Comparison of imputation and imputation-free methods for statistical analysis of mass spectrometry data with missing data.

Missing values are common in high-throughput mass spectrometry data. Two strategies are available to address missing values: (i) eliminate or impute the missing values and apply statistical methods that require complete data and (ii) use statistical methods that specifically account for missing values without imputation (imputation-free methods). This study reviews the effect of sample size and percentage of missing values on statistical inference for multiple methods under these two strategies. With increasing missingness, the ability of imputation and imputation-free methods to identify differentially and non-differentially regulated compounds in a two-group comparison study declined. Random forest and k-nearest neighbor imputation combined with a Wilcoxon test performed well in statistical testing for up to 50% missingness with little bias in estimating the effect size. Quantile regression imputation accompanied with a Wilcoxon test also had good statistical testing outcomes but substantially distorted the difference in means between groups. None of the imputation-free methods performed consistently better for statistical testing than imputation methods.

Altered behavior, brain structure, and neurometabolites in a rat model of autism-specific maternal autoantibody exposure.

Maternal immune dysregulation is a prenatal risk factor for autism spectrum disorder (ASD). Importantly, a clinically relevant connection exists between inflammation and metabolic stress that can result in aberrant cytokine signaling and autoimmunity. In this study we examined the potential for maternal autoantibodies (aAbs) to disrupt metabolic signaling and induce neuroanatomical changes in the brains of exposed offspring. To accomplish this, we developed a model of maternal aAb exposure in rats based on the clinical phenomenon of maternal autoantibody-related ASD (MAR-ASD). Following confirmation of aAb production in rat dams and antigen-specific immunoglobulin G (IgG) transfer to offspring, we assessed offspring behavior and brain structure longitudinally. MAR-ASD rat offspring displayed a reduction in pup ultrasonic vocalizations and a pronounced deficit in social play behavior when allowed to freely interact with a novel partner. Additionally, longitudinal in vivo structural magnetic resonance imaging (sMRI) at postnatal day 30 (PND30) and PND70, conducted in a separate cohort of animals, revealed sex-specific differences in total and regional brain volume. Treatment-specific effects by region appeared to converge on midbrain and cerebellar structures in MAR-ASD offspring. Simultaneously, in vivo 1H magnetic resonance spectroscopy (1H-MRS) data were collected to examine brain metabolite levels in the medial prefrontal cortex. Results showed that MAR-ASD offspring displayed decreased levels of choline-containing compounds and glutathione, accompanied by increased taurine compared to control animals. Overall, we found that rats exposed to MAR-ASD aAbs present with alterations in behavior, brain structure, and neurometabolites; reminiscent of findings observed in clinical ASD.

Validation of a new calculator for predicting success of vaginal birth after cesarean delivery.

Race as a variable in a predictive model for a successful vaginal birth after cesarean delivery has been challenged as contributing to health inequity. In May 2022, the National Institute of Child Health and Development released a modified calculator that removed race as a variable. The aim of this study was to externally validate the revised calculator amongst a cohort at our institution. We reviewed all patients who underwent a trial of labor after cesarean delivery in 2018-2020 at a tertiary academic medical center and calculated the predicted probability of successful vaginal birth after cesarean delivery for each patient using both original and revised classification calculators and compared these to observed birth outcomes. The area under the receiver operating characteristic curve was calculated for each model. From the cohort of 225 patients that fit inclusion criteria, 37% (n=83) identified as African-American or Hispanic. The vaginal birth after cesarean delivery success rate was 75% for the entire population, and 76% among African-American and/or Hispanic patients. The area under the receiver operating characteristic curve of the original calculator was 0.71, compared to 0.74 for the new calculator. For African-American and/or Hispanic patients, the average predicted success rates between the models rose from 60 to 69%. Our review confirmed that African-American and Hispanic patients were calculated to have a lower prediction score for a successful vaginal birth after cesarean delivery based on the original calculator as compared to the revised calculator. Our results also suggest that race/ethnicity did not significantly contribute to classification ability of the calculator in our patient population.

Time to completion of radiation treatment in locally advanced squamous cell carcinoma of the vulva and the impact on survival.

