Math versus meaning in MAIHDA: A commentary on multilevel statistical models for quantitative intersectionality.

RATIONALE: Intersectionality has been increasingly adopted as a theoretical framework within quantitative research, raising questions about the congruence between theory and statistical methodology. Which methods best map onto intersectionality theory, with regard to their assumptions and the results they produce? Which methods are best positioned to provide information on health inequalities and direction for their remediation? One method, multilevel analysis of individual heterogeneity and discriminatory accuracy (MAIHDA), has been argued to provide statistical efficiency for high-dimensional intersectional analysis along with valid intersection-specific predictions and tests of interactions. However, the method has not been thoroughly tested in scenarios where ground truth is known. METHOD: We perform a simulation analysis using plausible data generating scenarios where intersectional effects are present. We apply variants of MAIHDA and ordinary least squares regression to each, and we observe how the effects are reflected in the estimates that the methods produce. RESULTS: The first-order fixed effects estimated by MAIHDA can be interpreted neither as effects on mean outcome when interacting variables are set to zero (as in a correctly-specified linear regression model), nor as effects on mean outcome averaged over the individuals in the population (as in a misspecified linear regression model), but rather as effects on mean outcome averaged over an artificial population where all intersections are of equal size. Furthermore, the values of the random effects do not reflect advantage or disadvantage of different intersectional groups. CONCLUSIONS: Because first-order fixed effects estimates are the reference point for interpreting random effects as intersectional effects in MAIHDA analyses, the random effects alone do not provide meaningful estimates of intersectional advantage or disadvantage. Rather, the fixed and random parts of the model must be combined for their estimates to be meaningful. We therefore advise caution when interpreting the results of MAIHDA in quantitative intersectional analyses.

Estimating New Quantities from Longitudinal Test Scores to Improve Forecasts of Future Performance.

Psychometric models for longitudinal test scores typically estimate quantities associated with single-administration tests, like ability at each time-point. However, models for longitudinal tests have not considered opportunities to estimate new quantities that are unavailable from single-administration tests. Specifically, we discuss dynamic measurement models - which combine aspects of longitudinal IRT, nonlinear growth models, and dynamic assessment - to directly estimate capacity, defined as the expected future score once the construct has fully developed. After discussing the history and connecting these areas into a single framework, we apply the model to verbal test scores from the Intergenerational Studies, which follow 494 people from 3 to 72 years old. The goal is to predict adult verbal scores (Age ≥ 34) from adolescent scores (Age ≤ 20). We held-out the adult data for prediction and compared predictions from traditional longitudinal IRT ability scores and proposed dynamic measurement capacity scores from models fit to the adolescent data. Results showed that the R2 from capacity scores were 2.5 times larger than the R2 from longitudinal IRT ability scores (43% vs. 16%), providing some evidence that exploring new quantities available from longitudinal testing could be worthwhile when an interest in testing is forecasting future performance.

Variability in Palliative Care Use after Intracerebral Hemorrhage at US Hospitals: A Multilevel Analysis.

BACKGROUND: Palliative care (PC) is an essential component of comprehensive care of patients with intracerebral hemorrhage (ICH). In the present study, we sought to characterize the variability of PC use after ICH among US hospitals. METHODS: ICH admissions from hospitals with at least 12 annual ICH cases were identified in the Nationwide Inpatient Sample between 2008 and 2011. We used multilevel logistic regression modeling to estimate between-hospital variance in PC use. We calculated the intraclass correlation coefficient (ICC), proportional variance change, and median OR after accounting for individual-level and hospital-level covariates. RESULTS: Among 26,791 ICH admissions, 12.5% received PC (95% CI 11.5-13.5). Among the 629 included hospitals, the median rate of PC use was 9.1 (interquartile range 1.5-19.3) per 100 ICH admissions, and 150 (23.9%) hospitals had no recorded PC use. The ICC of the random intercept (null) model was 0.274, suggesting that 27.4% of the overall variability in PC use was due to between-hospital variability. Adding hospital-level covariates to the model accounted for 25.8% of the between-hospital variance observed in the null model, with 74.2% of between-hospital variance remaining unexplained. The median OR of the fully adjusted model was 2.62 (95% CI 2.41-2.89), indicating that a patient moving from 1 hospital to another with a higher intrinsic propensity of PC use had a 2.63-fold median increase in the odds of receiving PC, independent of patient and hospital factors. CONCLUSIONS: Substantial variation in PC use after ICH exists among US hospitals. A substantial proportion of this between-hospital variability remains unexplained even after accounting for patient and hospital characteristics.

