Discriminant analysis for longitudinal data with multiple continuous responses and possibly missing data

Summary .  Multiple outcomes are often used to properly characterize an effect of interest. This article discusses model-based statistical methods for the classification of units into one of two or more groups where, for each unit, repeated measurements over time are obtained on each outcome. We relate the observed outcomes using multivariate nonlinear mixed-effects models to describe evolutions in different groups. Due to its flexibility, the random-effects approach for the joint modeling of multiple outcomes can be used to estimate population parameters for a discriminant model that classifies units into distinct predefined groups or populations. Parameter estimation is done via the expectation-maximization algorithm with a linear approximation step. We conduct a simulation study that sheds light on the effect that the linear approximation has on classification results. We present an example using data from a study in 161 pregnant women in Santiago, Chile, where the main interest is to predict normal versus abnormal pregnancy outcomes.     

Effects of mother's dietary exposure to acesulfame-K in Pregnancy or lactation on the adult offspring's sweet preference

This study investigates whether mother's exposure to the artificial sweetener acesulfame-K (AK) during pregnancy or lactation affected her adult offspring's sweet preference. It was found that mother's dietary exposure to AK in pregnancy or lactation decreased the preference thresholds for AK and sucrose solutions in the adult offspring, whereas the preference pattern and the most preferred concentration for AK or sucrose solution were unchanged. Furthermore, the preference scores in the exposure groups were increased significantly when compared with the control group at a range of concentrations for AK or sucrose solution. The existence of AK and its dynamic changes within 24 h in amniotic fluid during pregnancy or in mother's milk during lactation after a single oral infusion of AK solution were revealed by the methods of reversed-phase high-performance liquid chromatography and mass spectrometry. Our data suggest that AK can be ingested by the prenatal or postnatal mice through their mother's amniotic fluid or breast milk, producing a long-dated function on the adult's sweet preference.     

Comparison of pregnancy outcome in mares among methods used to evaluate and select spermatozoa for insemination.

An artificial insemination dose for mares consisting of 500 million progressively motile spermatozoa is considered "standard" by most clinicians. However, little information is available directly comparing pregnancy outcome among methods of evaluating and selecting spermatozoa for insemination. The objective of this study was to determine if the method of spermatozoal evaluation and selection influences fertility as measured by pregnancy outcome. Mares were inseminated with 100 or 500 million spermatozoa that were selected for progressive motility, normal morphology, hypoosmotic swelling or absolute number regardless for evaluation method or quality. Thirty-two breeding cycles were tested for each treatment group and at each spermatozoal dose. Pregnancy outcomes were 44 and 41%, 55 and 41%, 39 and 31%, and 45 and 41%, for the 100 and 500 million progressively motile, morphologically normal, hypoosmotic swelling positive and absolute number treatment groups, respectively. Pregnancy outcome did not differ among methods of spermatozoal evaluation and selection for artificial insemination in the 100 (P=0.52) or 500 (P=0.78) million spermatozoa groups. Also the total number of spermatozoa and the absolute number of progressively motile, morphologically normal or hypoosmotic swelling positive spermatozoa inseminated, were not closely associated with pregnancy outcome in the 100 (P=0.24, 0.29, 0.33 and 0.38, respectively) or 500 (P=0.20, 0.84, 0.50 and 0.74, respectively) million spermatozoa groups. In this study, we found that the method of spermatozoal evaluation did not offer an advantage for pregnancy when used to select spermatozoa for insemination at the doses tested. These results were surprising, as we expected there would be differences among the evaluation methods. Instead, we found that evaluating spermatozoa offered no advantage for pregnancy over simply inseminating with a specified number of spermatozoa not selected for any particular characteristic under the conditions of our experiment.     

Fertility of prostaglandin-induced oestrus compared to normal post-partum oestrus.

