Propensity weighting: how to minimise comparative bias in non-randomised studies?

Are Physical and Depressive Symptoms Different Between Women and Men With Heart Failure?: An Exploration Using Two Analytic Techniques

BACKGROUND

Adults with heart failure (HF) experience a constellation of symptoms; however, understanding of gender differences in HF symptoms remain elusive. The aim of this study was to determine whether there are gender differences in physical and depressive symptoms and symptom patterns in HF using 2 different analytic techniques.

METHODS

We performed a secondary analysis of combined data from 6 studies of adults with HF. Physical symptoms were measured with the HF Somatic Perception Scale, and depressive symptoms were measured with the Patient Health Questionnaire-9. First, we performed propensity matching with the nearest neighbor to examine the average treatment effect for HF Somatic Perception Scale and Patient Health Questionnaire-9 in the matched sample of women and men. Next, we used the entire data set in a latent class mixture model to determine patterns of symptoms. Finally, we calculated predictors of class membership with multinomial logistic regression.

RESULTS

The sample (n = 524, 86.5% systolic HF) was 37% women with a mean age of 58.3 ± 13.9 years and mean number of years with HF of 6.9 ± 6.9. Three hundred sixty-six participants were matched on propensity scores; there were no significant gender differences in symptom scores between matched women (n = 183) and men (n = 183). Among all 524 participants, 4 distinct latent classes of symptom patterns indicate that many patients with HF are fatigued, some have more depressive symptoms, and others have significantly more edema or cough. Gender did not predict membership to any symptom pattern.

CONCLUSIONS

There were significant gender differences in sociodemographics, health behaviors, and clinical characteristics, but not HF symptoms or symptom patterns, using either analytic technique.

De Ritis Ratio to Predict Clinical Outcomes of Intermediate- and High-Risk Pulmonary Embolisms

Background: Abnormal liver function tests can identify severe cardiopulmonary failure. The aspartate transaminase/alanine transaminase (AST/ALT) ratio, or the De Ritis ratio, is commonly used to evaluate acute liver damage. However, its prognostic value in pulmonary embolism (PE) is unknown. Methods: Two cohorts, including patients with intermediate- and high-risk PEs, were established: one with an abnormal baseline AST/ALT ratio (>1) and another with a normal baseline AST/ALT ratio (<1). The primary outcome was a 60-day mortality. Secondary outcomes included peak N-terminal pro-brain-natriuretic-peptide (NT-proBNP) levels, complications, and the need for critical care treatment. To assess the effect of abnormal AST/ALT ratios, inverse probability weighted (IPW) analyses were performed. Results: In total, 230 patients were included in the analysis, and 52 (23%) had an abnormal AST/ALT ratio. After the IPW correction, patients with an abnormal AST/ALT ratio had a significantly higher mortality rate and peak NT-proBNP levels. The relative risks of 60-day mortality, shock development, use of inotropes/vasopressors, mechanical ventilation, and extracorporeal life support were 9.2 (95% confidence interval: 3.3-25.3), 10.1 (4.3-24), 2.7 (1.4-5.2), 2.3 (1.4-3.7), and 5.7 (1.4-23.1), respectively. Conclusions: The baseline AST/ALT ratio can be a predictor of shock, multiorgan failure, and mortality in patients with a pulmonary embolism.

Hitting the (bio)mark part 2: analyzing, interpreting, and reporting biomarker data in cardiovascular research

Incorporating biomarkers into cardiovascular studies, including nursing research, is a common approach when identifying underlying mechanisms and providing targets for intervention. However, effective utilization of biomarker data demands careful consideration. In the analysis, interpretation, and reporting phase, there are many facets to consider, including non-normality of the data, normalization procedures, and potential confounding influences of other clinical data. Furthermore, as many studies focus on patient-reported outcomes, it is important that the analysis and interpretation of biomarkers in relation to patient-reported outcomes is rigorous and reproducible. In this article, part 2 of 2, we provide an overview of considerations for the analysis, interpretation, and reporting phases of biomarker studies. We also provide an example of these steps.

