Stata Command Research Articles

On the risks of secure attachment in infancy: Childhood irritability and adolescent depression predicted by secure attachment and high approach behaviours at 14-months towards a caregiver living with inter-parental violence

Introduction The secure infant turns to a caregiver for comfort when distressed or threatened. Does this level of openness create vulnerability where the caregiver is unsupportive or aggressive? In this study we examined prospectively whether secure attachment in infancy, and approach behaviours on reunion with a parent, were associated with childhood emotional dysregulation (irritability) and adolescent depression among children exposed to inter-parental violence. Methods We followed 219 families recruited from the general population during pregnancy (members of the Wirral Child Health and Development Study; WCHADS), with attachment assessments (Strange Situation Procedure; SSP) at 14 months, maternal interviews about inter-parental violence at 2.5, 5 and 7 years, and parent and teacher rated irritability at 9 years (CBCL). At age 13 years, 199 young people rated their levels of depression (SMFQ). In addition to the standard SSP classification, a latent variable reflecting approach behaviours during reunions was generated from the SSP dimensional scores and a factor score extracted. Analyses used path analysis using the gsem command in Stata. Results There were interactions between attachment security and inter-parental violence for age 9 irritability ( p = .084) and age 13 depression ( p = .039) which arose from greater levels of psychopathology among secure children exposed to inter-parental violence. Similarly, higher approach behaviours during SSP reunions among children exposed to inter-parental violence were associated with irritability (interaction term p = .004) and depression (interaction term p < .001). Among children who were not exposed to partner violence higher approach behaviours in the SSP were associated with lower irritability and depression. Conclusion Infant behaviours characteristic of attachment security in the Strange Situation Procedure may not equip children to deal with exposure to inter-parental violence and associated parental negativity.

Changes in risk factors for food sensitization in early life: Analysis over a period of 10 years.

Although epidemiological trends of childhood food sensitization (FS) in IgE-mediated food allergy were reported in China, few studies have examined at changes in its risk factors. To investigate the change in early-life risk factors associated with childhood food sensitization during 2009-2019 in China. Data from two cross-sectional surveys conducted in 2009 and 2019 (401 and 513 children, respectively) were analyzed. The results of skin prick tests and information on food sensitization-related risk factors in children were summarized, including family history of atopic disease (FHA), demographic characteristics, method of delivery, feeding patterns, sibship size, pet ownership, and vitamin D supplementation. Binary logistic regression was used to calculate the odds ratio and the regression coefficient β-value of risk factors in the 2009 and 2019 surveys separately. Then, coefficient β-value differences between the two surveys were analyzed by the bdiff command in STATA to describe the change in risk factors over 10 years. The 2009 survey revealed that FHA, age, only child, and feeding patterns were associated with food sensitization. The 2019 survey showed that food sensitization was affected by age, sex, and feeding patterns. However, from 2009 to 2019, the probability of food sensitization in the only-child group significantly increased by 226.0% (β-value difference = 0.81, P = 0.024) and decreased by 65.0% in female children (β-value difference = -1.06, P = 0.008). The effect of age on food sensitization decreased by 50.0% (β-value difference = -0.69, P < 0.001) over 10 years. The effect of FHA and common lifestyle factors on food sensitization did not significantly change during 2009-2019. However, the influence of demographic characteristics on food sensitization has changed since 2009; that is, older age, male gender, and only child are more likely to develop food sensitization, which needs to be considered in future epidemiological surveys. http://www.chictr.org.cn/, identifier ChiCTR1900024338.

