985. Safety of Guidelines Recommending LAIV for Routine Use in Children and Adolescents With Asthma

Background

Asthma is the most common chronic medical condition in children. Prior observational studies of live attenuated influenza vaccine (LAIV) safety in asthmatic children have been limited due to confounding by indication, with LAIV restricted to patients with mild asthma. To minimize bias, we evaluated safety of LAIV in children with asthma using a natural experiment in which two medical groups, within a single health system, serving similar populations, differed in vaccination guidelines. Prior to 2010 both groups recommended inactivated influenza vaccine (IIV). Starting in 2010, one group recommended LAIV for children with asthma.

Methods

Asthmatic children age 2–18 years with visits to two large medical groups in the upper Midwest from 2007 to 2015 were identified and classified by severity and control using validated algorithms. Primary outcomes were lower respiratory events (LRE) occurring within 21 and 42 days after influenza immunization. Multiple records per subject were included when children received influenza vaccines in more than one season. The analysis was intention to treat with each medical group’s subjects analyzed as a group. A pre-/post-ratio of ratios (ROR) approach was used to estimate the LAIV guideline impact using a generalized linear model with a Poisson distribution, accounting for multiple records per subject and adjusting for age and asthma classification. Analyses were for the overall population, and stratified by age group: 2–4 and 5–18 years.

Results

A total of 7,959 observations from 4,824 unique asthmatic children were analyzed, with 1,896 from the IIV guideline and 6,061 from the LAIV guideline medical groups. Postimplementation, 67% received LAIV. Age and asthma classification adjusted ROR showed no increase in LREs using the LAIV guideline: overall ROR (95% CI): 0.79 (0.46–1.37) for LRE 21 days and 0.82 (0.56–1.20) for 42 days; age 2–4: 1.07 (0.40–2.83) for 21 days and 1.0 (0.53–1.90) for 42 days; and age 5–18: 0.72 (0.37–1.41) for 21 days and 0.75 (0.46–1.21) for 42 days.

Conclusion

A guideline recommending LAIV rather than IIV for asthmatic children did not result in more LREs following vaccination in children age 2–18. Guidelines for influenza vaccination in asthmatic children should be based on effectiveness studies.

Disclosures

All authors: No reported disclosures.

Uptake and safety of Hepatitis B vaccination during pregnancy: A Vaccine Safety Datalink study

INTRODUCTION

Hepatitis B virus (HBV) infection acquired during pregnancy can pose a risk to the infant at birth that can lead to significant and lifelong morbidity. Hepatitis B vaccine (HepB) is recommended for anyone at increased risk for contracting HBV infection, including pregnant women. Limited data are available on the safety of HepB administration during pregnancy.

OBJECTIVES

To assess the frequency of maternal HepB receipt among pregnant women and evaluate the potential association between maternal vaccination and pre-specified maternal and infant safety outcomes.

METHODS

We examined a retrospective cohort of pregnancies in the Vaccine Safety Datalink (VSD) resulting in live birth outcomes from 2004 through 2015. Eligible pregnancies in women aged 12-55 years who were continuously enrolled from 6 months pre-pregnancy to 6 weeks postpartum in VSD integrated health systems were included. We compared pregnancies with HepB exposure to those with other vaccine exposures, and to those with no vaccine exposures. High-risk conditions for contracting HBV infection were identified up to one-year prior to or during the pregnancy using ICD-9 codes. Maternal and fetal adverse events were also evaluated according to maternal HepB exposure status.

RESULTS

Among over 650,000 pregnancies in the study period, HepB was administered at a rate of 2.1 per 1000 pregnancies (n = 1399), commonly within the first 5 weeks of pregnancy. Less than 3% of the HepB-exposed group had a high-risk ICD-9 code indicating need for HepB; this was similar to the rate among HepB unvaccinated groups. There were no significant associations between HepB exposure during pregnancy and gestational hypertension, gestational diabetes, pre-eclampsia/eclampsia, cesarean delivery, pre-term delivery, low birthweight or small for gestational age infants.

CONCLUSIONS

Most women who received maternal HepB did not have high-risk indications for vaccination. No increased risk for the adverse events that were examined were observed among women who received maternal HepB or their offspring.

Performance of the Net Reclassification Improvement for Nonnested Models and a Novel Percentile-Based Alternative

The net reclassification improvement (NRI) is a widely used metric used to assess the relative ability of 2 risk models to distinguish between low- and high-risk individuals. However, the validity and usefulness of the NRI have been questioned. Criticism of the NRI focuses on its use comparing nested risk models, whereas in practice it is often used to compare nonnested risk models derived from distinct data sources. In this study, we evaluated the performance of the NRI in a nonnested context by using it to compare competing cardiovascular risk-prediction models. We explored the NRI's sensitivity to variations in risk categories and to the calibration of the compared models. We found that the NRI was very sensitive to changes in the definition of risk categories, especially when at least 1 model was miscalibrated. To address these shortcomings, we describe a novel alternative to the usual NRI that uses percentiles of risk instead of cutoffs based on absolute risk. This percentile-based NRI demonstrates the relative ability of 2 models to rank patient risk. It displays more stable behavior, and we recommend its use when there are no established risk categories or when models are miscalibrated.

Cost-Effectiveness of a Community-Based Diabetes Prevention Program with Participation Incentives for Medicaid Beneficiaries

OBJECTIVE

To examine the cost-effectiveness of a community-based Diabetes Prevention Program (DPP) for Medicaid beneficiaries from the perspective of the health care sector.

DATA SOURCES/STUDY SETTING

A total of 847 Medicaid enrollees at high risk for type 2 diabetes participating in a community-based DPP.

STUDY DESIGN

Pre- and post clinical outcome and cost data were used as inputs into a validated diabetes simulation model. The model was used to evaluate quality-adjusted life years (QALYs) and health care costs over a 40-year time horizon from the perspective of the health care sector.

DATA COLLECTION/EXTRACTION METHODS

Clinical outcome and cost data were derived from a study examining the effect of financial incentives on weight loss.

