Revisiones sistemáticas y metaanálisis: bases conceptuales e interpretación

doi:10.1016/j.recesp.2011.03.029

Revista Española de Cardiología

Volume 64, Issue 8, August 2011, Pages 688-696

https://doi.org/10.1016/j.recesp.2011.03.029 Get rights and content

Resumen

Las revisiones sistemáticas son investigaciones científicas en las cuales la unidad de análisis son los estudios originales primarios. Constituyen una herramienta esencial para sintetizar la información científica disponible, incrementar la validez de las conclusiones de estudios individuales e identificar áreas de incertidumbre donde sea necesario realizar investigación. Además, son imprescindibles para la práctica de una medicina basada en la evidencia y una herramienta fundamental en la toma de decisiones médicas. Sin embargo, la realización de una revisión sistemática de calidad no es una tarea sencilla, como en ocasiones tampoco lo es su interpretación. En este artículo especial se presentan las bases conceptuales para la realización y la interpretación de revisiones sistemáticas, poniendo especial énfasis en los puntos clave durante su ejecución mediante un ejemplo hipotético.

Abstract

Systematic reviews represent a specific type of medical research in which the units of analysis are the original primary studies. They are essential tools in synthesizing available scientific information, increasing the validity of the conclusions of primary studies, and identifying areas for future research. They are also indispensable for the practice of evidence-based medicine and the medical decision-making process. However, conducting high quality systematic reviews is not easy and they can sometimes be difficult to interpret. This special article presents the rationale for carrying out and interpreting systematic reviews and uses a hypothetical example to draw attention to key-points.

Full English text available from: www.revespcardiol.org

Section snippets

Introducción

Has finalizado una semana agotadora. Al final de la semana, te sientas tranquilo y reflexionas sobre las decisiones que has tenido que tomar. Entre otras, has indicado una intervención quirúrgica a un paciente con enfermedad de tres vasos, tuviste que decidir si se fibrinolizaba o se trasladaba a tu centro para angioplastia primaria a un paciente de 82 años con infarto inferior de 70 min de duración, y en consultas externas decidiste anticoagular a una paciente con fibrilación auricular.

Aunque

CONCEPTO Y NOMENCLATURA

Las RS son investigaciones científicas en las que la unidad de análisis son los estudios originales primarios, a partir de los cuales se pretende contestar a una pregunta de investigación claramente formulada mediante un proceso sistemático y explícito. Por eso se las considera investigación secundaria («investigación sobre lo investigado»). Por el contrario, a las revisiones que no siguen un proceso sistemático, denominadas revisiones narrativas, no se las puede considerar un proceso formal de

ETAPAS DE UNA REVISIÓN SISTEMÁTICA

De forma resumida, una RS sigue las siguientes etapas:

•
Definición de la pregunta clínica de interés y los criterios de inclusión y exclusión de los estudios.
•
Localización y selección de los estudios relevantes.
•
Extracción de datos de los estudios primarios.
•
Análisis y presentación de los resultados.
•
Interpretación de los resultados.

REVISIONES SISTEMÁTICAS Y PRÁCTICA CLÍNICA

Por mucho que los hallazgos de una RS sean consistentes y convincentes, al final, es el clínico quien tiene que tomar una decisión sobre un paciente en particular. No deben tomarse los hallazgos de la RS como normas fijas e invariables fruto de la «ortodoxia evidencialista». Dicho de otra manera, hay que adaptar los hallazgos de una RS al paciente, y no al revés. En este sentido, antes de tomar una decisión sobre un paciente basándose en una RS, se recomienda la siguiente reflexión³¹:

Conclusiones

Las RS constituyen una herramienta esencial para sintetizar la información científica disponible, incrementar la validez de las conclusiones de estudios individuales e identificar áreas de incertidumbre donde sea necesario realizar investigación. Además, son imprescindibles para la práctica de una medicina basada en la evidencia. Sin embargo, la realización de una RS se debe llevar a cabo siguiendo una estricta metodología y un control de calidad para evitar conclusiones sesgadas. En última

Conflicto de intereses

Ninguno.

