The COVID-19-12O-score has been validated to determine the risk of respiratory failure in patients hospitalized for COVID-19. Our study aims to assess whether the score is effective in patients with SARS-CoV-2 pneumonia discharged from a hospital emergency department (HED) to predict readmission and revisit.
MethodRetrospective cohort of patients with SARS-CoV-2 pneumonia discharged consecutively from an HUS of a tertiary hospital, from January 7 to February 17, 2021, where we applied the COVID-19-12O -score, with a cut-off point of 9 points to define the risk of admission or revisit. The primary outcome variable was revisit with or without hospital readmission after 30 days of discharge from HUS.
ResultsWe included 77 patients, with a median age of 59 years, 63.6% men and Charlson index of 2. 9.1% had an emergency room revisit and 15.3% had a deferred hospital admission. The relative risk (RR) for emergency journal was 0.46 (0.04–4.62, 95% CI, p=0.452), and the RR for hospital readmission was 6.88 (1.20–39.49, 95% CI, p<0.005).
ConclusionsThe COVID-19-12O -score is effective in determining the risk of hospital readmission in patients discharged from HED with SARS-CoV-2 pneumonia, but is not useful for assessing the risk of revisit.
La escala COVID-19-12O se ha validado para determinar el riesgo de insuficiencia respiratoria en pacientes hospitalizados por COVID-19. Nuestro estudio pretende evaluar si la escala es efectiva en pacientes con neumonía por SARS-CoV-2 dados de alta desde un servicio de urgencias hospitalario (SUH) para predecir el reingreso y revisita.
MétodoCohorte retrospectiva de pacientes con neumonía por SARS-CoV-2 dados de alta de forma consecutiva desde un SUH de un hospital terciario, del 7 de enero al 17 de febrero de 2021, donde aplicamos la escala COVID-19-12O, con un punto de corte de 9 puntos para definir el riesgo de ingreso o revisita. La variable de resultado principal fue la revisita con o sin reingreso hospitalario tras los 30 días de su alta desde el SUH.
ResultadosSe incluyeron 77 pacientes, con una edad mediana de 59 años, 63,6 % hombres e índice Charlson de 2. El 9,1 % tuvieron revisita a urgencias y en el 15,3 % se produjo un ingreso hospitalario diferido. El riesgo relativo (RR) para revista de urgencias fue 0,46 (0,04−4,62, IC 95 %, p=0,452), y el RR para el reingreso hospitalario de 6,88 (1,20–39,49, IC 95 %, p<0,005).
ConclusionesLa escala COVID-19-12O es efectiva en determinar el riesgo de reingreso hospitalario en pacientes dados de alta desde el SUH con neumonía por SARS-CoV-2, pero no es útil para valorar el riesgo de revisita.
The clinical spectrum of SARS-CoV-2 infection in adults is highly varied, ranging from asymptomatic and mild disease manifestations to severe illness with pneumonia, acute respiratory distress syndrome (ARDS), or multiple organ dysfunction1,2.
By identifying the risk factors, various tools have been developed and implemented to detect patients who are at higher risk of falling ill/being admitted to hospital or to predict their prognosis, among other aspects3–8.
The COVID-19-12O score, proposed by Lalueza et al., has been validated for the Spanish population and has been implemented quickly and easily (“at bedside”). In addition, it is capable of differentiating, early on and with excellent precision, among patients hospitalized due to COVID-19, who are at higher risk of developing respiratory failure, thereby enabling provision of the most appropriate therapeutic measures and optimal use of the available resources5,9.
However, to date it is not yet known whether the score is capable of predicting revisits to the Hospital Emergency Department (HED) in patients with pneumonia who have been discharged from this service. This is an important question given that 12% of patients with pneumonia who are discharged end up being readmitted/revisiting the HED due to a worsening clinical picture related to their SARS-CoV-2 infection10, indicating a failure in outpatient management or poor initial classification due to unforeseeable progression of the disease in these cases. All of this translates into higher associated expenditure and unnecessary prolongation of the healthcare process.
Therefore, this study aims to evaluate whether the score is effective at predicting readmission/revisits to the emergency department in patients with SARS-CoV-2 pneumonia who are discharged from an HED.
Material and methodsStudy design and patientsRetrospective cohort with patients diagnosed with SARS-CoV-2 pneumonia in the HED of Hospital General Universitario Dr. Balmis de Alicante, between 7 January and 17 February 2021, when the B.1.1.7 (ALPHA) variant was dominant11, who did not meet admission criteria. These patients were classified within a mild-moderate risk category according to the COVID-19-12O score12. An optimum cut-off point of 9 points was used because it offered greater discriminating capacity with a sensitivity of 82.66%, specificity of 71.96%, and area under the curve of 0.84.
This study follows the criteria established by the Declaration of Helsinki and the World Medical Association and the European Union Guideline for Good Clinical Practice. The study was approved by the hospital’s Ethics Committee (record no. 2020-8) and informed consent was not required due to the observational design.