PRINCIPAL FINDINGS

Study participants lost an average of 4.2 lb (p < .001) and increased high-density lipoprotein cholesterol by 1.75 mg/dl (p = .002). Intervention costs, which included financial incentives for participation and weight loss, were $915 per participant. The incremental cost-effectiveness ratio was estimated to be $14,011 per QALY but was sensitive to the time horizon studied.

CONCLUSIONS

Widespread adoption of community-based DPP has the potential to reduce diabetes and cardiovascular-related morbidity and mortality for low-income persons at high risk for diabetes and may be a cost-effective investment for Medicaid programs.

Development and Validation of a Novel Pediatric Appendicitis Risk Calculator (pARC)

OBJECTIVES

We sought to develop and validate a clinical calculator that can be used to quantify risk for appendicitis on a continuous scale for patients with acute abdominal pain.

METHODS

The pediatric appendicitis risk calculator (pARC) was developed and validated through secondary analyses of 3 distinct cohorts. The derivation sample included visits to 9 pediatric emergency departments between March 2009 and April 2010. The validation sample included visits to a single pediatric emergency department from 2003 to 2004 and 2013 to 2015. Variables evaluated were as follows: age, sex, temperature, nausea and/or vomiting, pain duration, pain location, pain with walking, pain migration, guarding, white blood cell count, and absolute neutrophil count. We used stepwise regression to develop and select the best model. Test performance of the pARC was compared with the Pediatric Appendicitis Score (PAS).

RESULTS

The derivation sample included 2423 children, 40% of whom had appendicitis. The validation sample included 1426 children, 35% of whom had appendicitis. The final pARC model included the following variables: sex, age, duration of pain, guarding, pain migration, maximal tenderness in the right-lower quadrant, and absolute neutrophil count. In the validation sample, the pARC exhibited near perfect calibration and a high degree of discrimination (area under the curve: 0.85; 95% confidence interval: 0.83 to 0.87) and outperformed the PAS (area under the curve: 0.77; 95% confidence interval: 0.75 to 0.80). By using the pARC, almost half of patients in the validation cohort could be accurately classified as at <15% risk or ≥85% risk for appendicitis, whereas only 23% would be identified as having a comparable PAS of <3 or >8.

CONCLUSIONS

In our validation cohort of patients with acute abdominal pain, the pARC accurately quantified risk for appendicitis.

Infant Hospitalizations and Mortality After Maternal Vaccination

BACKGROUND

The Advisory Committee on Immunization Practices currently recommends pregnant women receive influenza and tetanus toxoid, reduced diphtheria toxoid, and acellular pertussis (Tdap) vaccines. There are limited studies of the long-term safety in infants for vaccines administered during pregnancy. We evaluate whether maternal receipt of influenza and Tdap vaccines increases the risk of infant hospitalization or death in the first 6 months of life.

METHODS

We included singleton, live birth pregnancies in the Vaccine Safety Datalink between 2004 and 2014. Outcomes were infant hospitalizations and mortality in the first 6 months of life. We performed a case-control study matching case patients and controls 1:1 and used conditional logistic regression to estimate odds ratios for maternal exposure to influenza and/or Tdap vaccines in pregnancy.

RESULTS

There were 413 034 live births in our population. Of these, 25 222 infants had hospitalizations and 157 infants died in the first 6 months of life. We found no association between infant hospitalization and maternal influenza (adjusted odds ratio: 1.00; 95% confidence interval [CI]: 0.96-1.04) or Tdap (adjusted odds ratio: 0.94; 95% CI: 0.88-1.01) vaccinations. We found no association between infant mortality and maternal influenza (adjusted odds ratio: 0.96; 95% CI: 0.54-1.69) or Tdap (adjusted odds ratio: 0.44; 95% CI: 0.17-1.13) vaccinations.

CONCLUSIONS

We found no association between vaccination during pregnancy and risk of infant hospitalization or death in the first 6 months of life. These findings support the safety of current recommendations for influenza and Tdap vaccination during pregnancy.

Predicting neutropenia risk in patients with cancer using electronic data

Objectives

Clinical guidelines recommending the use of myeloid growth factors are largely based on the prescribed chemotherapy regimen. The guidelines suggest that oncologists consider patient-specific characteristics when prescribing granulocyte-colony stimulating factor (G-CSF) prophylaxis; however, a mechanism to quantify individual patient risk is lacking. Readily available electronic health record (EHR) data can provide patient-specific information needed for individualized neutropenia risk estimation. An evidence-based, individualized neutropenia risk estimation algorithm has been developed. This study evaluated the automated extraction of EHR chemotherapy treatment data and externally validated the neutropenia risk prediction model.

Materials and Methods

A retrospective cohort of adult patients with newly diagnosed breast, colorectal, lung, lymphoid, or ovarian cancer who received the first cycle of a cytotoxic chemotherapy regimen from 2008 to 2013 were recruited from a single cancer clinic. Electronically extracted EHR chemotherapy treatment data were validated by chart review. Neutropenia risk stratification was conducted and risk model performance was assessed using calibration and discrimination.

Results

Chemotherapy treatment data electronically extracted from the EHR were verified by chart review. The neutropenia risk prediction tool classified 126 patients (57%) as being low risk for febrile neutropenia, 44 (20%) as intermediate risk, and 51 (23%) as high risk. The model was well calibrated (Hosmer-Lemeshow goodness-of-fit test = 0.24). Discrimination was adequate and slightly less than in the original internal validation (c-statistic 0.75 vs 0.81).

Conclusion

Chemotherapy treatment data were electronically extracted from the EHR successfully. The individualized neutropenia risk prediction model performed well in our retrospective external cohort.

Reducing diabetes distress and improving self-management with mindfulness

Stress associated with diabetes makes managing diabetes harder. We investigated whether mindfulness-based stress reduction (MBSR) could reduce diabetes distress and improve management. We recruited 38 participants to complete an MBSR program. Surveys and lab values were completed at baseline and post-intervention. Participants showed significant improvement in diabetes-related distress (Cohen's d -.71, p < .002), psychosocial self-efficacy (Cohen's d .80, p < .001), and glucose control (Cohen's d -.79, p < .001). Significant improvements in depression, anxiety, stress, coping, self-compassion, and social support were also found. These results suggest that MBSR may offer an effective method for helping people better self-manage their diabetes and improve mental health.