Bibliografía (31)

A. Liberati et al.
The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration
J Clin Epidemiol
(2009)
D.J. Cook et al.
Methodologic guidelines for systematic reviews of randomized control trials in health care from the Potsdam Consultation on Meta-Analysis
J Clin Epidemiol
(1995)
D. Moher et al.
Completeness of reporting of trials published in languages other than English: implications for conduct and reporting of systematic reviews
Lancet
(1996)
G. Gregoire et al.
Selecting the language of the publications included in a meta-analysis: is there a Tower of Babel bias?
J Clin Epidemiol
(1995)
P.J. Easterbrook et al.
Publication bias in clinical research
Lancet
(1991)
D. Moher et al.
Assessing the quality of randomized controlled trials: an annotated bibliography of scales and checklists
Control Clin Trials
(1995)
D.J. Cook et al.
Systematic reviews: synthesis of best evidence for clinical decisions
Ann Intern Med
(1997)
D. Moher et al.
Epidemiology and reporting characteristics of systematic reviews
PLoS Med
(2007)
D.F. Stroup et al.
Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group
JAMA
(2000)
Stewart LA, Tierney JF, Clarke M. Reviews of individual patient data. En: Higgins JP, Green S, editores. Cochrane...

S. Straus et al.

Registering systematic reviews

CMAJ

(2010)

W.S. Richardson et al.

The well-built clinical question: a key to evidence-based decisions

ACP J Club

(1995)

X. Sun et al.

Is a subgroup effect believable? Updating criteria to evaluate the credibility of subgroup analyses

BMJ

(2010)

K. Dickersin et al.

Meta-analysis: state-of-the-science

Epidemiol Rev

(1992)

T.C. Chalmers et al.

Meta-analysis of clinical trials as a scientific discipline. I: Control of bias and comparison with large co-operative trials

Stat Med

(1987)

Cited by (142)