Hospital mortality of blunt abdominal aortic injury (BAAI): a systematic review and meta-analysis

BackgroundStudies on the mortality of blunt abdominal aortic injury (BAAI) are rare and have yielded inconsistent results. In the present study, we aimed to quantitatively analyse the retrieved data to more accurately determine the hospital mortality of BAAI.MethodsThe Excerpta Medica Database, PubMed, Web of Science and Cochrane Library databases were searched to identify relevant publications without date restrictions. The overall hospital mortality (OHM) of BAAI patients was set as the primary outcome measure. English publications with data that met the selection criteria were included. The quality of all included studies was assessed by the Joanna Briggs Institute checklist and the American Agency for Health Care Quality and Research’s cross-sectional study quality evaluation items. After data extraction, a meta-analysis of the Freeman–Tukey double arcsine transformation of data was performed using the Metaprop command in Stata 16 software. Heterogeneity was assessed and reported as a percentage using the I2 index value and as a P value using the Cochrane Q test. Various methods were used to determine the sources of heterogeneity and to analyse the sensitivity of the computation model.ResultsOf the 2147 references screened, 5 studies that involved 1593 patients met the selection criteria and were included. There were no low-quality references after assessment. One study that only included 16 juvenile BAAI patients was excluded from the meta-analysis of the primary outcome measure due to high heterogeneity. Due to the low heterogeneity (I2 = 47.6%, P = 0.126 for Q test) that was observed after using the random effects model, the fixed model was subsequently used to pool the effect sizes of the remaining four studies, thus yielding an OHM of 28.8% [95% confidence interval (CI) 26.5–31.1%]. The stability of the model was verified by sensitivity analysis, and Egger’s test (P = 0.339) indicated a low level of publication bias. In addition, we also performed meta-analyses and obtained a pooled hospital mortality of operation (13.5%, 95% CI 8.0–20.0%), a pooled hospital mortality of non-operation (28.4%, 95% CI 25.9–31.0%), and a pooled rate of aortic rupture (12.2%, 95% CI 7.0–18.5%) of BAAI.ConclusionsThe present study indicated that BAAI has an OHM of 28.8%, indicating that this disease deserves more attention and research.

DS-HECK: double-lasso estimation of Heckman selection model

We extend the Heckman (1979) sample selection model by allowing for a large number of controls that are selected using lasso under a sparsity scenario. The standard lasso estimation is known to under-select causing an omitted variable bias in addition to the sample selection bias. We outline the required adjustments needed to restore consistency of lasso-based estimation and inference for vector-valued parameters of interest in such models. The adjustments include double lasso for both the selection equation and main equation and a correction of the variance matrix. We also connect the estimator with results on redundancy of moment conditions. We demonstrate the effect of the adjustments using simulations and we investigate the determinants of female labor market participation and earnings in the US using the new approach. The paper comes with dsheckman, a dedicated Stata command for estimating double-selection Heckman models.

Family resilience and neighborhood factors affect the association between digital media use and mental health among children: does sleep mediate the association?

The associations between digital media use and mental well-being among children and adolescents have been inconclusive. We examined (i) the associations between digital media use and mental health outcomes, anxiety, depression, and ADHD, (ii) whether family resilience and neighborhood factors attenuate the associations, and (iii) whether sleep mediates these associations. We used the National Survey of Children’s Health data from 2019 to 2020. A total of 45,989 children’s (6–17 years) data were analyzed in this study. Multivariate logistic regression was used to assess the associations between digital media use and anxiety, depression, and ADHD. Path models and Paramed command in STATA were used to test the role of sleep as a mediator of these associations. The prevalence of heavy digital media users (who spent 4 or more hours per day) among the analytic sample was 30.52%, whereas anxiety was 13.81%, depression was 5.93%, and ADHD was 12.41%. Children in the heavy media user group had 63% increased odds of anxiety (95% CI: 1.32–2.01) and 99% increased odds of depression (95% CI: 1.35–2.94) after adjusting for sociodemographic factors, compared to the children in light media user group (who spent < 2 h per day), and these relations were significant at 0.01 level. However, family resilience and community factors significantly attenuated the effect of digital media use on anxiety and depression. Sleep did not mediate the associations between digital media use and anxiety or depression. Conclusions: Family resilience and neighborhood factors protect against the harmful effects of digital media use. Further research is needed to examine the relationships of media contents, the presence of electronic devices in bedrooms, and sleep quality with mental health.What is Known:• Spending long hours on digital media may adversely affect children and adolescents' health and development. However, the mediating role of sleep in the association between digital media use and mental health outcomes is inconclusive.What is New:• Digital media use has detrimental effects on anxiety and depression. However, family resilience and neighborhood factors attenuated the association. The study highlights the importance of positive family functioning and neighborhood conditions reducing the harmful effects of digital media use.

