A practical guide to propensity score analysis for applied clinical research

Observational studies are often the only viable options in many clinical settings, especially when it is unethical or infeasible to randomly assign participants to different treatment régimes. In such case propensity score (PS) analysis can be applied to accounting for possible selection bias and thereby addressing questions of causal inference. Many PS methods exist, yet few guidelines are available to aid applied researchers in their conduct and evaluation of a PS analysis. In this article we give an overview of available techniques for PS estimation and application, balance diagnostic, treatment effect estimation, and sensitivity assessment, as well as recent advances. We also offer a tutorial that can be used to emulate the steps of PS analysis. Our goal is to provide information that will bring PS analysis within the reach of applied clinical researchers and practitioners.

National weighting of data from the Behavioral Risk Factor Surveillance System (BRFSS)

Background

The Behavioral Risk Factor Surveillance System (BRFSS) is a network of health-related telephone surveys--conducted by all 50 states, the District of Columbia, and participating US territories—that receive technical assistance from CDC. Data users often aggregate BRFSS state samples for national estimates without accounting for state-level sampling, a practice that could introduce bias because the weighted distributions of the state samples do not always adhere to national demographic distributions.

Methods

This article examines six methods of reweighting, which are then compared with key health indicator estimates from the National Health Interview Survey (NHIS) based on 2013 data.

Results

Compared to the usual stacking approach, all of the six new methods reduce the variance of weights and design effect at the national level, and some also reduce the estimated bias. This article also provides a comparison of the methods based on the variances induced by unequal weighting as well as the bias reduction induced by raking at the national level, and recommends a preferred method.

Conclusions

The new method leads to weighted distributions that more accurately reproduce national demographic characteristics. While the empirical results for key estimates were limited to a few health indicators, they also suggest reduction in potential bias and mean squared error. To the extent that survey outcomes are associated with these demographic characteristics, matching the national distributions will reduce bias in estimates of these outcomes at the national level.

A comparison of spatial smoothing methods for small area estimation with sampling weights

Small area estimation (SAE) is an important endeavor in many fields and is used for resource allocation by both public health and government organizations. Often, complex surveys are carried out within areas, in which case it is common for the data to consist only of the response of interest and an associated sampling weight, reflecting the design. While it is appealing to use spatial smoothing models, and many approaches have been suggested for this endeavor, it is rare for spatial models to incorporate the weighting scheme, leaving the analysis potentially subject to bias. To examine the properties of various approaches to estimation we carry out a simulation study, looking at bias due to both non-response and non-random sampling. We also carry out SAE of smoking prevalence in Washington State, at the zip code level, using data from the 2006 Behavioral Risk Factor Surveillance System. The computation times for the methods we compare are short, and all approaches are implemented in R using currently available packages.

Design and Weighting Methods for a Nationally Representative Sample of HIV-infected Adults Receiving Medical Care in the United States-Medical Monitoring Project

Background:

Health surveys of the general US population are inadequate for monitoring human immunodeficiency virus (HIV) infection because the relatively low prevalence of the disease (<0.5%) leads to small subpopulation sample sizes.

Objective:

To collect a nationally and locally representative probability sample of HIV-infected adults receiving medical care to monitor clinical and behavioral outcomes, supplementing the data in the National HIV Surveillance System. This paper describes the sample design and weighting methods for the Medical Monitoring Project (MMP) and provides estimates of the size and characteristics of this population.

Methods:

To develop a method for obtaining valid, representative estimates of the in-care population, we implemented a cross-sectional, three-stage design that sampled 23 jurisdictions, then 691 facilities, then 9,344 HIV patients receiving medical care, using probability-proportional-to-size methods. The data weighting process followed standard methods, accounting for the probabilities of selection at each stage and adjusting for nonresponse and multiplicity. Nonresponse adjustments accounted for differing response at both facility and patient levels. Multiplicity adjustments accounted for visits to more than one HIV care facility.