First Trimester Influenza Vaccination and Risks for Major Structural Birth Defects in Offspring

OBJECTIVE

To examine risks for major structural birth defects in infants after first trimester inactivated influenza vaccine (IIV) exposures.

STUDY DESIGN

In this observational study, we used electronic health data from 7 Vaccine Safety Datalink sites to examine risks for selected major structural defects in infants after maternal IIV exposure. Vaccine exposures for women with continuous insurance enrollment through pregnancy who delivered singleton live births between 2004 and 2013 were identified from standardized files. Infants with continuous insurance enrollment were followed to 1 year of age. We excluded mother-infant pairs with other exposures that potentially increased their background risk for birth defects. Selected cardiac, orofacial or respiratory, neurologic, ophthalmologic or otologic, gastrointestinal, genitourinary and muscular or limb defects were identified from diagnostic codes in infant medical records using validated algorithms. Propensity score adjusted generalized estimating equations were used to estimate prevalence ratios (PRs).

RESULTS

We identified 52 856 infants with maternal first trimester IIV exposure and 373 088 infants whose mothers were unexposed to IIV during first trimester. Prevalence (per 100 live births) for selected major structural birth defects was 1.6 among first trimester IIV exposed versus 1.5 among unexposed mothers. The adjusted PR was 1.02 (95% CI 0.94-1.10). Organ system-specific PRs were similar to the overall PR.

CONCLUSION

First trimester maternal IIV exposure was not associated with an increased risk for selected major structural birth defects in this large cohort of singleton live births.

Maternal Tdap vaccination and risk of infant morbidity

INTRODUCTION

An increased risk of diagnosed chorioamnionitis in women vaccinated with Tdap during pregnancy was previously detected at two Vaccine Safety Datalink (VSD) sites. The clinical significance of this finding related to infant outcomes remains uncertain.

METHODS

Retrospective cohort study of singleton live births born to women who were continuously insured from 6months prior to their last menstrual period through 6weeks postpartum, with ≥1 outpatient visit during pregnancy from January 1, 2010 to November 15, 2013 at seven integrated United States health care systems part of the VSD. We re-evaluated the association between maternal Tdap and chorioamnionitis and evaluated whether specific infant morbidities differ among infants born to mothers who did and did not receive Tdap during pregnancy. We focused on 2 Tdap exposure windows: the recommended 27-36weeks gestation or anytime during pregnancy. We identified inpatient diagnostic codes for transient tachypnea of the newborn (TTN), neonatal sepsis, neonatal pneumonia, respiratory distress syndrome (RDS), and newborn convulsions associated with an infant's first hospitalization. A generalized linear model with Poisson distribution and log-link was used to estimate propensity score adjusted rate ratios (ARR) with 95% confidence intervals (CI).

RESULTS

The analyses included 197,564 pregnancies. Chorioamnionitis was recorded in 6.4% of women who received Tdap vaccination any time during pregnancy and 5.2% of women who did not (ARR [95% CI]: 1.23 [1.17, 1.28]). Compared with unvaccinated women, there were no significant increased risks (ARR [95% CI]) for TTN (1.04 [0.98, 1.11]), neonatal sepsis (1.06 [0.91, 1.23]), neonatal pneumonia (0.94 [0.72, 1.22]), RDS (0.91 [0.66, 1.26]), or newborn convulsions (1.16 [0.87, 1.53]) in infants born to Tdap-vaccinated women.

CONCLUSIONS AND RELEVANCE

Despite an observed association between maternal Tdap vaccination and maternal chorioamnionitis, we did not find increased risk for clinically significant infant outcomes associated with maternal chorioamnionitis.

Use and Customization of Risk Scores for Predicting Cardiovascular Events Using Electronic Health Record Data

Background

Clinicians who are using the Framingham Risk Score (FRS) or the American College of Cardiology/American Heart Association Pooled Cohort Equations (PCE) to estimate risk for their patients based on electronic health data (EHD) face 4 questions. (1) Do published risk scores applied to EHD yield accurate estimates of cardiovascular risk? (2) Are FRS risk estimates, which are based on data that are up to 45 years old, valid for a contemporary patient population seeking routine care? (3) Do the PCE make the FRS obsolete? (4) Does refitting the risk score using EHD improve the accuracy of risk estimates?

Methods and Results

Data were extracted from the EHD of 84 116 adults aged 40 to 79 years who received care at a large healthcare delivery and insurance organization between 2001 and 2011. We assessed calibration and discrimination for 4 risk scores: published versions of FRS and PCE and versions obtained by refitting models using a subset of the available EHD. The published FRS was well calibrated (calibration statistic K=9.1, miscalibration ranging from 0% to 17% across risk groups), but the PCE displayed modest evidence of miscalibration (calibration statistic K=43.7, miscalibration from 9% to 31%). Discrimination was similar in both models (C‐index=0.740 for FRS, 0.747 for PCE). Refitting the published models using EHD did not substantially improve calibration or discrimination.

Conclusions

We conclude that published cardiovascular risk models can be successfully applied to EHD to estimate cardiovascular risk; the FRS remains valid and is not obsolete; and model refitting does not meaningfully improve the accuracy of risk estimates.

Identifying birth defects in automated data sources in the Vaccine Safety Datalink

PURPOSE

The Vaccine Safety Datalink (VSD), a collaboration between the Centers for Disease Control and Prevention and several large healthcare organizations, aims to monitor safety of vaccines administered in the USA. We present definitions and prevalence estimates for major structural birth defects to be used in studies of maternal vaccine safety.

METHODS

In this observational study, we created and refined algorithms for identifying major structural birth defects from electronic healthcare data, conducted formal chart reviews for severe cardiac defects, and conducted limited chart validation for other defects. We estimated prevalence for selected defects by VSD site and birth year and compared these estimates to those in a US and European surveillance system.