This study was undertaken to determine if fertility could be improved by increasing the interval from foaling to breeding. Forty-two mares, not bred during normal post-partum oestrus, were injected with a prostaglandin analogue on Day 6 or 7 following ovulation. Mares were mated artifically with antibiotic-treated semen during the resulting oestrus and, if necessary, for the following 4 cycles. Their fertility was compared, by cycles/pregnancy and rate of fetal loss, to mares bred by the same methods on 86 normal post-partum oestrous periods. The interval from foaling to the onset of breeding and the duration of oestrus were longer in the injected mares than the intervals observed in the untreated mares. Both groups averaged 1.3 breeding periods/pregnancy. The rate of fetal loss was not significantly different between the groups. Increasing the interval from foaling to breeding did not improve fertility.     

Aggression in pigtailed macaque (Macaca nemestrina) breeding groups affects pregnancy outcome

Past research has shown that aggressive behaviors can affect female reproductive outcome in nonhuman primate captive breeding programs. In this study, aggressive behaviors were recorded in a colony of pigtailed macaque monkeys (Macaca nemestrina) and related to pregnancy outcome. For 22 weeks, behavioral data were collected from nine breeding groups, consisting of zero to one male (some males were removed after a cycle of conceptions for husbandry reasons) and four to eight females. Observations included all occurrences of 11 aggressive behaviors during 15 min observation sessions, 1-3 times a week. Mean weekly aggression levels during the study period were determined for each group as well as for each pregnancy. Aggression data were summarized with Principal Components Analyses. Results indicate that pigtailed macaque aggression falls into five distinctive categories: warn, engage, threaten, pursue, and attack. Breeding groups differed in their levels of aggression, even after controlling for group size, presence of a sire, and group stability. Levels of the five aggression categories were found to affect the probability that a pregnancy ended in either a natural birth of a live infant, a clinical intervention producing a live infant, or a nonviable outcome. The predictive value of aggression was significant when clinical interventions were included as possible reproductive outcomes. Behavioral observation of captive groups could identify "risk" conditions affecting pregnancy outcome and the requirement for clinical intervention.     

基础窦卵泡数对IVF／ICSI-ET超排卵周期卵巢反应的预测价值

目的：窦卵是女性宫腔内的泡状细胞,与女性卵巢的发育及功能密切相关。本研究针对体外受精（IVF）周期中基础窦卵泡数（antral follicle count, AFC）的变化情况,探讨AFC 对卵巢反应的预测价值,为临床研究提供理论基础。方法：回顾性分析2012 年1月至2012 年12 月在我院生殖医学中心接受体外受精- 胚胎移植的157 例患者的临床资料,根据基础窦卵泡数将所选病例分为A 组（bAFC逸10）和B组（bAFC＜10）。对比并分析两组研究对象的促性腺激素（Gn）的用量及使用时间、获卵数、人绒毛膜促性腺激素（HCG）注射日的血清中雌激素（E2）水平、受精率、临床妊娠率等。结果：A 组患者的Gn 用量、HCG日E2 值、获卵数及临床妊娠率均显著高于B 组,差异具有统计学意义（P〈0.05）;两组患者的年龄和基础卵泡刺激素水平差异显著且具有统计学意义（P〈0.05）;两组患者的不孕原因、不孕年限及助孕方式无明显差异（P〉0.05）。结论：基础窦卵泡对体外受精女性促排卵周期的卵巢反应及助孕结局具有一定的预测价值,其数量的多少可作为促排卵过程中评价卵巢反应性的参考指标,应在临床进一步推广。     

Factors affecting pregnancy loss from gestation Day 38 to 90 in lactating dairy cows from a single herd