Evaluation of a nurse-led multi-component transition program for adolescents with congenital heart disease

OBJECTIVE

To evaluate the effectiveness of the transition program for adolescents with congenital heart disease (CHD) 'Transition With a Heart' (TWAH) on disease-related knowledge, quality of life (QoL), transition experiences, and gaps in follow-up.

METHODS

A study with pre-posttest and control group (post-test) using consecutive sampling, including adolescents with moderate to severely complex CHD, without intellectual disability, aged≥ 12 y, and parents. After weighting, t-tests were performed. A multivariable regression analysis explored the outcomes' determinants.

RESULTS

In the intervention group, 28 adolescents and 25 parents were included, and 53 adolescents and 18 parents as controls. Adolescents' knowledge significantly increased after completing TWAH (from 59.8% to 75.7%;p < 0.01). Their knowledge was positively correlated with TWAH (β = +13.3;p < 0.01). Adolescents' transition experiences were also positively related to TWAH (general experience: β = +5.5;p < 0.01; transfer satisfaction: β = +0.8; p < 0.01). Adolescents' QoL was mainly determined by CHD complexity and not by TWAH. No one showed gaps in follow-up. TWAH was not associated with parents' transition experiences.

CONCLUSION

Implementing TWAH substantially improved adolescents' disease-related knowledge and transition experiences.

PRACTICE IMPLICATIONS

The results regarding transition experiences need to be confirmed by further research. The TWAH design with the person-tailored educational program, skills training, and the transition coordinator can be used in settings with other chronic diseases.

Comparing youth with and without type 1 diabetes on perceived parenting and peer functioning: a propensity weighting approach

The premise of this study was to gain more insight into whether type 1 diabetes (T1D) can impact how youth perceive parents and peers. To address limitations of previous observational studies comparing youth with T1D to control youth, propensity weighting was used to mimic a randomized controlled trial. A total of 558 youth with T1D and 426 control youth (14-26y) completed questionnaires on parental responsiveness, psychological control, overprotection, friend support, extreme peer orientation, and a host of background and psychological functioning variables. The groups were statistically weighted to become as comparable as possible except for disease status. The analysis plan and hypotheses were preregistered on the open science framework. Youth with T1D perceived their mothers to be more overprotective, perceived fewer friend support, and were less extremely oriented toward peers than control youth. There were no group differences for paternal overprotection and paternal and maternal responsiveness and psychological control. Mothers of youth with T1D seem at risk to practice overprotective parenting and clinicians could play an important role in making mothers aware of this risk. However, the absence of group differences for the maladaptive parenting dimension of psychological control and adaptive dimension of responsiveness are reassuring and testify to the resilient nature of youth with T1D and their families. Additionally, there is accumulating evidence that T1D could interfere with engaging in supportive friendships.

Childbearing Performances and Outcomes of Female Patients with Rheumatic Mitral Valve Diseases after Different Mitral Interventions

PURPOSE

This study aimed to illustrate how percutaneous balloon mitral valvuloplasty (PBMV) and mitral valve (MV) surgeries influence women of childbearing age with rheumatic mitral valve diseases (RMVDs) from two aspects, including clinical outcomes and their postoperative childbearing performances.

METHODS

Female patients with RMVD who were of childbearing age and underwent MV interventions between 2007 and 2019 at Beijing Anzhen Hospital were identified. Outcomes included all-cause deaths, repeated MV interventions, and atrial fibrillation. A survey about childbearing attempts and complications during pregnancy was also performed during follow-up.

RESULTS

A total of 379 patients were involved in this study, consisting of 226 cases of mitral valve replacements, 107 cases of mitral valve repairs (MVrs), and 46 cases of PBMVs. PBMV was associated with higher possibilities of repeated MV interventions (P <0.05). Postoperative childbearing attempts were more frequently observed among bioprosthesis, MVr, and PBMV (P <0.05). However, PBMV and MVr showed a higher incidence of cardiac complications during pregnancy as compared to prosthesis replacement (P <0.05).

CONCLUSIONS

MVr and PBMV are not recommended to young female patients for higher incidences of postoperative complications. Safe pregnancy is more likely to be present among patients with biological prosthesis.