RESULTS

We developed algorithms to enumerate >50 major structural birth defects from standardized administrative and healthcare data based on utilization patterns and expert opinion, applying criteria for number, timing, and setting of diagnoses. Our birth cohort included 497 894 infants across seven sites. The period prevalence for all selected major birth defects in the VSD from 2004 to 2013 was 1.7 per 100 live births. Cardiac defects were most common (65.4 per 10 000 live births), with one-fourth classified as severe, requiring emergent intervention. For most major structural birth defects, prevalence estimates were stable over time and across sites and similar to those reported in other population-based surveillance systems.

CONCLUSIONS

Our algorithms can efficiently identify many major structural birth defects in large healthcare datasets and can be used in studies evaluating the safety of vaccines administered to pregnant women. Copyright © 2017 John Wiley & Sons, Ltd.

Predictors of Poor Response to Depression Treatment in Primary Care

OBJECTIVE

Depression is pervasive and costly, and the majority of depression is treated in primary care. The objective of this study was to identify patient characteristics predictive of poor depression outcomes in primary care clinics.

METHODS

This observational study followed 792 patients receiving usual care for depression in 83 clinics across Minnesota for at least six months between 2008 and 2010. The primary outcome was an ordinal outcome of remission or response without remission ("response") six months after the start of treatment. The outcome was assessed via telephone administration of the Patient Health Questionnaire-9. Associations of patient characteristics with the primary outcome were assessed by using ordinal logistic regression.

RESULTS

The majority of patients were female, Caucasian, and employed, and most had some college education and good, very good, or excellent self-rated health. At baseline, 32% had mild depression, 40% moderate depression, 20% moderately severe depression, and 8% severe depression. One-third of patients had psychotherapy or psychiatric care in addition to antidepressant medications. At six months, only 47% of patients obtained depression remission or response. Patients were significantly less likely to experience remission or response if they rated their health as poor or fair or if they were unemployed and were more likely to achieve remission or response if they were younger or had mild depression.

CONCLUSIONS

Patients with poor or fair health or who were unemployed were less likely to respond to usual depression care and may be good candidates for limited, but potentially more effective, intensive treatment resources for depression.

The effects of patient-centered depression care on patient satisfaction and depression remission

BACKGROUND

While health systems are striving for patient-centered care, they have little evidence to guide them on how to engage patients in their care, or how this may affect patient experiences and outcomes.

OBJECTIVE

To explore which specific patient-centered aspects of care were best associated with depression improvement and care satisfaction.

METHODS

Design: observational.

SETTING

83 primary care clinics across Minnesota.

SUBJECTS

Primary care patients with new prescriptions for antidepressants for depression were recruited from 2007 to 2009.

OUTCOME MEASURES

Patients completed phone surveys regarding demographics and self-rated health status and depression severity at baseline and 6 months. Patient centeredness was assessed via a modified version of the Patient Assessment of Chronic Illness Care. Differences in rates of remission and satisfaction between positive and negative responses for each care process were evaluated using chi-square tests.

RESULTS

At 6 months, 37% of 792 patients ages 18-88 achieved depression remission, and 79% rated their care as good-to-excellent. Soliciting patient preferences for care and questions or concerns, providing treatment plans, utilizing depression scales and asking about suicide risk were patient-centered measures that were positively associated with depression remission in the unadjusted model; these associations were mildly weakened after adjustment for depression severity and health status. Nearly all measures of patient centeredness were positively associated with care ratings.

CONCLUSION

The patient centeredness of care influences how patients experience and rate their care. This study identified specific actions providers can take to improve patient satisfaction and depression outcomes.

Who Must We Target Now to Minimize Future Cardiovascular Events and Total Mortality?: Lessons From the Surveillance, Prevention and Management of Diabetes Mellitus (SUPREME-DM) Cohort Study

BACKGROUND

Examining trends in cardiovascular events and mortality in US health systems can guide the design of targeted clinical and public health strategies to reduce cardiovascular events and mortality rates.

METHODS AND RESULTS

We conducted an observational cohort study from 2005 to 2011 among 1.25 million diabetic subjects and 1.25 million nondiabetic subjects from 11 health systems that participate in the Surveillance, Prevention and Management of Diabetes Mellitus (SUPREME-DM) DataLink. Annual rates (per 1000 person-years) of myocardial infarction/acute coronary syndrome (International Classification of Diseases-Ninth Revision, 410.0–410.91, 411.1–411.8), stroke (International Classification of Diseases-Ninth Revision, 430–432.9, 433–434.9), heart failure (International Classification of Diseases-Ninth Revision, 428–428.9), and all-cause mortality were monitored by diabetes mellitus (DM) status, age, sex, race/ethnicity, and a prior cardiovascular history. We observed significant declines in cardiovascular events and mortality rates in subjects with and without DM. However, there was substantial variation by age, sex, race/ethnicity, and prior cardiovascular history. Mortality declined from 44.7 to 27.1 (P<0.0001) for those with DM and cardiovascular disease (CVD), from 11.2 to 10.9 (P=0.03) for those with DM only, and from 18.9 to 13.0 (P<0.0001) for those with CVD only. Yet, in the [almost equal to]85% of subjects with neither DM nor CVD, overall mortality (7.0 to 6.8; P=0.10) and stroke rates (1.6–1.6; P=0.77) did not decline and heart failure rates increased (0.9–1.15; P=0.0005).

CONCLUSIONS

To sustain improvements in myocardial infarction, stroke, heart failure, and mortality, health systems that have successfully focused on care improvement in high-risk adults with DM or CVD must broaden their improvement strategies to target lower risk adults who have not yet developed DM or CVD.

