Prehospital early warning scores for predicting clinical deterioration of COVID-19 patients: An integrative review

Abstract

Introduction

Triage, and in particular scales, are a tool that allows patients with clinical risk to be managed for early, effective and efficient care.

Objective

To identify the most precise and specific prehospital score for the detection of clinical worsening risk in COVID-19 patients.

Methods

The protocol followed for the integrative review was the PRISMA method 2020. A literature search was performed in five databases: Scopus, Cochrane Library, Pubmed, Embase, Prospero and Lit-covid-NIH-NLM. Based on 19 keywords, 5 inclusion and 5 exclusion points. Finally, 22 articles were selected.

Results

Twenty-two studies were identified that addressed effective outcomes for early measures such as telephone triage, web, protocols or tools such as scales. We compared the functionality of 12 scales in patients with COVID-19, showing that the most important variables for this early assessment of clinical worsening were systolic blood pressure, temperature, oxygen saturation and the need for oxygen supplementation. The best predictive value for clinical deterioration and mortality was obtained by NEWS score, with sensitivities and specificities ranging from 77 to 88%.

Conclusions

Prehospital scales are still under development, with few research studies and a relative confidence in their statistical values. Nonetheless, it has been observed that the scale that best fit the COVID-19 was NEWS with an optimal prediction for patients. This could pave the way for its use under other relevant clinical scenarios, such as acute respiratory infections, exacerbations of chronic diseases or future health emergencies.

Resumen

Introducción

El triaje, y en particular las escalas, son una herramienta que permite manejar a los pacientes con riesgo clínico para una atención temprana, eficaz y eficiente.

Objetivo

Identificar el puntaje prehospitalario más preciso y específico para la detección del riesgo de empeoramiento clínico en pacientes COVID-19.

Métodos

El protocolo seguido para la revisión integrativa fue el método PRISMA 2020. Se realizó una búsqueda bibliográfica en cinco bases de datos: Scopus, Cochrane Library, PubMed, Embase, Prospero y Lit covid-nih-nlm. Basado en 19 palabras clave, 5 puntos de inclusión y 5 puntos de exclusión. Finalmente, se seleccionaron 22 artículos.

Resultados

Se identificaron 22 estudios que abordaron resultados efectivos para medidas tempranas como el triaje telefónico, la web, los protocolos o herramientas como las escalas. Comparamos la funcionalidad de 12 escalas en pacientes con COVID-19, mostrando que las variables más importantes para esta evaluación temprana del empeoramiento clínico fueron la presión arterial sistólica, la temperatura, la saturación de oxígeno y la necesidad de suplementación con oxígeno. El mejor valor predictivo para el deterioro clínico y la mortalidad se obtuvo mediante la puntuación NEWS, con sensibilidades y especificidades que oscilaron entre el 77 y el 88%.

Conclusiones

Las escalas prehospitalarias aún están en desarrollo, con pocos estudios de investigación y una relativa confianza en sus valores estadísticos. No obstante, se ha observado que la escala que mejor se ajustaba al COVID-19 era NEWS con una predicción óptima para los pacientes. Esto podría allanar el camino para su uso en otros escenarios clínicos relevantes, como infecciones respiratorias agudas, exacerbaciones de enfermedades crónicas o futuras emergencias sanitarias.

What is known/What it contributes

• There are adequate pre-hospital early risk scales for COVID-19. Predictor variables include systolic pressure, temperature, saturation and O2 need.

Implications of the study

• The number of studies carried out at the pre-hospital level is still scarce.

Introduction

The pandemic caused by the SARS-CoV-2 virus, of the Coronaviridae family, has posed challenges to healthcare systems worldwide. While coronavirus disease 2019 (COVID-19) has been reduced from a global health emergency, new viral variants continue to emerge today, making COVID-19 a global health threat.^1,2

The disease can cause a wide range of symptoms, including fever, dry cough, headache, throat, and muscle pain. In addition, it can cause anosmia, ageusia (characteristic of SARS-CoV-2), fatigue and respiratory distress. These manifestations can appear mild, although they can worsen and become potentially dangerous for people with certain risk factors, causing pneumonia, multi-organ failure and in the worst cases death. The most serious complications that develop with COVID-19 are due to factors such as advanced age, and/or the presence of comorbidities. In this sense, chronic diseases are a cofactor that includes conditions such as heart, lung, kidney disease, diabetes, and cancer. In addition to conditions that the patient presents, conditions such as obesity or pregnancy increase this risk, as well as immunosuppression states due to treatments or pathologies. These factors encourage patients to have a higher risk of developing complications, and therefore of clinical worsening, when contracting SARS-CoV-2 infection.^1–3

