Beruflich Dokumente
Kultur Dokumente
Abstract
Background
La infeccin con Clostridium difficile ha llegado a ser un problema epidemiolgico
global tanto para pacientes hospitalizados como para pacientes ambulatorios.
Los medicamentos ms comnmente usados para tratar CDI son el metronidazol
y la vancomicina. El propsito de este estudio fue comparar la eficacia y
seguridad de la monoterapia de metronidazol con la monoterapia de vancomicina
y la terapia combinada en pacientes CDI.
Methods
A comprehensive search without publication status or other restrictions was
conducted.
Studies comparing metronidazole monotherapy with vancomycin monotherapy or
combination
therapy in patients with CDI were considered eligible. Meta-analysis was
performed
using the Mantel-Haenszel fixed-effects model, and odds ratios (ORs) with 95%
confidence
intervals (95% CIs) were calculated and reported.
Se llev a cabo una bsqueda exhaustiva sin restricciones de status de
publicacin u otras restricciones. Se consideraron elegibles los estudios que
comparan la monoterapia de metronidazol con la monoterapia de vancomicina o
con la terapia combinada en pacientes con CDI. El metanlisis se realiz
utilizando el modelo de efectos fijos de Mantel-Haenszel, y se calcularon e
informaron odds ratios (OR) con intervalos de confianza del 95% (IC del 95%).
Results
Of the 1910 records identified, seventeen studies from thirteen articles (n = 2501
patients)
were included. No statistically significant difference in the rate of clinical cure was
found
between metronidazole and vancomycin for mild CDI (OR = 0.67, 95%CI (0.45,
1.00), p =
0.05) or between either monotherapy and combination therapy for CDI (OR =
1.07, 95%CI
(0.58, 1.96), p = 0.83); however, the rate of clinical cure was lower for
metronidazole than for
vancomycin for severe CDI (OR = 0.46, 95%CI (0.26, 0.80), p = 0.006).
De los 1910 registros identificados, se incluyeron 17 estudios llevados a cabo en
los 13 artculos elegidos (ms de un estudio en algunos artculos) (n=2501
pacientes). No se encontr diferencia estadsticamente significativa en la tasa de
curacin clnica entre el metronidazol y la vancomicina para CDI leve ((OR =
0.67, 95%CI (0.45, 1.00), p = 0.05) o entre cualquier monoterapia y la terapia
combinada para CDI. (OR = 1.07, 95%CI (0.58, 1.96), p = 0.83), sin embargo la
tasa de curacin clnica fue ms baja con el metronidazol que con la vancomicina
para CDI severa. (OR = 0.46, 95%CI (0.26, 0.80), p = 0.006).
Conclusions
Metronidazole and vancomycin are equally effective for the treatment of mild CDI,
but vancomycin is superior for the treatment of severe CDI. Combination therapy
is not superior to monotherapy because it appears to be associated with an
increase in the rate of AEs.
El metronidazol y la vancomicina son igualmente efectivos para el tratamiento de
CDI leve, pero la vancomicina es superior para el tratamiento de CDI severa. La
terapia combinada no es superior a la monoterapia ya que parece estar asociada
con un incremento en la tasa de eventos Adversos.
Introduction
Clostridium difficile is an anaerobic, gram-positive, spore-forming bacillus[1]. The
incidence
and severity of C. difficile infection (CDI) appear to be on the rise[2], and CDI has
become an epidemiological problem for both hospitalized patients and outpatients
worldwide[2].
Clostridium difficile es un bacilo anaerobio, gram positivo, formador de esporas.
La incidencia y severidad de CID (infeccin por C. difficile) parece estar en
aumento y CDI ha llegado a ser un problema epidemiolgico tanto para los
pacientes hospitalizados como los pacientes ambulatorios (***Un paciente
ambulatorio es aquel que debe acudir regularmente a un centro de salud por
razones de diagnstico o tratamiento pero que no necesita pasar la noche all (es
decir, no queda internado). Por esta razn, tambin se conoce al paciente
ambulatorio como diurno o de da***)
Oral metronidazole is recommended for patients undergoing initial therapy and for
those
with mild infections, whereas oral vancomycin is recommended for seriously ill
patients.
