Desarrollo de un modelo predictivo para el índice de esfuerzo del cuidador / Developing a predictive model for the caregiver strain index

Fundamentos. La atención al paciente pluripatológico en el hogar es un hecho cada vez más frecuente. El índice de esfuerzo del cuidador es un instrumento en forma de cuestionario que está diseñado para medir la carga percibida en aquellas personas que cuidan a sus familiares. El objetivo fue la construcción de un nomograma diagnóstico de sobrecarga en el cuidador informal mediante el cuestionario del índice de esfuerzo del cuidador con los datos de un modelo predictivo. Métodos. El modelo se confeccionó mediante regresión logística binaria, siendo incluidos los ítems del cuestionario del índice de esfuerzo del cuidador como variables predictoras dicotómicas. La variable dependiente fue la puntuación final obtenida mediante el cuestionario realizándose la categorización referenciada por la bibliografía: valores entre 0 y 6 fueron considerados como no existencia de estrés del cuidador y los iguales o superiores a 7 como existencia de estrés del cuidador. Se utilizó el programa estadístico R versión 3.1.1. Para construir los intervalos de confianza de la curva ROC se utilizaron 2.000 repeticiones bootstrap. Resultados. Sobre una muestra de 67 cuidadores se confeccionó un nomograma diagnóstico, con su gráfica de calibración (índice de Brier escalado = 0,686; R2 de Nagelkerke=0,791) y con la correspondiente curva ROC (área bajo la curva de 0,962). Conclusiones. El modelo predictivo generado mediante regresión logística binaria y su nomograma contienen cuatro variables predictoras (los ítems 1, 4, 5 y 9 del cuestionario). El área bajo la curva ROC (0,96; IC al 95%: 0,994-0,941) muestra un valor alto y discriminativo. La calibración del nomograma también presenta valores altos de bondad de ajuste por lo que estimamos que puede tener utilidad clínica en la consultas de enfermería comunitaria, de gestión de casos, de medicina de familia y de geriatría (AU)

Background. Patient homecare with multiple morbidities is an increasingly common occurrence. The caregiver strain index is tool in the form of questionnaire that is designed to measure the perceived burden of those who care for their families. The aim of this study is to construct a diagnostic nomogram of informal caregiver burden using data from a predictive model. Methods. The model was drawn up using binary logistic regression and the questionnaire items as dichotomous factors. The dependent variable was the final score obtained with the questionnaire but categorised in accordance with that in the literature. Scores between 0 and 6 were labelled as 'no' (no caregiver stress) and at or greater than 7 as 'yes'. The version 3.1.1R statistical software was used. To construct confidence intervals for the ROC curve 2000 boot strap replicates were used. Results. A sample of 67 caregivers was obtained. A diagnosing nomogram was made up with its calibration graph (Brier scaled = 0.686, Nagelkerke R2=0.791), and the corresponding ROC curve (area under the curve=0.962). Findings. The predictive model generated using binary logistic regression and the nomogram contain four items (1, 4, 5 and 9) of the questionnaire. R plotting functions allow a very good solution for validating a model like this. The area under the ROC curve (0.96; 95% CI: 0.994-0.941) achieves a high discriminative value. Calibration also shows high goodness of fit values, suggesting that it may be clinically useful in community nursing and geriatric establishments (AU)

[Developing a predictive model for the caregiver strain index]. / Desarrollo de un modelo predictivo para el índice de esfuerzo del cuidador.

BACKGROUND: Patient homecare with multiple morbidities is an increasingly common occurrence. The caregiver strain index is tool in the form of questionnaire that is designed to measure the perceived burden of those who care for their families. The aim of this study is to construct a diagnostic nomogram of informal caregiver burden using data from a predictive model. METHODS: The model was drawn up using binary logistic regression and the questionnaire items as dichotomous factors. The dependent variable was the final score obtained with the questionnaire but categorised in accordance with that in the literature. Scores between 0 and 6 were labelled as "no" (no caregiver stress) and at or greater than 7 as "yes". The version 3.1.1R statistical software was used. To construct confidence intervals for the ROC curve 2000 boot strap replicates were used. RESULTS: A sample of 67 caregivers was obtained. A diagnosing nomogram was made up with its calibration graph (Brier scaled = 0.686, Nagelkerke R2=0.791), and the corresponding ROC curve (area under the curve=0.962). FINDINGS: The predictive model generated using binary logistic regression and the nomogram contain four items (1, 4, 5 and 9) of the questionnaire. R plotting functions allow a very good solution for validating a model like this. The area under the ROC curve (0.96; 95% CI: 0.994-0.941) achieves a high discriminative value. Calibration also shows high goodness of fit values, suggesting that it may be clinically useful in community nursing and geriatric establishments.

