How Expensive Are Hospitalizations by COVID-19? Evidence From Colombia.

OBJECTIVES: This study aimed to estimate the direct medical costs due to hospitalizations by COVID-19 in Colombia and to identify their cost drivers in Colombia. METHODS: This is a retrospective cost-of-illness study of COVID-19 in Colombia. We estimated direct medical costs using data from patients insured to a Benefit Plan Administrator Company, between March 15, 2020 and May 29, 2020. Absolute and relative frequencies, averages, medians, and interquartile ranges (IQRs) were used to characterize the population and estimate the costs of hospitalized patients with COVID-19. We stratified the cost analysis by sex, age groups, comorbidities, and type of hospitalization (general ward and intensive care unit [ICU]). Cost drivers were calculated from a generalized linear model. RESULTS: We studied 113 confirmed patients, 51.3% men. On average, the hospital length of stay was 7.3 (± 6.2) days. A person hospitalized with COVID-19 reported median costs of $1688 (IQR 788-2523). In women, this cost was $1328 (IQR 463-2098); in men, this was 1.4 times greater. The median cost for ICU was $4118 (IQR 2069-5455), 3 times higher than those hospitalized only in the general ward. Admission to the ICU, having 1 comorbidity, length of stay, high blood pressure, having 5 comorbidities, and being treated in the city of Cartagena were statistically significant with direct medical costs. CONCLUSIONS: Our study provides an idea of the magnitude of costs needed to hospitalize a COVID-19 case in Colombia. Other studies in Colombia have assessed the costs of hospitalization for infectious diseases such as influenza, costs significantly lower than those described here.

Determinantes de estancia prolongada de neonatos en una unidad de cuidados intensivos / Determinants of Prolonged Stay of Neonates in Intensive Care Unit / Determinantes de estadia prolongada de neonatos em uma unidade de cuidados intensivos

Resumen Objetivo : determinar los factores asociados con estancias prolongadas en uci neonatal. Materiales y métodos : estudio de tipo retrospectivo, de corte transversal, descriptivo y analítico a partir de los Registros Individuales de Prestación de Servicios (RIPS) y la facturación de una Empresa Prestadora de Servicios de Salud (EPS) de Colombia. Se estimó un modelo logístico binomial tomando como variable dependiente estancias prolongadas. Resultados : la mediana de la duración en estancia en uci de los 947 neonatos incluidos en el análisis fue de 3 días y una estancia promedio de 4.8 días (amplitud intercuartílica de 1-5 días). Respecto a los determinantes, la edad gestacional, el peso al nacer y la edad de la madre mantienen una relación inversa con la probabilidad de generar días estancias, residir en zonas urbanas y contar con un buen control prenatal se convierten en factores protectores. Conclusión : el control prenatal es una intervención eficiente y efectiva para la gestión del riesgo de la salud, así como disminuir los embarazos en las mujeres jóvenes (<18 años) y mayores (>35 años) cumplirían un papel fundamental en la reducción de estancias prolongadas en uci neonatal.

Abstract Objective : To determine the factors associated with prolonged stays in neonatal icu. Materials and Methods : This is a retrospective, cross-sectional, descriptive and analytical study based on the Individual Service Delivery Registries (RIPS) and the billing of a Health Services Provider Company (EPS) of Colombia. A binomial logistic model was estimated using prolonged stays as a dependent variable. Results : The median length of stay in the icu of the 947 neonates included in the analysis was three days with an average sojourn of 4.8 days (interquartile range of 1-5 days). Regarding the determinants: gestational age, birth weight and age of the mother, they have an inverse relationship with the probability of ge nerating stays, while living in urban areas and having accurate prenatal control become protective factors. Conclusion : Prenatal control is an efficient and effective intervention for managing health risk, also reducing pregnancies in young (<18 years) and older women (> 35 years) would play a key role in decreasing prolonged stays in neonatal icu.