In Spain, there was a great spread in the first wave, which meant that the demand for medical care skyrocketed, causing an overload of the health system and, with it, a deficit in the care of the sick. To alleviate hospital saturation, new methods of out-of-hospital care had to be designed.⁴ Triage, in its various forms, has made it possible to effectively and efficiently manage the early care of patients at clinical risk. In this context, the use of scales to detect the risk of clinical worsening was essential. However, currently, there are no gold standard scales or scores (both are used indistinctly throughout the text) for COVID-19 disease, and it is necessary to use those that are adequately adjusted to the possible complications, risk factors, and symptoms of the patients.^5,6

Therefore, it is highlighted that the pandemic has represented a challenge in health care, especially in the identification of patients at risk of complications. The scales evaluated by the nursing staff generally analyze patients with different clinical criteria, including vital signs, respiratory symptoms, and risk factors, facilitating the detection of severe symptoms and prioritization in each particular case. The assessment must be accurate, effective, simple and clear, so that its parameters can be applied quickly. Thus, thanks to these scales, the care of the most critical patients in the out-of-hospital environment has been improved, positively influencing the saturation of hospital and emergency services. All this has favored a better recovery, avoiding possible deaths.⁷ The National Early Warning Score (NEWS)⁸ or its modified version NEWS2,⁹ which includes a more specific assessment for hypercapnic respiratory failure, is one of the most well-known scales. Other scales were designed for specific pathologies, such as the qSOFA (Quick Sequential Sepsis-related Organ Failure Assessment)¹⁰ scale used to identify sepsis, or the mREMS scale¹¹ used to classify the severity of trauma, but has also been used for COVID.

However, due to the wide variety of scales used in this context, it would be of great relevance to identify the most effective ones. Therefore, the main objective of this integrative review is to identify prehospital scales that help detect the clinical deterioration of patients with suspected or confirmed COVID-19, even when their condition is not yet clearly differentiated.

Methods

This integrative review of the literature follows the recommendations of the PRISMA Declaration (supplementary data p1), allowing the unification of the data analyzed. It begins with the asking of the “PICO” question as suggested by¹²:

• P – Patients with COVID-19

• I – Prehospital scales performed on patients with COVID-19

• C – Comparison between the different scales used (the results compare the reliability and validity data between scales or instruments, such comparison is key to detecting the most reliable scale)

• O – Finding the most reliable scales for detecting the risk of clinical worsening

To establish a selection criterion, a total of 10 points were determined to be taken into account for inclusion and exclusion.

Inclusion criteria

• 1.

  Studies in which the population consists of patients with COVID-19.

• 2.

  Studies carried out and published between 2020 and 2024.

• 3.

  Studies in which pre-hospital scales have been carried out to detect the risk of clinical worsening.

• 4.

  Studies conducted in an out-of-hospital setting only.

• 5.

  Systematic, narrative, observational, experimental, prospective, retrospective, qualitative, and quantitative review studies.

Exclusion criteria

• 1.

  Studies in which the population is patients with COVID-19 in the hospital setting.

• 2.

  Studies with prehospital interventions related to cardiac arrest and cardiovascular accident.

• 3.

  Studies related to the consequences of COVID-19.

• 4.

  Studies that cannot be accessed in full text.

• 5.

  Interventions performed on patients with COVID-19 as a result of other pathologies.

Search strategy

The following keywords were determined to perform the MeSH/EMTREE in the databases: novel coronavirus; early warning score; pre-alert; COVID; prognostic; COVID-19; predictive; biomarker; new coronavirus; score; pre-hospital; SARS-CoV-2; scale; prehospital; SARS CoV-2; Triage; out of hospital; nCoV. In particular, we searched articles in 5 different databases (Pubmed, Scopus, Embase, Lit-covid-NIH-NLM, Cochrane Library and Prospero) and applying date filters from December 2019 to September 2024 and full articles. The search string was composed of three parts: (i) “novel coronavirus” OR “COVID” OR “COVID-19” OR “new coronavirus” OR “SARS- CoV-2” OR “SARS-CoV-2” OR “nCoV”. (ii) “early warning score” OR “prognostic” OR “predictive” OR “score” OR “scale” OR “Triage” OR “pre-alert” OR “biomarker” OR “biomarkers”. (iii) “pre-hospital” OR “prehospital” OR “out of hospital”. Depending on the database, the format of the search string had to be adjusted (Table 1).