Sobrecarga sentida en personas cuidadoras informales de pacientes pluripatológicos en una zona urbana / Overload in the informal caregivers of patients with multiple comorbidities in an urban area

Objetivos: Determinar el perfil de la persona cuidadora familiar de pacientes pluripatológicos, identificar factores asociados a la sobrecarga sentida y manifestaciones de cansancio. Construir modelos predictivos mediante los ítems del Índice de Esfuerzo del Cuidador (IEC). Método Estudio descriptivo transversal. Población de estudio: personas cuidadoras de pacientes pluripatológicos de centro de salud urbano. Recogida de datos desde la historia clínica y mediante cuestionarios (índice de Barthel, índice de Pfeiffer, IEC). Análisis estadístico mediante medidas de centralización, de dispersión y mediante la construcción de modelos multivariantes con regresión logística binaria (RLB) con los ítems del IEC como predictoras (programa R versión 2.14.0). Resultados Muestra total de 67 personas cuidadoras, con una edad media de 64,69 años (desviación estándar=12,71; mediana 62 años); un 74,6% mujeres, un 35,8% esposas e hijas un 32,8%. El nivel de dependencia de las personas cuidadas fue total/severa del 77,6% y moderada del 12% (Barthel); un 47,8% tenían algún nivel de deterioro cognitivo (Pfeiffer). IEC>7 en el 47,8% de cuidadoras, identificándose en >40% inconvenientes para la vida, restricción de vida social, esfuerzo físico, molestias ante cambios, comportamientos molestos, cambios emocionales personales y familiares, y trastornos del sueño. El ítem n.° 4 del IEC que analiza la restricción social fue el que mostró una mayor significación en el modelo predictivo del estrés del cuidador. El ítem n.° 12 (sobrecarga económica) fue el más significativo en los pacientes con deterioro cognitivo. Conclusiones Las mujeres suelen adoptar el rol de cuidadora a una edad más temprana que los hombres en el entorno urbano estudiado, y con relación al IEC, son los ítems n.° 4 (restricción social) y n.° 12 (sobrecarga económica) los que tienen más significación en los modelos predictivos construidos con RLB (AU)

Objectives: The aim of the study was, to determine the profile of the family caregiver of patients with multiple pathologies, identify factors associated with overload, and construct predictive models using items from the Caregiver Strain Index (CSI). Method: A cross-sectional study of caregivers of patients with multiple comorbidities who attended an urban health centre. Data were collected from health records and questionnaires (Barthel index, Pfeiffer index, and CSI). Statistical analysis was performed using measures of central tendency and dispersion, and by building multivariate models with binary logistic regression with the CSI items as predictors (program R version 2.14.0). Results: The sample included 67 caregivers, with a mean age of 64.69 years (standard deviation = 12.71, median 62 years), of whom 74.6% were women, 35.8% were wives, and 32.8% were daughters. The level of dependence of the patients cared for was total/severe in 77.6%, and moderate in 12% (Barthel), and 47.8% had some level of cognitive impairment (Pfeiffer). A CSI equal or greater than 7 was seen in 47.8% of caregivers, identifying life problems in more than 40% of them such as, restriction of social life, physical exertion, discomfort with change, bad behaviour, personal and family emotional changes, and sleep disturbances. Item 4 of the CSI, analysing the social restriction, was the one that showed a greater significance in the predictive multivariate model. Item 12 (economic burden) was the most signifi can't with age in patients with cognitive impairment. Conclusions: Women tend to take the role of caregiver at an earlier age than men in the urban environment studied, and items from CSI showed that items 4 (social restrictions) and 12 (economic burden) have more significance in the predictive models constructed with Binary Logistic Regression (AU)

[Overload in the informal caregivers of patients with multiple comorbidities in an urban area]. / Sobrecarga sentida en personas cuidadoras informales de pacientes pluripatológicos en una zona urbana.