El metronidazol oral se recomienda para pacientes sometidos a tratamiento inicial
y para aquellos con infecciones leves, mientras que la vancomicina oral se
recomienda para pacientes gravemente enfermos.
Furthermore, in patients with recurrent CDI, infusion of donor feces has resulted
in improved treatment outcomes. In particular, patients with multiple relapses of
CDI have benefited from this approach [15]. Unfortunately, prospective clinical
trials of most of the above-mentioned antibiotics evaluating their applicability for
general use have not yet been performed[4], and several questions regarding
fecal transplantation remain unanswered; for example, the optimal protocol for
donor feces infusion is unknown[15]. Therefore, the treatment of CDI has relied
primarily on metronidazole and vancomycin[16].
Adems, en pacientes con CDI recurrente, la infusin de heces de donante ha
dado mejores resultados del tratamiento. En particular, pacientes con mltiples
recadas de CDI se han beneficiado de esta aproximacin. Desafortunadamente,
todava no se han realizado ensayos clnicos prospectivos de la mayora de los
antibiticos antes mencionados que evalen su aplicabilidad para uso general y
varias preguntas sobre el trasplante fecal permanecen sin respuesta; Por
ejemplo, se desconoce el protocolo ptimo para la infusin de heces de
donantes. Por lo tanto, el tratamiento del CDI se ha basado principalmente en
metronidazol y vancomicina.
Methods
Our study protocol and analysis were planned in accordance with the Preferred
Reporting
Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines[19], and
no ethical
approval was needed.
Nuestro protocolo de estudio y anlisis fueron planeados de acuerdo con las
directrices de PRISMA (tems de reporte preferidos para Revisiones sistemticas
y Meta-anlisis) y no fue necesaria una aprobacin tica.
Search strategy
A systematic literature search of the following electronic databases was
conducted to identify relevant literature published in English or Chinese before
November 2014: PubMed, Embase, Web of Science, and Cochrane Library.
Three Chinese language databasesChina National Knowledge Infrastructure
(CNKI; available at www.cnki.net), Chinese Scientific Journals Database (VIP;
available at www.cqvip.com), and WANFANG DATA (available at www.
wanfangdata.com.cn)were also searched. The search strategy included the
following medical subject headings and key words: pseudomembranous
enterocolitis, antibiotic-associated diarrhea, antibiotic-associated colitis,
Clostridium difficile, metronidazole, and vancomycin.
Se realiz una bsqueda bibliogrfica sistemtica de las siguientes bases de
datos electrnicas para identificar la literatura relevante publicada en ingls o
chino antes de noviembre de 2014: PubMed, Embase, Web of Science, y la
biblioteca Cochrane Library. Adems, se busc en 3 bases de datos en chino
(China National Knowledge Infrastructure (CNKI; available at www.cnki.net),
Chinese Scientific Journals Database (VIP; available at www.cqvip.com), and
WANFANG DATA (available at www. wanfangdata.com.cn)). La estrategia de
bsqueda incluy los siguientes encabezamientos mdicos y palabras clave:
"enterocolitis pseudomembranosa", "diarrea asociada a antibiticos", "colitis
asociada a antibiticos", "Clostridium difficile", "metronidazol" y "vancomicina".
These search terms were combined using Boolean logic as follows: both the
medical subject headings and the terms related to the patient population of
interest (Clostridium difficile, pseudomembranous colitis, antibiotic-associated
diarrhea, and antibiotic-associated colitis) were combined with those describing
interventions (metronidazole OR vancomycin). The complete search strategy
used to search each database is described in S1 Table. Furthermore, the
references in the initially identified articles, including relevant reviews, were
manually searched and reviewed. Academic dissertations and abstracts
presented at scientific conferences were also searched using ProQuest
Dissertation & Theses Database and The Conference and Academic Dissertation
Database of CNKI, respectively, to ensure that no relevant study was missed (up
to May 2015).