Resumo Objetivo : determinar os fatores associados a estadias prolongadas na uci neonatal. Materiais e métodos : estudo de tipo retrospectivo, de corte transversal, descritivo e analítico a partir dos Registros Individuais de prestação de serviços (RIPS) e a faturação de uma Empresa Prestadora de Serviços de Saúde (EPS) da Colômbia. Se estimou um modelo logístico binomial tomando como variável dependente estadias pro longadas. Resultados : a mediana da duração em estadias na uci dos 947 neonatos incluídos na análise foi de 3 dias e uma estadia média de 4.8 dias (amplitude interquartílica de 1-5 dias). Respeito aos deter minantes, a idade gestacional, o peso ao nascer e a idade da mãe mantêm uma relação inversa com a probabilidade de gerar dias estadias, residir em zonas urbanas e contar com um bom controle pré-natal tornam-se em fatores protetores. Conclusão : o controle pré-natal é uma intervenção eficiente e efetiva para a gestão do risco da saúde, assim como diminuir as gravidezes nas mulheres jovens (<18 anos) e maiores (>35 anos) teriam um papel fundamental na redução de estadias prolongadas na uci neonatal.

Concentración del gasto sanitario en una aseguradora colombiana del régimen subsidiado, 2014 / Concentration of Health Spending in a Subsidized-scheme Colombian Insurance Company, 2014

Resumen Objetivo: Estimar la concentración del gasto sanitario según grupos de patologías y tipos de servicios en una aseguradora subsidiada que afilia personas pobres (estratos socioeconómicos 1 y 2) al Sistema de Salud de Colombia en 2014. Métodos: Se analizó el gasto sanitario en 1 666 477 afiliados, por grupos de patología y tipos de servicios, en el año 2014. Para la descripción de la concentración del gasto se utilizó el coeficiente de Gini y la curva de Lorenz. Resultados: El gasto sanitario de la aseguradora fue de usd 418 millones. De este, el 81 % se aplicó a servicios contenidos en el Plan Obligatorio de Salud (pos). Los grupos de patologías que concentraron el 43,4 % del gasto sanitario fueron cardiovascular (14,3 %), cáncer (7,8 %), enfermedades respiratorias (7,3 %), enfermedades urinarias (7 %) y traumatismos (6,9 %). Los servicios diagnósticos, curativos y de rehabilitación representaron el 77,8 % del gasto sanitario. La hospitalización fue el grupo de servicio que más impactó el costo (47 %), y las consultas, el más utilizado. Las curvas de Lorenz demostraron que el 70 % del gasto en salud se concentra en aproximadamente el 20 % de los afiliados, generando un coeficiente de Gini de 0,58. Conclusión: El diagnóstico y el tratamiento de las enfermedades crónicas no transmisibles concentran una alta proporción del gasto sanitario, lo cual genera una competencia de recursos para servicios preventivos y de promoción de la salud.

Abstract Objective: to estimate the concentration of health spending depending on pathology groups and types of services in a subsidized insurance company which enrolled low-income people (social economic strata 1 and 2) in the Colombian Healthcare System in 2014. Methodology: Health spending was analyzed in 1 666 477 members, set up by pathologies and types of services during 2014. To describe the concentration of health spending, researchers used the Gini coefficient and the Lorenz curve. Results: the health spending of the insurance company was US$418 million. Out of this, 81 % was used in service contained in the mandatory Health Plan (in Spanish, Plan Obligatorio de Salud -POS). The pathology groups that concentrated 43.4% of health spending were cardiovascular (14.3%), cancer (7.8%), respiratory diseases (7.3%), urinary diseases (7%) and trauma (6.9%). Diagnostic, healing and rehabilitation services represented 77.8% of health spending. Hospitalization was the service group with the highest impact on costs (47%), and consultations, the most used. The Lorenz curves showed that 70% of the health spending is concentrated in approximately 20% the Affiliated people, resulting in a 0.58 Gini coefficient. Conclusion: Diagnostic and treatment of chronic non-transmittable diseases concentrate a vast part of health spending, which produces a competition of resources for preventive services and healthcare promotion.