Association of HSD17B13 and PNPLA3 With Liver Enzymes and Fibrosis in Hispanic/Latino Individuals of Diverse Genetic Ancestries

BACKGROUND & AIMS

Genetic variants affecting liver disease risk vary among racial and ethnic groups. Hispanics/Latinos in the United States have a high prevalence of PNPLA3 I148M, which increases liver disease risk, and a low prevalence of HSD17B13 predicted loss-of-function (pLoF) variants, which reduce risk. Less is known about the prevalence of liver disease-associated variants among Hispanic/Latino subpopulations defined by country of origin and genetic ancestry. We evaluated the prevalence of HSD17B13 pLoF variants and PNPLA3 I148M, and their associations with quantitative liver phenotypes in Hispanic/Latino participants from an electronic health record-linked biobank in New York City.

METHODS

This study included 8739 adult Hispanic/Latino participants of the BioMe biobank with genotyping and exome sequencing data. We estimated the prevalence of Hispanic/Latino individuals harboring HSD17B13 and PNPLA3 variants, stratified by genetic ancestry, and performed association analyses between variants and liver enzymes and Fibrosis-4 (FIB-4) scores.

RESULTS

Individuals with ancestry from Ecuador and Mexico had the lowest frequency of HSD17B13 pLoF variants (10%/7%) and the highest frequency of PNPLA3 I148M (54%/65%). These ancestry groups had the highest outpatient alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels, and the largest proportion of individuals with a FIB-4 score greater than 2.67. HSD17B13 pLoF variants were associated with reduced ALT level (P = .002), AST level (P < .001), and FIB-4 score (P = .045). PNPLA3 I148M was associated with increased ALT level, AST level, and FIB-4 score (P < .001 for all). HSD17B13 pLoF variants mitigated the increase in ALT conferred by PNPLA3 I148M (P = .006).

CONCLUSIONS

Variation in HSD17B13 and PNPLA3 variants across genetic ancestry groups may contribute to differential risk for liver fibrosis among Hispanic/Latino individuals.

A new method for analyzing clinical trials in depression based on individual propensity to respond to placebo estimated using artificial intelligence

One of the major reasons for trial failures in major depressive disorders (MDD) is the presence of unpredictable levels of placebo response as the individual baseline propensity to respond to placebo is not adequately controlled by the current randomization and statistical methodologies. The individual propensity to respond to any treatment or intervention assessed at baseline was considered as a major non-specific prognostic and confounding effect. The objective of this paper was to apply the propensity score methodology to control for potential imbalance at baseline in the propensity to respond to placebo in clinical trials in MDD. Individual propensity was estimated using artificial intelligence (AI) applied to observations collected in two pre-randomization occasions. Cases study are presented using data from two randomized, placebo-controlled trials to evaluate the efficacy of paroxetine in MDD. AI models were used to estimate the individual propensity probability to show a treatment non-specific placebo effect. The inverse of the estimated probability was used as weight in the mixed-effects analysis to assess treatment effect. The comparison of the results obtained with and without propensity weight indicated that the weighted analysis provided an estimate of treatment effect and effect size significantly larger than the conventional analysis.

A comparison of clinical outcomes and costs between robotic and open ventral hernia repair

BACKGROUND

As robotic ventral hernia repair(VHR) adoption increases, real-world evidence is needed to ensure appropriate utilization.

METHODS

Data for open and robotic VHR(OVHR, RVHR) was retrospectively analyzed. Outcomes and costs were compared via inverse probability treatment weighting using propensity scores to estimate the average treatment effect on the treated for RVHR.

RESULTS

675 open and 609 robotic ventral hernia repairs were included. Demographics and hernia characteristics were comparable. Complications rates were lower in RVHR(p < 0.001). Clavien-Dindo grade-III complications were lower in RVHR(13.2% vs. 4.9%, p < 0.001). RVHR resulted in fewer surgical site events(21.5% vs. 12.2%, p < 0.001). Recurrence rates were greater in OVHR(8.9% vs. 2.8%, p < 0.001). The higher RVHR hospital costs (Δ = $2456, p = 0.005) were balanced by the lower post-discharge costs, compared to OVHR(Δ = $799, p = 0.023). Total costs did not differ(Δ = $1656 p = 0.081).

CONCLUSION

Although hospital costs were higher, post-discharge expenses favored RVHR due to the lower postoperative complications, which lead to comparable total costs to OVHR.