Development of a Stata Command for Pediatric Risk of Mortality Calculation.

Azamfirei, Razvan MS1,2; Mennie, Colleen BSN, RN1; Fackler, James MD1,3; Kudchadkar, Sapna R. MD, PhD, FCCM1,3–5 Author Information

A note on creating inset plots using graph twoway

Inset plots can be used to “zoom in” on densely populated areas of a graph or to add extra relevant data in the form of, for example, distribution plots. However, the standard Stata command for combining plots, graph combine, does not permit this type of seamless integration. Each plot within a graph combine object is allocated a grid cell that cannot be placed within another grid cell— at least not without certain (invariably unwanted) graphical complications. We present a fairly simple work-around to this issue using reproducible examples. The main idea is to plot insets along a second axis and then artificially modify the range of this axis to constrain the inset plot within a specified area of the main graph. Additional tips are included for producing more intricate, multilayered inset graphs.

Projecting wheat demand in China and India for 2030 and 2050: Implications for food security.

The combined populations of China and India were 2.78 billion in 2020, representing 36% of the world population (7.75 billion). Wheat is the second most important staple grain in both China and India. In 2019, the aggregate wheat consumption in China was 96.4 million ton and in India it was 82.5 million ton, together it was more than 35% of the world's wheat that year. In China, in 2050, the projected population will be 1294-1515 million, and in India, it is projected to be 14.89-1793 million, under the low and high-fertility rate assumptions. A question arises as to, what will be aggregate demand for wheat in China and India in 2030 and 2050? Applying the Vector Error Correction model estimation process in the time series econometric estimation setting, this study projected the per capita and annual aggregate wheat consumptions of China and India during 2019-2050. In the process, this study relies on agricultural data sourced from the Food and Agriculture Organization of the United States (FAO) database (FAOSTAT), as well as the World Bank's World Development Indicators (WDI) data catalog. The presence of unit root in the data series are tested by applying the augmented Dickey-Fuller test; Philips-Perron unit root test; Kwiatkowski-Phillips-Schmidt-Shin test, and Zivot-Andrews Unit Root test allowing for a single break in intercept and/or trend. The test statistics suggest that a natural log transformation and with the first difference of the variables provides stationarity of the data series for both China and India. The Zivot-Andrews Unit Root test, however, suggested that there is a structural break in urban population share and GDP per capita. To tackle the issue, we have included a year dummy and two multiplicative dummies in our model. Furthermore, the Johansen cointegration test suggests that at least one variable in both data series were cointegrated. These tests enable us to apply Vector Error Correction (VEC) model estimation procedure. In estimation the model, the appropriate number of lags of the variables is confirmed by applying the "varsoc" command in Stata 17 software interface. The estimated yearly per capita wheat consumption in 2030 and 2050 from the VEC model, are multiplied by the projected population in 2030 and 2050 to calculate the projected aggregate wheat demand in China and India in 2030 and 2050. After projecting the yearly per capita wheat consumption (KG), we multiply with the projected population to get the expected consumption demand. This study found that the yearly per capita wheat consumption of China will increase from 65.8 kg in 2019 to 76 kg in 2030, and 95 kg in 2050. In India, the yearly per capita wheat consumption will increase to 74 kg in 2030 and 94 kg in 2050 from 60.4 kg in 2019. Considering the projected population growth rates under low-fertility assumptions, aggregate wheat consumption of China will increase by more than 13% in 2030 and by 28% in 2050. Under the high-fertility rate assumption, however the aggregate wheat consumption of China will increase by 18% in 2030 and nearly 50% in 2050. In the case of India, under both low and high-fertility rate assumptions, aggregate wheat demand in India will increase by 32-38% in 2030 and by 70-104% in 2050 compared to 2019 level of consumption. Our results underline the importance of wheat in both countries, which are the world's top wheat producers and consumers, and suggest the importance of research and development investments to maintain sufficient national wheat grain production levels to meet China and India's domestic demand. This is critical both to ensure the food security of this large segment of the world populace, which also includes 23% of the total population of the world who live on less than US $1.90/day, as well as to avoid potential grain market destabilization and price hikes that arise in the event of large import demands.