A Systematic Review of the Effectiveness of Non-Pharmacological Therapies Used by Nurses in Children Undergoing Surgery
2024, Pain Management Nursing
The aim of this study was to identify non-pharmacological pain relief therapies in children undergoing surgery.
Using keywords extracted from Medical Subject Headings and “Descriptores en Ciencias de la Salud” we searched for articles in the Web of Science, Scopus, Cuiden, PubMed, and CINHAL databases from the last five years, and performed a reverse search. We assessed the documentary quality of the articles using various standardized instruments.
The final review included eleven studies. In terms of cognitive-behavioral techniques, there is evidence that both music and video therapy are effective in reducing postoperative pain in children in seven studies, and therapeutic play in five studies. Other methods used less frequently but found to be effective included laughter therapy in one study and deep breathing in another. Regarding physical methods of pain relief, massage was found to be an effective non-pharmacological therapy for reducing pediatric postoperative pain in two studies and ineffective in another.
In this study, we highlight the importance of non-pharmacological therapies in pediatric postoperative pain management. Cognitive-behavioral techniques, especially music therapy, video therapy, and therapeutic play, reduce pediatric postoperative pain. They are therefore effective therapies that nurses can use in this area. Further research into the effectiveness of storytelling is necessary, as the evidence is not entirely conclusive. More evidence is also needed on physical methods of pain relief, particularly massage.
An approach for searching genes in signaling pathways or gene-gene interaction networks related to Hypertension in the Mexican population
2024, Gene
The selection of single nucleotide polymorphisms (SNPs) to evaluate the genetic susceptibility in complex traits is often conducted in isolation, without considering the entire set of genes. Incorporating signaling pathways or gene-gene interaction search may provide a more comprehensive approach to selecting SNP candidates for further study.
To propose a systematic procedure for identifying SNPs candidates with complex traits such as hypertension and blood pressure.
Sequential stages to SNPs selection: 1) literature review to identify SNPs, following the PRISMA methodology, 2) identification and selection of signaling pathways and selection of gene-gene interaction networks using the STRING software, and 3) application of specific criteria for SNPs candidates, including: a) SNPs with minor allele frequency > 5% in the target population, b) SNPs located within genes involved in three or more signaling pathways, and c) SNPs that are not in linkage disequilibrium, with a D’or r² value < 0.8.
Stage 1) A total of 44 publications were selected, providing information on 230 genes evaluated with blood pressure. Stage 2) Using the STRING software, we selected 7 signaling pathways with a false discovery rate < 0.0001 and strength ≥ 0.8; and we identified 16 genes belonging to gene-gene interaction networks, six of them share ≥ 3 signaling pathways. Stage 3) Finally, 7 SNPs were selected for genotyping in the Health Workers Cohort Study. We observed a positive association between SNPs with hypertension incidence in males (rs1130214, rs3807989) and females (rs5051, rs2493123).
Our methodological proposal may be a reliable way for selecting SNPs candidates to study complex traits.
Meta-QTL analysis for mining of candidate genes and constitutive gene network development for fungal disease resistance in maize (Zea mays L.)
2023, Crop Journal
Citation Excerpt :
Validation and fine mapping of the mapped QTL can be potentially utilized to verify and narrow down the genomic regions governing target traits. meta-QTL (MQTL) analysis is a potent approach for the identification of consensus regions using the information on the QTL for different traits across studies [23]. In maize, MQTL analysis has been successfully attempted for various traits like gray leaf spot resistance [24], consensus synthetic QTL for disease resistance [18], root traits [25] and popping traits [26].
The development of resistant maize cultivars is the most effective and sustainable approach to combat fungal diseases. Over the last three decades, many quantitative trait loci (QTL) mapping studies reported numerous QTL for fungal disease resistance (FDR) in maize. However, different genetic backgrounds of germplasm and differing QTL analysis algorithms limit the use of identified QTL for comparative studies. The meta-QTL (MQTL) analysis is the meta-analysis of multiple QTL experiments, which entails broader allelic coverage and helps in the combined analysis of diverse QTL mapping studies revealing common genomic regions for target traits. In the present study, 128 (33.59%) out of 381 reported QTL (from 82 studies) for FDR could be projected on the maize genome through MQTL analysis. It revealed 38 MQTL for FDR (12 diseases) on all chromosomes except chromosome 10. Five MQTL namely 1_4, 2_4, 3_2, 3_4, and 5_4 were linked with multiple FDR. Total of 1910 candidate genes were identified for all the MQTL regions, with protein kinase gene families, TFs, pathogenesis-related, and disease-responsive proteins directly or indirectly associated with FDR. The comparison of physical positions of marker-traits association (MTAs) from genome-wide association studies with genes underlying MQTL interval verified the presence of QTL/candidate genes for particular diseases. The linked markers to MQTL and putative candidate genes underlying identified MQTL can be further validated in the germplasm through marker screening and expression studies. The study also attempted to unravel the underlying mechanism for FDR resistance by analyzing the constitutive gene network, which will be a useful resource to understand the molecular mechanism of defense-response of a particular disease and multiple FDR in maize.
Factors associated with the inappropriate use of the pediatric emergency department. A systematic review
2023, Journal of Pediatric Nursing
Pediatric emergency department (PED) admissions have risen in recent years, a trend not justified by the severity of the pathologies presented. The aim of this study is to analyse factors related to the inappropriate use of pediatric emergency departments.
This is a systematic review reported in accordance with the PRISMA statement. We searched the PubMed, Web of Science and Science Direct databases, using keywords extracted from MeSH, and conducted a reverse search using Google Scholar and Open Grey, for the period January 2017 to August 2022. The quality of the papers was assessed using STROBE, CASPe, AMSTAR-2, GRADE, Levels Of Evidence and Grades Of Recommendation.
A total of 20 studies were selected. Factors related to inappropriate use included the younger age of children, black caregivers, lower socioeconomic status, lower parental educational attainment, perceived urgent demand for care, parental emotions in response to their children's health problems, psychological distress, the ineffective exercise of the parental role, the advantages of the PED and the nature of health insurance.
The results illustrate the heterogeneous nature of the phenomenon under investigation. Gaining an understanding of the factors related to the inappropriate demand of PEDs, from the perspective of health professionals, can help in developing interventions to reduce unnecessary consultations and relieve pressure on these healthcare services.
Information and digital technology-assisted interventions to improve intercultural competence: A meta-analytical review
2023, Computers and Education
Citation Excerpt :
As alerted by McAuley, Tugwell, and Moher (2000), the exclusion of grey literature from meta-analysis could result in an overestimate of an intervention effect by an average of 12%. Although the increase of amount and types of selected articles for meta-analysis made the sample size more representative (González, Urrútia, & Alonso-Coello, 2011), we carefully assessed the quality of such grey literature for being included in our review. Some conference papers were excluded either because (a) they lacked sufficient statistics for analyses, (b) no technology-based interventions were assessed, or (c) qualitative methods were used.
This study synthesized 56 studies published during year 2000–2021 that measured the use of technology to develop students' intercultural competence (IC). Meta-analysis, a statistical process that combines the quantitative findings of multiple relevant studies, was used to analyze the effectiveness of the interventions to improve IC. The findings suggested that technology-based interventions generally produced a positive and intermediate effect (with an effect size of .35) on the students’ overall and subdimensions of IC. Subsequent publication bias analyses and sensitivity analyses revealed that our findings were not affected. Based on the high heterogeneity between the reviewed studies, the types of intervention and IC measurement tools were identified as moderating variables. Among the four intervention types, computer simulation had the largest effect size (0.40) and blended learning environment had the smallest effect size (0.22). The effect sizes of digital materials/tools and computer-mediated communication were 0.38 and 0.35, respectively. Compared to studies using self-developed scales (0.38) and previously validated scales (0.33), studies that employed self-developed tests to assess IC produced a larger effect size (0.66). The impact of the technology-based interventions at different levels of moderating variables was discussed. Implications of the findings were provided.
Coronary artery bypass, percutaneous coronary intervention, and mortality: a reflection on methods
2023, Revista Espanola de Cardiologia

