Mostrando las entradas con la etiqueta The New England Journal of Medicine. Mostrar todas las entradas
Mostrando las entradas con la etiqueta The New England Journal of Medicine. Mostrar todas las entradas

18 junio, 2013

Consumo de Chocolate e Conquista de Prêmios Nobel

Via: Medicina Baseada em Evidencias



Na semana passada Franz Messerli publicou New England Journal of Medicine um intrigante trabalho demonstrando que quanto maior o consumo de chocolate de uma nação, maior o número de prêmios Nobel conquistados pela nação.

A figura acima representa esta análise. Primeiro, vamos aproveitar para revisar como interpretar uma análise de correlação.

Associação

1)    Este é um gráfico de correlação, ou seja, mostra associação linear entre duas variáveis numéricas. A variável preditora (eixo x) é o consumo de chocolate anual per capita. A variável de desfecho (eixo y) é o número de prêmios Nobel a cada 10 milhões de pessoas no país.
2)    O coeficiente de regressão (b) do consumo de chocolate foi de 2.5, indicando que o cada 1.0 Kg de consumo per capita, a nação conquista 2.5 prêmios Nobel. Mostra o grau de influência do chocolate nos prêmios.
3)    O coeficiente de correlação (r) foi de 0.79. Este mostra a força de associação linear, quanto mais próximo de 1, mais forte é a associação. Neste caso, podemos interpretar como associação moderada a forte.
4)    Se elevarmos o r ao quadrado, obteremos o coeficiente de determinação (R2 = 0.62), que representa o quanto das conquistas de prêmios Nobel pode ser explicado pelo consumo de chocolate. Neste caso, 62% dos prêmios podem ser explicados pelo consumo de chocolate.
5)    O valor de P desta associação foi < 0.0001, indicando que é muito pequena a probabilidade destes achados serem devido ao acaso.

Observem que do ponto de vista estatístico, a análise é bastante convincente. De fato, há uma verdadeira associação linear entre chocolate per capita e prêmio Nobel a cada 10 milhões de pessoas.

Em segundo lugar, devemos nos perguntar se esta associação ocorre porque o consumo de chocolate provoca conquistas de prêmios Nobel.

Causalidade

Quando a análise chega neste ponto, o pensamento deixa de ser estatístico e passa a ser baseado na lógica da causalidade. Neste caso, devemos analisar os critérios de causalidade. Em 1965, Sir Austin Bradford Hill, professor Emérito de Bioestatística da Universidade de Londres publicou o artigo “The Environment and Disease: Association or Causation?” na revista Proceedings of the Royal Society of Medicine (58:295-30), onde ele explica 9 critérios que devem ser analisados para avaliar se uma associação é causal. Este se tornaram conhecidos como Critérios de Hill. Nesta postagem citarei os 5 critérios mais importantes.

1) Força e independência da associação: diz-se que quanto mais forte é uma associação, mas provável esta ser causal. Uma associação muito forte tem menos possibilidade de ser mediada por fator de confusão. Isto porque se uma associação forte é mediada por um fator de confusão, este fator ficará muito aparente.

Além da força da associação, esta deve ser independente de fatores de confusão.

Mas o que é mesmo fator de confusão? É uma terceira variável que intermedia a relação entre a exposição (chocolate) e o desfecho (prêmios Nobel), porém essa mediação não faz parte de uma seqüência causal, ou seja, não é o chocolate levando a alguma coisa que por sua vez provoca a conquista dos prêmios. Na verdade, essa variável (fator de confusão) seria associada a chocolate e associada aos prêmios, porém essas duas associações seriam desconectadas. Sendo assim, de maneira artificial, o fator de confusão promoveria uma associação não causal entre a exposição e o desfecho. 

Esse pode ser o caso em questão. Pode ser que o consumo de chocolate esteja associado ao nível sócio-econômico do país. Por sua vez, o nível sócio-econômico está associado a melhor educação e investimento em pesquisa, predispondo ao ganho de prêmios Nobel. Observem que esse link não é causal, pois não é o consumo de chocolate que melhora do nível sócio-econômico, para que este promova os prêmios. Neste caso, o nível sócio-econômico pode ser o fator de confusão desta análise.

Mas como podemos saber se a associação é ou não mediada por este fator de confusão? Precisaríamos fazer uma análise multivariada, onde colocamos chocolate, nível sócio-econômico e prêmios Nobel no mesmo modelo estatístico, e verificamos se a associação entre chocolate e prêmio independe do nível sócio-econômico. Se for independente, sugere causalidade (mas não garante). Essa análise não está presente no estudo, portanto está faltando um critério essencial para se demonstrar causalidade.

Portanto, força de associação fala a favor de causalidade, mas não é uma condição necessária. Já independência da associação é uma condição necessária para causalidade, porém não suficiente. Outros critérios devem ser obedecidos.

2) Plausibilidade: além de existir uma associação independente, deve haver uma razão lógica para que a variável preditora esteja causando o desfecho. Por exemplo, colesterol possui associação independente com infarto e há plausibilidade para se acreditar que esta relação é causal. Claro, é provavelmente o acúmulo do colesterol na parede do vaso que faz surgir a placa aterosclerótica, que será responsável pelo infarto.

Por outro lado, há muitas associações que carecem de lógica causal.