Association between low health literacy and adverse health behaviors in North Carolina, 2016.

Health literacy (HL) is an urgent public health challenge facing the U.S. HL is a critical factor in health inequities and exacerbates underlying social determinants of health. This study assesses the association between low HL (LHL) and adverse health behaviors, which contribute to poor health. Researchers used North Carolina's 2016 Behavioral Risk Factor Surveillance System data, namely, the Health Literacy optional module which asks respondents to rate how difficult it is for them to get health-related advice or to understand medical information (verbal or written). Health behaviors analyzed were excessive alcohol consumption, lack of adequate exercise and sleep, and irregular medical and dental check-ups. The sample was divided into four age categories (18-49, 50-64, and 65-75, and 76 and older) for statistical comparisons. Stata 15 and a user-written Stata command, - psacalc-, were used to examine the relationships by addressing omitted variable bias in OLS regressions. Findings indicate that LHL has a direct robust relationship with not exercising, inadequate sleep, irregular health and dental checkup, and health screenings across different age groups. Among women, LHL is associated with getting a Pap test in 3 years as opposed to more than 3 years. The adverse behaviors can explain the mechanisms underlying the link between LHL and adverse health outcomes. Further research on the causal relationship between LHL and adverse health behaviors using longitudinal data on a broader geographic region is warranted.

Audit_cc: A Stata command for the analysis of matched case-control audits of cervical cancer screening

Background and objectivesCervical screening programmes are crucial for the early diagnosis and prevention of cancer of the cervix. Regular auditing is vital for ensuring that these programmes achieve their full potential and meet their objectives in practice. Unfortunately, the time and skills required for the statistical analysis of the data collected are often important limiting factors. Comparisons across countries and over time have also been particularly difficult due to a lack of standardized definitions and methodology. We aimed to overcome these problems. MethodsUsing the statistical software Stata, we developed a new command called audit_cc for the analysis of matched case-control audits of cervical cancer screening. Analyses are reported for two measures of screening history: time since last test and time since last negative test. ResultsThe command carries out the data manipulation which is required for the analysis and allows to save the resulting data set in an external file for further investigations. It promotes consistent evaluations of screening programmes over time and across studies and facilitates the creation of automatic publication-quality reports, which are especially useful in the context of routine audits. Conclusionsaudit_cc is a valid tool that not only simplifies the analysis and reporting of cervical screening audits but also allows meaningful international comparisons. Although it is specific for cervical cancer, it can be seen as an example of how the standardisation of exposure definitions and key methodological issues can enable consistent and comparable evaluations of screening programmes across different countries and settings.

Global Health Security Index not a proven surrogate for health systems capacity to respond to pandemics: The case of COVID-19

