Introduction

Patients considered serious or potentially serious are referred for hospitalization in hospital units and require specific and accurate care, often in Intensive Care Units (ICU).¹ In these units, they may undergo physiological deterioration during hospitalization, usually characterized by change in heart rate, respiratory rate, blood pressure, temperature, and level of consciousness. If identified early, physiological deterioration can be prevented and, in turn, decrease hospital mortality.²

In order to measure the quality and effectiveness of the care provided to critically ill patients during the period of hospitalization, scales were created by the end of the 20th century to meet these objective.³ Severity scales or scores aim to provide resources that establish a parameter to identify the clinical evolution of patients and indicate signs of instability.²

In 1997, the first clinical severity assessment scale was developed in the United Kingdom, called the Early Warning Score (EWS) and consisting of scores that measured the level of physiological deterioration of patients.⁴

Based on the original version of the EWS, the Modified Early Warning Scoring (MEWS), an easy-to-manage bedside instrument, was designed for fast identification of patients at an alarming state and at risk of catastrophic physiological deterioration.⁴ The instrument has values ranging from zero to 14 points that indicate the patient's level of consciousness, blood pressure, heart rate, respiratory rate and body temperature.^5-7

The literature indicates that 79% of patients with cardiopulmonary arrest presented significant change in vital signs before the arrest, and that 54% of these patients died, indicating that this clinical outcome could have been avoided in most patients, in the in-hospital environment, especially if health professionals used the MEWS, a fundamental tool for identifying physiological deterioration in patients allowing the implementation of preventive actions.^7-8 The use of MEWS by nurses, doctors, physiotherapists and nursing technicians to identify the physiological deterioration of patients as early as in the first day in the ICU results in a remarkable survival rate and significant improvement in hospital discharge, as these professionals have the chance to plan and anticipate the care.^9-11

Despite the importance of adopting scales such as MEWS for early detection of physiological deterioration, and the fact that the prognosis of patients admitted to the ICU is directly related to the severity of their condition, the use of this scale is little explored by health care team and studies about its efficacy are incipient.^9-12 Given the above, the aim of the present study was to evaluate the predictive power of a modified early warning scale to identify clinical deterioration in critically ill patients.

Method

This is a descriptive, prospective study with quantitative approach conducted through observational analysis in a university hospital in the state of Minas Gerais that assists the municipal and regional population, with availability of 220 beds. Specifically, the adult ICU has a specialized care structure, consisting of 20 beds, all with multiparametric monitors for accurate real-time assessment of the patient’s clinical condition.¹³

The study population consisted of 214 critically ill ICU patients, according to the sample calculation duly performed by a statistician, which considered the average monthly hospitalizations of this population, with a sampling power of 80% and a significance level of 5%. .

We included clinical patients aged 18 years or older and coming from the urgency and emergency unit and nursing wards. Patients who were sedated, who had undergone surgery, and who had no family members or guardians present at visiting hours to agree by signing the Informed Consent Form were excluded.

Data collection took place from March to November 2018. The clinical data used to apply the MEWS were obtained from electronic medical records of ICU patients, prioritizing two moments: admission to the ICU, to identify the MEWS value at the entry into the sector, and the last note of the desired clinical parameters, before the patient's outcome, discharge or death.

Although the MEWS is not yet validated for the national context, the similarity with the context of the original validation study¹⁴ and its use in other Brazilian studies^4,8 were considered as reasons for its adoption in the present study.

The researchers were present five days a week in the ICU during the collection period, taking turns, and establishing 2 hours of evaluation per day. This resulted in an average of 5 to 6 patients between clinical diagnoses admitted to the ICU and evaluations of their outcomes.

Because the use of the MEWS is not usual in the reality of the ICU analyzed in this research, to meet the objective of evaluating the predictive power of this tool regarding physiological deterioration of patients, the researchers created an instrument and had its face and content validated by three professionals who held a PhD in the area of urgency and emergency, with due expertise in the proposed theme.

The instrument had two parts; the first addressed the characterization of individuals, including sex, age, main body system affected, date of admission to the ICU and length of stay in this sector. The second indicated the registration of clinical characteristics such as level of consciousness, systolic blood pressure, heart rate, respiratory rate and body temperature.

MEWS value of zero indicates that the patient has no physiological deterioration and should continue to be monitored and observed. Score 14 indicates that physiological deterioration is installed and determines the referral to ICU.^5-7 However, score 3 or above 3 is considered an important indicator of physiological deterioration.^5-6 As the MEWS score increases, the need for urgent care also increases, with a proportionality between these variables.⁶

The values corresponding to each variable present in the MEWS were summed for further evaluation of its predictive power of clinical deterioration of critically ill patients.