Artificial intelligence approach for the analysis of placebo-controlled clinical trials in major depressive disorders accounting for individual propensity to respond to placebo

Treatment effect in clinical trials for major depressive disorders (RCT) can be viewed as the resultant of treatment specific and non-specific effects. Baseline individual propensity to respond non-specifically to any treatment or intervention can be considered as a major non-specific confounding effect. The greater is the baseline propensity, the lower will be the chance to detect any treatment-specific effect. The statistical methodologies currently applied for analyzing RCTs doesn't account for potential unbalance in the allocation of subjects to treatment arms due to heterogenous distributions of propensity. Hence, the groups to be compared may be imbalanced, and thus incomparable. Propensity weighting methodology was used to reduce baseline imbalances between arms. A randomized, double-blind, placebo controlled, three arms, parallel group, 8-week, fixed-dose study to evaluate efficacy of paroxetine CR 12.5 and 25 mg/day is presented as a cases study. An artificial intelligence model was developed to predict placebo response at week 8 in subjects assigned to placebo arm using changes from screening to baseline of individual Hamilton Depression Rating Scale items. This model was used to predict the probability to respond to placebo in each subject. The inverse of the probability was used as weight in the mixed-effects model applied to assess treatment effect. The analysis with and without propensity weight indicated that the weighted analysis provided an estimate of treatment effect and effect-size about twice larger than the non-weighted analysis. Propensity weighting provides an unbiased strategy to account for heterogeneous and uncontrolled placebo effect making patients' data comparable across treatment arms.

Associations between Objectively Measured Patterns of Sedentary Behavior and Arterial Stiffness in Chinese Community-Dwelling Older Women

AIMS

Arterial stiffness is an important risk factor of arteriosclerosis and cardiovascular events. Sedentary behavior (SB) is associated with increased risk for chronic cardiovascular diseases and poor health outcomes. This study aims to investigate the associations of objectively measured patterns of SB with arterial stiffness in Chinese community-dwelling older women.

METHODS AND RESULTS

Cross-sectional data were derived from the baseline survey of the Physical Activity and Health in Older Women Study (PAHIOWS). Arterial stiffness was evaluated through cardio-ankle vascular index (CAVI), CAVI ≥ 9 was defined as cutoff point. SB patterns including sedentary time in SB bouts of ≥10, 30 and 60 min, number of SB bouts ≥ 10, 30 and 60 min, were measured via tri-axial accelerometers. Multivariate logistic regression was used to investigate associations of different SB patterns with arterial stiffness. Of the total 1125 women aged 60-70, the prevalence of CAVI ≥ 9 was 25.8%. After adjustment for confounding factors, only sedentary time in SB bout ≥ 30 min and 60 min, number of SB bouts ≥ 60 min were associated with arterial stiffness, with an adjusted odds ratio (95% confidence interval) of 1.06 (1.00-1.12) and 1.11 (1.03-1.20) per 30 min increase per day, 1.35 (1.05-1.74) per bout per day, respectively.

CONCLUSIONS

Longer duration of prolonged SB periods was associated with higher level of arterial stiffness. Shortening and interrupting accumulated periods of SB may be an achievable strategy to reduce risk of cardiovascular disease in self-care and cardiovascular nursing.

Absence from work or school in young adults with congenital heart disease: is illness identity associated with absenteeism?

AIMS

Absence from work or school in patients with congenital heart disease (CHD) is an understudied domain. Illness identity, which is the degree to which a chronic illness is integrated into one's identity, could be hypothesized to be a predictor for absenteeism. This study determined the proportion of young adults with CHD absent from work or school and examined if illness identity dimensions were related to absenteeism.

METHODS AND RESULTS

In a cross-sectional study, we included 255 patients with CHD. Data on absence from work or school over the past 12 months were obtained using self-report. The Illness Identity Questionnaire was used, which comprises four illness identity dimensions: rejection, engulfment, acceptance, and enrichment. Linear models with doubly robust estimations were computed after groups were balanced using propensity weighting. Absence from work/school occurred in 69% of young adults with CHD. Absence because of CHD specifically was present in 15% of the patients. Engulfment was significantly related to both all-cause absence and absence for CHD reasons.