Participant selectionInclusion criteria: a) confirmed diagnosis of COVID-19 in the HED; b) presence of pneumonia without respiratory failure, with or without alterations in one or multiple analytical risk markers; c) patients discharged from the HED under outpatient care via a specialised, high-resolution consultation unit.
Infection was confirmed via reverse transcription polymerase chain reaction testing for SARS-CoV-2 in nasopharyngeal aspiration. The absence of respiratory failure was defined as: respiratory rate (RR) of fewer than 22 breaths per minute, baseline oxygen saturation at rest of ≥95% or negative walking test (oxygen saturation >94% after walking 50m on flat ground). Pneumonia was defined as the presence of radiological alveolar-interstitial infiltrates confirmed by the diagnostic radiology service.
Exclusion criteria: oral intolerance, pregnant individuals, severe immunosuppression (transplant of solid organs or hematopoietic tissue in the previous six months), difficulty maintaining outpatient follow-up or lack of adequate social support.
Exploratory variables and measurement instrumentThe exploratory variables were: demographic data, underlying comorbidities, and lab results from the electronic medical histories. The laboratory variables were dichotomized according to the reference values established at the centre.
The measuring instrument applied to the patients prior to discharge from the HED was that proposed by Lalueza et al. The COVID-19-12O score9 is based on 5 variables: age (under 55 years: 0 points, between 55 and 75: 2, older than 75: 3), lymphocytes (less than 500/μL: 4 points, between 500 and 100: 3, over 1000: 0), 02 saturation (less than 92%: 9 points, between 92 and 96%: 1, higher than 96%: 0), lactate dehydrogenase (LDH) (less than 280 U/L: 0 points, between 280 and 380: 1, higher than 380: 4) and C-reactive protein (CRP) (less than 4mg/dL: 0 points, between 4 and 14: 1, higher than 12: 3); the scoring varies between 0 and 23 points, with a lower score denoting a lower estimated probability of respiratory failure and vice versa (Table 1).
Predictive score proposed by Lalueza et al.9.
| Variables | Cut-off pointsAssociated points | Points | Probability of respiratory failure | ||
|---|---|---|---|---|---|
| AGE | <55 years | 55−75 years | 75 years | 0 | 1.1% |
| 0 points | 2 points | 3 points | 1 | 1.6% | |
| 2 | 2.6% | ||||
| LYMPHOCYTES | <500 | 500−1000 | >1000 | 3 | 4.0% |
| cel/μL | 4 points | 3 points | 0 points | 4 | 6.0% |
| 5 | 9.1% | ||||
| SpO2% | <92 | 92−96 | >96 | 6 | 13.5% |
| 9 points | 1 point | 0 points | 7 | 19.4% | |
| 8 | 27.3% | ||||
| LDH | <280 | 280−380 | >380 | 9 | 36.8% |
| U/L | 0 points | 1 point | 4 points | 10 | 47.5% |
| 11 | 58.4% | ||||
| CRP | <4 | 4−12 | >12 | 12 | 68.6% |
| mg/dL | 0 points | 1 point | 3 points | 13 | 77.2% |
| 14 | 84.0% | ||||
| 15 | 89.1% | ||||
| 16 | 92.7% | ||||
| 17 | 95.2% | ||||
| 18 | 96.8% | ||||
| 19 | 97.9% | ||||
| 20 | 98.6% | ||||
| >20 | 99.1% | ||||
Revisit to the HED without hospital admission and revisit with hospital admission in the 30 days following discharge from the HED under outpatient monitoring.
Statistical analysisQualitative variables: absolute and relative frequency in percentages of each one of the categories. Qualitative variables: median and 25th and 75th percentiles. For the analysis of the association of the explanatory variables and revisit to the emergency department, we used the chi-squared test. To quantify the magnitude of the association, we used relative risk (RR) with a 95% confidence interval (95% CI). For hypothesis testing, a statistical significance level of p<0.05 was used.
The data analysis was performed using the program IBM SPSS Statistics v25 (Armonk, NY, USA).
ResultsA total of 77 patients were included with SARS-CoV-2 pneumonia who were discharged from the HED. The mean age was 59 years (interquartile range [IQR] 48−66), 63.6% were male and the median Charlson Index was 2. The rest of the sociodemographic and clinical variables are displayed in Table 2.
Characteristics of patients with SARS-CoV-2 pneumonia discharged from the hospital emergency department.
| Total, n=77Median (IQR) | Reference values | |
|---|---|---|
| Demographics | ||
| Age | 59 (48−66) years | |
| Male | 57.9% | |
| Comorbidities | ||
| Charlson Comorbidity Index | 2 (1−3) | |
| Estimated 10-year survival | 90.2% (77.4−95.8) | |
| Analytical data | ||
| Oximetry-room air | 97% (95−98) | |
| PaO2/FiO2 | 395 (350−447) | |
| Lymphocytes | 1310 mm3 (930−1720) | 1000−3700 mm3 |
| C-reactive protein | 2.2mg/dL (0.9−4.9) | <0.5mg/dL |
| Procalcitonin | 0.07ng/mL (0.04−0.07) | <0.5ng/mL |
| Ferritin | 390mg/L (178−679) | 30−400mg/L |
| Lactate dehydrogenase | 245 U/L (208−302) | <250 U/L |
| D-dimer | 0.6mg/mL (0.4−0.8) | <0.5mg/mL |
| Troponin T | 7 ng/L (1−11.5) | <14ng/L |
| Potassium | 4 mmol/L (3.8−4.3) | 3.5−4.5mmol/L |
| Urea | 34mg/dL (26−40) | 17−49mg/dL |
| Glucose | 114mg/dL (100−134) | 74−106mg/dL |
IQR: interquartile range.