Acreditar que consumir chocolate promove um aumento da função cognitiva capaz de influenciar na produção de prêmios Nobel, com esta força de associação, é querer demais. Portanto, não há plausibilidade biológica para tal, até porque se isso fosse verdade chocólatras seriam evidentes gênios e podemos perceber exemplos contrários à nossa volta. Sendo assim, a conclusão do trabalho carece do segundo critério, o de plausibilidade.

3) Temporalidade: para que consideremos uma relação causal, o desenho do estudo deve nos garantir que a exposição veio antes do desfecho, pois a causa deve vir antes da conseqüência. Isso é o que ocorre em estudos de coorte prospectiva. Por exemplo, a coorte de Framingham demonstrou que níveis elevados de colesterol medidos no início do estudo (quando ninguém tinha tido o desfecho) se associaram à ocorrência do desfecho infarto ao longo do seguimento de longo prazo.

Porém isso não ocorre com nosso estudo do chocolate. As duas informações (consumo e Nobel) foram colhidas no mesmo momento. Pode ser até que o padrão de consumo maior de chocolate tenha surgido depois das conquistas de prêmios Nobel, se tornando impossível que aquele seja o causador dos Nobel. Na ausência dos critérios de temporalidade, pode existir o fenômeno de causalidade reversa: o ganho dos prêmios Nobel induzindo a uma maior consumo de chocolate – embora isso também careça de plausibilidade.

Portanto, este é mais um critério de causalidade que está ausente no trabalho.

4) Gradiente Dose-resposta: quanto maior o nível do preditor, maior a intensidade do desfecho. Esta é uma relação que se observa nesse estudo, visto que há demonstração de correlação positiva.

5) Reversibilidade Experimental: este é o critério que representa a validação final de uma relação causal, e deve ser proveniente de ensaios clínicos randomizados. Significa que ao controlar o preditor em um experimento controlado (ensaio clínico), o desfecho reduz. Isso ocorre no caso do colesterol. Ensaios clínicos randomizados indicam que na medida em que estatinas reduzem o colesterol, o risco de infarto diminui. Isso valida o colesterol como causador da doença coronária.

Se acreditamos que chocolate promove aumento da função cognitiva a ponto de gerar mais prêmios Nobels, ao reduzir o chocolate de uma nação, seus prêmios iriam reduzir ao longo do tempo. Não é esse tipo de evidência que se faz presente neste estudo, que é puramente observacional.

Há casos de variáveis que não fecham o ciclo de validação como fatores de risco, visto que seu controle não promove redução de risco. O caso mais recente é o do HDL-colesterol baixo, cujas drogas que promovem aumento em sua concentração não resultam em benefício clínico. É bem provável que este não seja um fator de risco, apenas seja uma marcador de risco.

Na recente década de 90 acreditava-se que infecção na placa aterosclerótica agravava a doença coronária. Porém, o uso de antibióticos em ensaios clínicos não modificou a incidência de desfechos coronários (versus placebo), fazendo cair por terra a hipótese infecciosa.

Sendo assim, uma breve análise dos critérios de causalidade nos sugerem que este trabalho está longe de provar uma associação causal entre consumo de chocolate e conquistas de um prêmio Nobel. Na verdade, o trabalho não passa de um brincadeira do autor que usou esta correlação para chamar a atenção de que associação e causalidade não são a mesma coisa. Uma brincadeira tão inteligente que resultou em um artigo no New England Journal of Medicine.

A Verdade

Por que o gráfico mostra que países como Noruega, Áustria, Dinamarca, Finlândia possuem um desempenho científico muito melhor do que os Estados Unidos? Tem algo errado, pois sabemos que os Estados Unidos são o país de maior produção científica do mundo. O erro está em ajustar o número de prêmios Nobel para cada 10 milhões de habitantes, pois a grande maioria destes habitantes não são cientistas e não estão contribuindo nada com isso. Os Estados Unidos são os que possuem a maior população e a maioria de qualquer população não é cientista. Como a população fica no denominador da fração, isso reduz o índice de prêmios Nobel dos Estados Unidos. O correto seria ajustar o número de prêmios para o número de cientistas do país e não para a população geral. Ao fazer errado, a medida de premiação beneficiou países pequenos da Europa, que coincidentemente consomem muito chocolate. Tudo não passa de um viés de análise de dados.
Enhanced by Zemanta

08 febrero, 2013

Estudio ONTARGET: Telmisartan, ramipril o ambos en pacientes con alto riesgo de eventos cardiovasculares

Structural diagram of ramipril. Created using ...
Structural diagram of ramipril. Created using ACD/ChemSketch 8.0 and Inkscape. (Photo credit: Wikipedia)
Fuente: Galo Sanchez

Silvia Redondo Muñoz[1] ha hecho el resumen de una evaluación GRADE, que hemos puesto a disposición de los interesados en la web de la Oficina (evalmed.es), en la pestaña “FORMACIÓN”, cuyas recomendaciones (válidas para este estudio) se recuadran más abajo.

Estudio ONTARGET: Telmisartan, ramipril o ambos en pacientes con alto riesgo de eventos cardiovasculares.

Yusuf S., Phil. D., Teo K.K., Pogue J., Dyal L., Copland I. and the ONTARGET Investigators. Telmisartan, ramipril, or both in patients at high risk for vascular events. N Engl J Med 2008;358:1547-59.