OBJECTIVE: To assess whether radiation completion within a planned timeframe in locally advanced squamous cell vulvar cancer impacts overall survival (OS). METHODS: The National Cancer Database from 2004 to 2017 was used to identify women ≥18 years old with stage II-IVA squamous cell vulvar cancer. We included women who received radiation alone (RT) or concurrent chemoradiation (CRT) for initial vulvar cancer treatment. Primary outcome was overall survival associated with time of delay in radiation completion. RESULTS: There were 2378 women identified (n = 856 RT and n = 1522 CRT). Median age was 67 (IQR 56-78), majority (88.35%) were white with advanced stage III or IVA (72.29%) disease. Median radiation dose was 5720 c-Gray (IQR 5040-6300). Radiation completion with delay ≥7 days resulted in reduction in survival compared to delay of <7 days (unadjusted HR 1.183 [95%CI: 1.066-1.313], p = 0.0016). When delays extended to ≥14 days compared to <14 days there was increased hazard of death (unadjusted HR: 1.263 [95%CI:1.126-1.416], p < 0.0001). Survival improved for patients with <7 versus ≥7 days delay whether treatment was with RT (median OS: 34.9 months versus 21.6 months, p < 0.01) or CRT (Median OS:58 months versus 41.3 months, p < 0.01). Stage IVA disease was associated with the greatest increase in hazard of death (HR 1.759 [95%CI 1.517-2.039], p < 0.0001) compared to stage II. CONCLUSION: Radiation completion with <7 days delay is associated with improved overall survival, independent of concurrent chemotherapy. This suggest that strategies to minimize delays in radiation are crucial in locally advanced vulvar cancer.

Postoperative Opioid Prescribing and Consumption after Hysterectomy: A Prospective Cohort Study.

STUDY OBJECTIVE: To examine opioid prescribing and consumption patterns after hysterectomy and identify factors associated with postoperative opioid consumption. DESIGN: Prospective cohort study. SETTING: Single university medical center. PATIENTS: Women undergoing hysterectomy for benign, nonobstetric indications. INTERVENTIONS: Participant preoperative and surgical characteristics were obtained through chart review and patient report of baseline pain score. During the third postoperative week, participants completed a telephone interview, including a direct count of remaining opioid pills and assessment of satisfaction with pain management. We assessed factors associated with opioid consumption in oral morphine equivalents (OME) using a linear regression model. MEASUREMENTS AND MAIN RESULTS: Of the 129 participants, 113 (88%) completed the postoperative survey after hysterectomy: 16 vaginal, 43 robotic-assisted, 42 conventional laparoscopic, and 12 abdominal hysterectomies. The median amount of opioid prescribed was 150 OME (interquartile range [IQR] 113-200), while the median amount consumed was 75 (IQR 10-135), reflecting an average consumption of about 50% of the prescription. Opioid prescription size was associated with consumption; for every additional oral morphine equivalent prescribed, on average, an additional 0.5 was consumed (p <.001). If the indication for hysterectomy was related to pain, participants consumed 25.3 additional OME (p = .04). The amount of opioid prescribed was inversely correlated with pain management satisfaction; every additional point on a 1 through 5 Likert scale of increasing satisfaction was associated with 44 fewer OME prescribed (standard error 9 OME, p <.001). For the 1464 total unused pills among the 104 participants with leftover opioids, only 20% reported an Food and Drug Administration -compliant opioid disposal plan. CONCLUSION: Gynecologic surgeons can respond to the opioid epidemic by reducing excess opioid pills after hysterectomy by providing both the smallest effective prescription size and concrete resources for safe opioid disposal. These actions may contribute to a reduction in opioid use disorder cases or overdose deaths.

Factors affecting umbilical cord blood stem cell suitability for transplantation in an in utero collection program.