IntroductionGlobal Health Security borders on prevention, detection and response to public health threats like the novel coronavirus disease 2019 (COVID-19). Global Health Security Index (GHSI) of 2019 and 2021 revealed the world remains ill-prepared to deal with future pandemics, evident in the historic impact of COVID-19 on countries. As at 7th December 2022, COVID-19 has infected over 600 million people and claimed over six million lives, mostly in countries with higher GHSI scores. ObjectiveDetermine whether the GHSI scores of countries have a correlation with COVID-19 cases, deaths and vaccination coverage, while adjusting for country level dynamics. MethodsThis paper utilizes GHSI database of 195 countries. Data consists of 171 questions grouped into 37 indicators across six overarching categories on health security and COVID-19. Multivariate multiple regression analysis with robust standard errors was conducted to test the hypothesis that high GHSI ratings do not guarantee better COVID-19 outcomes like cases, deaths and vaccination coverage. Also, avplots STATA command was used to check outliers with potential negative effect on outcome and predictor variables. ResultsGlobal average GHSI score for all 195 countries was 38.9. United States of America recorded the highest GHSI score of 75.9 but also recorded one of the highest COVID-19 cases and deaths; Somalia recorded the worst GHSI score of 16.0 and one of the lowest COVID-19 cases and deaths. High GHSI scores did not associate positively with reduction in COVID-19 cases (Coef=157133.4, p-value=0.009, [95%CI 39728.64 274538.15]) and deaths (Coef=1405.804, p-value=0.047, [95%CI 18.1 2793.508]). However, high GHSI ratings associated with increases in persons fully vaccinated per 100 population (Coef=0.572, p-value=0.000, [95%CI.272.873]). ConclusionIt appears the world might still not be adequately prepared for the next major pandemic, if the narrative remains unchanged. Countries that recorded higher GHSI scores, counter-intuitively, recorded higher COVID-19 cases and deaths. Countries need to invest more in interventions towards attaining Universal Health Coverage (UHC) including integrated health systems and formidable primary health care to enhance preparedness and response to pandemics.

1132: MAKING PRISM EASY: DEVELOPMENT OF A STATA COMMAND FOR RISK-OF-MORTALITY CALCULATION

Introduction: PRISM scores have been used in pediatric research to estimate mortality risk and control for illness severity. A major barrier to the usage of this score is the complexity of the calculation and the lack of publicly available implementations. The single publicly available calculator is inconsistently accurate depending on units used and supports data entry for only one patient at a time. Usage of the score over large datasets requires developing custom code to assign scores to physiological variables and calculate the probability of mortality, a resource-intensive and error-prone process. Custom implementations duplicate efforts and are not guaranteed to be accurate. Here we describe the development of a STATA command that calculates PRISM III and PRISM IV scores. Methods: A STATA command that performs PRISM III and PRISM IV score calculations was developed on STATA/MP 17.0. The command implemented the scoring described in the original papers while adding robust error-proofing, data validation, the ability to use both SI and US laboratory value units, and a graphical user interface. Command validation was accomplished by performing unit tests and calculating scores based on deidentified data. Spot checks were performed to ensure the accuracy of the calculations. The command was additionally tested by generating datasets of 10,000,000 simulated patient records containing random realistic data with predetermined PRISM scores and comparing the command output with the predetermined score. To test the run-time of the command, the command was run over 5 distinct datasets using STATA/MP 17.0 2-cores. Results: The STATA command achieved 100% accuracy on both the spot checks and the generated data. The median run-time for PRISM III calculations over 10,000,000 simulated patient records is 50.48 seconds (IQR 46.74 - 50.83). The median run-time for PRISM IV calculations over 10,000,000 simulated patient records is 51.52 seconds (IQR 50.46 - 53.44). Conclusions: The prismscore STATA command performs PRISM III and PRISM IV calculations accurately and efficiently over datasets of all sizes. The tool improves the reproducibility of research involving the scores, streamlines data analysis for pediatric critical care datasets, and offers significant time and resource savings to the research community.

Power swgee: GEE-based power calculations in stepped wedge cluster randomized trials.

Stepped wedge cluster randomized trials are increasingly being used to evaluate interventions in medical, public health, educational, and social science contexts. With the longitudinal and crossover nature of a SW-CRT, complex analysis techniques are often needed which makes appropriately powering SW-CRTs challenging. In this paper, we introduce a newly-developed SW-CRT power calculator, embedded within the power command in Stata. The power calculator assumes a marginal model (i.e., generalized estimating equations [GEE]) for the primary analysis of SW-CRTs, for which other currently available SW-CRT power calculators may not be suitable. The program accommodates complete cross-sectional and closed-cohort designs, and includes multilevel correlation structures appropriate for such designs. We discuss the methods and formulae underlying our SW-CRT calculator, and provide illustrative examples of the use of power swgee. We provide suggestions about the choice of parameters in power swgee, and conclude by discussing areas of future research which may improve the program.