Effective workgroups: the role of diversity and culture / Grupos de trabajo eficaces: el papel de la diversidad y de la cultura

The purpose of this study was to contribute to the clarification of the conditions under which teams can be successful. To attain this goal, the direct and interactive effects of diversity and of the team’s cultural orientation towards learning on team outcomes (team performance, team members’ satisfaction) were analyzed. Data were obtained from a survey among 73 teams from different industrial and service companies, which perform complex and non-routine tasks. In order to test the hypotheses, multilevel analysis and hierarchical regression analysis were conducted. The results show a significant (although marginal) effect of diversity on members’ satisfaction with the team. The team orientation towards learning presented positive effects on both team performance and members’ satisfaction with the team. No interactive effects were identified. Although the positive impact of a learning culture on organizational effectiveness has already been studied and is well established in the literature, this is one of the first studies that provides empirical evidence of the impact of this kind of culture at the team level. At an intervention level, this study points to managers who want to create successful teams that they may be advised to enhance the levels of the team orientation towards learning, creating conditions in the team to promote and support the acquisition of knowledge (AU)

El objetivo de este estudio ha sido contribuir a aclarar las condiciones que hacen que los equipos sean eficaces. A tal fin se analizaron los efectos interactivos directos e indirectos de la diversidad y de la orientación cultural del equipo sobre el desempeño del mismo (rendimiento, satisfacción de sus miembros). Se obtuvieron datos de una encuesta aplicada a 73 equipos de diferentes empresas industriales y de servicios que llevan a cabo tareas complejas y no rutinarias. Para la puesta a prueba de las hipótesis se llevó a cabo un análisis multinivel y regresión jerárquica. Los resultados señalan un efecto significativo (aunque marginal) de la diversidad en la satisfacción de los miembros en el equipo. La orientación de éste hacia el aprendizaje tenía efectos positivos tanto para el desempeño del equipo como para la satisfacción de los miembros con el equipo. No se observaron efectos interactivos. A pesar de que ya se haya estudiado la repercusión positiva de una cultura de aprendizaje en la eficacia de la organización y que esté bien consolidada en las publicaciones, este estudio es uno de los primeros en aportar pruebas empíricas de las consecuencias de este tipo de cultura al nivel de equipo. Al nivel de intervención, el estudio señala que puede aconsejarse a los directivos que deseen crear equipos eficaces que mejoren el nivel de orientación del equipo hacia el aprendizaje, sentando las condiciones en el equipo que favorezcan y respalden la adquisición de conocimientos (AU)

Multilevel bootstrap analysis with assumptions violated / Análisis bootstrap multinivel con supuestos incumplidos

Background: Likelihood-based methods can work poorly when the residuals are not normally distributed and the variances across clusters are heterogeneous. Method: The performance of two estimation methods, the non-parametric residual bootstrap (RB) and the restricted maximum likelihood (REML) for fitting multilevel models are compared through simulation studies in terms of bias, coverage, and precision. Results: We find that (a) both methods produce unbiased estimates of the fixed parameters, but biased estimates of the random parameters, although the REML was more prone to give biased estimates for the variance components; (b) the RB method yields substantial reductions in the difference between nominal and actual confidence interval coverage, compared with the REML method; and (c) for the square root of the mean squared error (RMSE) of the fixed effects, the RB method performed slightly better than the REML method. For the variance components, however, the RB method did not offer a systematic improvement over the REML method in terms of RMSE. Conclusions: It can be stated that the RB method is, in general, superior to the REML method with violated assumptions (AU)

Antecedentes: los métodos basados en la verosimilitud pueden trabajar con dificultad cuando los errores no se distribuyen normalmente y las varianzas a través de los grupos son heterogéneas. Método: el desempeño de dos métodos de estimación, el bootstrap residual (BR) no paramétrico y el de la máxima verosimilitud restringida (MVR), para ajustar modelos multinivel es comparado mediante estudios de simulación en términos de sesgo, cobertura y precisión. Resultados: encontramos que: (a) ambos métodos proporcionan estimaciones no sesgadas de los efectos fijos, pero sesgadas de los efectos aleatorios, aunque el método MVR es más propenso a generar estimaciones sesgadas para los componentes de la varianza; (b) el método BR depara diferencias más pequeñas entre las tasas de cobertura real y nominal de los intervalos de confianza que el método MVR; y (c) los valores de la raíz del error cuadrático medio (RECM) para los efectos fijos son algo más pequeños bajo el método BR que bajo el método REML. Sin embargo, en lo referido a los componentes de la varianza, el método de BR no ofrece una mejora sistemática sobre el método MVR en términos de RECM. Conclusiones: en general, se puede afirmar que el método BR resulta superior al método MVR con supuestos incumplidos (AU)

Identificación de algunas características del médico de atención primaria que influyen en el gasto farmacéutico / Identification of some characteristics of Primary Health Care physicians that influence pharmaceutical costs