OBJECTIVES: The aim of the study was, to determine the profile of the family caregiver of patients with multiple pathologies, identify factors associated with overload, and construct predictive models using items from the Caregiver Strain Index (CSI). METHOD: A cross-sectional study of caregivers of patients with multiple comorbidities who attended an urban health centre. Data were collected from health records and questionnaires (Barthel index, Pfeiffer index, and CSI). Statistical analysis was performed using measures of central tendency and dispersion, and by building multivariate models with binary logistic regression with the CSI items as predictors (program R version 2.14.0). RESULTS: The sample included 67 caregivers, with a mean age of 64.69 years (standard deviation=12.71, median 62 years), of whom 74.6% were women, 35.8% were wives, and 32.8% were daughters. The level of dependence of the patients cared for was total/severe in 77.6%, and moderate in 12% (Barthel), and 47.8% had some level of cognitive impairment (Pfeiffer). A CSI equal or greater than 7 was seen in 47.8% of caregivers, identifying life problems in more than 40% of them such as, restriction of social life, physical exertion, discomfort with change, bad behaviour, personal and family emotional changes, and sleep disturbances. Item 4 of the CSI, analysing the social restriction, was the one that showed a greater significance in the predictive multivariate model. Item 12 (economic burden) was the most significant with age in patients with cognitive impairment. CONCLUSIONS: Women tend to take the role of caregiver at an earlier age than men in the urban environment studied, and items from CSI showed that items 4 (social restrictions) and 12 (economic burden) have more significance in the predictive models constructed with Binary Logistic Regression.

[Overload felt by the figure of the main caregiver in a cohort of patients with multiple pathologies]. / Sobrecarga sentida por la figura del cuidador principal en una cohorte de pacientes pluripatológicos.

OBJECTIVE: To determine the profile of the main caregiver (MC) and the factors associated with her/his care burden, in a multi-centre cohort of patients with multiple pathologies (PMP). DESIGN: Multi-centre cross-sectional study. SETTING: Four health districts in the Virgen del Rocío University Hospitals Health Area, Seville, Spain. PARTICIPANTS: The PMP cohort was created by checking all the patients who satisfied the health department criteria for PMP (2002): patients suffering from chronic diseases in 2 or more of the 7 clinical categories defined. MAIN MEASUREMENTS: The profile of PMP caregiver was determined for all patients. The caregiver strain index (CSI) was determined by the index of care stress (ICS). Predictive factors were analysed by the Student t, ANOVA, and Pearson's tests. Multivariate analysis was performed by a forward stepwise linear regression model. RESULTS: The interview was attended by 461 (69%) out of 662 eligible PMP. Of these, 293 (63.6%) had an MC whose mean age was 62 (15) years; 80% of them were women. First-degree relatives made up 88% of caregivers, with spouses 49.7% of them (n=146). In 41.5%, the CSI was >7 points (mean CSI was 5.35 [3.5]). This was higher in those caring for PMP with neurological illnesses (7 [3.2 vs 4.5 [3.3]; P=.0001). The CSI was compared directly with the medical vulnerability of the PMP (R=0.37; P=.001), cognitive deterioration on the Pfeiffer scale (PS) (R=0.4; P=.0001), and inversely with functional status on Barthel's scale (BS) (R=-0.67; P=.0001). Patient's age (P=.03), his/her medical vulnerability (P=.016) and functional (P< .0001) and cognitive (P=.019) deterioration were independently associated with the CSI. CONCLUSIONS: The profile of the MC of the PMP cohort corresponded mainly to first-degree female relatives around sixty years old. The burden of care was high in more than a third of them. Predictive factors were age, medical vulnerability, and the functional and cognitive deterioration of the PMP.