Los IECA reducen la mortalidad, IAM, ACV e insuficiencia cardiaca y disfunción del ventrículo izquierdo en pacientes con enfermedad CV previa o en diabéticos de alto riesgo. Pero no ejercen un bloqueo completo de la producción de toda la angiotensina II y se cree que su bloqueo completo podría ser más efectivo. Además los IECA sea asocian con tos y angioedema como consecuencia reducir la degradación de la bradiquinina.
     El ONTARGET pretende averiguar si en esta población de alto riesgo telmisartán puede ser una alternativa a ramipril, dado su conocido comportamiento en el estudio HOPE.

RECOMENDACIONES GRADE (VÁLIDAS SÓLO PARA ESTE ESTUDIO)

Para pacientes de 66 años (DE 7,2) con al menos una de las siguientes condiciones: enfermedad coronaria, enfermedad arterial periférica, enfermedad cerebrovascular o diabetes con afectación de órganos diana, basándonos en lo observado en este estudio, en la calidad de la evidencia y la magnitud y precisión de sus resultados, hacemos:

1º Una recomendación débil en contra de utilizar telmisarán en lugar de ramipril, salvo que estos pacientes presenten una tos que motive el abandono del IECA.

Justificación:
A) BENEFICIOS Y RIESGOS AÑADIDOS: En los pacientes sin tos, telmisartán no ofrece ningún beneficio frente a ramipril en las variables de resultados en salud. En los efectos adversos que motivan el abandono del tratamiento, telmisartán presenta en 4,66 años: a) más casos de hipotensión que ramipril, con un NND 106 (72 a 201), cuya magnitud de efecto estimamos baja a muy baja; y b) menos casos de angioedema, con un NNT 573 (322 a 3847), cuya magnitud de efecto estimamos es aún más baja.
B) INCONVENIENTES: No hay diferencias en tomar una pastilla de uno u otro fármaco.
C) COSTES: Telmisartán 80 mg: 260 euros/año; Ramipril 10 mg: 137 euros/año.

2º Una recomendación fuerte en contra de utilizar la terapia combinada de “telmisartán más ramipril” en lugar de “ramipril solo”.

Justificación:
A) BENEFICIOS Y RIESGOS AÑADIDOS: La terapia combinada de telmisartán más ramipril  no ofrece ningún beneficio frente a ramipril en las variables de resultados en salud. En 4,66 años se encontró un aumento del riesgo en el abandono de tratamiento por: a) Deterioro renal, NND 30 (23 a 42); b) Insuficiencia renal que requiere diálisis, NND 246 (145 a 867); c) Hipotensión, NND 33 (28 a 40); y d) Síncope, NND 602 (313 a 37968).
B) INCONVENIENTES: En contra de la terapia combinada al tener que tomar una pastilla adicional.
C) COSTES: Telmisartán 80 mg más Ramipril 10 mg: 397 euros/año. Ramipril 10 mg: 137 euros/año.

NOTA ACLARATORIA: Reseñamos que estamos haciendo una recomendación para lo sucedido en este ensayo clínico, y no una generalización propia de una revisión sistemática.


Enhanced by Zemanta

09 septiembre, 2011

Eficacia de una intervención sobre el tabaquismo mediante SMS


Free C, Knight R, Robertson S, Whittaker R, Edwards P, Zhou W et alSmoking cessation support delivered via mobile phone text messaging (txt2stop): a single-blind, randomised trial. Lancet 2011; 378: 49-55.    TC (s)   PDF (s)

Introducción

La aparición de las nuevas tecnologías de información y comunicación (TIC) permite llegar a un elevado número de pacientes de forma automatizada, a un coste muy bajo, sin que éstos se tengan que desplazar a la consulta y estén donde estén. Se han publicado estudios en los que la utilización de un programa de tratamiento del tabaquismo vía SMS es eficaz a corto plazo.

Objetivo

Evaluar la eficacia de un programa de cesación tabáquica automatizado vía SMS sobre la abstinencia mantenida durante 6 meses.

Perfil del estudio

Tipo de estudio: Ensayo clínico
Área del estudio: Tratamiento
Ámbito del estudio: Comunitario

Métodos

Se invitó a participar en el estudio txt2stop a fumadores >16 años de edad, propietarias de un teléfono móvil y que tuviesen interés en dejar de fumar a un mes vista. Los pacientes se ponían en contacto con el equipo de investigación y daban su conformidad a participar en el estudio vía SMS. Los participantes eran distribuidos aleatoriamente al grupo intervención o al grupo control. Los asignados al grupo control recibían mensajes SMS periódicos sobre la importancia de participar en el estudio. Los del grupo intervención recibían un total de 186 mensajes (seleccionados en función de sus características de una base de datos de 713) a una tasa de 5 al día durante las 5 primeras semanas y de 3 a la semana durante las 26 semanas restantes. Si enviaban un SMS con el texto crave (ansias [de fumar]) o lapse (recaída) el sistema enviaba automáticamente un mensaje adecuado a estas situaciones. Los participantes podían utilizar cualquier otro sistema de cesación tabáquica y, de hecho se les proporcionaba el número de teléfono del servicio de cesación tabáquica del National Health Service.
La variable de resultado principal era la tasa de cesación mantenida durante 6 meses autodeclarada por el participante y verificada bioquìmicamente mediante la medición de la cotinina salivar o una medición del CO espirado. Se consideró que se había alcanzado el resultado si no fumado ≤ 5 cigarrillos semanales la cuarta semana y ≤ 5 cigarrillos desde el inicio del periodo de abstinencia. Si un paciente no se sometía a la prueba de verificación de la abstinencia se le consideró fumador.
Las variables de resultado secundarias fueron la abstinencia puntual (ausencia de consumo en los últimos 7 días) valorada a las 4 semanas y a los 6 meses, la abstinencia autodeclarada desde el inicio, la abstinencia durante los últimos 28 días, la implicación en algún accidente de tráfico, la presencia de dolor en el pulgar cuando se escribían SMS y la utilización de otros servicios de cesación tabáquica. Se llevó a cabo un análisis por intención de tratar.