The present study was designed to establish whether factors such as previous estrus synchronization, corpus luteum and embryo number at the time of pregnancy diagnosis, changes in body condition score, milk production, clinical disease (mastitis or lameness) and the inseminating bull affect pregnancy loss from 38 to 90 days of gestation. We derived data from 601 pregnant lactating dairy cows from a single herd. Pregnancy diagnosis was performed by ultrasonography between Day 38 and 44 following insemination. We also recorded corpus luteum and embryo number at this time. Pregnancy loss was defined as a negative pregnancy diagnosis on the second palpation per rectum undertaken between 90 and 96 days after insemination. Data were analyzed using multiple logistic regression methods. Cows that had an additional corpus luteum were eight times less likely to miscarry. The risk of pregnancy loss was 3.1 times higher in cows bearing twins. A one unit reduction in body condition score from previous partum to 30 days postpartum resulted in a 2.4-fold increase in pregnancy loss. We noted a higher incidence of pregnancy loss in cows inseminated using semen from one of the six bulls used. This particular bull led to a 3.4-fold increase in the rate of pregnancy loss. Logistic regression analysis showed no significant effects of previous estrus synchronization, milk production, clinical disease, body condition at previous partum or at pregnancy diagnosis, or body condition change between previous partum and pregnancy diagnosis. Our findings indicate a positive relationship between the presence of an additional corpus luteum and the maintenance of gestation. Risk factors for pregnancy loss were twin pregnancy, reduced body condition after previous parturition and the inseminating bull.     

Fitting semiparametric random effects models to large data sets

For large data sets, it can be difficult or impossible to fit models with random effects using standard algorithms due to memory limitations or high computational burdens. In addition, it would be advantageous to use the abundant information to relax assumptions, such as normality of random effects. Motivated by data from an epidemiologic study of childhood growth, we propose a 2-stage method for fitting semiparametric random effects models to longitudinal data with many subjects. In the first stage, we use a multivariate clustering method to identify G相似文献   

The beta-geometric distribution applied to comparative fecundability studies

A convenient measure of fecundability is time (number of menstrual cycles) required to achieve pregnancy. Couples attempting pregnancy are heterogeneous in their per-cycle probability of success. If success probabilities vary among couples according to a beta distribution, then cycles to pregnancy will have a beta-geometric distribution. Under this model, the inverse of the cycle-specific conception rate is a linear function of time. Data on cycles to pregnancy can be used to estimate the beta parameters by maximum likelihood in a straightforward manner with a package such as GLIM. The likelihood ratio test can thus be employed in studies of exposures that may impair fecundability. Covariates are incorporated in a natural way. The model is illustrated by applying it to data on cycles to pregnancy in smokers and nonsmokers, with adjustment for covariates. For a cross-sectional study, when length-biased sampling is taken into account, the pre-interview attempt time is shown to follow a beta-geometric distribution, so that the same methods of analysis can be applied even though all of the available data are right-censored. For a cohort followed prospectively, there will be some couples enrolled whose fecundability is effectively 0, and for such applications, the beta could be considered to be contaminated by a distribution degenerate at 0. The mixing parameter (proportion sterile) can be estimated by application of the expectation-maximization (EM) algorithm. This, too, can be carried out using GLIM.     

腹腔镜术后联合GnRH-a治疗卵巢型内异症并不孕的疗效观察

目的:研究腹腔镜保守术后联合短期促性腺激素释放激素激动剂(GnRH-a)对卵巢型子宫内膜异位症合并不孕患者的治疗效果。方法:回顾性分析2008年5月-2011年5月在我院治疗内异症合并不孕的126例患者的临床资料,比较腹腔镜手术后联合GnRH-a治疗(用药组)和术后期待处理(对照组)的症状缓解率和妊娠情况。结果:用药组患者症状缓解率比对照组高,且差异有统计学意义(P0.05);用药组总妊娠率较对照组高,差异无统计学意义(P0.05),流产率比较无统计学差异(P0.05);治疗后患者1年内妊娠率最高,两组比较无统计学差异(P0.05),随访期间累积妊娠率比较,用药组高于对照组,无统计学意义(P0.05);用药组及对照组中Ⅰ-Ⅱ期及Ⅲ-Ⅳ期患者妊娠率均在50%以上,两组比较均无统计学差异(P0.05)。结论:此研究进一步明确腹腔镜术联合GnRH-a治疗卵巢子宫内膜异位症合并不孕患者能较好提高其症状缓解率和妊娠率,但妊娠率提高与期待治疗无差别。经手术或联合GnRH-a治疗后患者首次妊娠多发生在治疗后1年内,如果1年以上仍未妊娠,可根据实际情况选择其他助孕方法如辅助生育技术,以增加受孕机会。     