Health care costs and resource utilization for different asthma severity stages in Colombia: a claims data analysis.

BACKGROUND: Asthma is one of the most common chronic respiratory conditions worldwide. Asthma-related economic burden has been reported in Latin America, but knowledge about its economic impact to the Colombian health care system and the influence of disease severity is lacking. This study estimated direct medical costs and health care resource utilization (HCRU) in patients with asthma according to severity in Colombia. METHODS: This study identified all-age patients who had at least one medical event linked to an asthma diagnosis (CIE-10: J45-J46) between 2004 and 2014. Patients were selected if they had a continuous enrollment and uninterrupted insurance coverage between January 1-2015 and December 31-2015 and were categorized into 4 different severity levels using a modified algorithm based on Leidy criteria. Healthcare utilization and costs were estimated in a 1-year period after the identification period. A Generalized Linear Model (GLM) with gamma distribution and log link was used to analyze costs adjusting for patient demographics. RESULTS: A total of 20,410 patients were included: 69.5% had mild intermittent, 18.0% mild persistent, 6.9% moderate persistent and 5.5% severe persistent asthma; with mean costs (SD) of $67 (134), $482 (1506), $1061 (1983), $2235 (3426) respectively (p < 0.001). The mean total direct cost was estimated at $331 (1278) per patient. Medication and hospitalization had the higher proportion in total costs (46% and 31% respectively). General physician visits was the most used service (57.2%) and short-acting ß-2 agonists the most used medication (24%). CONCLUSIONS: Health services utilization and direct costs of asthma were highly related to disease severity. Nationwide health policies aimed at the effective control of asthma are necessary and would play an important role in reducing the associated economic impact.

Gasto de bolsillo y gasto catastrófico en salud en los hogares de Cartagena, Colombia / Out-of-pocket and catastrophic health expenditure in households of Cartagena, Colombia

RESUMEN Objetivo Estimar el gasto de bolsillo y la probabilidad de gasto catastrófico de los hogares y sus determinantes socioeconómicos en Cartagena, Colombia. Materiales y Métodos Estudio transversal en una muestra poblacional estratificada aleatoria de hogares de Cartagena. Se estimaron dos modelos de regresión cuyas variables dependientes fueron gasto de bolsillo y probabilidad de gasto catastrófico en salud de los hogares. Resultados El gasto de bolsillo promedio anual en hogares pobres fue 1 566 036 COP (US$783) (IC95% 1 117 597-2 014 475); en hogares de estrato medio 2 492 928 COP (US$1246) (IC95% 1 695 845-3 290 011) y en hogares ricos 4 577 172 COP (US$2 288) (IC95% 1 838 222-7 316 122). Como proporción del ingreso, el gasto de bolsillo en salud fue de 14,6% en los hogares pobres, de 8,2% en los hogares de estrato medio y de 7,0% en los hogares ricos. La probabilidad de gasto catastrófico en salud de los hogares pobres fue 30,6% (IC95% 25,6-35,5%), de los de estrato medio del 10,2% (IC95% 4,5-15,9%) y de los hogares de estrato alto del 8,6% (IC95% 1,823,0%). El estrato socioeconómico, la educación y la ocupación fueron los principales determinantes del gasto de bolsillo en salud y de la probabilidad de incurrir en gasto catastrófico en salud. Conclusiones En el sistema de salud persisten desigualdades en la protección financiera de los hogares contra el gasto de bolsillo y la probabilidad de gasto catastrófico. El presente estudio genera evidencia para revisar la política de protección social de los hogares socioeconómicamente más vulnerables.(AU)