CONCLUSIONS

Patients who strongly define themselves in terms of their heart disease were more likely to be absent from work or school than those who did not. If this finding can be confirmed in future research, it has the potential to be a target for intervention to influence work/school absence.

Metastatic Pattern Discriminates Survival Benefit of Type of Surgery in Patients With De Novo Stage IV Breast Cancer Based on SEER Database

Background: The role of surgery and surgery type in de novo stage IV breast cancer (BC) is unclear.

Methods: We carried out a retrospective cohort study that included the data of 4,108 individuals with de novo stage IV BC abstracted from SEER (Surveillance, Epidemiology, and End Results) data resource from 2010 to 2015. The patients were stratified into the non-surgery group, breast-conserving (BCS) surgery group, and mastectomy group. Inverse probability propensity score weighting (IPTW) was then used to balance clinicopathologic factors. Overall survival (OS), as well as the breast cancer-specific survival (BCSS), was assessed in the three groups using Kaplan–Meier analysis and COX model. Subgroups were stratified by metastatic sites for analysis.

Results: Of the 4,108 patients, 48.5% received surgery and were stratified into the BCS group (574 cases) and mastectomy group (1,419 cases). After IPTW balance demographic and clinicopathologic factors, BCS and mastectomy groups had better OS (BCS group: HR, 0.61; 95% CI: 0.49–0.75; mastectomy group: HR, 0.7; 95% CI: 0.63–0.79) and BCSS (BCS group: HR, 0.6; 95% CI, 0.47–0.75; mastectomy group: HR, 0.71; 95% CI, 0.63–0.81) than the non-therapy group. Subgroup analyses revealed that BCS, rather than mastectomy, was linked to better OS (HR, 0.66; 95% CI: 0.48–0.91) and BCSS (HR, 0.63; 95% CI: 0.45–0.89) for patients with bone-only metastasis. For patients with viscera metastasis or bone+viscera metastases, BCS achieved similar OS (viscera metastasis: HR, 1.05; 95% CI: 0.74–1.48; bone+viscera metastases: HR, 1.01; 95% CI: 0.64–1.61) and BCSS (viscera metastasis: HR, 0.94; 95% CI: 0.64–1.38; bone+viscera metastases: HR, 1.06; 95% CI: 0.66–1.73) in contrast with mastectomy.

Conclusions: Local surgery for patients with distant metastasis (DS) exhibited a remarkable survival advantage in contrast with non-operative management. BCS may have more survival benefits for patients with de novo stage IV BC with bone-only metastasis than other metastatic sites. Decisions on de novo stage IV BC primary surgery should be tailored to the metastatic pattern.

Abdominal obesity and myocardial infarction risk - We demonstrate the anthropometric and mathematical reasons that justify the association bias of the waist-to-hip ratio

Introduction

Antecedentes: el índice cintura-cadera (ICC) se utiliza ampliamente para evaluar la asociación de la obesidad abdominal con el infarto de miocardio (IM). Objetivo: nuestro propósito era determinar si el riesgo asociado a la ICC produce sesgo. Métodos: estudio de casos y controles en 252 varones. Usamos la estratificación como criterio para eliminar los efectos del sesgo. Creamos una covariable basal (ICC0,95-0,99) para una nueva muestra emparejada en el estrato de valores entre 0,95 y 0,99. Este estrato coincide con el área común de solapamiento de la distribución de puntos, donde todos los sujetos tienen un índice de propensión similar. Consideramos otra covariable (ICCS) condicionada en ICC < 1 y una circunferencia de cintura (CC) donde la asignación de riesgo fuera espúrea. Hipotetizamos que restando CC del valor de la cadera se calculaba otra variable aritmética (DCC) que podría ser esencial para evidenciar el efecto de confusion que genera el ICC. Resultados: IMC: ABC: 0,694, IC 95 % (0,628-0,760); OR: 3,8. CC: ABC: 0,743, IC 95 % (0,681-0,805); OR: 5,7. ICC: ABC: 0,798, IC 95 % (0,740-0,855); OR: 8,6. Índice cintura-talla (ICT): ABC: 0,782, IC 95 % (0,724-0,840); OR: 8,5. DCC: ABC: 0,204, IC 95 % (0,146-0,261); OR: 0,36. Prevalencia en los casos: ICC ≥ 0,95 (84,1 % vs. 38 %; OR: 8,6); ICC < 1 (36,3 % vs. 85,7 %; OR: 2,3); ICC ≥ 1 (63,4 % vs. 14,2 %; OR: 4,4); CC ≥ 94,4 (71,4 % vs. 30,1 %; OR: 5,7); DCC ≥ 2,2 (27,7 % vs. 75,3 %; OR: 7,9); ICCs (50 % vs. 25 %; OR: 2). Conclusiones: el ICC produce un sesgo de asociación en los casos de IM. Ello puede extrapolarse a otras poblaciones de estudio. El sesgo se explica por un error de concepto matemático que sobreestima el efecto protector de la cadera con respecto a la CC y la altura. El riesgo asociado al ICC por encima del de la CC o el ICT presenta inconsistencia antropométrica y sesgo, llegando a ser epidemiológicamente falso.