Resultados

Participaron en el estudio 5.800 personas (fig. 1). Las características de los participantes asignados a los dos grupos fueron simililares. La edad media fue de 36 años, un 55% eran varones, las tres cuartas partes habían llevado a cabo 1-5 intentos de abandono previos y un 40% tenían una puntuación en el test de Fagerström>5. Se dispuso de datos de seguimiento del 94% de los pacientes en el grupo intervención y del 97% en el grupo control.
Figura 1. Flujo de los participantes.
592 individuos informaron que estaban abstinentes a los 6 meses, de los cuales en 542 se pudo llevar a cabo la comprobación bioquímica de la misma. En un 28% de estos los valores obtenidos sugerían que seguían fumando. Los participantes asignados al grupo intervención obtuvieron mayores tasas de abandono que los asignados al grupo control (tabla 1).
Tabla 1. Principales variables de resultado del estudio.

Intervención (%)Control (%)RR (IC95%)
6 m


Abstinencia comprobada10,74,92,20 (1,80 a 2,68)
Abstinencia declarada los últimos 28 días19,813,51,47 (1,30 a 1,66)
Abstinencia declarada los últimos 7 días24,218,31,32 (1,19 a 1,47)
Implicación en accidentes de tráfico4,53,81,16 (0,89 a 1,51)
Dolor en el pulgar al escribir4,54,51,00 (0,78 a 1,28)
4 semAbstinencia declarada los últimos 7 días28,712,12,37 (2,11 a 2,66)
Los resultados fueron similares aunque se considerasen como no fumadores los pacientes que se habían perdido en el seguimiento. Los resultados fueron bastante homogéneos en los subgrupos analizados en función de la edad, tipo de empleo, grado de adicción o utilización de otros servicios de deshabituación tabáquica. No se detectaron efectos adversos de la intervención en relación con los SMS (accidentes de tráfico o dolor en los pulgares).

Conclusiones

Los autores concluyen que la intervención txt2stop es eficaz para promover el abandono del hábito tabáquico, lo que abre la puerta a que estas intervenciones puedan resultar útiles para modificar otros factores de riesgo conductuales.

Conflictos de interés

Ninguno declarado. Financiado por el UK Medical Research Council y la Primary Care Research Networks.

Comentario

No cabe duda de que el tabaquismo es un problema sanitario de primer orden en todo el mundo y se calcula que causa unos 5 millones de muertes al año. La lucha contra el tabaquismo debe llevarse a cabo mediante una combinación de medidas legislativas y de salud pública y de tratamiento de los fumadores.
El tratamiento del tabaquismo tradicionalmente ha recaído sobre la consulta, especialmente de atención primaria, abordando tanto los aspectos neurobiológicos como conductuales. La evolución de las TIC y la amplia difusión de los dispositivos móviles ha hecho que se empiece a explorar la factibilidad de su utilización en la atención sanitaria de los pacientes. En concreto, se ha planteado su utilización en tres áreas:
  • Mejorar el diagnóstico, la investigación el tratamiento, la monitorización y el abordaje de las enfermedades.
  • Tratar al paciente, promoción de la salud y mejorar la adherencia terapéutica.
  • Mejorar el proceso de atención sanitaria: recordatorios de visitas y vacunaciones y comunicación de resultados analíticos.
Los resultados de este estudio demuestran que un programa de tratamiento del tabaquismo basado en el envío de SMS personalizados a personas que desean abandonar el hábito dobla el porcentaje de éxitos del tratamiento.Estos resultados son coherentes con otro estudio publicado a más corto plazo (6 semanas) en el que la relación de éxitos entre el grupo intervención y el grupo control fueron similares a los de éste.
La metodología del estudio es rigurosa y se han utilizado los métodos de confirmación de la abstinencia habituales, pero sería conveniente disponer de los resultados a los 12 meses, que es el periodo utilizado habitualmente en los estudios de cesación tabáquica.