Table 1.

Search string according to database.

PUBMED: 131 results	((“novel coronavirus”[Title/Abstract] OR “COVID”[Title/Abstract] OR “COVID- 19”[Title/Abstract] OR “new coronavirus”[Title/Abstract] OR “SARS-CoV- 2”[Title/Abstract] OR “SARS-CoV-2”[Title/Abstract] OR “nCoV”[Title/Abstract]) AND (“early warning score”[Title/Abstract] OR “prognostic”[Title/Abstract] OR “predictive”[Title/Abstract] OR “score”[Title/Abstract] OR “scale”[Title/Abstract] OR “Triage”[Title/Abstract] OR “pre-alert”[Title/Abstract] OR “biomarker”[Title/Abstract] OR “biomarkers”[Title/Abstract]) AND (“pre-hospital”[Title/Abstract] OR “prehospital”[Title/Abstract] OR “out of hospital”[Title/Abstract]))
Cochrane: 61 results	‘Novel coronavirus’: ti,ab,kw OR ‘COVID’: ti,ab,kw OR ‘COVID-19’: ti,ab,kw OR ‘new coronavirus’: ti,ab,kw OR ‘SARS Cov-2’: ti,ab,kw OR ‘SARS-Cov-2’: ti,ab,kw OR ‘nCov’: ti,ab,kw AND ‘early warning score’: ti,ab,kw OR ‘prognostic’: ti,ab,kw OR ‘predictive’: ti,ab,kw OR ‘score’: ti,ab,kw OR ‘scale’: ti,ab,kw OR ‘Triage’: ti,ab,kw OR ‘pre-alert’: ti,ab,kw OR ‘biomarker’: ti,ab,kw OR ‘biomarkers’: ti,ab,kw AND ‘pre- hospital’: ti,ab,kw OR ‘prehospital’: ti,ab,kw OR ‘out of hospital’: ti,ab,kw
SCOPUS: 349 results	(TITLE-ABS-KEY(“novel coronavirus”) OR TITLE-ABS-KEY(“COVID”) OR TITLE-ABS-KEY(“new coronavirus”) OR TITLE-ABS-KEY(“SARS CoV-2”) OR TITLE-ABS-KEY(“SAR-Cov-2”) OR TITLE-ABS-KEY(“nCov”))AND(TITLE-ABS-KEY(“early warning score”) OR TITLE-ABS-KEY(“prognostic”) OR TITLE-ABS-KEY(“predictive”) OR TITLE-ABS-KEY(“score”) OR TITLE-ABS-KEY(“scale”) OR TITLE-ABS-KEY(“Triage”) OR TITLE-ABS-KEY(“pre-alert”) OR TITLE-ABS-KEY(“biomarker”) OR TITLE-ABS-KEY(“biomarkers”))AND(TITLE-ABS-KEY(“pre-hospital”) OR TITLE-ABS-KEY(“prehospital”) OR TITLE-ABS-KEY(“out of hospital”))
PROSPERO: 35 results	(“novel coronavirus” or “covid” or “covid-19” or “novel coronavirus” or “SARS-CoV-2” or “SARS-CoV-2” or “nCoV”) and (“early warning score” or “prognosis” or “predictive” or “score” or “scale” or “triage” or “pre-alert” or “biomarker” or “biomarkers”) and (“pre-hospital” or “pre-hospital” or “out-of-hospital”)
Lit COVID- NIH-NLM: 200 results	(“novel coronavirus” OR “COVID” OR “COVID-19” OR “new coronavirus” OR “SARS- CoV-2” OR “SARS-CoV-2” OR “nCoV”) AND (“early warning score” OR “prognostic” OR “predictive” OR “score” OR “scale” OR “Triage” OR “pre-alert” OR “biomarker” OR “biomarkers”) AND (“pre-hospital” OR “prehospital” OR “out of hospital”)
Embase: 323 results	(‘novel coronavirus’:ti,ab OR ‘COVID’:ti,ab OR ‘new coronavirus’:ti,ab OR ‘SARS CoV-2’:ti,ab OR ‘SAR-Cov-2’:ti,ab OR ‘nCov’:ti,ab)AND(‘early warning score’:ti,ab OR prognostic:ti,ab OR predictive:ti,ab OR score:ti,ab OR scale:ti,ab OR triage:ti,ab OR ‘pre-alert’:ti,ab OR biomarker:ti,ab OR biomarkers:ti,ab)AND(‘pre-hospital’:ti,ab OR prehospital:ti,ab OR ‘out of hospital’:ti,ab)