Results:

MMP used a multistage stratified probability sampling design that was approximately self-weighting in each of the 23 project areas and nationally. The probability sample represents the estimated 421,186 HIV-infected adults receiving medical care during January through April 2009. Methods were efficient (i.e., induced small, unequal weighting effects and small standard errors for a range of weighted estimates).

Conclusion:

The information collected through MMP allows monitoring trends in clinical and behavioral outcomes and informs resource allocation for treatment and prevention activities.

Association between residential greenness and gut microbiota in Chinese adults

BACKGROUND

A growing body of studies have reported the health benefits of greenness. However, less is known about the potential beneficial effects of residential greenness on gut microbiota, which is essential to human health. In this study, we aim to examine the association between residential greenness and gut microbiota in a population-based cohort study.

METHODS

We included 1758 participants based on the China Multi-Ethnic Cohort (CMEC) study and collected their stool samples for 16S sequencing to derive gut microbiota data. Residential greenness was estimated using the satellite-based data on enhanced vegetation index (EVI) and the normalized differential vegetation index (NDVI) in circular buffers of 250 m, 500 m, and 1000 m. The relationships between residential greenness levels and the composition of gut microbiota, measured by standardized α-diversity and taxonomic composition, were assessed using linear regression and Spearman correlation weighted by generalized propensity scores.

RESULTS

Higher greenness levels were significantly positively associated with standardized α-diversity. Per interquartile range (IQR) increase of EVI and NDVI in the circular buffer of 250 m were associated with the increments of 0.995(95% confidence interval (CI): 0.212-1.778) and 0.653(95% CI: 0.160-1.146) in the standardized Shannon index. For the taxonomic composition of gut microbiota, higher greenness levels were significantly correlated with 29 types of microbial taxonomic composition. NDVI in the circular buffer of 250 m was associated with increased Firmicutes (r = 0.102, adjusted p value = 0.004), which was the dominant composition in the gut microbiota.

CONCLUSIONS

Increased amounts of residential greenness may support healthy gut microbiota by benignly altering their composition. These findings suggested that green spaces should be designed to support diverse gut microbiota and ultimately optimize health benefits.

Data supporting midpoint-weighting life cycle assessment and energy forms of cumulative exergy demand for horticultural crops

With an increasing demand of horticultural crops, it is critical to examine environmental damages and exergy impacts and evaluate their potential in producing sustainable products of agricultural systems. As such, environmental midpoints of five dominated horticultural products, namely, hazelnut, watermelon, tea, kiwifruit, and citrus, are scrutinized using life cycle assessment approach in Guilan province, Iran. Each crop is considered under a separate scenario and 10 tons of yield is determined as the functional unit. ReCiPe2016, as a new approach, is used for computation of 17 midpoints. Moreover, a weighting analysis is undertaken to find the share of each input in environmental damages with dimensionless notation. In the second part of this paper, cumulative exergy demand (CExD) is applied for evaluation of energy forms in each scenario. Data are presented under two sectors in the main article. The first part is midpoint results of each crop and the second part depicts energy forms of CExD with input rate in each category. Besides, the supplementary files contain raw material of each input, midpoint physical rate, share of each input to contribute midpoint, raw data of weighted damages and share of each input in total weighted damages.

Principal of environmental life cycle assessment for medical waste during COVID-19 outbreak to support sustainable development goals