Physical activity in women with depression in the Spanish national health survey: a mediation analysis

Background: Promotion of physical activity (PA) is an important public health goal to reduce depression in women. Although several studies have reported an indirect relationship between physical activity and severity of depression in women across cultures and age groups, other studies have failed to establish this relationship. In this study we aim to analyse whether certain clinical and demographic variables can explain this discrepancy. Methods: Cross-sectional study with 1096 women aged 15 to 69 years who participated in the National Health Survey in Spain 2017. The chi-square test was used for qualitative variables and a logistic regression was used to determine association between PA and depression with sociodemographic characteristics. To assess possible mediation of each clinical and health variable in the association between depression and physical activity, we used the KHB command in Stata. Results: Both the crude model and models adjusted for clinical and socio-demographic variables indicated a positive association between low PA and depression. The KHB decomposition indicates that 62.21% of the association between low physical activity and depression is associated with mediational variables. Self-perception of health and the polypharmacy were the primary contributors to the mediation. Conclusion: This study supports the conclusion that depression is related to lower levels of physical activity in women. Self-perception of health and polypharmacy mediate this relationship. These results have implications for professionals who are developing interventions aimed at increasing physical activity in women experiencing depression.

MEM and MEM4PP: New Tools Supporting the Parallel Generation of Critical Metrics in the Evaluation of Statistical Models

This paper describes MEM and MEM4PP as new Stata tools and commands. They support the automatic reporting and selection of the best regression and classification models by adding supplemental performance metrics based on statistical post-estimation and custom computation. In particular, MEM provides helpful metrics, such as the maximum acceptable variance inflation factor (maxAcceptVIF) together with the maximum computed variance inflation factor (maxComputVIF) for ordinary least squares (OLS) regression, the maximum absolute value of the correlation coefficient in the predictors’ correlation matrix (maxAbsVPMCC), the area under the curve of receiving operator characteristics (AUC-ROC), p and chi-squared of the goodness-of-fit (GOF) test for logit and probit, and also the maximum probability thresholds (maxProbNlogPenultThrsh and maxProbNlogLastThrsh) from Zlotnik and Abraira risk-prediction nomograms (nomolog) for logistic regressions. This new tool also performs the automatic identification of the list of variables if run after most regression commands. After simple successive invocations of MEM (in a .do file acting as a batch file), the collectible results are produced in the console or exported to specially designated files (one .csv for all models in a batch). MEM4PP is MEM’s version for parallel processing. It starts from the same batch (the same .do file with its path provided as a parameter) and triggers different instances of Stata to parallelly generate the same results (one .csv for each model in a batch). The paper also includes some examples using real-world data from the World Values Survey (the evidence between 1981 and 2020, version number 1.6). They help us understand how MEM and MEM4PP support the testing of predictor independence, reverse causality checks, the best model selection starting from such metrics, and, ultimately, the replication of all these steps.