BACKGROUND: The greatest limiting factor in the use of umbilical cord blood (UCB) for transplantations is the number of stem cells available in any given collected unit. This study examines maternal, neonatal, and obstetric factors that influence the suitability for banking and transplantation of UCB units collected in utero. STUDY DESIGN AND METHODS: This study examined 7839 UCB units collected in utero at two hospitals from August 1, 2008, to August 31, 2011. The variables were collected from cord blood records, patient charts, delivery records, and data processed by StemCyte, Inc. (Covina, CA), our parent bank. The mean total nucleated cell (TNC) count between units selected for transplant and those banked but not selected for transplant were also compared. RESULTS: Of the 11 variables evaluated, six had a significant influence on the acceptable banking TNC count of at least 90 × 10(7) . Of these, cord blood volume was the best predictor. The other variables that were predictors of the acceptable TNC count of significance were gestational age, infant race, parity, birthweight, and infant sex. The minimal bankable TNC count for banking is considerably lower than the mean TNC count of units actually selected for transplantation. CONCLUSION: Maternal, neonatal, and obstetric factors can all influence the acceptability of a cord blood unit collected in utero for banking. Furthermore, units with high TNC counts are more likely to be selected for banking. Identifying these variables in potential donors can likely increase the number of adequate collections.

Variation of FMRP Expression in Peripheral Blood Mononuclear Cells from Individuals with Fragile X Syndrome.

Fragile X syndrome (FXS) is the most common heritable cause of intellectual disability and autism spectrum disorder. The syndrome is often caused by greatly reduced or absent protein expression from the fragile X messenger ribonucleoprotein 1 (FMR1) gene due to expansion of a 5'-non-coding trinucleotide (CGG) element beyond 200 repeats (full mutation). To better understand the complex relationships among FMR1 allelotype, methylation status, mRNA expression, and FMR1 protein (FMRP) levels, FMRP was quantified in peripheral blood mononuclear cells for a large cohort of FXS (n = 154) and control (n = 139) individuals using time-resolved fluorescence resonance energy transfer. Considerable size and methylation mosaicism were observed among individuals with FXS, with FMRP detected only in the presence of such mosaicism. No sample with a minimum allele size greater than 273 CGG repeats had significant levels of FMRP. Additionally, an association was observed between FMR1 mRNA and FMRP levels in FXS samples, predominantly driven by those with the lowest FMRP values. This study underscores the complexity of FMR1 allelotypes and FMRP expression and prompts a reevaluation of FXS therapies aimed at reactivating large full mutation alleles that are likely not capable of producing sufficient FMRP to improve cognitive function.

Longitudinal follow-up of metformin treatment in Fragile X Syndrome.

Introduction: Metformin has been used as a targeted treatment to potentially improve cognition and slow the typical IQ decline that occurs during development among individuals with fragile X syndrome (FXS). In this follow-up study, we are following the trajectory of IQ and adaptive behavior changes over 1 to 3 years in individuals with FXS who are clinically treated with metformin in an open label trial. Method: Individuals with FXS ages 6 to 25 years (mean 13.15 ± 5.50) and nonverbal IQ mean 57.69 (±15.46) were treated for 1-3 years (1.88 ± 0.63). They all had a baseline IQ test using the Leiter-III non-verbal cognitive assessment and the Vineland-III adaptive behavior assessment before the start of metformin. Repeat Leiter-III and Vineland-III were completed after at least 1 year of metformin (500-1,000 mg/dose given twice a day). Result: There were no significant changes in non-verbal IQ or in the adaptive behavior measurements at FDR < 0.05. The findings thus far indicate that both IQ and adaptive behavior are stable over time, and we did not see a significant decline in either measure. Conclusion: Overall, the small sample size and short follow-up duration limit the interpretation of the effects of metformin on cognitive development and adaptive functioning. There is individual variability but overall for the group there was no significant decline in IQ or adaptive behavior.

Complex Family Planning fellowship graduates' intended practice plans and barriers to practicing in areas of unmet need.