Disposal of medical waste (MW) must be considered as a vital need to prevent the spread of pandemics during Coronavirus disease of the pandemic in 2019 (COVID-19) outbreak in the globe. In addition, many concerns have been raised due to the significant increase in the generation of MW in recent years. A structured evaluation is required as a framework for the quantifying of potential environmental impacts of the disposal of MW which ultimately leads to the realization of sustainable development goals (SDG). Life cycle assessment (LCA) is considered as a practical approach to examine environmental impacts of any potential processes during all stages of a product's life, including material mining, manufacturing, and delivery. As a result, LCA is known as a suitable method for evaluating environmental impacts for the disposal of MW. In this research, existing scenarios for MW with a unique approach to emergency scenarios for the management of COVID-19 medical waste (CMW) are investigated. In the next step, LCA and its stages are defined comprehensively with the CMW management approach. Moreover, ReCiPe2016 is the most up-to-date method for computing environmental damages in LCA. Then the application of this method for defined scenarios of CMW is examined, and interpretation of results is explained regarding some examples. In the last step, the process of selecting the best environmental-friendly scenario is illustrated by applying weighting analysis. Finally, it can be concluded that LCA can be considered as an effective method to evaluate the environmental burden of CMW management scenarios in present critical conditions of the world to support SDG.

Longitudinal drop-out and weighting against its bias

Background

The bias caused by drop-out is an important factor in large population-based epidemiological studies. Many studies account for it by weighting their longitudinal data, but to date there is no detailed final approach for how to conduct these weights.

Methods

In this study we describe the observed longitudinal bias and a three-step longitudinal weighting approach used for the longitudinal data in the MoMo baseline (N = 4528, 4–17 years) and wave 1 study with 2807 (62%) participants between 2003 and 2012.

Results

The most meaningful drop-out predictors were socioeconomic status of the household, socioeconomic characteristics of the mother and daily TV usage. Weighting reduced the bias between the longitudinal participants and the baseline sample, and also increased variance by 5% to 35% with a final weighting efficiency of 41.67%.

Conclusions

We conclude that a weighting procedure is important to reduce longitudinal bias in health-oriented epidemiological studies and suggest identifying the most influencing variables in the first step, then use logistic regression modeling to calculate the inverse of the probability of participation in the second step, and finally trim and standardize the weights in the third step.

Including Life Cycle Assessment for decision-making in controlling wastewater nutrient removal systems

This paper focuses on the use of Life Cycle Assessment (LCA) to evaluate the performance of seventeen control strategies in wastewater treatment plants (WWTPs). It tackles the importance of using site-specific factors for nutrient enrichment when decision-makers have to select best operating strategies. Therefore, the LCA evaluation is repeated for three different scenarios depending on the limitation of nitrogen (N), phosphorus (P), or both, when evaluating the nutrient enrichment impact in water bodies. The LCA results indicate that for treated effluent discharged into N-deficient aquatic systems (e.g. open coastal areas) the most eco-friendly strategies differ from the ones dealing with discharging into P-deficient (e.g. lakes and rivers) and N&P-deficient systems (e.g. coastal zones). More particularly, the results suggest that strategies that promote increased nutrient removal and/or energy savings present an environmental benefit for N&P and P-deficient systems. This is not the case when addressing N-deficient systems for which the use of chemicals (even for improving N removal efficiencies) is not always beneficial for the environment. A sensitivity analysis on using weighting of the impact categories is conducted to assess how value choices (policy decisions) may affect the management of WWTPs. For the scenarios with only N-limitation, the LCA-based ranking of the control strategies is sensitive to the choice of weighting factors, whereas this is not the case for N&P or P-deficient aquatic systems.

Calibrated propensity score method for survey nonresponse in cluster sampling

Weighting adjustment is commonly used in survey sampling to correct for unit nonresponse. In cluster sampling, the missingness indicators are often correlated within clusters and the response mechanism is subject to cluster-specific nonignorable missingness. Based on a parametric working model for the response mechanism that incorporates cluster-specific nonignorable missingness, we propose a method of weighting adjustment. We provide a consistent estimator of the mean or totals in cases where the study variable follows a generalized linear mixed-effects model. The proposed method is robust in the sense that the consistency of the estimator does not require correct specification of the functional forms of the response and outcome models. A consistent variance estimator based on Taylor linearization is also proposed. Numerical results, including a simulation and a real-data application, are presented.