Count data models for outpatient health services utilisation

BackgroundCount data from the national survey captures healthcare utilisation within a specific reference period, resulting in excess zeros and skewed positive tails. Often, it is modelled using count data models. This study aims to identify the best-fitting model for outpatient healthcare utilisation using data from the Malaysian National Health and Morbidity Survey 2019 (NHMS 2019) and utilisation factors among adults in Malaysia.MethodsThe frequency of outpatient visits is the dependent variable, and instrumental variable selection is based on Andersen’s model. Six different models were used: ordinary least squares (OLS), Poisson regression, negative binomial regression (NB), inflated models: zero-inflated Poisson, marginalized-zero-inflated negative binomial (MZINB), and hurdle model. Identification of the best-fitting model was based on model selection criteria, goodness-of-fit and statistical test of the factors associated with outpatient visits.ResultsThe frequency of zero was 90%. Of the sample, 8.35% of adults utilized healthcare services only once, and 1.04% utilized them twice. The mean-variance value varied between 0.14 and 0.39. Across six models, the zero-inflated model (ZIM) possesses the smallest log-likelihood, Akaike information criterion, Bayesian information criterion, and a positive Vuong corrected value. Fourteen instrumental variables, five predisposing factors, six enablers, and three need factors were identified. Data overdispersion is characterized by excess zeros, a large mean to variance value, and skewed positive tails. We assumed frequency and true zeros throughout the study reference period. ZIM is the best-fitting model based on the model selection criteria, smallest Root Mean Square Error (RMSE) and higher R2. Both Vuong corrected and uncorrected values with different Stata commands yielded positive values with small differences.ConclusionState as a place of residence, ethnicity, household income quintile, and health needs were significantly associated with healthcare utilisation. Our findings suggest using ZIM over traditional OLS. This study encourages the use of this count data model as it has a better fit, is easy to interpret, and has appropriate assumptions based on the survey methodology.

Comparative efficacy and acceptability of non-pharmacological interventions in fibromyalgia: Protocol for a network meta-analysis.

IntroductionAlthough several non-pharmacological interventions have been tested in the management of Fibromyalgia (FM), there is little consensus regarding the best options for the treatment of this health condition. The purpose of this network meta-analysis (NMA) is to investigate the comparative efficacy and acceptability of non-pharmacological interventions for FM, in order to assist clinical decision making through a ranking of interventions in relation to the most important clinical outcomes in these patients.Methods and analysisWe will perform a systematic search to identify randomised controlled trials of non-pharmacological interventions endorsed in guidelines and systematic reviews. Information sources searched will include major bibliographic databases without language or date restrictions (MEDLINE, Cochrane Library, EMBASE, AMED, PsycINFO and PEDro). Our primary outcomes will be pain intensity, patient-reported quality of life (QoL), and acceptability of treatment will be our secondary outcome. Risk of bias of the included trials will be assessed using the Cochrane risk of bias tool (RoB2). For each pairwise comparison between the different interventions, we will present mean differences (MDs) for pain intensity and QoL outcomes and Relative Risks (RRs) for acceptability, both with respective 95% confidence intervals (CIs). Initially, standard pairwise meta-analyses will be performed using a DerSimonian-Laird random effects model for all comparisons with at least two trials and then we will perform a frequentist NMA using the methodology of multivariate meta-analysis assuming a common heterogeneity parameter, using the mvmeta command and network suite in STATA. In the NMA, two different types of control group, such as placebo/sham and no intervention/waiting list will be combined as one node called “Control”. The competing interventions will be ranked using the P-score, which is the frequentist analogue of surface under the cumulative ranking curve (SUCRA) for the outcomes of interest at immediate- (intervention duration of up to 2 weeks), short- (over 2 weeks up to 12 weeks) and long-terms (over 12 weeks). The confidence in the results from NMA will be assessed using the Confidence in Network Meta‐analysis (CINeMA) framework.Ethics and disseminationThis work synthesises evidence from previously published studies and does not require ethics review or approval. A manuscript describing the findings will be submitted for publication in a peer-reviewed scientific journal.RegistrationOSF (DOI: 10.17605/OSF.IO/7MS25) and registered in the PROSPERO database (CRD42020216374).

Mediating role of homebound status between depressive symptoms and cognitive impairment among community-dwelling older adults in the USA: a cross-sectional analysis of a cohort study

ObjectiveDepressive symptoms are known modifiable factors of cognitive impairment in older adults. However, the pathway through which depressive symptoms lead to cognitive impairment is not well understood. This study aimed...