OBJECTIVES: To describe practice patterns and challenges encountered by Complex Family Planning (CFP) fellowship graduates. STUDY DESIGN: We invited all 110 obstetrics and gynecology physicians who graduated from the CFP fellowship from 2017-2020 via email to complete an anonymous online survey. We inquired about demographics, intended and obtained postfellowship positions, and successes and challenges in obtaining jobs. We used Fisher's exact test to assess if the proportion of graduates who grew up, attended residency, and completed fellowship in a US region (Northeast, Midwest, South, and West) and practiced in that same region differed. RESULTS: Ninety-nine (90.0%) graduates completed the survey. When entering fellowship, most (n = 92 [92.9%]) expected to practice in an academic environment. About half (n = 49 [49.5%]) pursued fellowship with the intent to practice in a location with an unmet need for abortion providers, of which 22 (44.9%) did so. Forty-nine (49.5%) respondents did not practice after fellowship where they initially intended, citing common challenges of job availability, family-related concerns, safety concerns, and relationship status changes. We found associations between regions where graduates completed residency and currently practice (p = 0.004), driven primarily by higher associations in the South (76.9%) and West (70.6%) and a lower association in the Midwest (22.7%). We found no association between current practice region and where graduates grew up (p = 0.15) or completed fellowship (p = 0.23). CONCLUSIONS: CFP fellowship graduates from 2017-2020 primarily intended to practice in academic environments with half planning to practice in underserved locations. However, more than half of those who entered fellowship hoping to fill an unmet need for abortion providers did not do so. IMPLICATIONS: About half of CFP fellowship graduates from 2017-2020 intended to obtain positions in areas they defined as having an unmet need for abortion provision. Personal life and job barriers prevented many from serving in such positions after fellowship. Practice location intentions and outcomes may be different in a post-Dobbs environment.

Breastfeeding Perceptions and Behavior Among Postpartum Women Initiating Different Hormonally Systemic Contraceptive Methods.

BACKGROUND: There continues to be controversy regarding the simultaneous encouragement of both breastfeeding and immediate postpartum contraception. RESEARCH AIM: To explore postpartum women's perspectives about breastfeeding and their breastfeeding behaviors, while using one of three different hormonally systemic contraceptive methods immediately postpartum over a 6 month period of time. METHODS: This was a retrospective, longitudinal, three group comparative, secondary analysis of a prospective cohort study (N = 471) of immediate postpartum contraception. Breastfeeding, for this study, was defined as providing any human milk to the infant. Participants who chose one of three different hormonally systemic forms of contraception immediately postpartum (a long-acting hormonal reversible contraceptive (n = 200), depot medroxyprogesterone acetate 150 mg (n = 98), or a non-hormonal method (n = 173)) were compared at hospital discharge, 6 weeks, 3 months, and 6 months postpartum. The primary outcome was any breastfeeding at 6 months. Secondary outcomes included any and exclusive breastfeeding, concerns about breastfeeding while using contraception, and reasons for breastfeeding discontinuation. RESULTS: There was no significant difference in the rate of any breastfeeding between the two hormonal and the non-hormonal contraceptive groups at 6 months postpartum (long-acting hormonal 20.1%, non-hormonal 21.7%, depot medroxyprogesterone acetate 13.9%, p = .77, 0.28, respectively). The number of participants who reported stopping breastfeeding due to decreased milk supply was not significantly different between any groups at all time points (total number who discontinued at 6 months postpartum was long-acting hormonal 24.7%, non-hormonal 25.1%, depot medroxyprogesterone acetate 19.3%, p = .30). CONCLUSIONS: Breastfeeding perspectives and behavioral outcomes over the first 6 months postpartum were not influenced by participants chosen form of immediate postpartum contraception.

Clinical interventions are more accurate than quantitative measurements for defining hemorrhage with dilation and evacuation.