Data were double-typed into Microsoft Excel® 2013 spreadsheets by two different researchers and analyzed in the Statistical Package for the Social Sciences (SPSS). For statistical analysis, mean, median, standard deviation, minimum and maximum values were calculated for quantitative variables, and the distribution of absolute and relative frequencies was calculated for qualitative variables. A paired t-test was used to compare the means of the MEWS scores of patients at the moment of admission and before the main outcome (death), considering a significance level of 5% (α = 0.05).

This study was submitted to the Ethics Committee to comply with the requirements of Resolution 466/12 of the National Health Council, with protocol 2424.370. The consent form authorizing the access to the intensive care unit and research and manipulation of medical records was obtained from the hospital. To participate in the study, the family members were consulted about their interest and they signed two copies of the Informed Consent Form. Likewise, the patients, who were oriented at the time of the approach, also signed the Informed Consent Form, after being informed of all the peculiarities of the research.

Results

This research aimed to evaluate the predictive power of the MEWS regarding the occurrence of physiological deterioration in ICU patients. An average age of 69.23 years was found, ranging from 18 to 99 years. The shortest hospital stay of patients in this sector was 1 day and the longest was 78 days, and the average length of stay was 10.42 days. Table 1 shows the sociodemographic profile of the patients included in this study.

Table 1 - Characterization of patients according to sex, main body system affected, and patient outcome in the Intensive Care Unit. Uberaba (MG), Brazil, 2018.

Variables	n	%
Sex
Male	96	44.9
Female	118	55.1

Main body system affected
Neurological	35	16.4
Respiratory	59	27.6
Cardiovascular	42	19.6
Gastrointestinal	17	7.9
Renal	16	7.5
Hematological	6	2.8
Endocrine	1	0.5
Immunological	38	17.7
Outcome
Discharge	136	63.6
Death	78	36.4

Most patients were female, representing 118 individuals (55.1%). Respiratory system involvement (27.6%) was the main reason for admission to the adult ICU, followed by cardiovascular (19.6%), immunological (17.7%), and neurological (16.4%) involvement.

Of the 214 patients, 136 (63.6%) were discharged from the ICU and 78 (36.4%) died in the same circumstance. Tables 2 and 3 below show, respectively, death as outcome, showing the interpretation of the MEWS, and then discharge, and the respective mean of the MEWS for comparison.

Table 2 - Modified Early Warning Scoring (MEWS) values of the population that presented death as outcome in the Intensive Care Unit, n = 78. Uberaba (MG), Brazil, 2018

MEWS	Mean	Standard deviation	Standard error of the mean	p value
Admission	3.410	2.0667	0.2340
Death	5.000	3.3342	0.3775	0.001
Difference between admission and death	-1.5897	3.7605	0.4258

The mean MEWS score at admission of the 78 patients (36.4%) who had death as outcome in the ICU was 3.410, and the mean MEWS before their death was 5,000, with a statistically significant p value for physiological deterioration (0.001).

Table 3 - Modified Early Warning Scoring (MEWS) values of the population that presented discharge as outcome in the Intensive Care Unit, n = 136. Uberaba (MG), Brazil, 2018.

MEWS	Mean	Standard deviation	Standard error of the mean	p value
Admission	2.9635	1.80036	.15382
Discharge	2.9781	2.51310	.21471	0.952
Difference between admission and discharge	-.0146	2.85169	.24364

The mean MEWS value of patients who were discharged from the ICU as an outcome during the study period was lower than 3.

Discussion

Researches that show sociodemographic and epidemiological characteristics of ICU patients help to consolidate changes in care strategies. They result in improved unit management, whether from a human, structural, process, or even care management perspective.¹⁵

The present research identified a mean age of 69.23 years in the studied population. This age is similar to an epidemiological, descriptive and retrospective study conducted with 695 patients in the adult ICU, where more than half of hospitalizations (51.2%) were of people aged between 40 and 69 years.¹⁵

When compared to the literature, the average length of stay of patients in this study, 10.42 days, was longer than in other intensive care units, which is usually ≤ 6 days. Significantly shorter stays than expected may indicate the presence of cost-reducing efforts through premature discharge of patients or greater severity of cases (high mortality in the first days/hours); on the other hand, significantly longer stay than expected may indicate deficiencies and difficulties in the care provided.¹⁶

Regarding sex, the result of the present study contradicts the findings of the literature, as the majority of the population was female (55.1%). Demographic characteristics in ICUs generally converge to male predominance, of more than 50%.¹⁷