Properties of chromatin of peripheral blood lymphocytes in women with threatened abortion

The authors studied the properties of peripheral blood lymphocyte chromatin from donors (non-pregnant women), from women with normal pregnancy and during threatened abortion. Using three independent cytochemical methods that characterize the DNA accessibility for acridine orange and 3H-actinomycin D and the presence in histones of free amino groups, it has been established that lymphocyte chromatin from women during threatened abortion is repressed to a greater degree than in those with normal pregnancy. The data presented demonstrate the prospectiveness of the use of the cytochemical tests for chromatin properties with a purpose of diagnosing whether the pregnancy will be full-term and treatment control.     

子宫内膜异位症合并不孕的原因及治疗方案探讨

目的：探讨子宫内膜异位症并不孕的原因及治疗方案的临床意义。方法：随机选取2007年6月至2012年8月就诊于我院的不孕症或疑似子宫内膜异位症（EMS）经检查确诊为EMS的患者216例,按照用药区别平均分为六组,分别为三苯氧胺组、米非司酮组、丹那唑组、诺雷德组、无效成分安慰剂组及两阶段复合治疗组,记录上述患者经治疗后随访1年内受孕率情况。结果：无效成分安慰剂组无疗效,各方面与其他组比较有明显差异（P〈0.05）,其余各组疗效相比无明显差异（P〉0.05）;三苯氧胺组与丹那唑组不良反应发生率高于其他组,比较有明显差异（P〈0.05）;两阶段复合治疗组的复发率最低,受孕率最高,除安慰剂组外其他组之间比较无明显差异（P〉0.05）。结论：此次研究的药物治疗效果近似,但是副作用、复发率及受孕率具有差异,临床应根据患者具体情况施治。     

Use of open-text responses to recode categorical survey data on postpartum contraception use among women in the United States: A mixed-methods inquiry of Pregnancy Risk Assessment Monitoring System data

BackgroundPostpartum contraception prevents unintended pregnancies and short interpregnancy intervals. The Pregnancy Risk Assessment Monitoring System (PRAMS) collects population-based data on postpartum contraception nonuse and reasons for not using postpartum contraception. In addition to quantitative questions, PRAMS collects open-text responses that are typically left unused by secondary quantitative analyses. However, abundant preexisting open-text data can serve as a resource for improving quantitative measurement accuracy and qualitatively uncovering unexpected responses. We used PRAMS survey questions to explore unprompted reasons for not using postpartum contraception and offer insight into the validity of categorical responses.Methods and findingsWe used 31,208 categorical 2012 PRAMS survey responses from postpartum women in the US to calculate original prevalences of postpartum contraception use and nonuse and reasons for contraception nonuse. A content analysis of open-text responses systematically recoded data to mitigate survey bias and ensure consistency, resulting in adjusted prevalence calculations and identification of other nonuse themes. Recoded contraception nonuse slightly differed from original reports (21.5% versus 19.4%). Both calculations showed that many respondents reporting nonuse may be at a low risk for pregnancy due to factors like tubal ligation or abstinence. Most frequent nonuse reasons were not wanting to use birth control (27.1%) and side effect concerns (25.0%). Other open-text responses showed common themes of infertility, and breastfeeding as contraception. Comparing quantitative and qualitative responses revealed contradicting information, suggesting respondent misinterpretation and confusion surrounding the term “pregnancy prevention.” Though this analysis may be limited by manual coding error and researcher biases, we avoided coding exhaustion via 1-hour coding periods and validated reliability through intercoder kappa scores.ConclusionsIn this study, we observed that respondents reporting contraception nonuse often described other methods of pregnancy prevention and contraception barriers that were not included in categorical response options. Open-text responses shed light on a more comprehensive list of pregnancy prevention methods and nonuse options. Our findings contribute to survey questions that can lead to more accurate depiction of postpartum contraceptive behavior. Additionally, future use of these qualitative methods may be used to improve other health behavior survey development and resulting data.