Objetivo. Envejecimiento y medicalización de la sociedad suponen un crecimiento del gasto farmacéutico. Interesa desarrollar estudios que evalúen este gasto para dar soporte a políticas para su contención. El objetivo del estudio es determinar elementos clave que influyen en el gasto farmacéutico inducido por médicos de atención primaria del Área de Salud de Cuenca. Material y métodos. Se desarrolla un estudio observacional longitudinal retrospectivo mediante modelos multinivel del gasto farmacéutico realizado por médicos entre 2006-2009, teniendo en cuenta características particulares del médico y del cupo de pacientes. Resultados. El gasto farmacéutico medio médico/persona aumenta 2,97 €/año al incrementar un punto la tasa de derivaciones a especializada, reduciéndose 2,54 €/año al incrementar 1% el porcentaje de genéricos prescritos. Estas variables y el porcentaje de pensionistas del cupo, explican la mayor parte de la variabilidad del gasto farmacéutico. Características particulares del médico como edad, formación, plaza en propiedad, antigüedad, no parecen influir significativamente en dicho gasto. Conclusiones. Este trabajo justifica desarrollar intervenciones para fomentar la prescripción por principio activo, fundamentalmente en médicos con alto porcentaje de pensionistas en su cupo. Además, muestra una relación importante entre primaria y especializada en cuanto al gasto farmacéutico sobre la que habría que desarrollar nuevos estudios (AU)

Objective. Aging and medicalisation are leading to a progressive growth in pharmaceutical expenditure, which is difficult to measure. It is important to develop social studies of this drug spending in order to support the policies of containment. The aim of this study is to identify elements that influence pharmaceutical expenditure of Primary Care physicians in the Health Area of Cuenca (Spain). Material and methods. A retrospective observational study using multilevel models of pharmaceutical expenditures by physicians between 2006 and 2009. It took into account the particular characteristics of each doctor and their patients. Results. The average pharmaceutical expenditure by doctor and patient was € 277.13 year. In addition, a one-point of increase in the rate of referrals to specialised care increases pharmaceutical expenditure by 2.97 € per year and an increase in the percentage of generic drugs prescribed reduced in 2.54 € per year. These two variables and the percentage of retired patients (pensioners) are the most important factors to explain the variability in pharmaceutical expenditure. In contrast, the characteristics of physicians such as education, tenure, seniority, or age does not influence pharmaceutical costs. Conclusions. This study supports the importance of developing interventions in prescription policies. It will be of particular interest to those physicians with higher quotas of elderly patients. It also shows a significant relationship in pharmaceutical expenditures between primary and specialist physicians, which needs to be clarified by future studies(AU)

Utilidad del análisis multinivel en las encuestas de satisfacción para evaluar las unidades clínicas y las especialidades médicas / Usefulness of multilevel analysis of satisfaction questionnaires on clinical units and medical specialties

Objetivos. Valorar la utilidad del análisis multinivel aplicado a las encuestas de satisfacción de los pacientes atendidos en consultas externas para evaluar las unidades clínicas y las especialidades médicas. Métodos. Encuestas telefónicas a 6.922 pacientes atendidos en consultas externas en el Servicio Navarro de Salud para evaluar la calidad percibida de 94 unidades clínicas y 37 especialidades médicas. Se ha realizado un análisis jerárquico con 3 niveles (pacientes, unidades clínicas y especialidades médicas). Resultados. El paciente explicó entre el 91,7 y el 99,3% de la variabilidad de las puntuaciones obtenidas en las encuestas. La percepción de la calidad de servicio también depende de la unidad clínica, ya que en 18 de los 20 ítems analizados, la varianza explicada por la unidad clínica fue estadísticamente significativa y fluctuó entre el 0,9 y el 8,3%. La especialidad médica consiguió explicar un 4,2% de la variabilidad en un solo ítem. Los pacientes menores de 65 años, los atendidos en primera consulta y los que acudieron a hospitales terciarios otorgaron puntuaciones más bajas. Conclusiones. El análisis multinivel es un instrumento útil para analizar y comparar los resultados de las encuestas de satisfacción, ya que permite discriminar de forma adecuada las distintas unidades clínicas de un servicio de salud, elemento esencial para la gestión de las mencionadas unidades. Además, permite conocer el porcentaje de variabilidad atribuida a cada nivel y qué variables son explicativas de la calidad de servicio obtenida(AU)

Objectives. Evaluate the usefulness of multilevel analysis applied to satisfaction surveys to assess the ambulatory care offered by the Clinical Units and Medical Specialties. Methods. Telephone surveys were conducted on 6,922 patients seen in outpatient visits in the Navarra Health System in order to assess 94 Clinical Units and 37 Medical Specialties. A hierarchical analysis was then performed at 3 levels (patients, Clinical Units and Medical Specialties). Results. The patient explained between 91.7% and 99.3% of the variability in the satisfaction survey scores. The patient perception of service quality also depends of the clinical unit, in 18 out of 20 items analysed, the variance related to the Clinical Unit was statistically significant, and ranged from 0.9% and 8.3%. The Medical Specialty explained 4.2% of the variability in only one item. Patients aged under 65 years, patients seen in first visit, and those who were seen at tertiary hospitals gave the lowest scores. Conclusions. Multilevel analysis is a useful tool to analyse and compare satisfaction survey results, as it can correctly distinguish the different Clinical Units of a health service, and this is a key element in the management of these Units. Moreover, it gives the percentage of variability due to each level and which variables predict the quality of service obtained(AU)