A new weighted injury severity scoring system: better predictive power for adult trauma mortality

Background

An accurate injury severity measurement is essential in the evaluation of trauma care and in outcome research. The traditional Injury Severity Score (ISS) does not consider the differential risks of the Abbreviated Injury Scale (AIS) from different body regions, and the three AIS involved in the calculation of ISS are given equal weights. The objective of this study was to develop a weighted injury severity scoring (wISS) system for adult trauma patients with better predictive power than the traditional Injury Severity Score (ISS).

Methods

The 2007-2014 National Trauma Data Bank (NTDB) Research Datasets were used. We identified adult trauma patients from the NTDB and then randomly split it into a study sample and a test sample. Based on the association between mortality and the Abbreviated Injury Scale (AIS) from each of the six ISS body regions in the study sample, we evaluated 12 different sets of weights for the component AIS scores used in the calculation of ISS and selected one best set of weights. Discrimination (areas under the receiver operating characteristic curve, sensitivity, specificity, positive predictive value, negative predictive value, concordance) and calibration were compared between the wISS and ISS.

Results

The areas under the receiver operating characteristic curves from the wISS and ISS are all 0.83, and 0.76 vs. 0.73 for patients with ISS = 16-74 and 0.68 vs. 0.53 for patients with ISS = 25-74. The wISS showed higher specificity, positive predictive value, negative predictive value, and concordance when they were compared at similar levels of sensitivity. The wISS had better calibration than the ISS.

Conclusions

By weighting the AIS from different body regions, the wISS had significantly better predictive power for mortality than the ISS, especially in critically injured adults.

The interplay of directional information provided by unpolarised and polarised light in the heading direction network of the diurnal dung beetle Kheper lamarcki

The sun is the most prominent source of directional information in the heading direction network of the diurnal, ball-rolling dung beetle Kheper lamarcki. If this celestial body is occluded from the beetle's field of view, the distribution of the relative weight between the directional cues that remain shifts in favour of the celestial pattern of polarised light. In this study, we continue to explore the interplay of the sun and polarisation pattern as directional cues in the heading direction network of K. lamarcki. By systematically altering the intensity and degree of the two cues presented, we effectively change the relative reliability of these directional cues as they appear to the dung beetle. The response of the ball-rolling beetle to these modifications allows us to closely examine how the weighting relationship of these two sources of directional information is influenced and altered in the heading direction network of the beetle. We conclude that the process in which K. lamarcki relies on directional information is very likely done based on Bayesian reasoning, where directional information conveying the highest certainty at a particular moment is afforded the greatest weight.

Genetic algorithm for assigning weights to gene expressions using functional annotations

A method, named genetic algorithm for assigning weights to gene expressions using functional annotations (GAAWGEFA), is developed to assign proper weights to the gene expressions at each time point. The weights are estimated using functional annotations of the genes in a genetic algorithm framework. The method shows gene similarity in an improved manner as compared with other existing methods because it takes advantage of the existing functional annotations of the genes. The weight combination for the expressions at different time points is determined by maximizing the fitness function of GAAWGEFA in terms of the positive predictive value (PPV) for the top 10,000 gene pairs. The performance of the proposed method is primarily compared with Biweight mid correlation (BICOR) and original expression values for the six Saccharomyces cerevisiae datasets and one Bacillus subtilis dataset. The utility of GAAWGEFA is shown in predicting the functions of 48 unclassified genes (using p-value cutoff 10^-13) from Saccharomyces cerevisiae microarray data where the expressions are weighted using GAAWGEFA and are clustered using k-medoids algorithm. The related code along with various parameters is available at http://sampa.droppages.com/GAAWGEFA.html.

Exploring Completeness of Adverse Event Reports as a Tool for Signal Detection in Pharmacovigilance

BACKGROUND

Completeness of adverse event (AE) reports is an important component of quality for good pharmacovigilance practices. We aimed to evaluate the impact of incorporating a measure of completeness of AE reports on quantitative signal detection.