View all citing articles on Scopus

View full text

Enfoque: Métodos contemporáneos en bioestadística (III)Revisiones sistemáticas y metaanálisis: bases conceptuales e interpretaciónSystematic Reviews and Meta-Analysis: Scientific Rationale and Interpretation

Resumen

Abstract

Section snippets

Introducción

CONCEPTO Y NOMENCLATURA

ETAPAS DE UNA REVISIÓN SISTEMÁTICA

REVISIONES SISTEMÁTICAS Y PRÁCTICA CLÍNICA

Conclusiones

Conflicto de intereses

J Clin Epidemiol

J Clin Epidemiol

Lancet

J Clin Epidemiol

Lancet

Control Clin Trials

Systematic reviews: synthesis of best evidence for clinical decisions

Ann Intern Med

Epidemiology and reporting characteristics of systematic reviews

PLoS Med

Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group

JAMA

Registering systematic reviews

CMAJ

The well-built clinical question: a key to evidence-based decisions

ACP J Club

Is a subgroup effect believable? Updating criteria to evaluate the credibility of subgroup analyses

BMJ

Meta-analysis: state-of-the-science

Epidemiol Rev

Meta-analysis of clinical trials as a scientific discipline. I: Control of bias and comparison with large co-operative trials

Stat Med

Enfoque: Métodos contemporáneos en bioestadística (III)
Revisiones sistemáticas y metaanálisis: bases conceptuales e interpretaciónSystematic Reviews and Meta-Analysis: Scientific Rationale and Interpretation