Estimating causal effects in the presence of competing events using regression standardisation with the Stata command standsurv

BackgroundWhen interested in a time-to-event outcome, competing events that prevent the occurrence of the event of interest may be present. In the presence of competing events, various estimands have been suggested for defining the causal effect of treatment on the event of interest. Depending on the estimand, the competing events are either accommodated or eliminated, resulting in causal effects with different interpretations. The former approach captures the total effect of treatment on the event of interest while the latter approach captures the direct effect of treatment on the event of interest that is not mediated by the competing event. Separable effects have also been defined for settings where the treatment can be partitioned into two components that affect the event of interest and the competing event through different causal pathways.MethodsWe outline various causal effects that may be of interest in the presence of competing events, including total, direct and separable effects, and describe how to obtain estimates using regression standardisation with the Stata command standsurv. Regression standardisation is applied by obtaining the average of individual estimates across all individuals in a study population after fitting a survival model.ResultsWith standsurv several contrasts of interest can be calculated including differences, ratios and other user-defined functions. Confidence intervals can also be obtained using the delta method. Throughout we use an example analysing a publicly available dataset on prostate cancer to allow the reader to replicate the analysis and further explore the different effects of interest.ConclusionsSeveral causal effects can be defined in the presence of competing events and, under assumptions, estimates of those can be obtained using regression standardisation with the Stata command standsurv. The choice of which causal effect to define should be given careful consideration based on the research question and the audience to which the findings will be communicated.

Prevalence and risk factors of malaria and anaemia and the impact of preventive methods among pregnant women: A case study at the Akatsi South District in Ghana.

AimThis study aimed to ascertain the prevalence and risk factors of malaria and anaemia as well as the impact of preventive methods among pregnant women at the Akatsi South District Hospital of Ghana.Subjects and methodsA hospital based cross-sectional study using simple random sampling technique was conducted among 200 pregnant women receiving antenatal care and laboratory services at the Akatsi District Hospital from May 2016 to July 2016. A semi-structured questionnaire was administered to obtain participants’ malaria preventive methods in addition to demographic and gestational details. Participants’ hemoglobin and malaria status were assessed using one milliliter (1 ml) whole blood collected from each participant following standard procedures. Factors that produced a p-value of ≤0.2 from the univariate model were included in the final model. Association between potential covariates and the outcomes was assessed using multivariate logistic regression. The Clopper-Pearson test statistic was used to determine the 95% confidence intervals of the outcome variables of interest. We also estimated the population attributable fraction (PAF) of anaemia due to malaria by substituting the adjusted relative risk estimates (RRi) (using the adjrr command in STATA) of anaemia due to malaria into the category-specific attributable formula. P-values of <0.05 were considered statistically significant.ResultsPrevalence of anaemia in pregnancy (AiP), malaria in pregnancy (MiP) and AiP/MiP comorbidity was 63.5% (95% CI:56.4–70.2), 11.0% (96% CI:7.0–16.2) and 10.5% (95% CI:6.6–15.6) respectively. Prevalence rates of AiP (66.7%) and MiP (18.5%) predominated among pregnant women aged < 20 years. PAF of AiP due to MiP was 34.5% (95% CI:23.8–43.6). High use of IPTp-SP, 64.0% (95% CI:56.9–70.6) and LLIN, 90.0% (95% CI:85.0–93.8) was observed in this study. Only 42.0% (95% CI:35.1–49.2) used repellent. Not being on the IPTp-SP program posed a 11.70 times risk of MiP (95% CI:2.32–58.96; p = 0.003) compared to pregnant women on the IPTp-SP program. Similarly, not sleeping under LLIN posed an 8.07 times risk of MiP (95% CI:1.98–32.2; p = 0.004) compared to pregnant women who slept under LLIN. Meanwhile, being positive for MiP posed a 12.10 times risk (95% CI:1.35–85.06; p = 0.025) of AiP compared to those negative for malaria whereas failure to attend ANC as scheduled posed 6.34 times risk (95% CI:1.81–22.19; p = 0.004) of AiP among the pregnant women studied.ConclusionThe prevalence of MiP and AiP among pregnant women in the Akatsi South District remains a great concern. High utilization of IPTp-SP and LLIN was observed with a resultant positive effect on malaria prevalence among pregnant women. Improved access to IPTp-SP and LLIN is hence encouraged to help further diminish the risk of malaria infection amongst pregnant women in the District.