Estos trminos de bsqueda se combinaron utilizando la lgica booleana de la
siguiente manera: se combinaron tanto los encabezamientos de las materias
mdicas como los trminos relacionados con la poblacin de pacientes de inters
(Clostridium difficile, colitis pseudomembranosa, diarrea asociada a antibiticos y
colitis asociada a antibiticos) (Metronidazol O vancomicina). La estrategia de
bsqueda completa utilizada para buscar cada base de datos se describe en la
Tabla S1. Adems, las referencias en los artculos inicialmente identificados,
incluidas las revisiones pertinentes, fueron buscadas manualmente y revisadas.
Tambin se realizaron bsquedas de tesis acadmicas y resmenes presentados
en conferencias cientficas utilizando: ProQuest Dissertation & Theses Database
y The Conference and Academic Dissertation Database of CNKI, respectivamente
para asegurar que no se perdi ningn estudio relevante (hasta mayo de 2015).
Study selection
Two reviewers (RL and LCL) assessed all potentially relevant studies and
reached a consensus on all items. They then independently searched the
literature and examined relevant studies to obtain data on the rates of clinical
cure, CDI recurrence and AEs. In the case of disagreement between the two
reviewers, the senior coauthor (XL) was consulted, and the disagreement was
resolved by consensus. Studies were included if they met the following criteria: (1)
contained original data, (2) contained data regarding the use of metronidazole or
vancomycin for the treatment of CDI, (3) contained data regarding clinical
therapeutic outcomes or AEs, and (4) reported on RCTs or case-control studies.
Studies were excluded if they (1) were unrelated to CDI, (2) were duplicate
reports, (3) were not written in English or Chinese, or (4) were not case-control
studies.
Dos revisores (RL y LCL) evaluaron todos los estudios potencialmente relevantes
y llegaron a un consenso sobre todos los tems.A continuacin, buscaron de
forma independiente la literatura y examinaron los estudios pertinentes para
obtener datos sobre las tasas de curacin clnica, recurrencia de CDI y EAs. En
caso de desacuerdo entre los dos revisores, se consult al coautor senior (XL), y
el desacuerdo se resolvi por consenso.
-Los estudios se incluyeron si cumplan los siguientes criterios:
(1) contenan datos originales, (2) contenan datos sobre el uso de metronidazol o
vancomicina para el tratamiento del CDI, (3) contenan datos sobre los resultados
clnicos teraputicos o EAs y (4 ) Informaron sobre ECA o estudios de casos y
controles.
- Se excluyeron los estudios si:
(1) no estaban relacionados con el CDI, (2) eran informes duplicados, (3) no
estaban escritos en ingls o chino, o (4) no eran estudios de casos y controles.
(4) mean age; (5) male (%); (6) follow-up duration; (7) number of enrolled
patients; (8) drug regimen; (9) rate of clinical cure; (10) rate of CDI recurrence;
(11) AEs related to the study medications; (12) case definitions; (13) quality of the
evidence; and (14) overall risk of bias. (4) edad media; (5) (%)masculino ; (6)
duracin del seguimiento; (7) nmero de pacientes inscritos; (8) rgimen de
frmaco; (9) tasa de curacin clnica; (10) tasa de recurrencia del CDI; (11) AEs
relacionados con los medicamentos del estudio; (12) definiciones de caso; (13)
calidad de la evidencia; Y (14) riesgo general de sesgo.
The extracted data (and specifically, the rates of clinical cure and CDI recurrence)
were analyzed based on the intention-to-treat population.
Los datos extrados (y especficamente, las tasas de curacin clnica y
recurrencia CDI) se analizaron sobre la base de la intencin de tratar la
poblacin.
