Palabras clave: evidence- based medicine, decision support techniques, review literature as topic, systematic reviews as topic, review [Publication Type]

Introduction

Systematic reviews are currently considered the best option to offer informed decisions[1] for clinicians and health decision-makers due to their ability to synthesize relevant scientific evidence on a subject, with high methodological standards in their process. However, the process is expensive and slow and can take over three years to complete[2], which can delay the questions posed for a long time, potentially affecting clinical decision-making and health policy design.

Rapid reviews arise to synthesize the knowledge with the components of a systematic review but are simplified or omitted to produce information in less time and fewer resources to meet the needs of decision-makers[3],[4]. These simplified processes are known as “shortcuts” and, while they could decrease reliability in the conclusions[5], studies concluded that, especially in therapeutic interventions, they do not change the result drastically[6],[7].

Without a universal definition, rapid reviews are presented as a heterogeneous set of products without standardization in their production[8]. Neither it is known which shortcuts would cause less impact to their quality[9]. There is an agreement to consider a rapid review as requiring from one to six months for production according to the needs of decision-makers[10], which has increased their demand and led organizations such as the World Health Organization to use these methods for the preparation of guides in a limited time[11],[12], and the Cochrane Collaboration to establish a working group on rapid reviews in 2015.

This article corresponds to the first in a methodological series of narrative reviews on general topics in biostatistics and clinical epidemiology that explore and summarize, in an accessible language, published articles available via the main databases and specialized consultation texts. The series is aimed at the training of undergraduate students. The Evidence-Based Medicine department from the School of Medicine of the Universidad de Valparaiso, Chile, in collaboration with the Research Department of Instituto Universitario Hospital Italiano de Buenos Aires, Argentina, and the Evidence Center UC, of the Universidad Católica, Chile have worked on the series. The main purpose of this article is to introduce the concept of rapid reviews, their similarities and differences with systematic reviews, and their usefulness as a tool to synthesize evidence with fewer resources.

General concepts of systematic reviews

Systematic reviews are the best evidence-synthesis design to achieve reliable conclusions[13], especially when designing public policies or clinical practice guidelines[14].

A good quality systematic review should have an explicit protocol established “a priori” and ideally kept unchanged during the process, which will allow evaluators to reproduce the process and make a critical analysis of its methodology[5]. This protocol should be enrolled in a specialized registration database, such as PROSPERO, or published in scientific journals.

The research question must be well structured (in PICO format)[15], clinically relevant, and possess a scope and complexity determined by the authors according to the needs of stakeholders[8]. The search strategy should be comprehensive and have high sensitivity, covering multiple electronic databases with published studies, documents not yet published in conventional channels (grey literature), journal publications or scientific societies, and sometimes even reaching out to experts to look for ongoing studies. This search requires many resources to access eligible studies[9].

The inclusion and exclusion of studies and the critical assessment of the risk of bias includes at least two independent reviewers[15]. Therefore, the time it takes to complete the process depends on the amount of evidence found, the studies selected by the reviewers that require full-text evaluation, and the extraction of their results.

We can outline the process in the following steps (Figure 1):

Figure 1. Steps to conduct a systematic review.

It is then necessary to know what stages of a systematic review can be simplified and what implications these shortcuts have in the conclusions achieved[13]. 

Rapid reviews

Currently, rapid reviews are used to develop public health policies in the Americas and Europe[5],[16]: health technology assessment reports conducted in Latin America are examples of the rise of this type of methodological design. Locally, the Epistemonikos[17] and LOVE[18],[19] initiative stand out as facilitators in the search for rapid reviews[20]. The ability to balance an abbreviated process with sufficient methodological rigor results in rapid reviews becoming a real alternative to evidence synthesis operating on moderate information confidence within limited timeframes[10].

Nonetheless, the authors must explain the methodological limitations and the consequent risk of design bias[21],[22]. Readers of these abbreviated syntheses should be cautious when analyzing the methodology and the shortcuts used by reviewers, taking into account the authors’ confidence in their conclusions[6].

Due to the lack of consensus on defining them, rapid reviews are challenging to identify if their titles are not explicit. Most of them are found in the grey literature[5]. Also, because they are directly commissioned by stakeholders such as ministries of health or policy-makers[5], many are not in the public domain.

