Measuring population health using health expectancy estimates from morbidity and mortality databases.

The progressive incorporation of quality of life indicators in health planning meets a critical need: The evaluation of the performance of health services, which are under stress by multiple causes, but in particular by an ageing population. In general, national health plans rely on health expectancies obtained using the Sullivan method. The Sullivan health expectancy index combines age-specific mortality rates and age-specific prevalence of healthy life, obtained from health surveys. The objective of this work is to investigate an equivalent estimation, using available information from morbidity and mortality datasets. Mortality and morbidity information, corresponding to years 2016 and 2017, was obtained for the population of the county of Baix Empordà (Catalonia), N = 91,130. Anonymized individual information on diagnoses, procedures and pharmacy consumption contained in the individual clinical record (ICD and ATC codes), were classified into health states. Based on the observed health transitions and mortality, life expectancies by health state were obtained from a multistate microsimulation model. Healthy life expectancies at birth and 65 years for females and males were respectively HLE0female = 39.94, HLE0male = 42.87, HLE65female = 2.43, HLE65male = 2.17. These results differed considerably from the Sullivan equivalents, e.g., 8.25 years less for HLE65female, 9.26 less for HLE65male. Point estimates for global life expectancies at birth and 65 years of age: LE0female = 85.82, LE0male = 80.58, LE65female = 22.31, LE65male = 18.86. Health indicators can be efficiently obtained from multistate models based on mortality and morbidity information, without the use of health surveys. This alternative method could be used for monitoring populations in the context of health planning. Life Expectancy results were consistent with the standard government reports. Due to the different approximation to the concept of health (data-based versus self-perception), healthy life expectancies obtained from multistate micro simulation are consistently lower than those calculated with the standard Sullivan method.

Ageing and healthcare expenditures: Exploring the role of individual health status.

In 1999, Zweifel, Felder, and Meiers questioned conventional wisdom on ageing and healthcare expenditure (HCE). According to these authors, the positive association between age and HCE is due to an increasing age-specific mortality and the high cost of dying. After a weighty academic debate, a new consensus was reached on the importance of proximity to death when analysing HCE. Nevertheless, the influence of individual health status remains unknown. The objective of our study is to analyse the influence individual health status has on HCE, when compared to proximity to death and demographic effects and considering a comprehensive view of healthcare services and costs. We examined data concerning different HCE components of N = 61,473 persons aged 30 to 95 years old. Using 2-part models, we analysed the probability of use and positive HCE. Regardless of the specific group of healthcare services, HCE at the end of life depends mainly on the individual health status. Proximity to death approximates individual morbidity when it is excluded from the model. The inclusion of morbidity generally improves the goodness of fit. These results provide implications for the analysis of ageing population and its impact on HCE that should be taken into account.

Evaluación de la efectividad de un programa de atención integrada y proactiva a pacientes crónicos complejos / Assessment of the effectiveness of a proactive and integrated healthcare programme for chronic complex patients

Objetivo: Evaluar la efectividad de un programa de atención integrada y proactiva para adecuar el uso de recursos sanitarios en pacientes crónicos complejos con riesgo de alto consumo según un modelo predictivo basado en morbilidad y utilización previa. Métodos: Ensayo clínico controlado aleatorizado con grupo control enmascarado, grupo de intervención parcial informado en la historia clínica y grupo de intervención total informado además a atención primaria, en una organización sanitaria integrada con 128.281 residentes en 2011. Variables dependientes: visitas de atención primaria, urgencias hospitalarias, hospitalización, coste farmacéutico y muerte. Variables independientes: grupo de intervención, edad, sexo, área básica, morbilidad (según grupo de riesgo clínico) y recurrencia como paciente crónico complejo. Análisis bivariado con ANOVA y t de Student, y multivariado mediante regresión logística y regresión lineal múltiple, con un nivel de confianza del 95%. Resultados: Se incluyeron 4.236 y 4.223 pacientes crónicos complejos en el primer y el segundo año de intervención, respectivamente. El 72% eran recurrentes. Edad media: 73,2 años. El 54,2% eran mujeres. Más del 70% tenían al menos dos enfermedades crónicas. El número de visitas a atención primaria fue significativamente mayor en el grupo de intervención total respecto al grupo de intervención parcial y el grupo control. La intervención solo tuvo un efecto significativo independiente en las estancias hospitalarias, que fueron menos en el grupo de intervención parcial. Este efecto diferencial se dio en el primer año y en los pacientes crónicos complejos nuevos del segundo año. Los indicadores asistenciales generales de la organización sanitaria integrada eran buenos, antes y durante la intervención. Conclusiones: Una buena situación general previa y mantenida, y una inevitable contaminación entre grupos, dificultaron la demostración de efectividad marginal del programa (AU)

