ADRs is very common. The aim of the current study was to explore the effect of an intervention program on the knowledge and attitudes among physicians and nurses regarding ADRs reporting. METHODS: A multicentre study consisted of three phases: filling out a questionnaire; an intervention program; filling out the same questionnaire again. The intervention program consisted of posters, lectures, and distant electronic learning. The questionnaire contained questions about personal/professional demographic variables, and statements regarding knowledge and attitudes regarding ADR reporting. RESULTS: The data revealed that the intervention program significantly elevated the "Objective knowledge" (P < 0.01) and "Practical knowledge" (P < 0.02) score as compared to the control group, while no significant differences were found regarding "Acquired knowledge" (P = 0.14). Seniority (P = 0.01) and experience in internal medicine (P = 0.05) were demonstrated as significant factors determining the knowledge of the staff. Obligation was the main motive for reporting in 80% of participants. After the intervention, no differences were found in the "Attitude related to the motive for reporting" or "Attitude related to the commitment to report", between the two groups. However, "Attitude related to the need to report" score significantly improved after the intervention (P = 0.04). CONCLUSIONS: The intervention program increased knowledge and attitudes regarding ADRs reports. Seniority had the most effect on the influence of the intervention program. The data from this study encourages the necessity to hold ongoing intervention programs in order to improve ADRs reporting rate.

The safety of amoxicillin and clavulanic acid use during the first trimester of pregnancy.

AIMS: The goal of the current study was to assess the risk for major congenital malformations following first-trimester exposure to amoxicillin, or amoxicillin and clavulanic acid (ACA). METHODS: A population-based retrospective cohort study was conducted, by linking 4 computerized databases: maternal and infant hospitalization records, drug dispensing database of Clalit Health Services in Israel and data concerning pregnancy terminations. Multivariate negative-binomial regression was used to assess the risk for major malformations following first-trimester exposure, adjusted for mother's age, ethnicity (Bedouin vs Jewish), parity, diabetes mellitus, lack of perinatal care, and the year of birth. RESULTS: The study included 101 615 pregnancies, of which 6919 (6.8%) were exposed to amoxicillin: 1045 (1.0%) to amoxicillin only and 6041 (5.9%) to ACA. No significant association was found, in the univariate and multivariate analyses, between first-trimester exposure to amoxicillin or ACA and major malformations in general (crude relative risk, 1.05 95% confidence interval 0.95-1.16; adjusted relative risk 1.09, 95% confidence interval 0.98-1.20), or for major malformations according to organ systems. No dose-response relationship was found between exposure in terms of the defined daily dose and major malformations. CONCLUSION: Exposure to amoxicillin and ACA during the first trimester of pregnancy was not associated with an increased risk of major congenital malformations.

Pregnancy outcome following opioid exposure: A cohort study.

INTRODUCTION: Opioids constitute a cornerstone of pain relief treatment. However, opioid safety during pregnancy has not been well established. Recent studies reported an association between in utero opioid exposure and spina bifida. METHODS: In order to further evaluate the association of opioids exposure during pregnancy with adverse pregnancy outcomes, we conducted a large historical cohort by linking four databases: medications dispensations, births, pregnancy terminations for medical reasons and infant hospitalizations during the years of 1999-2009. Confounders that were controlled for included maternal age, ethnicity, maternal diabetes, smoking status, parity, obesity, year and folic acid intake. A secondary analysis for total major malformations and for spina bifida was performed using propensity score matching for first trimester exposure. RESULTS: Of the 101,586 women included in the study, 3003 were dispensed opioids during the first trimester. Intrauterine exposure to opioids was not associated with overall major malformations (adjusted odds ratio (aOR) 0.97, 95% CI 0.83-1.13), cardiovascular malformations (aOR = 0.89, 95% CI 0.70-1.13) other malformations by systems or spina bifida in particular. However, the risk for spina bifida among newborns and abortuses who were exposed to codeine was four times higher than that of the unexposed (aOR = 4.42, 95% CI 1.60-12.23). This association remained significant in a secondary analysis using propensity score matching. Third trimester exposure to opioids was not associated with low birth weight (aOR = 1.08, 95% CI 0.77-1.52), perinatal death (aOR = 1.38, 95% CI 0.64-2.99) and other adverse pregnancy outcomes. CONCLUSIONS: These findings suggest that opioids exposure (as a homogenous group) is not a significant risk factor for overall major malformations. Exposure to codeine during the first trimester was found to be associated with increased risk of spina bifida. However, this finding was based on a small number of cases and need to be verified in future work.

Vaginal antimycotics and the risk for spontaneous abortions.