OBJECTIVES: To assess if quantitative blood loss (QBL) with dilation and evacuation (D&E) procedures correlated with clinically relevant outcomes or hemorrhage. STUDY DESIGN: We used a de-identified database to review D&E procedures performed at UC Davis Health from April 2019 through March 2020. Surgeons determined QBL during procedures and estimated blood loss, when excessive, during post-procedure recovery. We extracted patient demographic and procedure-related information. We defined clinically relevant bleeding as cases with bleeding-related interventions within 24 hours post-procedure including use of ≥2 uterotonics, tranexamic acid administration, cervical injury requiring repair, uterine balloon tamponade, blood transfusion, uterine artery embolization, hospitalization, or return to operating room; the latter 5 criteria defined hemorrhage. We used χ2 test for trend to evaluate bleeding outcomes. RESULTS: We evaluated 431 procedures with a mean gestational age of 19 weeks and 3 days. Clinically relevant bleeding outcomes occurred in 6/319 (2%), 15/97 (15%) and 7/12 (58%) patients with total blood loss <250mL, 250-500mL and >500mL, respectively (p<0.0001); 11 had bleeding related to cervical injuries. Hemorrhage occurred in 0, 4/97 (4%) and 5/12 (42%) patients, respectively (p<0.0001). Patients with relevant bleeding outcomes had QBLs ranging from 150-1800mL (median QBL 312.5mL, interquartile range [IQR] 250-550mL) while those without clinically relevant bleeding ranged from 10-900mL (median QBL 150mL, IQR 75-200mL). CONCLUSION: Most patients (75%) with clinically relevant bleeding outcomes had QBL ≤500mL. Although higher QBL correlates with clinical interventions, the need for significant interventions rather than a single blood loss amount should be used to define hemorrhage with D&E procedures. IMPLICATIONS: Clinical hemorrhage is best defined by the necessary clinical interventions required to manage bleeding rather than any quantified amount of blood loss.

Characteristics and outcomes of patients undergoing second-trimester dilation and evacuation for intrauterine fetal demise vs induced abortion.

OBJECTIVES: Patients with intrauterine fetal demise (IUFD) are at higher risk of complications when undergoing dilation and evacuation (D&E) compared to patients undergoing abortion for other indications. We aimed to compare baseline characteristics and describe outcomes, including frequencies of complications such as disseminated intravascular coagulation (DIC) and hemorrhage, in patients undergoing D&E for IUFD vs induced abortion, with a goal of identifying associated risk factors for complications. STUDY DESIGN: We conducted a retrospective matched cohort study of patients undergoing nonemergent D&Es for singleton ≥14-0/7-week IUFD January 1, 2019 to May 31, 2021, matched with two patients undergoing induced second-trimester D&Es by cesarean delivery history, patient age, and gestational age (GA). We collected demographics, history, GA, coagulation studies, quantitative blood loss (QBL), and complications. We calculated descriptive statistics and tested for association using chi-square, Fisher's exact, t, and Wilcoxon's rank sum tests. RESULTS: Of 1390 procedures, 64 patients with IUFD met inclusion criteria and were matched with 128 patients undergoing induced D&E. Eight (12.5%) patients with IUFD and six (4.7%) undergoing induced D&E had hemorrhage (odds ratio [OR] = 2.90, 95% confidence interval [0.96, 8.77]). Six (9.4%) patients with IUFD and none undergoing induced D&E had DIC (OR = 28.56 [1.58, 515.38]). Median QBL was 75.0 mL (50, 162.5) for patients with IUFD vs 110.0 mL (50, 200) for those undergoing induced D&E (p = 0.083). Twelve (18.8%) patients with IUFD vs seven (5.5%) undergoing induced D&E received at least one intervention due to bleeding complications (p = 0.004). CONCLUSIONS: We found a higher DIC frequency but no significant difference in hemorrhage or QBL in IUFD D&E compared to induced abortion. Our IUFD D&E complication frequency is higher than those previously published. IMPLICATIONS: Our results affirm current standards of care for D&E in patients with IUFD. Large referral centers may have higher proportions of complications compared to other sites.

Examining indicators of psychosocial risk and resilience in parents of autistic children.