ABSTRACT Objective To estimate out-of-pocket (OOP) health expenditure and the probability to incur in catastrophic health expenditure, as well as the socio-demographic determinants of households in Cartagena, Colombia. Materials and Methods Population-based cross-sectional study on a randomized stratified sample of Cartagena households. Two regression models were developed using OOP health expenditure and the probability to incur in catastrophic health expenditure as dependent variables. Results The average annual OOP health expenditure was $1 566 036 COP (US$783) (95%CI: $1 117 597 - $2 014 475) in poor households, $2 492 928 COP (US$1 246) (95%CI: $1 695 845 - $3 290 011) in middle class households, and $4 577 172 COP (uS$2 288) (95%CI: $1 838 222 - $7 316 122) in upper class households. Regarding the household income ratio, the OOP health expenditure was 14.6% in poor households, 8.2% in middle class, and 7.0% in upper class households. The probability to incur in catastrophic health expenditure was 30.6% (95%CI: 25.6% - 35.5%), 10.2% (95%CI: 4.5%-15.9%) and 8.6% (95%CI: 1.8%-23%) in the low, middle- and high-class households, respectively. Educational attainment, socioeconomic strata and employment were the main determinants of OOP and the probability to incur in catastrophic health expenditure. Conclusions The health system has persistent inequalities regarding the financial protection of households related to out-of-pocket expenditure and the probability of catastrophic expenditure. This study provides evidence to review the social protection policy for the most socio-economically vulnerable households.(AU)

[Out-of-pocket and catastrophic health expenditure in households of Cartagena, Colombia]. / Gasto de bolsillo y gasto catastrófico en salud en los hogares de Cartagena, Colombia.

OBJECTIVE: To estimate out-of-pocket (OOP) health expenditure and the probability to incur in catastrophic health expenditure, as well as the socio-demographic determinants of households in Cartagena, Colombia. MATERIALS AND METHODS: Population-based cross-sectional study on a randomized stratified sample of Cartagena households. Two regression models were developed using OOP health expenditure and the probability to incur in catastrophic health expenditure as dependent variables. RESULTS: The average annual OOP health expenditure was $1 566 036 COP (US$783) (95%CI: $1 117 597 - $2 014 475) in poor households, $2 492 928 COP (US$1 246) (95%CI: $1 695 845 - $3 290 011) in middle class households, and $4 577 172 COP (uS$2 288) (95%CI: $1 838 222 - $7 316 122) in upper class households. Regarding the household income ratio, the OOP health expenditure was 14.6% in poor households, 8.2% in middle class, and 7.0% in upper class households. The probability to incur in catastrophic health expenditure was 30.6% (95%CI: 25.6% - 35.5%), 10.2% (95%CI: 4.5%-15.9%) and 8.6% (95%CI: 1.8%-23%) in the low, middle- and high-class households, respectively. Educational attainment, socioeconomic strata and employment were the main determinants of OOP and the probability to incur in catastrophic health expenditure. CONCLUSIONS: The health system has persistent inequalities regarding the financial protection of households related to out-of-pocket expenditure and the probability of catastrophic expenditure. This study provides evidence to review the social protection policy for the most socio-economically vulnerable households.