BACKGROUND: Spontaneous abortions are the most common complication of pregnancy. Clotrimazole and miconazole are widely used vaginal-antimycotic agents used for the treatment of vulvovaginal candidiasis. A previous study has suggested an increased risk of miscarriage associated with these azoles, which may lead health professionals to refrain from their use even if clinically indicated. OBJECTIVE: The aim of the current study was to assess the risk for spontaneous abortions following first trimester exposure to vaginal antimycotics. STUDY DESIGN: A historical cohort study was conducted including all clinically apparent pregnancies that began from January 2003 through December 2009 and admitted for birth or spontaneous abortion at Soroka Medical Center, Clalit Health Services, Beer-Sheva, Israel. A computerized database of medication dispensation was linked with 2 computerized databases containing information on births and spontaneous abortions. Time-varying Cox regression models were constructed adjusting for mother's age, diabetes mellitus, hypothyroidism, obesity, hypercoagulable or inflammatory conditions, recurrent miscarriages, intrauterine contraceptive device, ethnicity, tobacco use, and the year of admission. RESULTS: A total of 65,457 pregnancies were included in the study: 58,949 (90.1%) ended with birth and 6508 (9.9%) with a spontaneous abortion. Overall, 3246 (5%) pregnancies were exposed to vaginal antimycotic medications until the 20th gestational week: 2712 (4.2%) were exposed to clotrimazole and 633 (1%) to miconazole. Exposure to vaginal antimycotics was not associated with spontaneous abortions as a group (crude hazard ratio, 1.11; 95% confidence interval, 0.96-1.29; adjusted hazard ratio, 1.11; 95% confidence interval, 0.96-1.29) and specifically for clotrimazole (adjusted hazard ratio, 1.05; 95% confidence interval, 0.89-1.25) and miconazole (adjusted hazard ratio, 1.34; 95% confidence interval, 0.99-1.80). Furthermore, no association was found between categories of dosage of vaginal antimycotics and spontaneous abortions. CONCLUSION: Exposure to vaginal antimycotics was not associated with spontaneous abortions.

The Fetal Safety of Enoxaparin Use During Pregnancy: A Population-Based Retrospective Cohort Study.

INTRODUCTION: Enoxaparin is widely used during pregnancy as pregnancy is a hypercoagulable state; however, its fetal safety has scarcely been investigated. OBJECTIVE: Our study aimed to examine fetal safety following enoxaparin exposure during pregnancy. METHODS: A population-based, retrospective cohort study was performed by linking computerized databases, including the drug dispensing registries of Clalit Health Services in Israel and maternal and infant hospital records, between 1998 and 2009. Multivariate logistic regression models were used to examine associations between first- and third-trimester exposure to enoxaparin, major malformations, and other adverse birth outcomes, adjusted for confounders. RESULTS: From a total of 109,473 singleton pregnancies, 418 and 572 were exposed to enoxaparin during the first and third trimesters, respectively. Exposure to enoxaparin during the first trimester of pregnancy was not associated with an increased risk of major congenital malformations [adjusted odds ratio (aOR) 1.1, 95% confidence interval (CI) 0.8-1.6], while exposure during the third trimester was not associated with an increased risk of low birth weight (aOR 1.1, 95% CI 0.8-1.4), low Apgar score (aOR 0.9, 95% CI 0.4-1.8), or risk of perinatal mortality (aOR 0.6, 95% CI 0.1-2.9). CONCLUSION: Exposure to enoxaparin during pregnancy was not associated with an increased risk of major malformations in general or according to organ systems. Nonetheless, risk for specific malformations cannot be ruled out.

Acute kidney injury in pancreatic surgery; association with urine output and intraoperative fluid administration.

BACKGROUND: Acute kidney injury (AKI) is a devastating postoperative complication. Intraoperative urine output is assumed to reflect patient's intravascular volume and kidney function. We thus evaluated the incidence of postoperative AKI and its association with intraoperative urine output and the volume of fluid administered. METHODS: A retrospective study on 153 consecutive patients admitted to Tel Aviv Medical Center for pancreatic surgery. RESULTS: The incidence of AKI in patients undergoing pancreatic surgery was 9.8%. Oliguria was not a predictor of AKI. There was no association between the amount of fluids administered and AKI. Pulmonary disease is an independent predictor of AKI. AKI is an independent predictor of mortality. CONCLUSIONS: AKI is common in patients undergoing pancreas surgeries and is associated with high mortality. Neither urine output, nor the volume of fluids administered correlate with postoperative AKI. Low diuresis is therefore not a sole marker for fluid administration.

NSAIDs and spontaneous abortions - true effect or an indication bias?

AIM: The aim of the study was to characterize the extent of indication bias resulting from the excessive use of NSAIDs on the days preceding a spontaneous abortion to relieve pain. METHODS: We used data from a retrospective cohort study assessing the risk for spontaneous abortions following exposure to NSAIDs. Three definitions of exposure for cases of spontaneous abortions were compared, from the first day of pregnancy until the day of spontaneous abortion and until 3 and 2 days before a spontaneous abortion. Statistical analysis was performed using multivariate time programmed Cox regression. RESULTS: A sharp increase was observed in the dispensation of indomethacin, diclofenac and naproxen, and a milder increase was found in the use of ibuprofen during the week before a spontaneous abortion. Non- selective COX inhibitors in general and specifically diclofenac and indomethacin were found to be associated with spontaneous abortions when the exposure period was defined until the day of spontaneous abortion (hazard ratio (HR) 1.15, 95% confidence interval (CI) 1.04, 1.28; HR 1.31, 95% CI 1.08, 1.59 and HR 3.33, 95% CI 2.09, 5.29, respectively). The effect disappears by excluding exposures occurring on the day before the spontaneous abortion for non-selective COX inhibitors and on the last week before the spontaneous abortion for indomethacin. In general, decreasing HRs were found with the exclusion of exposures occurring on the days immediately before the spontaneous abortion. CONCLUSIONS: The increased use of NSAIDs during the last few days that preceded a spontaneous abortion to relieve pain associated with the miscarriage could bias studies assessing the association between exposure to NSAIDs and spontaneous abortions.