In a mixed-methods inquiry, Nicole Richards and colleagues investigate how open-text responses can provide insight into categorical reasons for postpartum contraception use and non-use among women in the US.     

Efficacy and long term effects of antenatal prophylaxis with anti-D immunoglobulin.

OBJECTIVE--To measure the safety and efficacy of antenatal treatment with anti-D immunoglobulin. DESIGN--Open study with historical controls. SETTING--Multicentre study in 17 hospitals in West Yorkshire. PATIENTS--1238 Rh negative women who delivered Rh positive infants after 34 weeks in their first pregnancy in 1980-1 (group 1) and 2000 similar primigravidas from 1978-9 (group 2). Obstetric data were collected for 616 women in group 1 who had a subsequent pregnancy, 536 similar women in group 2, and 410 Rh positive but otherwise similar primigravidas who delivered in the same hospitals in 1978-81 (group C). INTERVENTIONS--Anti-D immunoglobulin 100 micrograms intramuscularly was given at 28 and 34 weeks to the mothers in their first pregnancy who delivered in 1980-1. END POINTS--Detection of anti-D antibody in the first or any subsequent pregnancy in groups 1 and 2. For all three groups having subsequent pregnancies gestation at delivery, birth weight, fetal survival at one month, pre-eclampsia defined as blood pressure greater than 140/90 on two occasions more than 12 hours apart, and proteinuria greater than 0.25 milligram. MEASUREMENTS AND MAIN RESULTS--Antenatal immunisation to Rh(D) occurred in six mothers in group 1 and 32 group 2. Most immunisations occurred in the first or second pregnancy. The rates of abortion, gestation at delivery, birth weight, and fetal survival were not significantly different among the three groups. The incidence of pre-eclampsia was lower in mothers given antenatal anti-D immunoglobulin, but the difference was not significant. CONCLUSIONS--Antenatal prophylaxis with anti-D immunoglobulin is effective, and the effect of giving it in the first pregnancy persists into at least the second pregnancy. It seems to be safe for the fetus in the index and subsequent pregnancies.     

Toxoplasma gondii: The effects of infection at different stages of pregnancy on the offspring of mice

Congenital toxoplasmosis can cause fetal damage in humans and domestic animals. This study was focused on the effects of Toxoplasma gondii (Prugniaud strain) infection at different stages of pregnancy on the offspring of mice. Results showed that newborn mice from all infected groups were significantly lower in weight than those from the control group but significant difference was not found among these groups at day 60 after birth. The survival rate of the offspring from the group of mice infected at the earlier stage of pregnancy was significantly lower than those of infected and control groups. The positive offspring (with cysts found in their brain tissues) born from the mice infected at the earlier and intermediate stages of pregnancy showed a shorter latency and greater number of errors in the step-through passive avoidance test than those born from the mice infected at the late stage of pregnancy, the control group and the negative offspring from the infected groups. The number of cysts in the brain tissue was significantly higher in the offspring born from the groups of mice infected at the earlier and intermediate stages of pregnancy than those from the group of mice infected at the late stage of pregnancy. In addition, our results indicated that a high congenital transmission rate (90%) occurred in this NIH mouse model. In conclusion, the earlier and intermediate maternal infection of T. gondii can result in severe congenital toxoplasmosis, exhibiting conditions such as stillbirth or non-viability, and learning or memory capability damage in this mouse model. These results not only provide useful data for better understanding the effects of T. gondii infection on the offspring of mice infected at different stages of pregnancy but also for better consideration of the effect of this infection on other mammalian hosts including humans.     