Objective: To assess the effectiveness of a proactive and integrated care programme to adjust the use of health resources by chronic complex patients (CCP) identified as potential high consumers according to a predictive model based on prior use and morbidity. Methods: Randomized controlled clinical trial with three parallel groups of CCP: a blinded control group (GC), usual care; a partial intervention group (GIP) reported in the EMR; a total intervention group (GIT), also reported to primary care (PC). Conducted in an integrated health care organization (IHCO), N=128,281 individuals in 2011. Dependent variables: PC visits, emergency attention, hospitalizations, pharmaceutical cost and death. Independent variables: intervention group, age, sex, area of residence, morbidity (by clinical risk group) and recurrence as CCP. Statistical analysis: ANOVA, student's t test; logistic and multiple linear regressions at the 95% confidence level. Results: 4,236 CCP included for the first intervention year and 4,223 for the second; recurrence as CCP 72%. Mean age 73.2 years, 54.2% women and over 70% with 2 or more chronic diseases. The number of PC visits was significantly higher for GIT than for GIP and GC. The hospital stays were significantly lower in GIP. This effect was observed in the first year and in the second year only in the new CCP. The general indicators of the IHCO were good, before and during the intervention. Conclusions: A high standard of quality, previous and during the study, and the inevitable contamination between groups, hindered the assessment of the marginal effectiveness of the program (AU)

[Assessment of the effectiveness of a proactive and integrated healthcare programme for chronic complex patients]. / Evaluación de la efectividad de un programa de atención integrada y proactiva a pacientes crónicos complejos.

OBJECTIVE: To assess the effectiveness of a proactive and integrated care programme to adjust the use of health resources by chronic complex patients (CCP) identified as potential high consumers according to a predictive model based on prior use and morbidity. METHODS: Randomized controlled clinical trial with three parallel groups of CCP: a blinded control group (GC), usual care; a partial intervention group (GIP) reported in the EMR; a total intervention group (GIT), also reported to primary care (PC). Conducted in an integrated health care organization (IHCO), N=128,281 individuals in 2011. Dependent variables: PC visits, emergency attention, hospitalizations, pharmaceutical cost and death. INDEPENDENT VARIABLES: intervention group, age, sex, area of residence, morbidity (by clinical risk group) and recurrence as CCP. STATISTICAL ANALYSIS: ANOVA, student's t test; logistic and multiple linear regressions at the 95% confidence level. RESULTS: 4,236 CCP included for the first intervention year and 4,223 for the second; recurrence as CCP 72%. Mean age 73.2 years, 54.2% women and over 70% with 2 or more chronic diseases. The number of PC visits was significantly higher for GIT than for GIP and GC. The hospital stays were significantly lower in GIP. This effect was observed in the first year and in the second year only in the new CCP. The general indicators of the IHCO were good, before and during the intervention. CONCLUSIONS: A high standard of quality, previous and during the study, and the inevitable contamination between groups, hindered the assessment of the marginal effectiveness of the program.

Analysing the Costs of Integrated Care: A Case on Model Selection for Chronic Care Purposes.

BACKGROUND: The objective of this study is to investigate whether the algorithm proposed by Manning and Mullahy, a consolidated health economics procedure, can also be used to estimate individual costs for different groups of healthcare services in the context of integrated care. METHODS: A cross-sectional study focused on the population of the Baix Empordà (Catalonia-Spain) for the year 2012 (N = 92,498 individuals). A set of individual cost models as a function of sex, age and morbidity burden were adjusted and individual healthcare costs were calculated using a retrospective full-costing system. The individual morbidity burden was inferred using the Clinical Risk Groups (CRG) patient classification system. RESULTS: Depending on the characteristics of the data, and according to the algorithm criteria, the choice of model was a linear model on the log of costs or a generalized linear model with a log link. We checked for goodness of fit, accuracy, linear structure and heteroscedasticity for the models obtained. CONCLUSION: The proposed algorithm identified a set of suitable cost models for the distinct groups of services integrated care entails. The individual morbidity burden was found to be indispensable when allocating appropriate resources to targeted individuals.