This contradiction stems possibly from the modern massive incentive of preventive strategies directed at the male population for the control of specific diseases and the consequent awareness and adherence of men to health programs, which has minimized the health problems and reduced their stay in ICUs.^18-19

Respiratory tract diseases were the main cause of hospitalization, followed by cardiovascular diseases. These findings are in line with the results of an epidemiological study conducted with 171 patients where cardiorespiratory diseases stood out as the main cause of ICU admission.²⁰

The clinical demands of ICU patients stand out against the surgical ones, possibly due to the aggravation of chronic problems. Acute clinical conditions and age over 80 years are associated with mortality in survival studies with follow-up of less than 30 days, and situations such as lowering of consciousness, use of mechanical ventilation and respiratory diseases are factors that worsen this outcome.²⁰

As for the clinical outcome, most patients in this study were discharged from the sector with an average MEWS value of less than 3, which shows no physiological deterioration,²¹ and yet, a considerable number of patient had death as outcome. Disease severity is generally used to predict mortality in ICU patients, with a high mortality rate in this setting, ranging from 5.4 to 33% in the international scenario.²¹ In the national scenario, the mortality rate is above international rates, from 36.9 to 50.74%.^22-24

Patients who presented death as clinical outcome in this study had a mean MEWS value at admission of 3,410, and 5,000 at the last assessment of clinical parameters before death. The results showed a statistically significant difference in the association between MEWS values and the outcome (mortality) of ICU patients, with a p value of 0.001.

Patients who scored ≥ 5 points on the MEWS were associated with a higher risk of death and ICU admission.²⁵ This fact reinforces the need for the use of the MEWS by health professionals in emergency wards to recognize early physiological deterioration before ICU admission, and to avoid it.²⁴

This scale has the main goal of facilitating the communication between health teams about the physiological deterioration of the patient so as to prevent it. In this sense, the use of this tool may avoid the transference of the patient to the ICU, and when transference is inevitable, it may ensure that this occurs without delay.^4,25

Patients admitted to the ICU have a variable morbidity and mortality and usually present warning signs a few days before hospitalization,⁸ which is shown in the present study by the already high MEWS value at the moment of admission (MEWS = 3.410). This shows that physiological changes, which indicate the worsening of clinical status, may early indicate potential critically ill patients, who will need special monitoring in nursing wards or emergency rooms.⁸

When patients come from the wards or emergency rooms, they usually have high morbidity at the moment of admission to the ICU, which is indicated by MEWS values greater than 3 points in more than 70% of the population, with an average of more than 5 points.⁸ These data were confirmed in the present study. Despite the lower mortality rate than that identified in the national scenario,²³ there was a strong association between the presence of critical scores (≥ 3 points) at admission and physiological deterioration, with a significant increase of the MEWS average to 5,000 in the majority of patients in their outcome.

Therefore, the adoption of the MEWS in the in-hospital health care context must be encouraged. Given its potential to predict physiological deterioration in patients, MEWS may prevent undesirable and irreversible outcomes, besides being easy to handle and to be interpreted by health professionals.⁸

Conclusion

MEWS values greater than or equal to 3 were related to clinical deterioration of patients in the ICU; this scale constitutes therefore an important indicator at the moment of admission. There was significant statistical relevance in the association of MEWS and death in the ICU, which shows the predictive power of this scale for irreversible and unwanted outcomes.

The limitation of this study was the fact that data collection took place in only one ICU, which hinders the generalization of the findings. The main contribution and implication for the practice is that the efficacy of the MEWS was demonstrated, when the physiological deterioration of patients in the in-hospital environment was detected, thus indicating the importance of the MEWS for the early and effective decision making regarding patients, with the main intention of enabling evidence-based practice to improve professional competence and patient safety.

The MEWS is considered a reliable and effective instrument to identify physiological deterioration in patients, and it is therefore recommended for the prevention of cardiorespiratory arrest in adults.

References

1. Dobkin C, Finkelstein A, Kluender R, Notowidigd MJ. The economic consequences of hospital admissions. Am Econ Rev [Internet]. 2018 [acesso em 2019 ago 23];108(2):308–52. Disponível em: https://doi.org/10.1257/aer.20161038

2. Miranda JO, Camargo CL, Nascimento Sobrinho CL, Portela DS, Monaghan A, Freitas KS, et al. Tradução e adaptação de um escore pediátrico de alerta precoce. Rev Bras Enferm [Internet]. 2016 [acesso em 2018 nov 12];69(5):833-41. Disponível em: http://dx.doi.org/10.1590/0034-7167-2015-0096

3. Silva LC, Nogueira LS, Settervall CH, Sousa RM, Padilha KG. Desempenho de índices de gravidade para estimar risco de morte em Unidades de Terapia Intensiva. Rev Esc Enferm USP. 2012;46(4):846-50.