Sobrecarga sentida por la figura del cuidador principal en una cohorte de pacientes pluripatológicos / Overload felt by the figure of the main caregiver in a cohort of patients with multiple pathologies

Objetivo. Conocer el perfil del cuidador principal (CP) y los factores relacionados con la sobrecarga sentida, en una cohorte multicéntrica de pacientes pluripatológicos (PP). Diseño. Estudio transversal, multicéntrico. Emplazamiento. Cuatro zonas básicas de salud del área sanitaria de los Hospitales Universitarios Virgen del Rocío, Sevilla. Participantes. La cohorte de PP se generó prospectivamente mediante el censado de todos los pacientes que cumplían los criterios de PP de la Consejería de Salud (2002): aquellos que tienen enfermedades crónicas de dos o más de las 7 categorías clínicas definidas. Mediciones principales. El perfil del cuidador se determinó a todos los PP. El cansancio del CP se determinó mediante el índice de esfuerzo del cuidador (IEC). Los factores predictores se analizaron mediante los tests de la t de Student, ANOVA y Pearson. Posteriormente se realizó una regresión lineal multivariable paso a paso hacia delante. Resultados. Accedieron a la entrevista 461 (69% de los 662 elegibles) PP; 293 (63,6%) pacientes tenían CP, que en el 88% eran familiares de primer grado (146 [49,7%] de ellos, el cónyuge), de 62 ± 15 años de edad, y el 80%, mujeres. El IEC fue > 7 puntos en el 41,5% y en general fue 5,35 ± 3,5, mayor en los que cuidaban de PP con enfermedades neurológicas (7 ± 3,2 frente a 4,5 ± 3,3; p < 0,0001). El IEC se correlacionó directamente con la vulnerabilidad clínica del PP (R = 0,37; p < 0,001), con el deterioro cognitivo por escala de Pfeiffer (R = 0,4; p < 0,0001), e inversamente con la situación funcional por índice de Barthel (R = ­0,67; p < 0,0001). La edad del paciente (p = 0,03), su vulnerabilidad clínica (p = 0,016) y el deterioro funcional (p < 0,0001) y cognitivo (p = 0,019) predijeron de forma independiente el IEC. Conclusiones. El perfil del CP de los PP se correspondió con mujeres familiares en primer grado de unos 60 años. Más de la tercera parte estaban sobrecargadas; los factores predictores fueron la edad, la vulnerabilidad clínica y el deterioro funcional y cognitivo del PP

Objective. To determine the profile of the main caregiver (MC) and the factors associated with her/his care burden, in a multi-centre cohort of patients with multiple pathologies (PMP). Design. Multi-centre cross-sectional study. Setting. Four health districts in the Virgen del Rocío University Hospitals Health Area, Seville, Spain. Participants. The PMP cohort was created by checking all the patients who satisfied the health department criteria for PMP (2002): patients suffering from chronic diseases in 2 or more of the 7 clinical categories defined. Main measurements. The profile of PMP caregiver was determined for all patients. The caregiver strain index (CSI) was determined by the index of care stress (ICS). Predictive factors were analysed by the Student t, ANOVA, and Pearson's tests. Multivariate analysis was performed by a forward stepwise linear regression model. Results. The interview was attended by 461 (69%) out of 662 eligible PMP. Of these, 293 (63.6%) had an MC whose mean age was 62 (15) years; 80% of them were women. First-degree relatives made up 88% of caregivers, with spouses 49.7% of them (n=146). In 41.5%, the CSI was >7 points (mean CSI was 5.35 [3.5]). This was higher in those caring for PMP with neurological illnesses (7 [3.2 vs 4.5 [3.3]; P=.0001). The CSI was compared directly with the medical vulnerability of the PMP (R=0.37; P=.001), cognitive deterioration on the Pfeiffer scale (PS) (R=0.4; P=.0001), and inversely with functional status on Barthel's scale (BS) (R=­0.67; P=.0001). Patient's age (P=.03), his/her medical vulnerability (P=.016) and functional (P<.0001) and cognitive (P=.019) deterioration were independently associated with the CSI. Conclusions. The profile of the MC of the PMP cohort corresponded mainly to first-degree female relatives around sixty years old. The burden of care was high in more than a third of them. Predictive factors were age, medical vulnerability, and the functional and cognitive deterioration of the PMP