A Narrative Review of Methods for Causal Inference and Associated Educational Resources

BACKGROUND AND OBJECTIVES

Root cause analysis involves evaluation of causal relationships between exposures (or interventions) and adverse outcomes, such as identification of direct (eg, medication orders missed) and root causes (eg, clinician's fatigue and workload) of adverse rare events. To assess causality requires either randomization or sophisticated methods applied to carefully designed observational studies. In most cases, randomized trials are not feasible in the context of root cause analysis. Using observational data for causal inference, however, presents many challenges in both the design and analysis stages. Methods for observational causal inference often fall outside the toolbox of even well-trained statisticians, thus necessitating workforce training.

METHODS

This article synthesizes the key concepts and statistical perspectives for causal inference, and describes available educational resources, with a focus on observational clinical data. The target audience for this review is clinical researchers with training in fundamental statistics or epidemiology, and statisticians collaborating with those researchers.

RESULTS

The available literature includes a number of textbooks and thousands of review articles. However, using this literature for independent study or clinical training programs is extremely challenging for numerous reasons. First, the published articles often assume an advanced technical background with different notations and terminology. Second, they may be written from any number of perspectives across statistics, epidemiology, computer science, or philosophy. Third, the methods are rapidly expanding and thus difficult to capture within traditional publications. Fourth, even the most fundamental aspects of causal inference (eg, framing the causal question as a target trial) often receive little or no coverage. This review presents an overview of (1) key concepts and frameworks for causal inference and (2) online documents that are publicly available for better assisting researchers to gain the necessary perspectives for functioning effectively within a multidisciplinary team.

CONCLUSION

A familiarity with causal inference methods can help risk managers empirically verify, from observed events, the true causes of adverse sentinel events.

Real-world impact of postmastectomy radiotherapy in T1-2 breast cancer with one to three positive lymph nodes

Background

The utility of postmastectomy radiotherapy (PMRT) in breast cancer patients with T1-2 (tumor size ≤5 cm) and N1 (one to three lymph nodes involved) disease remains controversial. The aim of this population-based study was to investigate the effectiveness of PMRT in this patient subset in the current clinical practice.

Methods

We included T1-2N1 breast cancer patients treated with mastectomy from 2004 to 2012 using the data form the Surveillance, Epidemiology, and End Results program. The association of PMRT administration with breast cancer-specific survival was determined using multivariable Cox analysis.

Results

We identified 10,248 patients of this study, including 3,725 (36.3%) received PMRT and 6,523 (63.7%) patients did not receive PMRT. Use of PMRT showed increase from 2008 onward; the percentage of patients receiving PMRT was 30.6% in 2004 and was 47.1% in 2012 (P<0.001). Patients diagnosis after 2008, aged <50 years, high tumor grade, T2 stage, and ≥2 positive lymph nodes were independently related to PMRT receipt. Multivariate analysis indicated that PMRT was not related to better breast cancer-specific survival compared to those without PMRT both before (P=0.186) and after propensity score matching (P=0.137).

Conclusions

In breast cancer with T1-2N1 disease, PMRT does not appear to improve survival in the era of modern systemic therapy.