Background: Parents of autistic children experience increased levels of caregiver strain and adverse mental health outcomes, even in comparison to parents of children with other neurodevelopmental disabilities. Previous studies have largely attributed these increased levels of mental health concerns to their child behavioral concerns and autism symptomatology, but less attention has been given to other potential child factors, such as child adaptive functioning. Additionally, little is known about potential protective factors, such as parents' emotion regulation (ER) abilities, that may ameliorate the experience of caregiver strain, anxiety, and depression. Objective: The current study examined the impact of child characteristics (restricted and repetitive behaviors, adaptive functioning and behavioral concerns) on parent mental health outcomes (caregiver strain, anxiety, depression and wellbeing). Additionally, we explore parents' ER abilities as a moderator of the impact of child characteristic on parents' mental health outcomes. Results: Results of linear mixed effect models indicated a significant relationship between parents' ER abilities and all four parent outcomes. Additionally, children's adaptive functioning abilities and repetitive behaviors (RRBs) were significant predictors of caregiving strain. Parents' ER abilities were a significant moderator of the effect of children's repetitive behaviors and adaptive functioning challenges on caregiver strain, such that better ER abilities mitigated the impact of child clinical factors on caregiver strain. Finally, a significant difference was detected for mothers' and fathers' mental health, with mothers reporting higher caregiver strain, and more symptoms of anxiety and depression than did fathers. Conclusion: This study leveraged a large sample of autistic children and their biological parents to examine the relationship between children's clinical characteristics and parents' psychological wellbeing. Results indicate that, although parents of autistic children do experience high rates of internalizing mental health concerns that relate to child adaptive functioning and RRBs, parent ER abilities act as a protective factor against parents' adverse mental health outcomes. Further, mothers in our sample reported significantly higher rates of depression, anxiety, and caregiver strain, as compared with fathers.

Intercorrelation of Molecular Biomarkers and Clinical Phenotype Measures in Fragile X Syndrome.

This study contributes to a greater understanding of the utility of molecular biomarkers to identify clinical phenotypes of fragile X syndrome (FXS). Correlations of baseline clinical trial data (molecular measures-FMR1 mRNA, CYFIP1 mRNA, MMP9 and FMRP protein expression levels, nonverbal IQ, body mass index and weight, language level, NIH Toolbox, adaptive behavior rating, autism, and other mental health correlates) of 59 participants with FXS ages of 6-32 years are reported. FMR1 mRNA expression levels correlated positively with adaptive functioning levels, expressive language, and specific NIH Toolbox measures. The findings of a positive correlation of MMP-9 levels with obesity, CYFIP1 mRNA with mood and autistic symptoms, and FMR1 mRNA expression level with better cognitive, language, and adaptive functions indicate potential biomarkers for specific FXS phenotypes. These may be potential markers for future clinical trials for targeted treatments of FXS.

Adaptive, behavioral, and cognitive outcomes in individuals with fragile X syndrome with varying autism severity.

This study aimed to determine the association between severity of autism spectrum disorder (ASD) and cognitive, behavioral, and molecular measures in individuals with fragile X syndrome (FXS). Study inclusion criteria included individuals with FXS and (1) age 6-40 years, (2) full-scale IQ < 84, and (3) language ≥3-word phrases. ASD symptom severity was determined by Autism Diagnostic Observation Schedule-2 (ADOS-2). Other measures identified non-verbal IQ, adaptive skills, and aberrant behaviors. Molecular measures included blood FMR1 and CYFIP1 mRNA levels, FMRP and MMP9 levels. Analysis of variance (ANOVA) and Spearman's correlations were used to compare ASD severity groups. Data from 54 individuals was included with no/mild (N = 7), moderate (N = 18), and severe (N = 29) ASD. Individuals with high ASD severity had lower adaptive behavior scores (47.48 ± 17.49) than the no/mild group (69.00 ± 20.45, p = 0.0366); they also had more challenging behaviors, lethargy, and stereotypic behaviors. CYFIP1 mRNA expression levels positively correlated with the ADOS-2 comparison score(r2  = 0.33, p = 0.0349), with no significant correlations with other molecular markers. In conclusion, autism symptom severity is associated with more adverse cognitive and adaptive skills and specific behaviors in FXS, whereas CYFIP1 mRNA expression levels may be a potential biomarker for severity of ASD in FXS.

Open-Label Sulforaphane Trial in FMR1 Premutation Carriers with Fragile-X-Associated Tremor and Ataxia Syndrome (FXTAS).