The shortcuts used by the authors of the different published rapid reviews are inconsistent, which creates a high heterogeneity in this type of evidence synthesis. Further studies are needed to properly assess which parts of a systematic review process can be simplified to have the least impact on their conclusions[10]. However, when a rapid review includes at least ten studies, it is less likely to lead to different conclusions than those achieved from a traditional systematic review[6].

As mentioned, rapid review methods are not standardized. Notwithstanding, PRISMA, a project that seeks to standardize the reporting of systematic reviews, has been updating and incorporating formats such as rapid reviews since 2018[23].

Review record

While databases such as PROSPERO accept the registration of rapid review protocols, not all consider this step in their production[21]. Failure to register or publish the protocol increases the risk of reporting bias[15].

Review question

Rapid reviews use a question with a PICO structure but should have a specific scope and not become a quick alternative to a systematic review[24]. The question must be specific or have a limited scope considering the stakeholders’ objectives[10]. The objective will largely determine the time the review takes. Some topics beyond the reach of rapid reviews require multiple or complex interventions[8].

Bibliographic search

As a first step, authors can look for a high-quality systematic review to avoid duplication of effort and redirect to a quick update of the review; for example, the authors can search from the review’s search date a small number of databases. This abbreviated search can be carried out in one of the leading electronic bibliographic sources, with PubMed/MEDLINE being the most widely used[25] given that it includes, on average, more than 80% of the studies usually included in systematic reviews[6].

Other shortcuts used to simplify the search include:

●        Removing manual search for evidence (non-electronic sources) and gray literature[5].

●        Limiting the search to studies in specific languages[5],[26], for example, only studies published in English.

●        Limiting the studies’ publication date search for publications in the last ten years, for example[5]. This shortcut is the most used by published rapid reviews[21].

●        Limiting the design of eligible studies by, for example, by limiting the search to randomized clinical trials[8].

These shortcuts may increase the risk of publication bias and selection bias[6],[10].

Study selection and data extraction

Authors may decide not to perform duplicate data collection and extraction[10]. This decision can speed up a rapid review; however, both steps will have a higher risk of error and bias[27],[28],[29]. Another shortcut used in this step is to include fewer outcomes when selecting the studies[9], for example, to consider only mortality as a relevant outcome. One disadvantage of this shortcut is that assessing potential selective reporting of outcomes of a primary study is lost.

Risk of bias assessment

Some rapid reviews omit the risk of bias assessment as a shortcut; however, this is not recommended due to the high impact on the reliability of the conclusions[30],[31]. By not conducting a critical appraisal, the quality of the evidence on which the conclusions are formulated is unknown[32],[33].

Synthesis of the evidence

Rapid reviews may choose not to conduct a meta-analysis, and only narratively describe quantitative findings; in fact, more than two-thirds of rapid reviews present their results this way[21].

Conclusions of a rapid review

Because available evidence is limited, few studies have evaluated the robustness of rapid reviews[9]. Some studies have found that the conclusions of rapid reviews do not differ substantially from a traditional systematic review[5],[24]. Due to the limitations that shortcuts may cause, the conclusions’ reliability is lower[9], and thus they must be interpreted by their consumers. Reliability will also depend on the quality of the studies included in each rapid review[32],[33].

Below is a comparative table with the main differences between a systematic review and a rapid review (Table 1).

Table 1. Comparison between a systematic review and a rapid review.

Notes

Roles and contributions
LTB: participated in conceptualization, methodology, investigation, resources, original writing and editing, and visualization. LVM: participated in conceptualization, investigation, resources, original writing and editing, and visualization. LG, LOM, CL y MV: participated in conceptualization, edition, visualization, supervision.

Conflicts of interest
The authors have completed the conflict of interest of ICMJE and declare that they have no conflict of interest.

Source of financial support
There was no funding for this review.

From the editors
The version of this manuscript that was peer-reviewed was submitted in Spanish. This English version was submitted by the authors and has been copyedited by the Journal.

Figure 1. Steps to conduct a systematic review.

Table 1. Comparison between a systematic review and a rapid review.

Esta obra de Medwave está bajo una licencia Creative Commons Atribución-NoComercial 3.0 Unported. Esta licencia permite el uso, distribución y reproducción del artículo en cualquier medio, siempre y cuando se otorgue el crédito correspondiente al autor del artículo y al medio en que se publica, en este caso, Medwave.