Bibliografía

  1. Rodgers A, Corbett T, Bramley D, Riddell T, Wills M, Lin RB, Jones MDo u smoke after txt? Results of a randomised trial of smoking cessation using mobile phone text messaging. Tobacco Control 2005; 14: 255-261.    PDF
  2. Brendryen H, Drozd F, Kraft PDigital smoking cessation program delivered through internet and cell phone without nicotine replacement (Happy Ending): randomized controlled trial. Journal of Medical Internet Research 2008; 10: e51    TC
  3. Organización Mundial de la Salud. Informe OMS sobre la epidemia mundial de tabaquismo, 2009. Consecución de ambientes libres de humo de tabaco. Ginebra: Organización Mundial de la Salud. 2010.
  4. Mataix J, Cabezas C, Lozano J, Camarelles F, Ortega G y Grupos de Abordaje del Tabaquismo (GAT) de semFYC y de Educación para la Salud del PAPPS-semFYC. Guía para el tratamiento del tabaquismo activo y pasivo. Barcelona: semFYC ediciones. 2008.

Autor

Manuel Iglesias Rodal. Correo electrónico: mrodal@menta.net.




06 julio, 2011

Antitrypanosomal Therapy for Chronic Chagas' Disease

Age-standardised disability-adjusted life year...Image via Wikipedia

Antitrypanosomal Therapy for Chronic Chagas' Disease

Caryn Bern, M.D., M.P.H.
Article
References
This Journal feature begins with a case vignette that includes a therapeutic recommendation. A discussion of the clinical problem and the mechanism of benefit of this form of therapy follows. Major clinical studies, the clinical use of this therapy, and potential adverse effects are reviewed. Relevant formal guidelines, if they exist, are presented. The article ends with the author's clinical recommendations.
A 42-year-old woman presents to her physician with a letter stating that after she made a recent blood donation, a serologic test of her donated blood was positive for Chagas' disease. The patient was born in El Salvador and moved to the United States when she was 18 years of age. Her three children are 8, 13, and 16 years of age. Her medical history is remarkable only for a cholecystectomy 2 years earlier; she reports no cardiac or gastrointestinal symptoms. Her physical examination is unremarkable. Electrocardiography (ECG) shows sinus rhythm at a rate of 72 beats per minute and a complete right bundle-branch block. An echocardiogram shows mild left ventricular segmental wall-motion abnormalities, but a normal ejection fraction and left ventricular diameter. The patient is referred to an infectious-disease consultant, who recommends antitrypanosomal therapy.

THE CLINICAL PROBLEM

Chagas' disease is caused by the protozoan parasite Trypanosoma cruzi.1 An estimated 8 million to 10 million people are infected with this parasite, primarily in the Americas.1-3 An estimated 300,000 infected persons live in the United States, most of whom are immigrants from areas of Latin America where the infection is endemic.4 Enzootic T. cruzi transmission has been reported across the southern half of the United States, but locally acquired cases in humans are rare.
Clinical Chagas' disease is classified into acute and chronic phases. During the acute phase of infection, most patients have mild, self-limited symptoms such as fever that do not come to medical attention. If inoculation of the parasite occurred through the conjunctiva, the patient may present with nonpainful unilateral edema of the upper and lower eyelids that lasts for several weeks; this is known as Romaña's sign. Severe acute infection with myocarditis, meningoencephalitis, or both is life-threatening; the acute-phase case fatality rate is estimated to be 0.25 to 0.50%.1
The acute phase lasts 4 to 8 weeks. The infection then enters the chronic phase, and without successful treatment, it is lifelong. Persons with chronic T. cruzi infection, but without signs or symptoms of Chagas' disease, are considered to have the “indeterminate” form of infection. An estimated 20 to 30% of people who initially have the indeterminate form have progression to clinically evident cardiac disease, gastrointestinal disease, or both over a period of years to decades.1
The earliest cardiac manifestations are usually conduction-system abnormalities and segmental left ventricular wall-motion abnormalities.1,5, Later manifestations include high-degree heart block, sustained and nonsustained ventricular tachycardia, sinus-node dysfunction leading to severe bradycardia, apical aneurysm (usually in the left ventricle), embolic phenomena due to thrombus formation in the dilated left ventricle or aneurysm, and progressive dilated cardiomyopathy with congestive heart failure. 5 These abnormalities are associated with palpitations, syncope, and a high risk of sudden death.6,7
Gastrointestinal Chagas' disease usually affects the esophagus, colon, or both.8,9 Advanced disease results in megaesophagus, megacolon, or both. Gastrointestinal involvement is much less common than Chagas' heart disease. This clinical form is seen predominantly in patients who are infected in the countries of the Southern Cone (Argentina, Bolivia, Chile, Paraguay, Uruguay, and parts of Brazil) and is rare in northern South America, Central America, and Mexico.8