Fragile X (FMR1) premutation is a common mutation that affects about 1 in 200 females and 1 in 450 males and can lead to the development of fragile-X-associated tremor/ataxia syndrome (FXTAS). Although there is no targeted, proven treatment for FXTAS, research suggests that sulforaphane, an antioxidant present in cruciferous vegetables, can enhance mitochondrial function and maintain redox balance in the dermal fibroblasts of individuals with FXTAS, potentially leading to improved cognitive function. In a 24-week open-label trial involving 15 adults aged 60-88 with FXTAS, 11 participants successfully completed the study, demonstrating the safety and tolerability of sulforaphane. Clinical outcomes and biomarkers were measured to elucidate the effects of sulforaphane. While there were nominal improvements in multiple clinical measures, they were not significantly different after correction for multiple comparisons. PBMC energetic measures showed that the level of citrate synthase was higher after sulforaphane treatment, resulting in lower ATP production. The ratio of complex I to complex II showed positive correlations with the MoCA and BDS scores. Several mitochondrial biomarkers showed increased activity and quantity and were correlated with clinical improvements.

Imputation of Missing Values for Multi-Biospecimen Metabolomics Studies: Bias and Effects on Statistical Validity.

The analysis of high-throughput metabolomics mass spectrometry data across multiple biological sample types (biospecimens) poses challenges due to missing data. During differential abundance analysis, dropping samples with missing values can lead to severe loss of data as well as biased results in group comparisons and effect size estimates. However, the imputation of missing data (the process of replacing missing data with estimated values such as a mean) may compromise the inherent intra-subject correlation of a metabolite across multiple biospecimens from the same subject, which in turn may compromise the efficacy of the statistical analysis of differential metabolites in biomarker discovery. We investigated imputation strategies when considering multiple biospecimens from the same subject. We compared a novel, but simple, approach that consists of combining the two biospecimen data matrices (rows and columns of subjects and metabolites) and imputes the two biospecimen data matrices together to an approach that imputes each biospecimen data matrix separately. We then compared the bias in the estimation of the intra-subject multi-specimen correlation and its effects on the validity of statistical significance tests between two approaches. The combined approach to multi-biospecimen studies has not been evaluated previously even though it is intuitive and easy to implement. We examine these two approaches for five imputation methods: random forest, k nearest neighbor, expectation-maximization with bootstrap, quantile regression, and half the minimum observed value. Combining the biospecimen data matrices for imputation did not greatly increase efficacy in conserving the correlation structure or improving accuracy in the statistical conclusions for most of the methods examined. Random forest tended to outperform the other methods in all performance metrics, except specificity.

Contraceptive plans before preoperative assessment and at procedure in surgical abortion patients.

OBJECTIVE: To describe changes in contraceptive method plans pre-appointment, after counseling, and post-procedure in patients having an abortion. STUDY DESIGN: We reviewed electronic medical records of University of California, Davis Health patients who had an operating room abortion from January 2015 to December 2016. We excluded persons with procedures for fetal anomaly or demise. We extracted patient demographics and contraceptive plans reported at each encounter (telephone intake, pre-operative appointment, and day of abortion). We evaluated individual contraceptive plans across the encounters, identified patient characteristics that contributed to plan change, and created a multivariable logistic regression model for predictors of contraception method plan change from telephone intake to post-procedure. RESULTS: The 747 patients had a mean gestational age of 16 4/7 ± 5 0/7 weeks with 244 (32.7%) <15 weeks and 235 (31.5%) ≥20 weeks. At telephone intake, 273 (36.4%) wanted a long-acting method (139 [50.9%] intrauterine device [IUD]; 99 [36.3%] implant; 35 [12.3%] unspecified), 11 (3.9%) permanent contraception, and 248 (33.2%) a less effective or no method; 215 (28.8%) stated they were undecided. Most (357/433 [82.4%]) patients who planned a reversible method based on the telephone intake obtained that or a similar method. Of the 273 patients planning a long-acting method, 258 (94.5%) received an IUD (158 [40.9%]) or implant (100 [36.6%]). Of the 215 undecided patients, 88 (40.9%) received an IUD and 55 (25.6%) an implant. No demographic factors predicted a change in method plan. CONCLUSIONS: Most patients will receive the method they initially identified at the telephone intake after an abortion, especially those planning an IUD or implant. Undecided patients are commonly open to discussing options.