METHODS

An internal safety database from a global pharmaceutical company was used in the analysis. vigiGrade, an index score of completeness, was derived for each AE report. Data from various patient support programs (PSPs) were categorized based on average vigiGrade score per PSP. Performance of signal detection was compared between: (1) weighting and not weighting by vigiGrade score; and, (2) well documented and poorly documented PSPs using sensitivity, specificity, area under the receiver operating characteristics curve (AUC) and time-to-signal detection.

RESULTS

The ability to detect signals did not differ significantly when weighting by vigiGrade score [sensitivity (50% vs. 45%, p = 1), specificity (82.8% vs. 82.8%, p = 1), AUC (0.66 vs. 0.63, p = 0.051) or time-to-signal detection (HR 0.81, p = 0.63)] compared to not weighting. Well documented PSPs were better at detecting signals than poorly documented PSPs (AUC 0.66 vs. 0.52; p = 0.041) but time-to-signal detection did not differ significantly (HR 1.54, p = 0.42).

CONCLUSION

Completeness of AE reports did not significantly impact the ability to detect signals when weighting by vigiGrade score or restricting the database based on the level of completeness. While the vigiGrade helps provide quality assessments of AE reports and prioritize cases for review, our findings indicate the tool might not be useful for quantitative signal detection when used by itself.

Feature-weighted survival learning machine for COPD failure prediction

Chronic obstructive pulmonary disease (COPD) yields a high rate of failures such as hospital readmission and death in the United States, Canada and worldwide. COPD failure imposes a significant social and economic burden on society, and predicting such failure is crucial to early intervention and decision-making, making this a very important research issue. Current analysis methods address all risk factors in medical records indiscriminately and therefore generally suffer from ineffectiveness in real applications, mainly because many of these factors relate weakly to prediction. Numerous studies have been done on selecting factors for survival analysis, but their inherent shortcomings render these methods inapplicable for failure prediction in the context of unknown and intricate correlation patterns among risk factors. These difficulties have prompted us to design a new Cox-based learning machine that embeds the feature weighting technique into failure prediction. In order to improve predictive accuracy, we propose two weighting criteria to maximize the area under the ROC curve (AUC) and the concordance index (C-index), respectively. At the same time, we perform a Dirichlet-based regularization on weights, making differences between factor relevance clearly visible while maintaining the model's high predictive ability. The experimental results on real-life COPD data collected from patients hospitalized at the Centre Hospitalier Universitaire de Sherbrooke (CHUS) demonstrate the effectiveness of our learning machine and its great promise in clinical applications.

Applied causal inference methods for sequential mediators

BACKGROUND

Mediation analysis aims at estimating to what extent the effect of an exposure on an outcome is explained by a set of mediators on the causal pathway between the exposure and the outcome. The total effect of the exposure on the outcome can be decomposed into an indirect effect, i.e. the effect explained by the mediators jointly, and a direct effect, i.e. the effect unexplained by the mediators. However finer decompositions are possible in presence of independent or sequential mediators.

METHODS

We review four statistical methods to analyse multiple sequential mediators, the inverse odds ratio weighting approach, the inverse probability weighting approach, the imputation approach and the extended imputation approach. These approaches are compared and implemented using a case-study with the aim to investigate the mediating role of adverse reproductive outcomes and infant respiratory infections in the effect of maternal pregnancy mental health on infant wheezing in the Ninfea birth cohort.

RESULTS

Using the inverse odds ratio weighting approach, the direct effect of maternal depression or anxiety in pregnancy is equal to a 59% (95% CI: 27%,94%) increased prevalence of infant wheezing and the mediated effect through adverse reproductive outcomes is equal to a 3% (95% CI: -6%,12%) increased prevalence of infant wheezing. When including infant lower respiratory infections in the mediation pathway, the direct effect decreases to 57% (95% CI: 25%,92%) and the indirect effect increases to 5% (95% CI: -5%,15%). The estimates of the effects obtained using the weighting and the imputation approaches are similar. The extended imputation approach suggests that the small joint indirect effect through adverse reproductive outcomes and lower respiratory infections is due entirely to the contribution of infant lower respiratory infections, and not to an increased prevalence of adverse reproductive outcomes.