PATHOPHYSIOLOGY AND EFFECT OF THERAPY

T. cruzi is carried in the gut of hematophagous triatomine bugs. Transmission most often occurs when the feces of an infected bug are inoculated through a bite wound or through intact mucous membranes (Figure 1FIGURE 1The Life Cycle ofTrypanosoma cruzi.). Thus, the great majority of cases of Chagas' disease occur in the zones of distribution of these vectors.
T. cruzi can also be transmitted through transfusion, through organ or tissue transplantation, and congenitally.1 Because of the risk of transmission through transfusion, most blood banks in the United States screen donated units of blood; however, such screening is voluntary and is not required by the Food and Drug Administration (FDA).
The incubation period after exposure to vectorborne T. cruzi is 1 to 2 weeks.1The acute phase of infection is characterized by active parasite replication and microscopically detectable parasitemia (Figure 2FIGURE 2Phases and Forms ofTrypanosoma cruziInfection.). After 4 to 8 weeks in the acute phase, replication is controlled by the host immune response, and parasitemia decreases to levels that are undetectable by means of microscopy. However, intracellular T. cruzi amastigotes remain in infected tissues, especially in cardiac and skeletal muscle.
Chagas' cardiomyopathy is characterized by a chronic inflammatory process causing damage to the myocardium of all four chambers of the heart as well as the conduction system.10 The pathogenesis may involve several mechanisms, including immunologically mediated tissue damage, cardiac autonomic dysfunction, and coronary microvascular disease.10 In the past, it was hypothesized that the chronic cardiomyopathy was solely attributable to cross-reacting autoantibodies to cardiac tissue and that the presence of the parasite was unnecessary for pathogenesis.11 This belief led to skepticism about the usefulness of antiparasitic treatment in chronic T. cruzi infection, which persisted until the late 1990s.12 More recently, a consensus has emerged that parasite persistence is essential to the development and progression of Chagas' cardiomyopathy.13-15 Experimental data from animal models also provide support for the hypothesis that elimination of the parasite reduces the risk of progression of cardiac disease.16 In contrast, gastrointestinal manifestations are thought to result from damage to intramural neurons that occurs primarily during the acute phase of the disease and is unmasked by neural attrition later in life.8,9
The only currently available drugs with proven efficacy against T. cruzi are nifurtimox and benznidazole,1,17-19,– which were developed empirically in the late 1960s and early 1970s, respectively.20 Benznidazole, a nitroimidazole derivative, is thought to act through covalent binding of nitroreduction intermediates to parasite molecules. 18,21 Nifurtimox, a nitrofuran compound, acts through production of reduced oxygen metabolites (e.g., superoxide and hydrogen peroxide), for which the parasites have lower detoxification capacity, as compared with vertebrate cells.12,19,22

CLINICAL EVIDENCE

In acute T. cruzi infection, treatment with either benznidazole or nifurtimox reduces the severity of symptoms and shortens the clinical course and duration of detectable parasitemia.1,23,24Parasitologic cure is reported in 60 to 85% of patients treated in the acute phase.1,17,23,24
Clinical trial data for the treatment of chronic T. cruzi infection are sparse. Uncertainty remains regarding the degree of efficacy because of the lack of a reliable test of cure.17,25-27 The diagnosis of chronic infection relies on serologic methods to detect IgG antibodies to T. cruzi. No single assay has sufficient sensitivity and specificity to be relied on alone; two tests based on different antigens, techniques (e.g., enzyme-linked immunosorbent assay [ELISA], immunofluorescence antibody assay, and immunoblot assay), or both are used in parallel to increase the accuracy of the diagnosis.28 Patients with discordant results from two serologic assays require further testing. In a small proportion of cases, the infection status remains difficult to resolve even after a third test because there is no accepted reference assay for the detection of chronic T. cruzi infection.27 Most clinical trials have therefore required that participants have positive results on three separate serologic tests at baseline.29,30
Early trials of nifurtimox for chronic infection used a technique called “xenodiagnosis,” in which laboratory-reared triatomine vectors were allowed to feed either directly on the patient's arm or leg or indirectly on blood from the patient, kept for 30 to 60 days, and examined for T. cruzi in the contents of the gut.31 However, only 30 to 60% of untreated patients with chronic T. cruzi infection have positive results of xenodiagnosis.32,33 Similar considerations affect the use of polymerase-chain-reaction (PCR) assays for the assessment of treatment response. Because the circulating parasite load is low in the chronic phase, PCR sensitivity in untreated patients is not high and varies depending on the gene target, methods, and population tested.34-36
These test limitations likewise have implications for assessing the outcome of therapy. The results of conventional serologic assays remain positive for years or even decades after successful treatment; the length of time for serologic assays to become negative is reported to be proportionate to the duration of the infection.37 Since they lack sensitivity for detecting chronic infection, xenodiagnosis and PCR are not reliable indicators of cure if the results are negative after treatment.
In the 1990s, two double-blind, randomized, placebo-controlled trials evaluated the efficacy of a 60-day course of benznidazole treatment in children with chronic T. cruzi infection.29,33 Each trial used a different, nonconventional serologic assay to document the response. In one of these trials, which enrolled 130 children, 58% of those who received benznidazole, as compared with 5% of those who received placebo, had seroconversion to negative titers with the nonconventional assay at 3 years.29In the second trial, which involved 106 children, negative seroconversion rates at 48 months with another nonconventional assay were 62% with benznidazole as compared with 0% with placebo, whereas rates of positive xenodiagnosis were 4.7% and 51.2%, respectively.33
Another small, randomized, placebo-controlled clinical trial compared the efficacy of 30 days of treatment with either nifurtimox or benznidazole with placebo in 77 adults with chronic T. cruziinfection.38 At the end of the 12-month follow-up period, all patients continued to have positive results on conventional serologic assays, but patients who had received benznidazole and nifurtimox were significantly less likely to have a positive xenodiagnosis than the placebo group (1.8% and 9.6%, respectively, vs. 34.3%).38
More recently, a nonrandomized, nonblinded cohort study compared 283 adults who received benznidazole with 283 untreated patients over a median follow-up period of 9.8 years.30 The analysis showed a significant decrease in the proportion of patients with progression of cardiomyopathy (4.2% of treated patients vs. 14.1% of untreated patients; adjusted hazard ratio, 0.24; P=0.002) and a trend toward decreased mortality (1.1% vs. 4.2%; adjusted hazard ratio, 0.2; P=0.09) in treated as compared with untreated patients.30 Conversion to negative serologic results occurred in 14.7% of treated and 5.7% of untreated patients (adjusted hazard ratio, 0.55; P<0.001); however, the median time to negative seroconversion was 11.7 years. Progression of cardiac disease was more frequent among patients with persistently positive results on serologic testing than among those with negative seroconversion (10.7% vs. 2.4%; adjusted hazard ratio, 4.88; P=0.009).
The Benznidazole Evaluation for Interrupting Trypanosomiasis (BENEFIT; ClinicalTrials.gov number, NCT00123916), a large, multicenter, double-blind, randomized, placebo-controlled trial of benznidazole for patients with mild-to-moderate Chagas' cardiomyopathy, is under way.25,39