Results (Resultados)
Search results (Resultados de bsqueda)
A total of 1910 records were identified (Fig 1). After excluding duplicates and
screening the
titles of the studies, 635 articles were reviewed. After screening the abstracts of
these potentially relevant articles, 36 were selected for full-text review based on
relevance to the study topic (Fig 1). Thirteen articles containing seventeen studies
(one article included three separate studies and two articles each included two
separate studies) and 2501 patients were included [10,13,20,2635]. In addition,
we attempted to contact the corresponding authors of five potentially relevant
studies via e-mail. The authors of four studies did not respond to our e-mail
communication [3639], and the corresponding author of the fifth study
communicated that we would require approval by their institutional review board
for additional individual patient data[40]. Therefore, all five studies had to be
excluded from our final set of included studies. Fig 1 shows the selection process
for the studies included in the meta-analysis.
Se identific un total de 1910 registros (Fig 1). Despus de excluir duplicados y
seleccionar los ttulos de los estudios, 635 artculos fueron revisados.Despus de
seleccionar los resmenes de estos artculos potencialmente relevantes, 36
fueron seleccionados para la revisin de texto completo basado en la relevancia
para el tema del estudio (Fig 1). Trece artculos que contenan diecisiete estudios
(un artculo incluy tres estudios separados y dos artculos, cada uno incluy dos
estudios separados) y 2501 pacientes fueron incluidos [10, 13, 20, 26 - 35].
Adems, tratamos de contactar a los autores correspondientes de cinco estudios
potencialmente relevantes por correo electrnico. Los autores de cuatro estudios
no respondieron a nuestra comunicacin por correo electrnico [36-39], y el autor
correspondiente del quinto estudio comunic que requeriramos la aprobacin de
su junta de revisin institucional para datos adicionales del paciente individual
[40]. Por lo tanto, los cinco estudios tuvieron que ser excluidos de nuestro
conjunto final de estudios incluidos Fig. 1 muestra el proceso de seleccin de los
estudios incluidos en el metanlisis.
Study quality assessment and risk of bias assessment (Evaluacin de la
calidad del estudio y evaluacin del riesgo de sesgo)
Using the modified Jadad score (S2 Table), the results of the quality assessment
of each included study indicated that five studies were of high quality
[10,13,26,27,33] and that the remaining eight studies were of moderate quality
[20,2832,34,35] (Table 1, S2 Table). The results of the assessment of the overall
risk of bias for each included study indicated that three reports exhibited a low risk
of bias [10,13,26] and that the remaining ten reports exhibited an unclear risk of
bias [20,2735] (Table 1, S1 Fig). The main characteristics of the included studies
are shown in Table 1.
Utilizando el puntaje de Jadad modificado (Tabla S2), los resultados de la
evaluacin de calidad de cada estudio incluido indicaron que cinco estudios
fueron de alta calidad [10,13,26,27,33] y que los restantes ocho estudios fueron
de calidad moderada [ 20,28 - 32,34,35] (Tabla 1, Tabla S2). Los resultados de la
evaluacin del riesgo global de sesgo para cada estudio incluido indic que tres
informes exhibieron un bajo riesgo de sesgo [10, 13, 26] y que los diez informes
restantes presentaron un riesgo poco claro de sesgo [20, 27-35 ] (Tabla 1, Fig.
S1). Las principales caractersticas de los estudios incluidos se muestran en la
Tabla 1.