CONCLUSIONS

The four methods revealed similar results of small mediating role of adverse reproductive outcomes and early respiratory tract infections in the effect of maternal pregnancy mental health on infant wheezing. The choice of the method depends on what is the effect of main interest, the type of the variables involved in the analysis (binary, categorical, count or continuous) and the confidence in specifying the models for the exposure, the mediators and the outcome.

An evaluation of the 1987 French Disabled Workers Act: better paying than hiring

This paper presents the first evaluation of the French Disabled Workers Act of 1987, which aimed to promote the employment of disabled people in the private sector. We use a panel data set, which includes both the health and the labour market histories of workers. We account both for unobserved heterogeneity and for the change in the disabled population over time. We find that the law had a negative impact on the employment of disabled workers in the private sector. This counterproductive effect likely comes from the possibility to pay a fine instead of hiring disabled workers.

Integrating causation in investigative ecological weight of evidence assessments

Weight of evidence (WOE) frameworks integrate environmental assessment data to reach conclusions regarding relative certainty of adverse environmental effects due to stressors, possible causation, and key uncertainties. Such studies can be investigative (i.e., determining whether adverse impact is occurring to identify a need for management) or retrospective (i.e., determining the cause of a detected impact such that management efforts focus on the correct stressor). Such WOE assessments do not themselves definitively establish causation; they provide the basis for subsequent follow-up studies to further investigate causation. We propose a modified investigative WOE framework that includes an additional weighting step, which we term "direction weighting." This additional step allows for the examination of alternative hypotheses and provides improved certainty regarding possible causation. To our knowledge, this approach has not been previously applied in investigative ecological WOE assessments. We provide a generic example of 2 conflicting hypotheses related to a mine discharging treated effluent to a freshwater lake: chemical toxicity versus nutrient enrichment. Integr Environ Assess Manag 2017;13:702-713. © 2016 SETAC.

Two weights make a wrong: Cluster randomized trials with variable cluster sizes and heterogeneous treatment effects

In cluster randomized trials (CRTs), the hierarchical nesting of participants (level 1) within clusters (level 2) leads to two conceptual populations: clusters and participants. When cluster sizes vary and the goal is to generalize to a hypothetical population of clusters, the unit average treatment effect (UATE), which averages equally at the cluster level rather than equally at the participant level, is a common estimand of interest. From an analytic perspective, when a generalized estimating equations (GEE) framework is used to obtain averaged treatment effect estimates for CRTs with variable cluster sizes, it is natural to specify an inverse cluster size weighted analysis so that each cluster contributes equally and to adopt an exchangeable working correlation matrix to account for within-cluster correlation. However, such an approach essentially uses two distinct weights in the analysis (i.e. both cluster size weights and covariance weights) and, in this article, we caution that it will lead to biased and/or inefficient treatment effect estimates for the UATE estimand. That is, two weights "make a wrong" or lead to poor estimation characteristics. These findings are based on theoretical derivations, corroborated via a simulation study, and illustrated using data from a CRT of a colorectal cancer screening program. We show that, an analysis with both an independence working correlation matrix and weighting by inverse cluster size is the only approach that always provides valid results for estimation of the UATE in CRTs with variable cluster sizes.