CLINICAL USE

Before therapy for presumed Chagas' disease is initiated, it is necessary to confirm the diagnosis with the use of appropriate testing. As noted above, the diagnosis of chronic infection requires two separate serologic tests; if the results are discordant, further testing should be performed. Options for T. cruzi serologic testing in the United States are relatively limited; several ELISA kits based on parasite lysate or recombinant antigens have been cleared by the FDA for diagnostic application. The Centers for Disease Control and Prevention (CDC) offers consultation to health care providers concerning diagnostic testing for Chagas' disease and acts as a reference laboratory for serologic testing and PCR assays to detect Chagas' disease; contact information is provided below.
On the basis of the pediatric trials reviewed above, antitrypanosomal drug treatment in children with chronic T. cruzi infection was accepted as the standard of care throughout Latin America by the late 1990s, followed in recent years by a growing movement to offer treatment to older patients.3,15,28,30,39 Most experts now believe that the majority of patients up to 50 years of age who have chronic T. cruzi infection, including those without symptoms and those with early manifestations of cardiomyopathy, should be offered antitrypanosomal treatment.1,3,15 Because treatment is expected to reduce the probability of congenital transmission, stronger consideration may be warranted for women of reproductive age. 40
In patients older than 50 years of age, treatment decisions should take into account the current lack of certainty about the benefit of such treatment, the need for a prolonged course, and the frequent side effects. 1,3 Patients with advanced Chagas' cardiomyopathy, especially those with poorly compensated congestive heart failure, are not considered to be candidates for antitrypanosomal treatment. Such treatment would not be expected to reverse the cardiac structural abnormalities, and the drugs are likely to be poorly tolerated.1,3,39
There is no evidence that antitrypanosomal treatment affects the progression of gastrointestinal Chagas' disease.1,3 In patients with such disease, treatment decisions should be based on the potential to decrease the progression of heart disease. In patients with megaesophagus, absorption may be impaired, and treatment should be delayed until after corrective surgery has been performed.
Neither benznidazole nor nifurtimox is approved by the FDA, but both can be obtained free of charge from the CDC and used under investigational protocols. Consultations and requests for drugs should be addressed to the Parasitic Diseases Public Inquiries Line (            770-488-7775 begin_of_the_skype_highlighting            770-488-7775      end_of_the_skype_highlighting      ; www.cdc.gov/parasites/chagas), the CDC Drug Service (            404-639-3670 begin_of_the_skype_highlighting            404-639-3670      end_of_the_skype_highlighting      ) or, in emergencies after business hours, on weekends, and on federal holidays, to the CDC Emergency Operations Center (            770-488-7100 begin_of_the_skype_highlighting            770-488-7100      end_of_the_skype_highlighting      ). Questions about access to drugs outside the United States can be addressed to the World Health Organization (www.who.int/neglected_diseases/diseases/chagas/en/index.html).
Because benznidazole is usually better tolerated, this drug is viewed by most experts as the first-line treatment.1,3 Nevertheless, some patients tolerate nifurtimox better than benznidazole. Both agents are administered orally on an outpatient basis. Both drugs are contraindicated in pregnant women and in patients with severe renal or hepatic dysfunction. Alcohol use should be avoided during treatment with either agent, since it has been suggested that a disulfiram-like effect may occur, although the data are limited.19,22
The treatment regimen for benznidazole in adults consists of 5 to 7.5 mg per kilogram of body weight per day, administered orally in two divided doses for 60 days. In children younger than 12 years of age, the recommended dose is higher (10 mg per kilogram per day).41 A complete blood count and levels of hepatic enzymes, bilirubin, serum creatinine, and blood urea nitrogen should be obtained before the start of treatment, and the complete blood count should be repeated every 2 to 3 weeks during treatment. Patients should be monitored weekly for dermatologic side effects (described below), beginning 9 to 10 days after the initiation of treatment.
The treatment regimen for nifurtimox in adults is 8 to 10 mg per kilogram per day, administered orally in three or four divided doses for 90 days. In children 11 to 16 years of age, the recommended dose is 12.5 to 15 mg per kilogram per day, and in children 10 years of age or younger, the recommended dose is 15 to 20 mg per kilogram per day.41 Laboratory testing (a complete blood count and levels of hepatic enzymes, bilirubin, serum creatinine, and blood urea nitrogen) should be carried out before the start of treatment, 4 to 6 weeks into the course of treatment, and at the end of treatment. Patients should be weighed and monitored for symptoms and signs of peripheral neuropathy (described below) every 2 weeks, especially during the second and third months of treatment.