Risk of Preterm or Small-for-Gestational-Age Birth After Influenza Vaccination During Pregnancy: Caveats When Conducting Retrospective Observational Studies

Vaccines are increasingly targeted toward women of reproductive age, and vaccines to prevent influenza and pertussis are recommended during pregnancy. Prelicensure clinical trials typically have not included pregnant women, and when they are included, trials cannot detect rare events. Thus, postmarketing vaccine safety assessments are necessary. However, analysis of observational data requires detailed assessment of potential biases. Using data from 8 Vaccine Safety Datalink sites in the United States, we analyzed the association of monovalent H1N1 influenza vaccine (MIV) during pregnancy with preterm birth (<37 weeks) and small-for-gestational-age birth (birth weight < 10th percentile). The cohort included 46,549 pregnancies during 2009-2010 (40% of participants received the MIV). We found potential biases in the vaccine-birth outcome association that might occur due to variable access to vaccines, the time-dependent nature of exposure to vaccination within pregnancy (immortal time bias), and confounding from baseline differences between vaccinated and unvaccinated women. We found a strong protective effect of vaccination on preterm birth (relative risk = 0.79, 95% confidence interval: 0.74, 0.85) when we ignored potential biases and no effect when accounted for them (relative risk = 0.91; 95% confidence interval: 0.83, 1.0). In contrast, we found no important biases in the association of MIV with small-for-gestational-age birth. Investigators conducting studies to evaluate birth outcomes after maternal vaccination should use statistical approaches to minimize potential biases.

Implementation of Electronic Clinical Decision Support for Pediatric Appendicitis

BACKGROUND AND OBJECTIVE

Computed tomography (CT) and ultrasound (US) are commonly used in patients with acute abdominal pain. We sought to standardize care and reduce CT use while maintaining patient safety through implementation of a multicomponent electronic clinical decision support tool for pediatric patients with possible appendicitis.

METHODS

We conducted a quasi-experimental study of children 3 to 18 years old who presented with possible appendicitis to the pediatric emergency department (ED) between January 2011 and December 2013. Outcomes were use of CT and US. Balancing measures included missed appendicitis, ED revisits within 30 days, appendiceal perforation, and ED length of stay.

RESULTS

Of 2803 patients with acute abdominal pain over the 3-year study period, 794 (28%) had appendicitis and 207 (26.1% of those with appendicitis) had a perforation. CT use during the 10-month preimplementation period was 38.8% and declined to 17.7% by the end of the study (54% relative decrease). For CT, segmented regression analysis revealed that there was a significant change in trend from the preimplementation period to implementation (monthly decrease -3.5%; 95% confidence interval: -5.9% to -0.9%; P = .007). US use was 45.7% preimplementation and 59.7% during implementation. However, there was no significant change in US or total imaging trends. There were also no statistically significant differences in rates of missed appendicitis, ED revisits within 30 days, appendiceal perforation, or ED length of stay between time periods.

CONCLUSIONS

Our electronic clinical decision support tool was associated with a decrease in CT use while maintaining safety and high quality care for patients with possible appendicitis.

Diagnostic Delay and Antibiotic Overuse in Acute Pulmonary Blastomycosis

The diagnosis of blastomycosis is often delayed. We identified 28 cases of pulmonary blastomycosis in a retrospective chart review. Most patients received multiple antibiotic courses before being diagnosed, and the sputum KOH smear was rarely used. Diagnostic delay can be decreased with higher suspicion for pulmonary blastomycosis and early use of the sputum KOH smear.

Adapting machine learning techniques to censored time-to-event health record data: A general-purpose approach using inverse probability of censoring weighting

Models for predicting the probability of experiencing various health outcomes or adverse events over a certain time frame (e.g., having a heart attack in the next 5years) based on individual patient characteristics are important tools for managing patient care. Electronic health data (EHD) are appealing sources of training data because they provide access to large amounts of rich individual-level data from present-day patient populations. However, because EHD are derived by extracting information from administrative and clinical databases, some fraction of subjects will not be under observation for the entire time frame over which one wants to make predictions; this loss to follow-up is often due to disenrollment from the health system. For subjects without complete follow-up, whether or not they experienced the adverse event is unknown, and in statistical terms the event time is said to be right-censored. Most machine learning approaches to the problem have been relatively ad hoc; for example, common approaches for handling observations in which the event status is unknown include (1) discarding those observations, (2) treating them as non-events, (3) splitting those observations into two observations: one where the event occurs and one where the event does not. In this paper, we present a general-purpose approach to account for right-censored outcomes using inverse probability of censoring weighting (IPCW). We illustrate how IPCW can easily be incorporated into a number of existing machine learning algorithms used to mine big health care data including Bayesian networks, k-nearest neighbors, decision trees, and generalized additive models. We then show that our approach leads to better calibrated predictions than the three ad hoc approaches when applied to predicting the 5-year risk of experiencing a cardiovascular adverse event, using EHD from a large U.S. Midwestern healthcare system.

Maternal Tdap vaccination: Coverage and acute safety outcomes in the vaccine safety datalink, 2007-2013

INTRODUCTION

Since October 2012, the combined tetanus toxoid, reduced diphtheria toxoid, acellular pertussis vaccine (Tdap) has been recommended in the United States during every pregnancy.

METHODS

In this observational study from the Vaccine Safety Datalink, we describe receipt of Tdap during pregnancy among insured women with live births across seven health systems. Using a retrospective matched cohort, we evaluated risks for selected medically attended adverse events in pregnant women, occurring within 42 days of vaccination. Using a generalized estimating equation, we calculated adjusted incident rate ratios (AIRR).

RESULTS

Our vaccine coverage cohort included 438,487 live births between January 1, 2007 and November 15, 2013. Across the coverage cohort, 14% received Tdap during pregnancy. By 2013, Tdap was administered during pregnancy in 41.7% of live births, primarily in the 3rd trimester. Our vaccine safety cohort included 53,885 vaccinated and 109,253 matched unvaccinated pregnant women. There was no increased risk for a composite outcome of medically attended acute adverse events within 3 days of vaccination. Similarly, across the safety cohort, over a 42 day window, incident neurologic events, thrombotic events, and new onset proteinuria did not differ by maternal receipt of Tdap. Among women receiving Tdap at 20 weeks gestation or later, as compared to their matched controls, there was no increased risk for gestational diabetes or cardiac events while venous thromboembolic events and thrombocytopenia were diagnosed within 42 days of vaccination at slightly decreased rates.

CONCLUSION

Tdap coverage during pregnancy increased from 2007 through 2013, but was still below 50%. No acute maternal safety signals were detected in this large cohort.

Prevalence of chronic kidney disease among individuals with diabetes in the SUPREME-DM Project, 2005-2011

AIMS

Diabetes is a leading cause of chronic kidney disease (CKD). Different methods of CKD ascertainment may impact prevalence estimates. We used data from 11 integrated health systems in the United States to estimate CKD prevalence in adults with diabetes (2005-2011), and compare the effect of different ascertainment methods on prevalence estimates.