Study selection

The method used for removal of duplicated studies the “deduplicate” function from the package synthesisr from R. Documents identified during the initial literature search were evaluated using the information contained in the title, abstract, and MeSH subject descriptor/heading (EMG and AS-G authors). We obtained the full texts of studies that were considered relevant or if the information in the title and abstract of the study was inconclusive. Any disagreements on eligibility were resolved through discussion and consensus with the participation of the other authors (CMM, RC-S, JLM-C, FM-R).

The risk of bias was assessed through a detailed review by the authors (CMM, RC-S, JLM-C, FM-R), who had not participated in the initial selection of the articles, following the Prediction model Risk Of Bias Assessment Tool (PROBAST) method.

Results

We obtained 586 results in the initial search. After applying the data extraction process, eliminating duplicate records and performing a comprehensive review according to the aforementioned criteria, 22 studies were finally included in the integrative literature review. The process is detailed in the flowchart (Fig. 1).

Figure 1.

Studies flowchart.

The included studies are mostly from Europe (57%), followed by Asia (24%) and the Americas (14% from North America and 5% from South America) (supplementary material Fig. S1), showing a higher percentage of European studies. In addition, these studies are varied in terms of their typology, with a predominance of cohorts with a retrospective, prospective, or observational view (Fig. 2).

Figure 2.

Description of items. Developed from the studies analyzed (further details in supplementary Table S1). Abbreviations for study types: retrospective (RTR), prospective (PRP), cohort (COH), cross-sectional (TNV), observational (OBV), multicenter (MLC), review (RVN), systematic (SSM), meta-analysis (MTA), longitudinal (LNG), diagnostic (DIG), analysis (ANL), narrative (NRTV). Others such as (pac) patient, triage (TRJ). Countries abbreviated according to international code: Norway (NOR), China (CHN), United States (EEUU), Cyprus (CYP), Belgium (BEL), Turkey (TUR), England (UK), Denmark (DNK), Spain (ESP), Peru (PER), Iran (IRN), Ireland (IRL).

Currently, the effectiveness of pre-hospital COVID-19 detection tools has been evidenced, with early detection in ambulances being more effective than in hospitals, according to some authors.^6,7,13–17 Adequate early detection is essential, since a delay in assessment could increase the risk of clinical worsening in these patients. In the out-of-hospital setting, this screening is an essential preliminary step for assessing the risk of clinical worsening. To this end, the use of different scales is essential, since there is a high heterogeneity in the selected studies.

The variables considered in each scales described in the analyzed studies are shown in Fig. 3 (for more details see supplementary material Table S1).

Figure 3.

Variables presented by the scales, developed from the studies analyzed) (for more details see supplementary data Table S1). Abbreviations of variables: systolic blood pressure (PAS), diastolic blood pressure (PAD), heart rate (FC), oxygen saturation (SatO2), respiratory rate (FT), temperature (T^a), scale name which is an acronym for alert, voice, pain, does not respond (AVPU), Glasgow scale (GCS).

In summary (Fig. 4), a clear prevalence of systolic blood pressure use is observed on all scales and a low use of diastolic blood pressure, present only in CURB-65, SEWS and REMS. Oxygen saturation is not found on all scales; only in three of the twelve. Oxygen support is analyzed by four scales (EWS, NEWS, TREWS, PRIEST). Respiratory rate and temperature are considered in all or most scales. The level of consciousness is a key factor, analyzed by the majority, using the AVPU scale or the Glasgow Coma Scale. Age and comorbidity also influence the assessment of patients. Therefore, the selected variables according to the most commonly used were: systolic blood pressure, heart rate, oxygen saturation, respiratory rate and AVPU scale.

Figure 4.

Scales and results. Developed from the articles analyzed (for more details see supplementary data Table S1). Abbreviations: area under the curve (AUC), sensitivity (S), specificity (E), odds ratio expressing probability of occurrence (OR), results based on percentages of patients (%), triage (TRJ), mortality (MRT), hospital (HSP, prehospital (PHP). Symbology on whether (*) or not (×) performs the analysis of the exposed parameters.