Sensitivity of operational and environmental benchmarks of retail stores to decision-makers' preferences through Data Envelopment Analysis

Within the framework of multi-criteria decision analysis (MCDA), weighting methods are typically used to capture decision-makers' preferences. In this regard, the increasing use of the combined LCA (Life Cycle Assessment) + DEA (Data Envelopment Analysis) methodology as an MCDA tool requires an in-depth analysis of how the preferences of decision-makers could affect the outcomes of LCA + DEA studies. This work revisits a case study of 30 retail stores/supply chains located in Spain by applying alternative weighted DEA approaches to evaluate the influence of decision-makers' preferences (weights) on the final outcomes, with a focus on efficiency scores and operational and environmental benchmarks. The ultimate goal is to effectively capture the view of stakeholders when applying LCA + DEA for the sound, sustainability-oriented management of multiple similar entities. Different weight vectors are separately applied to three types of DEA elements: operational inputs, time terms, and divisions. Besides, preferences from three alternative standpoints are considered: company manager through direct rating, and environmental policy-maker and local community through AHP (analytic hierarchy process). A significant influence on efficiency scores and sustainability benchmarks was found when weighting decision-makers' preferences on operational inputs. Additionally, a moderate influence was observed when weighting divisions according to a policy-maker or local community perspective. Although the results are case-specific, they lead to the general recommendation to enrich LCA + DEA studies by following not only an equal-weight approach but also approaches that include the preferences of the stakeholders effectively involved in the study.

Improved Coherent Plane-Wave Compounding Using Sign Coherence Factor Weighting for Frequency-Domain Beamforming

This study proposes a novel modified sign coherence factor (SCF) weighting adapted to the frequency-domain (FD) beamforming for ultrasound plane-wave imaging to achieve a high frame rate and better image quality. First, before beamforming, the sign components were extracted from the radiofrequency signals of aperture data. Second, the modified SCF was established using the FD beamformed sign components. Finally, the FD beamformed image was weighted by the modified SCF. To assess the performance of the proposed modified SCF for FD beamforming, the resolution, contrast, computation complexity and execution time of the generated images were evaluated. The results revealed that the FD-SCF could significantly improve the computational load compared with the classic delay-and-sum SCF on the premise of equal image quality improvement. Therefore, high image quality and low computational load have been successfully combined under the proposed weighting method.

Exploring the relationship between breastfeeding and the incidence of infant illnesses in Ireland: evidence from a nationally representative prospective cohort study

BACKGROUND

Ireland has one of the lowest BF rates in the world. This study investigates the association between breastfeeding and infant health in Ireland.

METHODS

A cross-sectional, secondary analysis of data collected from Growing Up in Ireland (GUI): the National Longitudinal Study of Children was conducted. The average morbidity for 2212. infants exclusively breastfed for at least 90 days (EBF90days) was compared to data for 3987 infants in the non-breastfed (Non-BF) group. Data were weighted using entropy balancing to ensure the comparability of groups. Sensitivity analyses considered alternative definitions of the breastfeeding group.

RESULTS

Infants who were EBF90days were significantly less likely to be admitted to hospital (CI: - 0.06 to - 0.03), spent less nights in hospital (CI: - 0.37 to - 0.11), and were less likely to develop respiratory diseases including asthma (CI: - 0.03 to - 0.01), chest infections (CI: - 0.12 to - 0.08), snuffles/common colds (CI: - 0.07 to - 0.02), ear infections (CI: - 0.08 to - 0.04), eczema (CI: - 0.08 to - 0.04), skin problems (CI: - 0.04 to - 0.00), wheezing or asthma (CI: - 0.06 to - 0.03), vomiting (CI: - 0.03 to - 0.00), and colic (CI: - 0.04 to - 0.01). Further outcomes such as current health of the infant at time of interview (CI: - 0.04 to - 0.00), feeding problems (CI: - 0.04 to - 0.02) and sleeping problems (CI: - 0.02 to - 0.00) indicated a protective effect of EBF90days versus Non-BF. However, these infants were also more likely to fail to gain weight (CI: 0.01 to 0.02) and were at a slightly higher risk of developing nappy rash (CI: 0.00 to 0.02).

CONCLUSION

Exclusive breastfeeding for 90+ days is associated with protection against childhood morbidity. Given the protective effect of breastfeeding on adverse health effects in infants, policy makers should prioritise policies that support, promote and protect exclusive breastfeeding.