Efficacy outcomes (Resultados de la eficacia)
Rate of clinical cure. Regarding the comparison of metronidazole with
vancomycin, the meta-analysis results suggested that there was no significant
difference in the rate of clinical cure for the treatment of mild CDI (5
studies[13,20,26,31,32], 703 patients, OR = 0.67, 95% CI (0.45, 1.00), p = 0.05,
I2 = 0%) (Fig 2A). However, the rate of clinical cure was lower for metronidazole
than for vancomycin for the treatment of severe CDI (4 studies[13,20,26,29], 324
patients, OR = 0.46, 95% CI (0.26, 0.80), p = 0.006, I2 = 0%) (Fig 2B). Regarding
the comparison of monotherapy with combination therapy, three articles were
included[10,33,35], and the cases were divided into three subgroups according to
therapeutic intervention used. The metaanalysis results did not show a significant
difference in the rate of clinical cure between monotherapy and combination
therapy (4 studies (one article included two separate studies[35]), 190 patients,
OR = 1.07, 95% CI (0.58, 1.96), p = 0.83, I2 = 0%) (Fig 2C).
Safety outcomes
AEs. The reported AEs from the included studies consisted of death, colectomy,
diarrhea,
any complication, ileus, colonic perforation, nausea and vomiting,
pseudomembranous colitis,
toxic megacolon, rash and severe enterocolitis. We performed subgroup analysis
according to
the AEs. The meta-analysis results did not show any significant difference in the
rate of AEs
between metronidazole and vancomycin (the results from 7 studies were
separated into six subgroups[
13,20,2630], 1330 patients, OR = 1.18, 95% CI (0.80, 1.74), p = 0.41, I2 = 0%)
(Fig 4A).
However, the rate of AEs was significantly lower for monotherapy than for
combination therapy (the results from 3 studies were separated into eight
subgroups [10,28,33], 439
patients, OR = 0.30, 95% CI (0.17, 0.51), p<0.0001, I2 = 0%) (Fig 4B); in fact, this
rate was
more than 4-fold higher for combination therapy than for monotherapy (46.9%
vs.11.1%).
Publication bias and sensitivity analyses
The shape of Beggs funnel plot and the Eggers test results (all p>0.05) did not
demonstrate
any evidence of publication bias (S1 File). Considering that meta-analysis showed
little heterogeneity
(I2 = 6%) between the studies for the comparison of the rate of CDI recurrence
between
metronidazole and vancomycin. Thus, we conducted sensitivity analyses of the
subgroups
from these studies according to explicit or inexplicit classifications of severity and
follow-up
duration to explore underlying sources of heterogeneity. These analyses showed
that the primary
results were not influenced by the inexplicit classification of severity or by the
follow-up
durations (S2 File).
Discussion
The aim of this study was to compare the efficacy and safety of metronidazole
monotherapy with vancomycin monotherapy and combination therapy in CDI
patients. Based on the analyses according to disease severity and the treatment
methods used, this study has revealed that metronidazole and vancomycin are
equally effective for the treatment of mild CDI but that vancomycin is superior for
severe CDI. Additionally, these findings suggest that combination therapy is not
superior to monotherapy.
El propsito de este estudio fue comparar la eficacia y la seguridad de la
monoterapia de metronidazol con la terapia de vancomicina y la terapia
combinada en pacientes con CDI. Basado en el anlisis de acuerdo a la
severidad de la enfermedad y los mtodos de tratamiento utilizados, este estudio
ha revelado que el metronidazol y la vancomicina son igualmente efectivos para
el tratamiento de CDI leve pero que la vancomicina es superior para CDI severa.
Adicionalmente, esos hallazgos sugieren que la terapia combinada no es superior
a la monoterapia.
Our study has several limitations. First, although the quality of the majority of the
studies was high or medium and although the total sample size was sufficient, the
sample size of each subgroup was relatively small and was thus susceptible to
false-positive or false-negative results. Second, although sensitivity analyses
indicated that the primary results were not influenced by the inexplicit
classification of severity or by the follow-up duration, we roughly determined
disease severity based on the proportion of patients, which potentially resulted in
bias. For example, if 77% of the patients had mild CDI[27], we categorized all of
the patients into the mild CDI group.