4. Rocha TF, Neves JG, Viegas K. Escore de alerta precoce modificado: avaliação de pacientes traumáticos. Rev Bras Enferm [Internet]. 2016 [acesso em 2018 nov 12];69(5):850-5. Disponível em: http://dx.doi.org/10.1590/0034-7167-2015-0145

5. Kruisselbrink R, Kwizera A, Crowther M, Robichaud AF, O’Shea T, Nakibuuka J, et al. Modified Early Warning Score (MEWS) identifies critical illness among ward patients in a resource restricted setting in Kampala, Uganda: a prospective observational study. PLos ONE [Internet]. 2016 [acesso em 2018 abr 22];11(3):e0151408. Disponível em: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0151408

6. Westphal GA, Lino AS. Rastreamento sistemático é a base do diagnóstico precoce da sepse grave e choque séptico. Rev Bras Ter Intensiva [Internet]. 2015 [acesso em 2018 abr 26];27(2):96-101. Disponível em: http://www.scielo.br/scielo.php?script=sci_arttext&pid=S0103-507X2015000200096

7. Freitas CM, Preto EP, Nascimento CAF. Intervenções de enfermagem na monitorização da deterioração clínica da pessoa em enfermaria hospitalar: uma revisão integrativa. Rev Enferm Referência [Internet]. 2017 [acesso em 2019 ago 22];4(14):121-32. Disponível em: http://www.scielo.mec.pt/scielo.php?script=sci_abstract&pid=S0874-02832017000300013&lng=pt&nrm=iso

8. Tavares RC, Vieira AS, Uchoa LV, Peixoto Júnior AA, Meneses FA. Validação de um escore de alerta precoce pré-admissão na unidade de terapia intensiva. Rev Bras Ter Intensiva. 2008;125(20):124-7.

9. Johnson S, Nileswar A. Effectiveness of Modified Early Warning Score (MEWS) in the outcome of in-hospital adult cardiac arrests in a tertiary hospital. J Pulm Respir Med [Internet]. 2015 [acesso em 2018 abr 20];5:4. Disponível em: http://dx.doi.org/10.4172/2161-105X.1000285

10. Salottolo K, Carrick M, Johnson J, Gamber M, Bar-Or B. A retrospective cohort study of the utility of the modified early warning score for interfacility transfer of patients with traumatic injury. BMJ Open [Internet]. 2017 [acesso em 2018 nov 14];7(5):e016143. Disponível em: https://bmjopen.bmj.com/content/bmjopen/7/5/e016143.full.pdf

11. Nishijima I, Domari SO, Domari S, Mae T, Risa I, Chisato K, et al. Características sociodemográficas e epidemiológicas de pacientes internados em uma Unidade de Terapia Intensiva. Rev Enferm UERJ. 2016;25:e7736.

12. Kyriacos U, Jelsma J, James M, Jordan S. Monitoring vital signs: development of a modified early warning scoring (MEWS) system for general wards in a developing country. PLoS One [Internet]. 2014 [acesso em 2018 out 11];9(1):e87073. Disponível em: https://www.ncbi.nlm.nih.gov/pubmed/24475226

13. Mário Palmério Hospital Universitário. Quem Somos [Internet]. Uberaba (MG); 2019 [acesso em 2019 maio 06]. Disponível em: http://www.hospitalmariopalmerio.uniube.br/quemSomos.php

14. Subbe CP, Kruger M, Rutherford P, Gemmel L. Validation of a modified early warnig score in medical admissions. QJM [Internet]. 2001 [acesso em 2018 ago 17];94(10):521-6. Disponível em: https://academic.oup.com/qjmed/article/94/10/521/1558977

15. Rodriguez AH, Bubi MB, Zandonadil OF, Zandonadill G, Rodriguezi MJ. Características epidemiológicas e causas de óbitos em pacientes internados em terapia intensiva. Rev Bras Enferm [Internet]. 2016 [acesso em 2018 nov 12];69(2):229-34. Disponível em: http://www.scielo.br/scielo.php?script=sci_arttext&pid=S0034-71672016000200229&lng=pt&tlng=pt