METHODS

We used the SUPREME-DM DataLink (n = 879,312) to estimate annual CKD prevalence. Methods of CKD ascertainment included: diagnosis codes alone, impaired estimated glomerular filtration rate (eGFR) alone (eGFR < 60 mL/min/1.73 m(2)), albuminuria alone (spot urine albumin creatinine ratio > 30 mg/g or equivalent), and combinations of these approaches.

RESULTS

CKD prevalence was 20.0% using diagnosis codes, 17.7% using impaired eGFR, 11.9% using albuminuria, and 32.7% when one or more method suggested CKD. The criteria had poor concordance. After age- and sex-standardization to the 2010 U.S. Census population, prevalence using diagnosis codes increased from 10.7% in 2005 to 14.3% in 2011 (P < 0.001). The prevalence using eGFR decreased from 9.7% in 2005 to 8.6% in 2011 (P < 0.001).

CONCLUSIONS

Our data indicate that CKD prevalence and prevalence trends differ according to the CKD ascertainment method, highlighting the necessity for multiple sources of data to accurately estimate and track CKD prevalence.

A Naive Bayes machine learning approach to risk prediction using censored, time-to-event data

Predicting an individual's risk of experiencing a future clinical outcome is a statistical task with important consequences for both practicing clinicians and public health experts. Modern observational databases such as electronic health records provide an alternative to the longitudinal cohort studies traditionally used to construct risk models, bringing with them both opportunities and challenges. Large sample sizes and detailed covariate histories enable the use of sophisticated machine learning techniques to uncover complex associations and interactions, but observational databases are often 'messy', with high levels of missing data and incomplete patient follow-up. In this paper, we propose an adaptation of the well-known Naive Bayes machine learning approach to time-to-event outcomes subject to censoring. We compare the predictive performance of our method with the Cox proportional hazards model which is commonly used for risk prediction in healthcare populations, and illustrate its application to prediction of cardiovascular risk using an electronic health record dataset from a large Midwest integrated healthcare system.

Preventable major cardiovascular events associated with uncontrolled glucose, blood pressure, and lipids and active smoking in adults with diabetes with and without cardiovascular disease: a contemporary analysis

OBJECTIVE

The objective of this study was to assess the incidence of major cardiovascular (CV) hospitalization events and all-cause deaths among adults with diabetes with or without CV disease (CVD) associated with inadequately controlled glycated hemoglobin (A1C), high LDL cholesterol (LDL-C), high blood pressure (BP), and current smoking.

RESEARCH DESIGN AND METHODS

Study subjects included 859,617 adults with diabetes enrolled for more than 6 months during 2005-2011 in a network of 11 U.S. integrated health care organizations. Inadequate risk factor control was classified as LDL-C ≥100 mg/dL, A1C ≥7% (53 mmol/mol), BP ≥140/90 mm Hg, or smoking. Major CV events were based on primary hospital discharge diagnoses for myocardial infarction (MI) and acute coronary syndrome (ACS), stroke, or heart failure (HF). Five-year incidence rates, rate ratios, and average attributable fractions were estimated using multivariable Poisson regression models.

RESULTS

Mean (SD) age at baseline was 59 (14) years; 48% of subjects were female, 45% were white, and 31% had CVD. Mean follow-up was 59 months. Event rates per 100 person-years for adults with diabetes and CVD versus those without CVD were 6.0 vs. 1.7 for MI/ACS, 5.3 vs. 1.5 for stroke, 8.4 vs. 1.2 for HF, 18.1 vs. 40 for all CV events, and 23.5 vs. 5.0 for all-cause mortality. The percentages of CV events and deaths associated with inadequate risk factor control were 11% and 3%, respectively, for those with CVD and 34% and 7%, respectively, for those without CVD.

CONCLUSIONS

Additional attention to traditional CV risk factors could yield further substantive reductions in CV events and mortality in adults with diabetes.

Evaluation of the association of maternal pertussis vaccination with obstetric events and birth outcomes

IMPORTANCE

In 2010, due to a pertussis outbreak and neonatal deaths, the California Department of Health recommended that the tetanus toxoid, reduced diphtheria toxoid, and acellular pertussis vaccine (Tdap) be administered during pregnancy. Tdap is now recommended by the Advisory Committee on Immunization Practices for all pregnant women, preferably between 27 and 36 weeks' gestation. Limited data exist on Tdap safety during pregnancy.

OBJECTIVE

To evaluate whether maternal Tdap vaccination during pregnancy is associated with increased risks of adverse obstetric events or adverse birth outcomes.

DESIGN AND SETTING

Retrospective, observational cohort study using administrative health care databases from 2 California Vaccine Safety Datalink sites.

PARTICIPANTS AND EXPOSURES

Of 123,494 women with singleton pregnancies ending in a live birth between January 1, 2010, and November 15, 2012, 26,229 (21%) received Tdap during pregnancy and 97,265 did not.

MAIN OUTCOMES AND MEASURES

Risks of small-for-gestational-age (SGA) births (<10th percentile), chorioamnionitis, preterm birth (<37 weeks' gestation), and hypertensive disorders of pregnancy were evaluated. Relative risk (RR) estimates were adjusted for site, receipt of another vaccine during pregnancy, and propensity to receive Tdap during pregnancy. Cox regression was used for preterm delivery, and Poisson regression for other outcomes.

RESULTS

Vaccination was not associated with increased risks of adverse birth outcomes: crude estimates for preterm delivery were 6.3% of vaccinated and 7.8% of unvaccinated women (adjusted RR, 1.03; 95% CI, 0.97-1.09); 8.4% of vaccinated and 8.3% of unvaccinated had an SGA birth (adjusted RR, 1.00; 95% CI, 0.96-1.06). Receipt of Tdap before 20 weeks was not associated with hypertensive disorder of pregnancy (adjusted RR, 1.09; 95% CI, 0.99-1.20); chorioamnionitis was diagnosed in 6.1% of vaccinated and 5.5% of unvaccinated women (adjusted RR, 1.19; 95% CI, 1.13-1.26).