Nuestro estudio tiene varias limitaciones. En primer lugar, aunque la calidad de la
mayora de los estudios fue alta o media y aunque el tamao total de la muestra
fue suficiente, el tamao de la muestra de cada subgrupo era relativamente
pequeo y, por lo tanto, fue susceptible a resultados falsos positivos o falsos
negativos. En segundo lugar, aunque los anlisis de sensibilidad indican que los
resultados primarios no estaban influenciados por la clasificacin implcita de la
severidad o por la duracin del seguimiento, nosotros determinamos
aproximadamente la gravedad de la enfermedad basndonos en la proporcin de
pacientes, lo cual potencialmente result en sesgo. Por ejemplo, si 77% de los
pacientes tenan CDI leve, se clasific a todos los pacientes en el grupo de CDI
leve.
Moreover, the follow-up duration for CDI treatment varied greatly (from 312
weeks) among the different studies, which may have resulted in bias. Third,
because very few articles presented research on combination therapy, we could
not compare the subgroups according to disease severity between monotherapy
and combination therapy; therefore, the general patient conditions may have
caused prescription bias.
Por otra parte, la duracin de seguimiento para el tratamiento de CDI vari en
gran medida (de 3 a 12 semanas) entre los diferentes estudios, lo que puede
haber llevado a sesgo. En tercer lugar, debido a que muy pocos artculos
presentaban investigacin sobre la terapia combinada, no pudimos comparar los
subgrupos de acuerdo a la severidad de la enfermedad entre la monoterapia y la
terapia combinada; Por lo tanto, las condiciones generales del paciente pueden
haber causado sesgo de prescripcin.
Fourth, only studies published in English were included, which might have
rendered the results vulnerable to bias related to language and ethnicity. Fifth,
additional data on combination therapy using other novel antibiotics are needed.
En cuarto lugar, solo los estudios publicados en ingls fueron incluidos, lo cual
podra haber rendido los resultados vulnerables a sesgos relacionados con el
idioma y la etnia. En quinto lugar, se necesitan datos adicionales sobre la terapia
de combinacin utilizando otros antibiticos nuevos.
In conclusion, these findings have important clinical implications and support the
recommendations of the current CDI treatment guidelines[3,4]. However,
validation of the routine use of combination therapy may require additional
evidence. To compensate for the shortcomings of our study, further large-scale
clinical trials and well-designed research studies are needed to identify more
effective therapies for patients with CDI.
En conclusin, estos hallazgos tienen importantes implicaciones clnicas y
apoyan las recomendaciones de las actuales directrices para el tratamiento de
CDI. Sin embargo, la validacin del uso rutinario de la terapia combinada puede
requerir evidencia adicional. Para compensar las deficiencias de nuestro estudio,
se necesitan ms ensayos clnicos a gran escala y estudios de investigacin bien
diseados para identificar terapias ms eficaces para los pacientes con CDI.
LINKS
https://www.google.com.co/search?q=Oxman+AD%2C+Guyatt+GH+et+al
%2C+Users%27+Guides+to+The+Medical+Literature
%2C+VI+How+to+use+an+overview.+(JAMA+1994%3B+272+(17)%3A+1367-
1371)&oq=Oxman+AD%2C+Guyatt+GH+et+al%2C+Users
%27+Guides+to+The+Medical+Literature%2C+VI+How+to+use+an+overview.+
(JAMA+1994%3B+272+(17)%3A+1367-
1371)&aqs=chrome..69i57.685j0j7&sourceid=chrome&ie=UTF-8
https://www.google.com.co/search?q=Mantel-Haenszel+fixed-
effects+model&oq=Mantel-Haenszel+fixed-
effects+model&aqs=chrome..69i57j0.794j0j7&sourceid=chrome&ie=UTF-8
http://www.elsevier.es/es-revista-medicina-clinica-2-articulo-declaracion-prisma-
una-propuesta-mejorar-S0025775310001454
http://archivos.evidenciasenpediatria.es/DetalleArticulo/_LLP3k9qgzIh7aNQBiadw
mQH1vtdJ7DW4Na2i472Rz8-4L4-0IUwCS0hea82EwEC2bTmsyscz5hU-
B6Aet2cU6g
https://es.wikipedia.org/wiki/Metaan%C3%A1lisis
http://accessmedicina.mhmedical.com/content.aspx?