Predicción del riesgo individual de alto coste sanitario para la identificación de pacientes crónicos complejos / Predicting individual risk of high healthcare cost to identify complex chronic patients

Objetivo: Construir y validar un modelo predictivo del riesgo de consumo de recursos sanitarios elevado, y evaluar su capacidad para identificar pacientes crónicos complejos. Métodos: Estudio transversal realizado en una organización sanitaria integrada sobre datos individuales de residentes 2 años consecutivos (88.795 personas). Variable dependiente: coste sanitario real superior al percentil 95 (P95), incluyendo todos los servicios de la organización sanitaria integrada y las recetas de farmacia. Variables predictoras: edad, sexo, morbilidad (según los clinical risk groups [CRG]) y datos seleccionados de utilización previa (uso de hospitalización, uso de medicación hospitalaria ambulatoria, gasto en recetas de farmacia). Análisis univariado descriptivo. Construcción de un modelo de regresión logística con nivel de confianza del 95%; análisis de validez mediante sensibilidad, especificidad, valor predictivo positivo (VPP) y área bajo la curva ROC (AUC). Resultados: Las personas con coste >P95 acumulan el 44% del coste sanitario total y se concentran en las categorías ACRG3 (aggregated CRG level 3) de enfermedades crónicas múltiples o graves. La carga de morbilidad aumenta con la edad. En el modelo, todas las variables fueron estadísticamente significativas excepto el sexo. Se obtuvo una sensibilidad del 48,4% (intervalo de confianza [IC]: 46,9%-49,8%), una especificidad del 97,2% (IC: 97,0%-97,3%), un VPP del 46,5% (IC: 45,0%-47,9%) y un AUC de 0,897 (IC: 0,892-0,902). Conclusiones: El consumo sanitario elevado se relaciona con la morbilidad crónica compleja. Un modelo basado en la edad, la morbilidad y la utilización previa es válido para predecir el riesgo de alto consumo, y así identificar la población diana de estrategias de atención proactiva para pacientes crónicos complejos (AU)

Objective: To develop a predictive model for the risk of high consumption of healthcare resources, and assess the ability of the model to identify complex chronic patients. Methods: A cross-sectional study was performed within a healthcare management organization by using individual data from 2 consecutive years (88,795 people). The dependent variable consisted of healthcare costs above the 95th percentile (P95), including all services provided by the organization and pharmaceutical consumption outside of the institution. The predictive variables were age, sex, morbidity-based on clinical risk groups (CRG)-and selected data from previous utilization (use of hospitalization, use of high-cost drugs in ambulatory care, pharmaceutical expenditure). A univariate descriptive analysis was performed. We constructed a logistic regression model with a 95% confidence level and analyzed sensitivity, specificity, positive predictive values (PPV), and the area under the ROC curve (AUC). Results: Individuals incurring costs >P95 accumulated 44% of total healthcare costs and were concentrated in ACRG3 (aggregated CRG level 3) categories related to multiple chronic diseases. All variables were statistically significant except for sex. The model had a sensitivity of 48.4% (CI: 46.9%-49.8%), specificity of 97.2% (CI: 97.0%-97.3%), PPV of 46.5% (CI: 45.0%-47.9%), and an AUC of 0.897 (CI: 0.892 to 0.902). Conclusions: High consumption of healthcare resources is associated with complex chronic morbidity. A model based on age, morbidity, and prior utilization is able to predict high-cost risk and identify a target population requiring proactive care (AU)

[Predicting individual risk of high healthcare cost to identify complex chronic patients]. / Predicción del riesgo individual de alto coste sanitario para la identificación de pacientes crónicos complejos.