CONCLUSIONS AND RELEVANCE

In this cohort of women with singleton pregnancies that ended in live birth, receipt of Tdap during pregnancy was not associated with increased risk of hypertensive disorders of pregnancy or preterm or SGA birth, although a small but statistically significant increased risk of chorioamnionitis diagnosis was observed.

Monovalent H1N1 influenza vaccine safety in pregnant women, risks for acute adverse events

OBJECTIVE

To assess risks for acute adverse events and pregnancy complications in pregnant women following monovalent 2009 H1N1 inactivated influenza (MIV) vaccination.

METHODS

Within the Vaccine Safety Datalink, we compared rates of pre-specified medically attended events (MAE) occurring within 42 days of MIV vaccination to those occurring in matched cohorts that at the same gestational age were either unvaccinated or received seasonal trivalent inactivated influenza (TIV) vaccine. Using generalized estimating equation method, with a Poisson distribution and log link, we calculated adjusted incident rate ratios (AIRR).

RESULTS

Among 9349 women receiving MIV in any trimester, only one MAE occurred 0-3 days following MIV, an allergic reaction. No cases of Guillain-Barré syndrome, Bell's palsy, or transverse myelitis occurred 1-42 days after MIV. Compared to women receiving TIV and to unvaccinated women, risks for acute MAEs were not increased following MIV for any outcome. Hyperemesis was the most common adverse event in the MIV, TIV, and unvaccinated groups, occurring at a rate of about 4% over a 42-day period in all groups. Over a 42-day window, among all groups, incident gestational diabetes occurred at a rate of 3% and thrombocytopenia occurred at a rate of approximately 0.3%. Among women receiving MIV during pregnancy, increased risks for these and other less common obstetric events were not detected.

CONCLUSION

In this large cohort of pregnant women no acute safety signals were identified within 6 weeks of receipt of MIV.

Initial persistence with antihypertensive therapies is associated with depression treatment persistence, but not depression

The purpose of this study was to examine the relationship between the presence of clinical depression and persistence to drug therapy treatment for depression with early nonpersistence to antihypertensive therapies in a large, diverse cohort of newly treated hypertension patients. Using a hypertension registry at Kaiser Permanente Northern California, the authors conducted a retrospective cohort study of 44,167 adults (18 years and older) with hypertension who were new users of antihypertensive therapy in 2008. We used multivariate logistic regression analysis to model the relationships between the presence of clinical depression and early nonpersistence (defined as failing to refill the first prescription within 90 days after the end of the first fill days' supply) to antihypertensive therapies, controlling for sociodemographic and clinical risk factors. Within the group of 1484 patients who had evidence of clinical depression in the 12 months prior to the initiation of antihypertensive therapy, the authors examined the relationship between drug therapy treatment for depression and 6-month persistence with antidepressant therapy with early nonpersistence with antihypertensive therapies. No association was found between the presence of clinical depression and early nonpersistence to antihypertensive therapies after adjustment for individual demographic and clinical characteristics and neighborhood-level socioeconomic status. However, among the subset of 1484 patients with documented evidence of clinical depression in the 12 months prior to the initiation of antihypertensive therapy, being prescribed and persistence with antidepressant therapy was strongly associated with lower odds of early nonpersistence to antihypertensive medications (odds ratio, 0.64; confidence interval, 0.42-0.96). In an integrated delivery system, the authors found that treatment for depression was associated with higher levels of antihypertensive persistence. Improving quality of depression care in patients with comorbid hypertension may be an important strategy in decreasing cardiovascular disease risk in these patients.

Safety of the yellow Fever vaccine: a retrospective study

BACKGROUND

Yellow fever (YF) vaccine is considered safe; however, severe illness and death following vaccination have been reported.

METHODS

Vaccine Safety Datalink (VSD) and US Department of Defense (DoD) data were used to identify adverse reactions following YF vaccination. Within the VSD, YF-vaccine-exposed subjects were compared to age-, site-, and gender-matched unexposed subjects. YF-vaccine-exposed DoD subjects were studied using a risk-interval design. For both cohorts, ICD-9 codes were analyzed for allergic and local reactions, mild systemic reactions, and possible visceral and neurologic adverse events (AEs).

RESULTS

The VSD cohort received 47,159 doses from 1991 through 2006. The DoD cohort received 1.12 million doses from 1999 through 2007. Most subjects received other vaccines simultaneously. In the VSD cohort, rates of allergic, local, and mild systemic reactions were not statistically different between YF-vaccine-exposed and -unexposed subjects. In the DoD, there was an increased risk for outpatient allergic events in the period following vaccination with YF and other vaccines rate ratios [RR 3.85, 95% confidence interval (CI) 3.35-4.41] but with no increased risk for inpatient allergic reactions. In both cohorts, inpatient ICD-9 codes for visceral events were significantly less common following vaccination; inpatient codes for neurologic events were less common in the VSD YF-vaccine-exposed adult cohort, but did not differ between exposed and unexposed periods in the DoD. In the DoD, one fatal case of YF-vaccine-associated viscerotropic disease (YF-vaccine-AVD) was detected. The estimated death rate was 0.89 for 1,000,000 YF vaccine doses (95% CI 0.12-6.31/1,000,000 doses). No YF vaccine-associated deaths occurred in the VSD.

CONCLUSIONS

In these closed cohorts we did not detect increased risk for visceral or neurologic events following YF vaccination. The death rate following YF vaccine was consistent with previous reports. These data support current recommendations for use of YF vaccine in young healthy individuals. These data are inadequate to judge safety of YF vaccines in elderly patients.

Mortality rates and cause-of-death patterns in a vaccinated population

BACKGROUND

Determining the baseline mortality rate in a vaccinated population is necessary to be able to identify any unusual increases in deaths following vaccine administration. Background rates are particularly useful during mass immunization campaigns and in the evaluation of new vaccines.

PURPOSE

Provide background mortality rates and describe causes of death following vaccination in the Vaccine Safety Datalink (VSD).