A novel weighting method for multi-linear MPC control of Hammerstein systems based on included angle

A novel included angle based weighting method is proposed for multi-linear model predictive control (MPC) of Hammerstein systems. It makes full use of the special structure of the Hammerstein models, and thus it is intuitive and simple. Moreover, there is only one tuning parameter and the weights can be calculated offline and stored in a look-up table. Therefore, online computational load is largely reduced. Most important of all, it schedules local controllers properly and effectively. A Lab-tank system which can be modeled into a Hammerstein model is investigated. Comparisons are made among the nonlinearity inversion control method, the proposed weighting method and traditional weighting methods, e.g., Trapezoidal and Gaussian weighting methods. Simulations confirm that the proposed weighting method is superior to traditional methods.

Impact of sampling and data collection methods on maternity survey response: a randomised controlled trial of paper and push-to-web surveys and a concurrent social media survey

BACKGROUND

Novel survey methods are needed to tackle declining response rates. The 2020 National Maternity Survey included a randomised controlled trial (RCT) and social media survey to compare different combinations of sampling and data collection methods with respect to: response rate, respondent representativeness, prevalence estimates of maternity indicators and cost.

METHODS

A two-armed parallel RCT and concurrent social media survey were conducted. Women in the RCT were sampled from ONS birth registrations and randomised to either a paper or push-to-web survey. Women in the social media survey self-selected through online adverts. The primary outcome was response rate in the paper and push-to-web surveys. In all surveys, respondent representativeness was assessed by comparing distributions of sociodemographic characteristics in respondents with those of the target population. External validity of prevalence estimates of maternity indicators was assessed by comparing weighted survey estimates with estimates from national routine data. Cost was also compared across surveys.

RESULTS

The response rate was higher in the paper survey (n = 2,446) compared to the push-to-web survey (n = 2,165)(30.6% versus 27.1%, difference = 3.5%, 95%CI = 2.1-4.9, p < 0.0001). Compared to the target population, respondents in all surveys were less likely to be aged < 25 years, of Black or Minority ethnicity, born outside the UK, living in disadvantaged areas, living without a partner and primiparous. Women in the social media survey (n = 1,316) were less representative of the target population compared to women in the paper and push-to-web surveys. For some maternity indicators, weighted survey estimates were close to estimates from routine data, for other indicators there were discrepancies; no survey demonstrated consistently higher external validity than the other two surveys. Compared to the paper survey, the cost saving per respondent was £5.45 for the push-to-web survey and £22.42 for the social media survey.

CONCLUSIONS

Push-to-web surveys may cost less than paper surveys but do not necessarily result in higher response rates. Social media surveys cost significantly less than paper and push-to-web surveys, but sample size may be limited by eligibility criteria and recruitment window and respondents may be less representative of the target population. However, reduced representativeness does not necessarily introduce more bias in weighted survey estimates.

Using Retrospective Sampling to Estimate Models of Relationship Status in Large Longitudinal Social Networks

Estimation of longitudinal models of relationship status between all pairs of individuals (dyads) in social networks is challenging due to the complex inter-dependencies among observations and lengthy computation times. To reduce the computational burden of model estimation, a method is developed that subsamples the "always-null" dyads in which no relationships develop throughout the period of observation. The informative sampling process is accounted for by weighting the likelihood contributions of the observations by the inverses of the sampling probabilities. This weighted-likelihood estimation method is implemented using Bayesian computation and evaluated in terms of its bias, efficiency, and speed of computation under various settings. Comparisons are also made to a full information likelihood-based procedure that is only feasible to compute when limited follow-up observations are available. Calculations are performed on two real social networks of very different sizes. The easily computed weighted-likelihood procedure closely approximates the corresponding estimates for the full network, even when using low sub-sampling fractions. The fast computation times make the weighted-likelihood approach practical and able to be applied to networks of any size.