OBJECTIVE: To develop a predictive model for the risk of high consumption of healthcare resources, and assess the ability of the model to identify complex chronic patients. METHODS: A cross-sectional study was performed within a healthcare management organization by using individual data from 2 consecutive years (88,795 people). The dependent variable consisted of healthcare costs above the 95th percentile (P95), including all services provided by the organization and pharmaceutical consumption outside of the institution. The predictive variables were age, sex, morbidity-based on clinical risk groups (CRG)-and selected data from previous utilization (use of hospitalization, use of high-cost drugs in ambulatory care, pharmaceutical expenditure). A univariate descriptive analysis was performed. We constructed a logistic regression model with a 95% confidence level and analyzed sensitivity, specificity, positive predictive values (PPV), and the area under the ROC curve (AUC). RESULTS: Individuals incurring costs >P95 accumulated 44% of total healthcare costs and were concentrated in ACRG3 (aggregated CRG level 3) categories related to multiple chronic diseases. All variables were statistically significant except for sex. The model had a sensitivity of 48.4% (CI: 46.9%-49.8%), specificity of 97.2% (CI: 97.0%-97.3%), PPV of 46.5% (CI: 45.0%-47.9%), and an AUC of 0.897 (CI: 0.892 to 0.902). CONCLUSIONS: High consumption of healthcare resources is associated with complex chronic morbidity. A model based on age, morbidity, and prior utilization is able to predict high-cost risk and identify a target population requiring proactive care.

Estimating lifetime healthcare costs with morbidity data.

BACKGROUND: In many developed countries, the economic crisis started in 2008 producing a serious contraction of the financial resources spent on healthcare. Identifying which individuals will require more resources and the moment in their lives these resources have to be allocated becomes essential. It is well known that a small number of individuals with complex healthcare needs consume a high percentage of health expenditures. Conversely, little is known on how morbidity evolves throughout life. The aim of this study is to introduce a longitudinal perspective to chronic disease management. METHODS: Data used relate to the population of the county of Baix Empordà in Catalonia for the period 2004-2007 (average population was N = 88,858). The database included individual information on morbidity, resource consumption, costs and activity records. The population was classified using the Clinical Risk Groups (CRG) model. Future morbidity evolution was simulated under different assumptions using a stationary Markov chain. We obtained morbidity patterns for the lifetime and the distribution function of the random variable lifetime costs. Individual information on acute episodes, chronic conditions and multimorbidity patterns were included in the model. RESULTS: The probability of having a specific health status in the future (healthy, acute process or different combinations of chronic illness) and the distribution function of healthcare costs for the individual lifetime were obtained for the sample population. The mean lifetime cost for women was €111,936, a third higher than for men, at €81,566 (all amounts calculated in 2007 Euros). Healthy life expectancy at birth for females was 46.99, lower than for males (50.22). Females also spent 28.41 years of life suffering from some type of chronic disease, a longer period than men (21.9). CONCLUSIONS: Future morbidity and whole population costs can be reasonably predicted, combining stochastic microsimulation with a morbidity classification system. Potential ways of efficiency arose by introducing a time perspective to chronic disease management.

[Measurement of morbidity attended in an integrated health care organization]. / La medida de la morbilidad atendida en una organización sanitaria integrada.

INTRODUCTION: Understanding the quality, costs and outcomes of healthcare services requires precise determination of the morbidity in a population. Measurement of morbidity in a population and its association with the services provided remains to be performed. The aim of this article was to present our experience of using clinical risk groups (CRGs) to measure morbidity in an integrated healthcare organization (IHO). METHODS: We studied the population attended by an IHO in a county (approximately 120,000 patients) from 2002 to 2005. CRGs were used to measure morbidity. A descriptive analysis was performed of the population's distribution in CRG categories and utilization rates. RESULTS: One or more chronic diseases was found in 15.5% of the population, significant acute illness was found in 9%, minor chronic diseases was found in 7% and very severe diseases was found in 0.5%. Between 2002 and 2005, the number of individuals with chronic disease increased by 8%. The burden of illness increased with age. However, at all ages, at least 40% of the population remained healthy. Comorbidity in chronic illnesses was a crucial factor in explaining healthcare resource utilization. CONCLUSIONS: The CRG grouping system aids analysis at different levels for clinical administration. Due to its composition, this system allows better understanding of the use, costs and quality of the set of services received by a population.

