NYU-EDA in Modelling the Effect of COVID-19 on Patient Volumes in A Finnish Emergency Department

doi:10.21203/rs.3.rs-102314/v1

Download PDF

Research article

NYU-EDA in Modelling the Effect of COVID-19 on Patient Volumes in A Finnish Emergency Department

https://doi.org/10.21203/rs.3.rs-102314/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 11 Dec, 2020

Read the published version in BMC Emergency Medicine →

You are reading this latest preprint version

Background: Emergency departments (EDs) worldwide have been in the epicentre of the novel coronavirus disease (COVID-19). However, the impact of the pandemic and national emergency measures on the number of non-COVID-19 presentations and the assessed acuity of those presentations remain uncertain.

Methods: We acquired a retrospective cohort containing all ED visits in a Finnish secondary care hospital during years 2018, 2019 and 2020. We compared the number of presentations in 2020 during the national state of emergency, i.e. from March 16 to June 11, with numbers from 2018 and 2019. Presentations were stratified by New York University Emergency Department Algorithm (NYU-EDA) in order to evaluate changes in presentations with different acuity levels.

Results: A total of 27,526 presentations were observed. Compared to previous two years, total daily presentations were reduced by 23 % (from 113 to 87, p<.001). In NYU-EDA classes, Non-Emergent visits were reduced the most by 42 % (from 18 to 10, p<.001). Emergent presentations were reduced by 19 to 28 % depending on the subgroup (p<.001). Number of injuries were reduced by 25 % (from 27 to 20, p<.001). The NYU-EDA distribution changed statistically significantly with 4 percent point reduction in Non-Emergent visits (from 16 % to 12 %, p<.001) and 0.9 percent point increase in Alcohol-related visits (from 1.6 % to 2.5 %, p<.001).

Conclusions: We observed a significant reduction in total ED visits in the course of national state of emergency. Presentations were reduced in most of the NYU-EDA groups irrespective of the assessed acuity. A compensatory increase in presentations was not observed in the course of the 3 month lockdown. This implies either reduction in overall morbidity caused by decreased societal activity or widespread unwillingness to seek required medical advice.

Critical Care & Emergency Medicine

COVID-19

SARS-COV-19

Emergency Department

NYU-EDA

COVID-19 is a an infectious disease caused by severe acute respiratory syndrome virus 2 (SARS-CoV-2) (1). The spectrum of the disease ranges from asymptomatic to severe, sometimes requiring prolonged treatment in intensive care unit (ICU). The estimated fatality rate is approximately 1.1% and the proportion of patients that require hospitalization range from 1.1–18.4% with hospitalization and mortality rates sharply increasing in older population (2). COVID-19 puts serious strain on ICU and inpatient capacity of healthcare systems and has been mitigated by suspending non-urgent care (3, 4). Emergency care however has remained fully operational as public health measures against COVID-19 are unlikely to have major effect on incidence of non-infectious emergencies. Even though the progression of COVID-19 pandemic is followed closely, the effect of the pandemic on non-COVID-19 -related emergency care has so far received less attention.

COVID-19 is the third epidemic caused by coronaviruses in the 21st century (5). Serious acute respiratory syndrome (SARS) epidemic emerged in China in 2003 and spread to several countries. Middle East respiratory syndrome virus (MERS) emerged in 2012 in Saudi Arabia and spread to 27 countries (5). The difference between COVID-19 and the previous coronavirus epidemics is that the clinical picture of COVID-19 is in most patients less severe, which makes it easier for the disease to spread as patients with minimal symptom remain active in the society (6). Even though the scale of previous epidemics has been considerably smaller, the literature from heavily affected areas provides valuable information on patient flow dynamics in the face of an epidemic. Reports from SARS outbreak in Taiwan showed a decline of the number of ED patients by approximately 30% while the proportions of different patient segments remained largely unchanged (7). A similar decline was reported in Hong Kong (8). In a Canadian study, SARS outbreak resulted in an overall decrease of ED visits where the reduction was mostly explained by the lower attendance of pediatric patients (9). On the contrary, SARS epidemic in Singapore resulted in an overall increase of ED presentations by about 30%. The increase was explained by increased number of severe patients brought by ambulance and patients seeking help for respiratory symptoms (10).