Field trial to compare pregnancy rates of bovine embryo cryopreservation methods: vitrification and one-step dilution versus slow freezing and three-step dilution

We designed and conducted a field trial to obtain accurate pregnancy rates of Day 7 bovine embryos after vitrification in PB1 containing 6.5 M glycerol and 6% BSA (w/v) and one-step dilution in 1 M sucrose compared with controlled slow freezing in 1.5 M glycerol and three-step dilution. Embryos were collected from superovulated donor cows, and Grade 1 and 2 morulae and blastocysts were randomly assigned to each cryopreservation treatment group. Dutch farmers were solicited to participate in the field trial by an advertisement that offered cryopreserved embryos at subsidized cost. Within a period of 11 wk, one of six technicians visited 150 farms. Standard nonsurgical methods were used to transfer a total of 728 cryopreserved embryos. Recipient cows, mainly multiparous and of various breeds, the so-called "bottom-end" of the national herd, received embryos either 6, 7 or 8 d after standing estrus during natural estrous cycles. We compiled a database on 22 factors that may influence establishment of pregnancy in order to check randomization of each factor over cryopreservation treatment groups and embryo transfer technicians and to perform the statistical tests. Overall pregnancy rates were 44.5% (n = 393) for vitrified embryos and 45.1% (n = 335) for slowly frozen embryos. Pregnancy rates were not significantly different (ANOVA, P = 0.79 or Chi- square analysis, P = 0.88). The registered data confirm that all factors were randomly distributed over cryopreservation methods and technicians. Technician was not a significant source of variation in pregnancy rate (analysis of variance, P = 0.79). Although three technicians performed better with the one-step procedure and the other three performed better using the three-step method, the interaction between the technician and cryopreservation method was not significant (Tukey's test for nonadditivity, P = 0.13). Our results indicate that 1) vitrification and one-step dilution can be successfully used in the field without significant reduction in the pregnancy rate and 2) the pregnancy rate obtained using the "bottom-end" of the herd is satisfactory for practical application.     

Levels of DNA damage in blood leukocyte samples from non-diabetic and diabetic female rats and their fetuses exposed to air or cigarette smoke

The objective of the present study was to evaluate DNA damage level in blood leukocytes from diabetic and non-diabetic female Wistar rats exposed to air or to cigarette smoke, and to correlate the findings with levels of DNA damage detected in blood leukocyte samples from their fetuses. A total of 20 rats were distributed into four experimental groups: non-diabetic (control; G1) and diabetic exposed to filtered air (G2); non-diabetic (G3) and diabetic (G4) exposed to cigarette smoke. Rats placed into whole-body exposure chambers were exposed for 30min to filtered air (control) or to tobacco smoke generated from 10 cigarettes, twice a day, for 2 months. Diabetes was induced by a pancreatic beta-cytotoxic agent, streptozotocin (40mg/kgb.w.). At day 21 of pregnancy, each rat was anesthetized and humanely killed to obtain maternal and fetal blood samples for genotoxicity analysis using the alkaline comet assay. G2, G3 and G4 dams presented higher DNA damage values in tail moment and tail length as compared to G1 group. There was a significant positive correlation between DNA damage levels in blood leukocyte samples from G2 and G3 groups (tail moment); G3 and G4 groups (tail length) and G3 group (tail intensity) and their fetuses. Thus, this study showed the association of severe diabetes and tobacco cigarette smoke exposure did not exacerbate levels of maternal and fetal DNA damages related with only diabetes or cigarette smoke exposure. Based on the results obtained and taking into account other published data, maternal diabetes requires rigid clinical control and public health and education campaigns should be increased to encourage individuals, especially pregnant women, to stop smoking.     