Discussion

The results obtained reflect the importance of identifying the most relevant parameters in the prehospital assessment of COVID-19 patients, such as systolic blood pressure and respiratory rate. These findings allow for a discussion on the differences in the use of scales and their applicability in contexts with limited resources or specific demands. Multiple scales, such as NEWS, MEWS, qSOFA, and others, have been compared with sensitivity, specificity, and AUC data, revealing the superiority of NEWS in predicting clinical deterioration and its applicability in different contexts.

It should be noted that some studies, in addition to prehospital scales, used other methods of assessing the risk of clinical worsening. In this way, some research also highlights the usefulness of scales based on artificial intelligence,^18,19 telephone or web triage,²⁰ as well as rapid physical tests, such as the “1-minute sitting and standing test”.²¹ In addition, conventional triage using colors was also analyzed, although without using specific scales.²²

It is also important to highlight that several scales included in this review, such as qSOFA and NEWS, were initially designed for sepsis but have been adapted due to their predictive value in detecting clinical deterioration in COVID-19 patients, especially those with overlapping symptoms, such as respiratory distress and multi-organ failure. Thus, based on the comparison and efficacy of the prehospital scales, Hu et al.²³ compared SEWS, NEWS, HEWS and MEWS, showing AUC mortality of 0.841, 0.809, 0.821 and 0.670 respectively, with variable sensitivities and specificities. Aygun and Eraybar²⁴ highlighted the MEWS in predicting in-hospital mortality, although it showed a lower performance than the TREWS. Saberian et al.²⁵ compared qSOFA, NEWS, and PRESEP, discarding their predictive power for ICU admission and mortality. Bourn et al.²⁶ analyzed the REMS scale, showing a high capacity to predict ICU admission and mortality and, therefore, the risk of clinical worsening. Martín-Rodríguez et al.²⁷ showed that the NEWS scale has good predictive capacity in patients with and without COVID-19, with better results for those without the disease. Also, Castillo et al.²⁸ and Myrstad et al.²⁹ confirmed the good prognostic value of NEWS for COVID mortality and prediction of clinical deterioration, being superior to qSOFA, SIRS and CRB-65. However, different authors, such as Thomas et al.³⁰ and Zhang et al.³¹ concluded that NEWS is the most effective scale to predict deterioration in patients with COVID-19.

It is important to note the flexibility in the modification of parameters of the scales used. These have adapted to the constant evolution of signs and symptoms associated with SARS-CoV-2, allowing scores to be adjusted for a more accurate detection of the disease. Some variables have been dismissed and changed for others, in order to be carried out externally and to be compatible with the emergency services. These changes have been made based on the health criteria of the moment, the particularities of each country and available resources.^25,29,31 Respiratory system variables such as respiratory rate,¹⁴ oxygen saturation¹⁷ or oxygen supply¹⁸ are highly relevant, as well as age.^20,26 However, oxygen saturation is not present on all scales, only on 3 of the 12. On the other hand, most scales use the AVPU scale to assess the level of consciousness, possibly because it is faster and more simplified than the Glasgow scale and is easier to use at the out-of-hospital level. Beyond the description of each study, it is necessary to reflect on the real applicability of these scales in prehospital nursing practice. The simplicity of tools such as AVPU versus the Glasgow scale, or the inclusion of variables such as oxygen saturation, allows for more agile implementation in resource-constrained environments. However, the low presence of certain critical variables on many scales suggests the need to adapt or develop specific tools for the out-of-hospital setting, for instance, accuracy of predictive models can be affected by lack of critical parameters.^32,33