A total of 25% of patients revisited the hospital or were readmitted: 7 patients (9.1%) revisited the emergency department within 30 days post-discharge from the HED and 12 (15.6%) revisited the emergency department and were readmitted to the hospital. One patient (1.3%) died following hospital readmission 7 days after assessment in the emergency department. Of the admitted patients, the average number of days between symptom onset and admission was 8.5 days (IQR 7.0–12.7). The reasons for revisit or admission were persistence and worsening of the clinical symptoms of SARS-CoV-2 infection.
The RR of a revisit to the emergency department without admission in patients with a score greater than or equal to 9 points was 0.46 (0.04–4.623, 95% CI, p= 0.452), while the RR of a revisit to the emergency department with hospital admission was 6.88 (1.20–39.49, 95% CI, p< 0.05).
DiscussionIn a pandemic setting with an overworked healthcare system, our study shows that, in a real life setting, the COVID-19-12O score is effective at predicting hospital readmission in patients with SARS-CoV-2 pneumonia who were discharged from an HED. Patients with a score that is greater than or equal to 9 points were observed to have a six-fold higher risk of readmission. However, we were not able to demonstrate its capacity to predict revisits to the emergency department. Nevertheless, the original score was developed based on hospitalized patients and this study focused on outpatient subjects with COVID-19 pneumonia who did not require admission.
Regarding the risk factors measured by the COVID-19-12O score, age is the most important factor for determining the probability of respiratory failure. While SARS-CoV-2 infection can occur at any age, middle-aged and older adults are the most commonly affected and have a higher probability of experiencing more serious clinical pictures, particularly male subjects. The cases with the most comorbidity (such as obesity, hypertension, or metabolic syndrome) presented a higher risk of readmission due to a greater predisposition to experiencing complications and higher mortality. Severe forms do not only affect older patients or those with associated comorbidities, but adults with few or no comorbidities as well. Age is a datum that is easier to obtain in the emergency department than calculating the Charlson Comorbidity Index, though both present a similar discriminating capacity12–14.
Analytical inflammation findings are particularly important to calculating the score: their alteration precedes respiratory decline and indicates a worse prognosis, despite generally being unspecific and common in pneumonia14,15. More pronounced lymphopenia has been associated with critical illness and higher mortality; similarly, elevated LDH and CRP values are linked to severe respiratory failure as they are a reflection of the inflammatory state. It is not known why older patients, male patients and those with comorbidities or underlying conditions “become more inflamed”, but the reality remains that they have a greater inflammatory response which triggers more severe clinical presentations, with more complications and increased mortality9,12,16,17. According to some recent studies, lymphocyte count presents the weakest association while CRP and procalcitonin (PCT) present the strongest. The PCT provided better prognostic discrimination18.
Lastly, when calculating the score for this scale, peripheral pulse oximetry is easier to measure than PaO2 (partial pressure of oxygen), and is a reliable alternative for obtaining an estimation of PaO2/FiO2 (fraction of inspired oxygen) and is very useful for measuring the progression of respiratory failure9.
Prognosis throughout the course of the disease is unknown14 in the case of patients who, at diagnosis, present with a high COVID-19-12O score (greater than or equal to 9 points) and are not suitable candidates for admission. It would be advisable to design an individualized follow-up strategy that consider the patient’s clinical presentation, associated comorbidities, decompensation of one or multiple of these, and their social situation. The following could be contemplated: admission to the emergency department for observation, establishing a care plan with enhanced continuity of care through primary care, the Home Hospitalisation Unit, or specific consultations implemented at other hospitals4. This would facilitate a uniform and adequate distribution of available resources.
LimitationsThis study has some limitations. Firstly, since it is a single-centre and retrospective study with a limited patient sample size, it has inherent limitations such as potential selection bias and population demographic characteristics bias. Likewise, we cannot ensure the validity of the study in the context of other SARS-CoV-2 variants. We have not studied other lower cut-off points that could have also defined a greater risk of readmission.
ConclusionsThe 5-variable score proposed by Lalueza et al. could be useful in HEDs to predict hospital readmission in discharged patients with SARS-CoV-2 pneumonia. It may also help to establish a therapeutic or follow-up plan to improve patient care and prognosis in an individual and equitable manner, thereby guaranteeing care continuity.
FundingThe authors state they have not received any funding to carry out this study.
Conflicts of interestThe authors declare that they do not have any conflicts of interest.