ADVERSE EFFECTS

Both benznidazole and nifurtimox have frequent side effects, especially in adults38,42-45 (Table 1TABLE 1Frequency of Adverse Effects Associated with Benznidazole and Nifurtimox in Adults.). Benznidazole causes dermatitis and photosensitization in one third to one half of patients.30,38,44,45 Mild rashes may respond to antihistamines, and moderate dermatitis can be managed with topical or low-dose oral glucocorticoids.46 Severe or exfoliative dermatitis or dermatitis associated with fever and lymphadenopathy should prompt immediate discontinuation of treatment. Benznidazole can also cause peripheral neuropathy (in up to 30% of patients), which is also an indication for discontinuation of treatment. The neuropathy is nearly always reversible, but it may take months to resolve. Bone marrow suppression is a rare side effect (in <1% of patients) that should prompt immediate discontinuation. In the two placebo-controlled trials of benznidazole in children, adverse effects were less frequent than in adults (12% of the children had rash and <5% had gastrointestinal symptoms in one study; <10% had moderate reversible side effects in the other).29,33
Nifurtimox causes gastrointestinal side effects in 50 to 75% of patients.31,38,43 These side effects include anorexia leading to weight loss, nausea, vomiting, and abdominal discomfort. Patients' weight should be monitored at regular intervals. Neurologic toxicity is common (occurring in up to 50% of patients), including irritability, insomnia, disorientation, mood changes, paresthesias, and, less often, tremors. Peripheral neuropathy is a rare, dose-dependent side effect that may appear late in the course of drug treatment and should prompt discontinuation. The condition is reversible, but complete resolution may require months.

AREAS OF UNCERTAINTY

The assessment of drug efficacy has been hampered by the lack of a sensitive, timely test of cure, the slowly progressive natural history of the disease, and the inability to predict which patients will have cardiomyopathy and which patients will remain asymptomatic. The trend toward offering treatment to adults with long-standing T. cruzi infection rests largely on the observation that treated patients are less likely than untreated patients to have progression of cardiomyopathy.1,3,30 Current data are insufficient to determine whether this effect is due to parasitologic cure or a reduced parasite burden.15
Physicians are often reluctant to treat patients with chronic T. cruzi infection because of both the frequency of side effects with the drugs that are currently available and the inability to confirm cure conclusively.43,47 New drugs with better safety profiles and proven efficacy would alter the risk–benefit balance considerably and lead to wider treatment. Data from in vitro studies and studies in animals suggest that several triazoles that inhibit ergosterol synthesis, including posaconazole and ravuconazole, have curative activity against T. cruzi.47,48 In a single case report, a patient with chronic T. cruzi infection and prior failure of benznidazole treatment had a good response to posaconazole according to PCR monitoring.49 A phase II trial of posaconazole (Clinical Trial for the Treatment of Chronic Chagas Disease with Posaconazole and Benznidazole; NCT01162967) is under way. Additional trials of the ravuconazole prodrug E1224 are planned.50

GUIDELINES

Consensus documents from the World Health Organization, the United States, and Brazil3,28,51strongly recommend antitrypanosomal treatment for acute, congenital, and reactivated T. cruziinfection, and for children (up to 12 or up to 18 years of age, depending on the publication) with chronic infection. Recommendations for adults with long-standing infection carry a lower evidence grade and strength because of the lack of data from randomized clinical trials; nevertheless, most recommendations published since 2000 include provisions to offer treatment for this group of patients.3,28,51,52

RECOMMENDATIONS

For the patient described in the clinical vignette, the diagnosis of Chagas' disease should be confirmed on the basis of positive results of at least two serologic tests. She should receive counseling, including the advice not to donate blood in the future. Assuming that the diagnosis is established, antitrypanosomal treatment should be offered, accompanied by a careful discussion of uncertainty regarding benefits and potential side effects. The patient has characteristic ECG signs of early Chagas' cardiomyopathy, indicating a higher risk of subsequent progression as compared with the risk among infected patients with normal ECG findings.30,53 Our best current understanding suggests that treatment will decrease the risk of further progression.25,30 Treatment may also decrease the risk of congenital transmission in subsequent pregnancies. The patient's children should undergo serologic testing for congenital Chagas' disease. Regular cardiologic follow-up should be scheduled to monitor the patient for disease progression and to optimize her cardiac care.
    The findings and conclusions in this article are those of the author and do not necessarily represent the views of the Centers for Disease Control and Prevention.
    Disclosure forms provided by the author are available with the full text of this article at NEJM.org.
    No potential conflict of interest relevant to this article was reported.
    I thank Susan Montgomery for helpful comments.

    SOURCE INFORMATION

    From the Division of Parasitic Diseases and Malaria, Center for Global Health, Centers for Disease Control and Prevention, Atlanta.
    Address reprint requests to Dr. Bern at the Division of Parasitic Diseases and Malaria, Centers for Disease Control and Prevention, 1600 Clifton Rd., Atlanta, GA 30333, or at .