OBJETIVO: Estimar el gasto de bolsillo y la probabilidad de gasto catastrófico de los hogares y sus determinantes socioeconómicos en Cartagena, Colombia. MATERIALES Y MÉTODOS: Estudio transversal en una muestra poblacional estratificada aleatoria de hogares de Cartagena. Se estimaron dos modelos de regresión cuyas variables dependientes fueron gasto de bolsillo y probabilidad de gasto catastrófico en salud de los hogares. RESULTADOS: El gasto de bolsillo promedio anual en hogares pobres fue 1 566 036 COP (US$783) (IC95% 1 117 597-2 014 475); en hogares de estrato medio 2 492 928 COP (US$1246) (IC95% 1 695 845-3 290 011) y en hogares ricos 4 577 172 COP (US$2 288) (IC95% 1 838 222-7 316 122). Como proporción del ingreso, el gasto de bolsillo en salud fue de 14,6% en los hogares pobres, de 8,2% en los hogares de estrato medio y de 7,0% en los hogares ricos. La probabilidad de gasto catastrófico en salud de los hogares pobres fue 30,6% (IC95% 25,6-35,5%), de los de estrato medio del 10,2% (IC95% 4,5-15,9%) y de los hogares de estrato alto del 8,6% (IC95% 1,823,0%). El estrato socioeconómico, la educación y la ocupación fueron los principales determinantes del gasto de bolsillo en salud y de la probabilidad de incurrir en gasto catastrófico en salud. CONCLUSIONES: En el sistema de salud persisten desigualdades en la protección financiera de los hogares contra el gasto de bolsillo y la probabilidad de gasto catastrófico. El presente estudio genera evidencia para revisar la política de protección social de los hogares socioeconómicamente más vulnerables.

[Economic and health impact of prenatal care in pregnant women with public health insurance in Colombia during 2014]. / Impacto económico y sanitario del control prenatal en embarazadas afiliadas al régimen subsidiado en Colombia durante 2014.

OBJECTIVE:: To estimate health care costs of live births and the impact of prenatal care visit (PCV) in women from poor households. MATERIALS AND METHODS:: A randomized sample of 9 244 pregnant women (out of total= 25 000). Mean differences and proportions were calculated to compare results in both groups of women. The costs were estimated in American Dollars (USD) 2014, from the payer's perspective. RESULTS:: 75% of women live in urban areas. The mean age was 23 years old (CI95% 23.5-23.8). The average cost with PCV was USD 609.1 (CI95%: 581-632.7) and without PCV was USD 857.8 (CI95%: 774.7-923.8) and 87% of women attended at least one PCV. The health care costs increased in 32% (CI95% 27.1-41) in women who did not attended PCV. CONCLUSION:: The PCV is an efficient and effective intervention for managing the risk of maternal health.

Impacto económico y sanitario del control prenatal en embarazadas afiliadas al régimen subsidiado en Colombia durante 2014 / Economic and health impact of prenatal care in pregnant women with public health insurance in Colombia during 2014

Resumen: Objetivo: Analizar el impacto sanitario y económico del control prenatal (CPN) en mujeres de hogares pobres. Material y métodos: Se tomó una muestra aleatoria de 9 244 mujeres embarazadas (población= 25 000). Se estimó la incidencia de mortalidad materna, los costos directos de atención y la diferencia de medias y proporciones. Los costos fueron estimados en dólares americanos (USD) de 2014, con perspectiva del tercer pagador. Resultados: El 75% de las mujeres viven en zonas urbanas. La media de edad fue de 23 años (IC95%: 23.5-23.8). El 87.5% asistió al menos a una cita de CPN. El costo medio de partos con CPN fue de USD$ 609.1 (IC95%: 581-632.7). Sin CPN fue de USD $857.8 (IC95%: 774.7-923.8). Los costos se incrementan 32% (IC95% 27.1-41) sin CPN. Conclusión: El CPN es una intervención eficiente y efectiva para la gestión del riesgo de la salud materna.

Abstract: Objective: To estimate health care costs of live births and the impact of prenatal care visit (PCV) in women from poor households. Materials and methods: A randomized sample of 9 244 pregnant women (out of total= 25 000). Mean differences and proportions were calculated to compare results in both groups of women. The costs were estimated in American Dollars (USD) 2014, from the payer’s perspective. Results: 75% of women live in urban areas. The mean age was 23 years old (CI95% 23.5-23.8). The average cost with PCV was USD 609.1 (CI95%: 581-632.7) and without PCV was USD 857.8 (CI95%: 774.7-923.8) and 87% of women attended at least one PCV. The health care costs increased in 32% (CI95% 27.1-41) in women who did not attended PCV. Conclusion: The PCV is an efficient and effective intervention for managing the risk of maternal health.