Vomiting and risk of endotracheal intubation related to preoperative doxycycline use for dilation and evacuation.

OBJECTIVE: To describe the rate of vomiting from oral doxycycline 200 mg given the night before second trimester dilation and evacuation (D&E), proportion of anesthesia modalities, and anesthetic complications. STUDY DESIGN: We conducted a single-institution retrospective cohort study of patients presenting for second trimester D&E (14-0/7 to 23-6/7 weeks gestation) July 1, 2019-June30, 2020 following their scheduled preoperative visit as identified by billing codes. We recorded vomiting within 30 minutes of ingestion, anesthetic modality, and anesthetic complications. We tested for associations using chi-square or Fisher's exact test for categorical variables and Wilcoxon-rank sum for non-normal numeric variables. RESULTS: We reviewed 702 charts, of which 461 (66%) met inclusion criteria and 420 (60%) took doxycycline as prescribed. Of those who took doxycycline as prescribed, 30 (7.14%) reported vomiting within 30 minutes of ingestion. Nulliparity, primigravida and age less than 30 were significantly associated with vomiting (p = 0.005, p < 0.001 and p = 0.03, respectively), but gestational age (p = 0.53), BMI (p = 0.93), and gastrointestinal conditions (p > 0.99) were not. Only gravidity (p < 0.001) and parity (p = 0.01) remained significant in each of their respective multivariate models. None of the 10 patients who received general endotracheal tube anesthesia (2.4%) had vomited from doxycycline preoperatively. We observed 5 (1.2%) anesthetic complications (postoperative nausea or vomiting, anaphylaxis, and aspiration) that occurred only in those without vomiting. CONCLUSIONS: Vomiting rates following doxycycline were lower than those previously published. We found no significant association between doxycycline-associated vomiting and increased need for general endotracheal tube anesthesia or anesthetic complications; however, our study is underpowered to draw further conclusions. IMPLICATIONS: The findings of this study are consistent with guidelines indicating deep sedation as an effective anesthetic modality with low complication rates. Nulliparous patients may benefit from administration of an antiemetic prior to doxycycline prophylaxis, but routine antiemetic use may not be necessary.

Complex family planning and pediatric hematology oncology integrated clinic for young people with blood disorders and heavy or abnormal menstrual bleeding.

OBJECTIVES: To describe practice patterns of an integrated complex family planning-pediatric hematology oncology clinic for patients with blood disorders STUDY DESIGN: We retrospectively evaluated the outcomes of patients who had an initial consultation for blood disorders impacting menstrual bleeding in an integrated complex family planning-pediatric hematology oncology clinic from October 2015 to September 2020. We reviewed all charts to extract medical and gynecologic history, blood disorder diagnosis, hormonal treatment prior to and following initial consultation, subsequent visits to the integrated clinic, and hormonal treatment up to 24 months after initial consultation. RESULTS: We saw 47 patients; their most common blood disorder diagnosis was protein defect (14 of 47, 30%). Most patients (30 of 47, 64%) were not using any hormonal treatment prior to their initial consultation. After the initial consultation, 26 (55%) elected to start, change, or discontinue hormonal treatment for abnormal menstrual bleeding, the most common treatment being combined hormonal contraception (CHC, 22 of 47, 47%), alone or as dual therapy. Over the study duration, 36 patients (77%) initiated, changed, or discontinued their hormone treatment, 22 (61%) of whom changed their treatment plan more than once. CHC usage decreased from 19 of 47 (40%) to 8 of 37 (22%) and hormonal device usage, particularly the implant, increased from 9 of 47 (19%) to 11 of 37 (30%) over the 24 months from initial consultation. CONCLUSION: Most patients in an integrated complex family planning-pediatric hematology oncology clinic will change their menstrual bleeding hormone treatment with initial consultation, although management may require multiple changes. The most common treatment 24 months following the initial consultation was hormonal devices. IMPLICATIONS: Patients with blood disorders affecting menstrual bleeding have complex needs that could be addressed by an integrated complex family planning-pediatric hematology oncology clinic. Most patients require multiple changes in treatment to achieve adequate control of their bleeding, and patients were more likely to choose hormonal devices for management over time.