METHODS

Analyses were conducted in 2012. Mortality rates were calculated at 0-1 day, 0-7 days, 0-30 days, and 0-60 days following vaccination for deaths occurring between January 1, 2005, and December 31, 2008. Analyses were stratified by age and gender. Causes of death were examined, and findings were compared to National Center for Health Statistics (NCHS) data.

RESULTS

Among 13,033,274 vaccinated people, 15,455 deaths occurred between 0 and 60 days following vaccination. The mortality rate within 60 days of a vaccination visit was 442.5 deaths per 100,000 person-years. Rates were highest in the group aged ≥85 years, and increased from the 0-1-day to the 0-60-day interval following vaccination. Eleven of the 15 leading causes of death in the VSD and NCHS overlap in both systems, and the top four causes of death were the same in both systems.

CONCLUSIONS

VSD mortality rates demonstrate a healthy vaccinee effect, with rates lowest in the days immediately following vaccination, most apparent in the older age groups. The VSD mortality rate is lower than that in the general U.S. population, and the causes of death are similar.

Are benefits from diabetes self-management education sustained?

OBJECTIVES

To evaluate whether outcomes from diabetes self-management education for patients with suboptimal control were sustained.

STUDY DESIGN

A randomized controlled trial of 623 adults with type 2 diabetes and glycated hemoglobin (A1C) > 7% assigned to receive conventional individual education (IE), group education (GE) using US Diabetes Conversation Maps, or usual care (UC) with no education.

METHODS

A1C tests, Problem Areas in Diabetes (PAID), Diabetes Self-Efficacy (DES), Recommended Food Score (RFS), physical activity, and medication use were quantified at baseline and 1 year of follow-up through electronic health records and quarterly mailed surveys. Short-term (mean 6.8 months) and long-term (12.8 months) outcomes were evaluated using linear mixed models. In addition, follow-up trajectories were plotted in a random effects generalized additive model with smooth splines.

RESULTS

Compared with UC, IE resulted in long-term improved DES and PAID scores (DES, +.11, P = .03 and PAID, -2.94, P = .04), but not significantly improved long-term RFS or physical activity change. The A1C trajectory declined more steeply in IE than GE and UC for the first 150 days post randomization. However, by 250 days, there was no treatment group A1C difference. The model fit likelihood ratio test for A1C intervention trends was significant for 3 distinct non-linear trajectories (P = .02).

CONCLUSIONS

Conventional IE (but not GE) resulted in significant and sustained improvements in self-efficacy and reduced diabetes distress compared with UC, but short-term improvements in A1C, nutrition, and physical activity were not sustained. Patients may need ongoing reinforcement to achieve lasting behavioral change and glucose control.

Benefits of early hypertension control on cardiovascular outcomes in patients with diabetes

OBJECTIVE

To assess the impact of early hypertension (HT) control on occurrence of subsequent major cardiovascular events in those with diabetes and recent-onset HT.

RESEARCH DESIGN AND METHODS

Study subjects were 15,665 adults with diabetes but no diagnosed coronary or cerebrovascular disease at baseline who met standard criteria for new-onset HT. Poisson regression models assessed whether adequate blood pressure control within 1 year of HT onset predicts subsequent occurrence of major cardiovascular events with and without adjustment for baseline Framingham Risk Score (FRS) and other covariates.

RESULTS

Mean age was 51.5 years, and mean blood pressure at HT onset was 136.8/80.8 mmHg. In the year after HT onset, mean blood pressure decreased to 131.4/78.0 mmHg and was <130/80 mmHg in 32.9% of subjects and <140/90 mmHg in 80.2%. Over a mean follow-up of 3.2 years, age-adjusted rates of major cardiovascular events in those with mean 1-year blood pressure measurements of <130/80, 130-139/80-89, and ≥140/90 mmHg were 5.10, 4.27, and 6.94 events/1,000 person-years, respectively (P = 0.004). In FRS-adjusted models, rates of major cardiovascular events were significantly higher in those with mean blood pressure ≥140/90 mmHg in the first year after HT onset (rate ratio 1.30 [95% CI 1.01-1.169]; P = 0.04).

CONCLUSIONS

Failure to adequately control BP within 1 year of HT onset significantly increased the likelihood of major cardiovascular events within 3 years. Prompt control of new-onset HT in patients with diabetes may provide important short-term clinical benefits.

Intergenerational photovoice projects: optimizing this mechanism for influencing health promotion policies and strengthening relationships

Intergenerational photovoice groups are promising for promoting health through the topic that is explored and through group dynamics that can foster healthy relationships and communication. To investigate the potential benefits of intergenerational photovoice projects, photovoice groups were conducted in urban Minnesota, United States, and in rural Morelos, Mexico, between 2009 and 2012 with Mexican-origin adults and their adolescent relatives. Seven photovoice groups of adult-adolescent dyads met for eight sessions and developed exhibits highlighting their views on health and migration and made policy recommendations, using messages conveyed through their words and photographs. Informal process evaluation and focus groups were used to elicit feedback about photovoice project participation. Photovoice project themes were descriptively analyzed. Focus group evaluation data were thematically summarized, and facilitator reflections were descriptively summarized to identify factors associated with intergenerational photovoice groups. Seventy-five participants were recruited. Photovoice themes represented effects of migration on health, family, and well-being. The following two evaluative themes were identified: (a) participant sentiments about the benefits of photovoice participation and (b) facilitator observations of intergenerational photovoice group benefits and challenges. Participants described opportunities to learn new things and barriers to healthy relationships that the project was eliminating by providing them with time to work together. Used in health promotion, photovoice is a valuable tool that contributes to understanding the complex underlying factors influencing behaviors and health.

Assessing the safety of influenza immunization during pregnancy: the Vaccine Safety Datalink

The influenza vaccine can reduce maternal and neonatal morbidity and mortality and thus is recommended for all pregnant women. However, concerns regarding safety of influenza vaccine remain a barrier to vaccination. We describe ongoing analyses of influenza vaccine safety during pregnancy within the Vaccine Safety Datalink that includes the evaluation of acute events, adverse pregnancy and birth outcomes, and congenital anomalies. In addition, we highlight unique challenges and strategies for the study of vaccine safety among pregnant women with the use of large linked databases.