16. Roque KE, Tonini T, Melo EC. Eventos adversos na unidade de terapia intensiva: impacto na mortalidade e no tempo de internação em um estudo prospectivo. Cad Saúde Pública [Internet]. 2016 [acesso em 2018 abr 12];32(10):e00081815. Disponível em: http://dx.doi.org/10.1590/0102-311X00081815

17. Perão OF, Bub MBC, Zandonadi GC, Martins MA. Características sociodemográficas e epidemiológicas de pacientes internados em uma Unidade de Terapia Intensiva de adultos. Rev Enferm UERJ [Internet]. 2017 [acesso em 2018 out 11];25:e7736. Disponível em: https://doi.org/10.12957/reuerj.2017.7736

18. Pereira SM, Lima FES, Gonçalves FIR, Silva NBS, Santos WQ, Sousa PV, et al. Acessibilidade dos homens a unidade básica de saúde. Rev Eletrônica Acervo Saúde [Internet]. 2019 [acesso em 2019 ago 23];11(2):e108. Disponível em: https://doi.org/10.25248/reas.e108.2019

19. Bibiano AMB, Moreira RS, Tenório MMGO, Silva VL. Fatores associados à utilização dos serviços de saúde por homens idosos: uma revisão sistemática da literatura. Ciênc Saúde Colet [Internet]. 2019 [acesso em 2019 ago 22];24(6):2263-78. Disponível em: http://www.cienciaesaudecoletiva.com.br/artigos/fatores-associados-a-utilizacao-dos-servicos-de-saude-por-homens-idosos-uma-revisao-sistematica-da-literatura/16407?id=16407

20. Silva LS, Mont’Alverne DG, Medeiros AI, Silva AG, Carvalho EM. Características dos pacientes sob assistência fisioterapêutica na UTI de um hospital universitário: estudo epidemiológico transversal. Rev Fisioter Saúde Funcional. 2015;5(1):50-8.

21. Bonfada D, Santos MM, Lima KC, Garcia-Altés A. Análise de sobrevida de idosos internados em Unidades de Terapia Intensiva. Rev Bras Geriatr Gerontol [Internet]. 2017 [acesso em 2018 dez 10];20(2):197-205. Disponível em: http://dx.doi.org/10.1590/1981-22562017020.160131

22. Hon KL, Luk MP, Fung WM, Li CY, Yeung KL, Liu PK, et al. Mortality, length of stay, bloodstream and respiratory viral infections in a pediatric intensive care unit. J Crit Care [Internet]. 2017 [acesso em 2018 abr 11];38:57-61. Disponível em: https://www.sciencedirect.com/science/article/pii/S0883944116305482. Epub 2016 set 30.

23. Gershengorn HB, Harrison DA, Garland A, Wilcox ME, Rowan KM, Wunsch H. Association of intensive care unit patient-to-intensivist ratios with hospital mortality. JAMA Intern Med [Internet]. 2017 [acesso em 2018 jun 05];177(3):388-96. Disponível em: https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2598511

24. Guia CM, Biondi RS, Sotero S, Lima AA, Almeida KJ, Amorim FF. Perfil epidemiológico e preditores de mortalidade de uma unidade de terapia intensiva geral de hospital público do Distrito Federal. Comun Ciênc Saúde. 2015;26(1-2):9-19.

25. Vieira AM, Parente EA, Oliveira LS, Queiroz AL, Bezerra ISM, Rocha HAL. Características de óbitos dos pacientes internados em uma unidade de terapia intensiva de hospital terciário. J Health Biol Sci [Internet]. 2019 [acesso em 2019 maio 12];7(1):26-31. Disponível em: http://docs.bvsalud.org/biblioref/2019/01/969718/5-1999.pdf

Corresponding author

Juliana da Silva Garcia Nascimento

E-mail: mestradounesp28@yahoo.com.br

Endereço: Rua Osório Joaquim Guimarães, nº 281, Ap: 203, Bloco 05, Parque São Geraldo. Uberaba, Minas Gerais. Brasil.

CEP: 38030-130

Author Contributions

1 – Juliana da Silva Garcia Nascimento

Research project design and planning, data collection, analysis and interpretation, writing and critical review.

2 – Gabriela de Oliveira Macedo

Research project design and planning, data collection, analysis and interpretation, writing and critical review.

3 – Giovanna Beirigo Borges

Research project design and planning, data collection, analysis and interpretation, writing and critical review.

How to cite this article

Nascimento JSG, Macedo GO, Borges GB. Predictive power of the early warning scale for clinical deterioration of critically ill patients. Rev. Enferm. UFSM. 2019 [Acesso em: Anos Mês Dia]; vol.10 e5: 1-14. DOI:https://doi.org/10.5902/2179769238300