bookid=1442§ionid=101158961
http://accessmedicina.mhmedical.com/book.aspx?bookid=1442
https://www.google.com.co/search?
biw=1366&bih=662&q=descargar+gratis+modelos+estadisticos+en+metaanalisis
+Tania+B.+Huedo-Medina%2CBlair+T.
+Johnson&oq=descargar+gratis+modelos+estadisticos+en+metaanalisis+Tania+
B.+Huedo-Medina%2CBlair+T.
+Johnson&gs_l=serp.3...48955.49590.0.51037.2.2.0.0.0.0.153.261.0j2.2.0....0...1
.1.64.serp..0.0.0.mypJD1gQmIs
http://www.um.es/metaanalysis/pdf/5003.pdf
http://ocw.upc.edu/sites/all/modules/ocw/estadistiques/download.php?
file=715001/2013/1/54994/t15_metanalisis-5228.pdf
http://www.elsevier.es/es-revista-semergen-medicina-familia-40-articulo-que-es-
el-analisis-por-S1138359300736197
https://www.imim.cat/media/upload/arxius/terminologia.pdf
http://www.scielo.cl/scielo.php?script=sci_arttext&pid=S0034-
98872004001200016
https://www.google.com.co/search?q=metaan
%C3%A1lisis+medidas+primarias+y+secundarias&oq=metaan
%C3%A1lisis+medidas+primarias+y+secundarias&gs_l=serp.3...27271.31149.0.3
1467.16.14.1.0.0.0.199.1470.3j8.11.0....0...1.1.64.serp..5.3.347...30i10k1.JRh-
Oyn6-UI
https://www.google.com.co/search?
q=the+primary+outcome+measure+in+meta+analysis&spell=1&sa=X&ved=0ahU
KEwju99qViaXTAhWGQyYKHUpnClAQvwUIHygA&biw=1366&bih=662
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3528446/
https://bmcobes.biomedcentral.com/articles/10.1186/s40608-014-0025-1
http://www.bmj.com/content/339/bmj.b2700
https://www.google.com.co/search?
q=analisis+de+subgrupos&oq=analisis+de+sub&aqs=chrome.2.69i57j0l5.5403j0j
7&sourceid=chrome&ie=UTF-8
http://www.scielo.cl/scielo.php?script=sci_arttext&pid=S0034-
98872012000500018
https://www.google.com.co/search?
q=I2+and+Q+statistics&oq=I2+and+Q+statistics&aqs=chrome..69i57j0l2.765j0j7&
sourceid=chrome&ie=UTF-8
http://www.statsdirect.com/help/meta_analysis/heterogeneity.htm
https://www.google.com.co/search?
q=que+son+los+estadisticos+I2+y+Q&oq=que+son+los+estadisticos+I2+y+Q&aq
s=chrome..69i57.10714j0j7&sourceid=chrome&ie=UTF-8
https://books.google.com.co/books?
id=61lyol_6jK0C&pg=PA22&lpg=PA22&dq=que+son+los+estadisticos+I2+y+Q&s
ource=bl&ots=2DP57pMCWh&sig=aGp9YyPIW4iWtVZFhalHY880ph0&hl=es&sa
=X&ved=0ahUKEwj_kM-
bmKXTAhWBRyYKHSNuDqQQ6AEIXDAI#v=onepage&q=que%20son%20los
%20estadisticos%20I2%20y%20Q&f=false
https://www.google.com.co/search?
q=I2+y+Q+estadistica&ei=SGnxWJDfD8nNmQGb_56IAQ&start=20&sa=N&biw=1
366&bih=662