Currently, the incidence and, above all, the consequences of the COVID infection are lower than those of the 2020–2024 time frame. However, it is important to note that currently, as reported by the Centers for Disease Control and Prevention (CDC) in its August 2025 report, its most predictable model estimates a maximum weekly rate of COVID hospitalization of between 3.8 and 5.9 hospitalizations per 100,000 over the next few months.³⁴ Hence, it continues to be a burden for healthcare. In particular, for certain pathologies,³⁵ age ranges,³⁶ or especially in nursing homes,³⁷ COVID infection is an additional risk to already vulnerable patients. This can also be reflected in the vaccination recommendations in Spain in 2025.³⁸ Therefore, emergency teams continue to face COVID,³⁹ so having tools such as early warning scales can help improve the characterization of patients with respiratory infections, and more so in the current context with rapid diagnostic methods for COVID and other respiratory infections. In this sense, the pandemic context has decreased in intensity and the usefulness of these scales should not be limited to COVID-19. The ability to detect early clinical deterioration is equally valuable in other respiratory infections, chronic decompensated pathologies or emergency situations in nursing homes, it has been shown that some scores presented similar performances for different respiratory conditions.⁴⁰ In this sense, the integration of scales such as NEWS in nursing action protocols can improve decision-making, optimize resources and reduce clinical variability.

Implications in emergency nursing practice

The use of early warning scales like NEWS and MEWS has had a profound impact on emergency nursing practice, especially in patients at risk of rapid clinical deterioration, such as those with COVID-19. During the pandemic, these tools allowed nurses to detect early signs of worsening, prioritizing high-risk patients and optimizing resources in a high-demand context. In particular, the scales helped mitigate the overload on healthcare systems by facilitating a quick, evidence-based response. Moreover, the inclusion of biomarkers into the scales could boost their performance.⁴¹ Additionally, by standardizing assessment and improving communication among healthcare professionals, these scales reduce variability in clinical decision-making. Therefore, their implementation in emergencies can significantly improve the quality of care, reduce errors, and save lives, demonstrating their value not only in managing COVID-19 but also in future emergency care.

Finally, limitations presented in this study are recognized. This conclusion should be interpreted with caution since this study is not free of limitations: (i) the heterogeneity of the included studies, (ii) the number of studies, (iii) the preliminary nature of some of them, (v) the possible methodological bias of the reviewed works. (iv) few studies of prehospital scales related to COVID-19, which is due to the recent origin of the disease to be assessed and the scarcity of out-of-hospital research compared to other disciplines. For all these reasons, it is necessary to increase research on these scales, since they are recent methods and many of their possible uses for reducing time and resources are unknown. The use of an integrative review instead of other review methods deserves justification. In this sense, evidence-based practice (EBP) initiatives have increased the need to develop different types of literature reviews. Although the most widely used methodologies – such as systematic review and meta-analysis – are valuable, they do not always address relevant nursing issues related to care or the impact of diseases and treatments. In this context, the integrative review,⁴² due to its flexible methodological approach, allows the inclusion of studies with different designs, which makes it a potentially key tool for EBP in nursing. Therefore, it is essential to distinguish it from other types of existing studies.

It is important to mention that while this review focuses on early warning scales related to COVID-19, other general triage systems, such as the Emergency Severity Index (ESI), the Canadian Triage and Acuity Scale (CTAS), and the Manchester Triage System (MTS), also play a role in assessing the severity of patients in emergency contexts.⁴³

In conclusion, the implementation of prehospital assessment scales such as NEWS can optimize early detection and management of the risk of clinical worsening in patients with COVID-19, highlighting the need for further research to improve and adapt these tools to the changing needs of health care.

Conclusion

Prehospital scales are in full development and show great potential to improve care in emergency services. Although in our study the evidence is mainly limited to the context of the COVID-19 pandemic, the results obtained confirm the usefulness of tools such as the NEWS scale to predict clinical deterioration in patients in the out-of-hospital setting. However, it is essential to recognize that these scales may have broader applicability. Its use could be extended to other relevant clinical scenarios, such as acute respiratory infections, exacerbations of chronic diseases or future health emergencies. In particular, could help in characterizing patients at risk of clinical worsening, therefore optimizing resources and improving health care. In this sense, continuing with the research and validation of these tools will allow their operation to be adapted to the changing needs of health systems and improve nursing care in contexts of high demand.

CRediT authorship contribution statement

Elena Moreta Gil and Cristina Rivera-Picón carried out the analysis of the data and wrote the first version of the manuscript. José Luis Martín-Conty, Francisco Martín-Rodríguez and Ancor Sanz-García conceptualized the project, coordinated and assisted in the design of the methodology, the analysis of the data and prepared the final manuscript. Rosa Conty-Serrano and Begoña Polonio-López helped with the coordination of the project, and reviewed the included articles and the final version of the manuscript.

Ethical considerations

There are no ethical considerations.

Funding

There is no funding associated with this work.

Conflicts of interest

There is no conflict of interest to declare.