La medida de la morbilidad atendida en una organización sanitaria integrada / Measurement of morbidity attended in an integrated health care organization

Introducción: La comprensión de la calidad, los costes y los resultados de los servicios de salud obliga a conocer con precisión la morbilidad de la población. La medida de la morbilidad atendida en una población y su relación con los servicios recibidos es una tarea pendiente. El objetivo de este artículo es presentar la experiencia de utilización de los grupos de riesgo clínico (clinical risk groups [CRG]) como sistema de medida de la morbilidad atendida en una organización sanitaria integrada (OSI). Métodos: Se estudia la población de una comarca (unas 120.000 personas) atendida por una OSI durante los años 2002¿2005. Se utilizan los CRG como sistema de medida de la morbilidad poblacional. Se efectúa un análisis descriptivo de las diferentes posibilidades de utilización que ofrece este sistema. Resultados: El 15,5% de la población presenta una o más enfermedades crónicas dominantes, un 9% presenta enfermedades agudas significativas, un 7% presenta enfermedades crónicas menores y un 0,5% enfermedades muy graves. Entre 2002 y 2005, en un 8% de la población se identifica la aparición de alguna enfermedad crónica. La carga de enfermedad se incrementa con la edad, pero en cualquier grupo de edad al menos un 40% de las personas permanecen sanas. La comorbilidad en enfermedades crónicas es un factor determinante en la explicación del consumo de recursos sanitarios. Conclusiones: Los CRG son una herramienta que facilita el análisis a diferentes niveles para la gestión clínica y, por su configuración, permiten una mejor comprensión de la utilización, los costes y la calidad del conjunto de servicios recibidos por una población(AU)

Introduction: Understanding the quality, costs and outcomes of healthcare services requires precise determination of the morbidity in a population. Measurement of morbidity in a population and its association with the services provided remains to be performed. The aim of this article was to present our experience of using clinical risk groups (CRGs) to measure morbidity in an integrated healthcare organization (IHO). Methods: We studied the population attended by an IHO in a county (approximately 120,000 patients) from 2002 to 2005. CRGs were used to measure morbidity. A descriptive analysis was performed of the population's distribution in CRG categories and utilization rates. Results: One or more chronic diseases was found in 15.5% of the population, significant acute illness was found in 9%, minor chronic diseases was found in 7% and very severe diseases was found in 0.5%. Between 2002 and 2005, the number of individuals with chronic disease increased by 8%. The burden of illness increased with age. However, at all ages, at least 40% of the population remained healthy. Comorbidity in chronic illnesses was a crucial factor in explaining healthcare resource utilization. Conclusions: The CRG grouping system aids analysis at different levels for clinical administration. Due to its composition, this system allows better understanding of the use, costs and quality of the set of services received by a population(AU)

Hybrid risk adjustment for pharmaceutical benefits.

This paper analyses the application of hybrid risk adjustment versus either prospective or concurrent risk adjustment formulae in the context of funding pharmaceutical benefits for the population of an integrated healthcare delivery organisation in Catalonia during years 2002 and 2003. We apply a mixed formula and find that, compared to prospective only models, a hybrid risk adjustment model increases incentives for efficiency in the provision for low risk individuals in health organisations, not only as a whole but also within each internal department, by reducing within-group variation of drug expenditures.

Predictability of drug expenditures: an application using morbidity data.

The growth of pharmaceutical expenditure and its prediction is a major concern for policymakers and healthcare managers. This paper explores different predictive models to estimate future drug expenses, using demographic and morbidity individual information from an integrated healthcare delivery organization in Catalonia for years 2002 and 2003. The morbidity information consists of codified health encounters grouped through the Clinical Risk Groups (CRGs). We estimate pharmaceutical costs using several model specifications, and CRGs as risk adjusters, providing an alternative way of obtaining high predictive power comparable to other estimations of drug expenditures in the literature. These results have clear implications for the use of risk adjustment and CRGs in setting the premiums for pharmaceutical benefits.

Políticas públicas y regulación de los seguros sanitarios / Public policies and regulation of private health insurance

No disponible