Effect of reproductive status on uptake of latex microparticles in rat small intestine

This study investigates whether pregnancy or lactation affects microparticle uptake across the small intestinal mucosal barrier, since aspects of gastrointestinal physiology such as motility may be altered in these conditions. It also reports on validation of the model by several methods and discusses the findings in relation to possible mechanisms. Anaesthetised, pregnant, lactating, virgin female or male adult rats were gavaged with fluorescent latex microparticles. The small intestine was removed and fixed either 5 or 30 min later and successive segments of equal length were examined with fluorescence microscopy. Minor adjustments were made to experimental methods to explore details of the uptake mechanism. Control sections contained no particles. All experimental samples showed luminal and surface particles and also contained particles within the tissue, most associated with villous absorptive enterocytes. Particle uptake was greatest at the 30-min time-point, when maximum uptake was usually in the proximal jejunum; although in the early lactating group, this was shifted distally. Total tissue uptake was increased in pregnant and early lactating groups, mainly at villous absorptive and mucus-secreting cells. Accumulation and progression of particles was reflected in increased numbers in the lamina propria. These data were validated by several methods, including particle detection in the blood and mesenteric lymph nodes in some groups. At both time-points, uptake profiles for pregnancy and early lactation differed from those of other groups, implying possible links between particle uptake and hormone levels, surface mucus and tight junction patency.     

Occupational exposure to anticancer drug--potential and real hazards

Many anticancer agents have been shown to be mutagenic, teratogenic and carcinogenic in experimental systems and second malignancies are known to be associated with several specific therapeutic treatments. Anticancer agents thus represent a class of occupational carcinogens, the handling of which should involve no unnecessary exposure. The available methodologies to detect possible exposures from ambient air and from biological samples are discussed, and the published data on results are reviewed. Analytical methods are available for the detection of most frequently used anticancer drugs from all groups, i.e., alkylating agents, mitotic inhibitors, antimetabolites and antibiotics. The ambient samples taken from sites of admixture of cytostatics have often shown detectable, but low concentrations of anticancer agents. Urine samples from patients under chemotherapy as well as from personnel handling the drugs occupationally in hospitals have been analyzed both chemically and for excreted mutagenicity. Both cisplatin and cyclophosphamide have been detected in the urine of patients; furthermore, cyclophosphamide was observed in the urine of nurses who formulate and deliver this drug. Urinary mutagenicity assays have given both positive and negative results in various groups of nursing and pharmacy personnel. Cytogenetic methods have, likewise, been applied for monitoring purposes. Most of the available data concerns chromosome aberrations (CA) or sister-chromatid exchanges (SCE) induced in peripheral blood lymphocytes of patients under chemotherapy. A few studies on groups occupationally exposed to anticancer drugs have given positive results, but also negative reports have appeared for these same cytogenetic parameters. No studies are as yet available on the possible carcinogenic effects of occupational handling of anticancer drugs. Two recent case-referent studies among hospital personnel have pointed to slightly increased risks of disorders in pregnancy outcome; one of the studies has shown an excess of spontaneous abortions and other malformations in children of females with a history of work with anticancer agents.     

Measurements of birth defect prevalence: Which is most useful as a comparator group for pharmaceutical pregnancy registries?

Pharmaceutical pregnancy registries document birth defects and other complications reported in pregnancies exposed to specific medications or diseases. A baseline estimate of birth defect prevalence is necessary for comparison. To identify potential teratogenic signals, the pregnancy registry must have a comparator that most closely matches the exposed population and data collection methodology, which are characteristics that vary among the multiplicity of birth defect surveillance systems. The system that yields the most accurate prevalence data may be different from that most closely matching the pregnancy registry methods. State public health programs have highly accurate and precise statistics, but their populations are broader than those of a pharmaceutical pregnancy registry. Large collaborative databases may have a more useful covered population, but there are secondary problems related to data precision. Health care databases enroll large numbers of patients and have good information about exposures and health problems, but the data can be difficult to access and lack useful detail. Exposure‐related databases are closer in population definition and collection methods, though the presence of different diseases and exposures can be problematic. Internal comparators are likely to be most useful in formal statistical analysis, but added cost and management burden and may require significantly increased registry enrollment. There is no ideal comparator, and this must be taken into account when planning a single‐exposure or single‐disease pregnancy registry. Birth Defects Research (Part A), 2009. © 2009 Wiley‐Liss, Inc.