Although the reduction in emergency department presentations amidst the COVID-19 pandemic has been observed widely, published data has been mostly anecdotal. In a recent report from Northern Italian tertiary center, the number of ED visits decreased by almost 50% compared to the previous year but the overall composition of patient segments seemed largely unchanged. A noteworthy observation is the increased hospitalization rate suggesting a more serious patient profile than usually (11). Recent Portuguese report cited a reduction of 30–50% across emergency and non-emergency presentations (12). A report from Israeli tertiary center reported a reduction of 32% in the volume of surgical patients amidst COVID-19 pandemic and there was a moderate negative correlation in emergency department visits and confirmed COVID-19 cases in Israel (13). United States nationwide report from Centers for Disease Control reported a 42% reduction in emergency department visits. The largest proportional reductions were in musculoskeletal disorders and the highest increases in mental disorders (14).

Finland has not been spared by the impact of COVID-19. Until September 13, 2020 there has been 8,512 confirmed cases and 337 deaths with total of 789,500 tests performed. To protect the population from the consequences of a widespread disease, the Government of Finland declared a national state of emergency on March 16, 2020. In practice this meant closing the schools, closure of nightlife establishments and restaurants, and restricting movement to and from densely populated Helsinki capital area to the rest of the country. In practice this meant closing the schools together with rapid introduction of e-learning, closure of nightlife establishments and restaurants, and restricting movement to and from the most infected and densely populated Helsinki capital area to the rest of the country. The state of emergency was lifted on June 16, 2020, but extensive restrictions to certain areas of life (travel, mass gatherings) are still applied.

The stratification of the urgency of presenting ED patients based on administrative data is not trivial since triage classification differs between institutions and practitioners (15). Direct utilization of ICD-10 classification (16) is also insufficient since many categories contain a wide range of disorders of the organ system ranging from benign to acutely life-threatening. New York University Emergency Department Algorithm (NYU-EDA) was developed to provide a systematic reference for estimation of urgency, preventability and level of care needed for patients with different diagnoses by conducting a full chart review on the Clinical Modification of 10th revision of the International Statistical Classification of Diseases and Related Health Problems (ICD-10-CM) based on 3,500 ED records (17). The reference contains statistical estimation of urgency for different diagnoses (e.g. for ICD-10-CM diagnosis R104 Unspecified chest pain, 68% of are deemed to require ED care and 32% of patients are deemed to be treatable by primary care physician).

The goal of this study is to assess the effect of state of emergency and COVID-19 pandemic on quality and quantity of presentations in the Emergency Department of a Finnish secondary care hospital.

2.1 Data

Kanta-Häme Central Hospital is a secondary care hospital with catchment population of 171 000 people. The emergency department of Kanta-Häme Central hospital provides acute and critical care from 44 000 to 47 000 patients every year and 16 to 17 per cent of the population uses its services yearly. For the purposes of this investigation, we reviewed all patients presenting at our ED between March 16 and June 11 in 2018, 2019 and 2020. When reporting the results, this period is used unless other specified. Mean from years 2018 and 2019 was used as a reference to evaluate the impact of COVID-19 on ED service demand during year 2020. In addition to evaluating the ED as a whole, we grouped the patients according to NYU-EDA to assess the effect on different levels of presentation acuity. We investigated absolute and proportional changes in the number of daily presentations stratified by NYU-EDA groups and also analyzed changes in proportions of respective groups. Institutional approval for the study was obtained. .

2.2 NYU-EDA and Conversion

Since NYU-EDA is based on clinical modification of Intenational Classification of Disease (ICD-10-CM) and Finnish healthcare providers use standard ICD-10, direct use of the classification algorithm was not possible. For this reason, diagnoses covering 95% of the visits in this study were selected for manual review where equivalency chart for ICD-10-CM and WHO ICD-10 was created with the goal of placing diagnoses to equivalent categories. The full chart is available from authors upon request. Since diagnosis to NYU-EDA conversion was performed to 95% of the presentations, the resulting counts were corrected by a factor of 0.95^− 1 to reflect the actual number of presentations in our ED.

2.3 Statistical tools

Mann-Whitney U test was performed for daily presentation incidence. The chi-square (χ2) test was used to compare proportions in different NYU-EDA groups. Statistical significance was specified as P < .05. Statistical tests were performed using Python 3.7.8.

A total of 27,526 episodes of care were included into the analysis. When comparing the mean of 2018 and 2019 with 2020, the total daily number of presentations declined by 23% (from 113 to 87, p < .001) (Fig. 1). Non-Emergent visits declined the most with 42% reduction (from 18 to 10, p < .001). Injuries, being the most common reason for presentation, were reduced by 25% (from 27 to 20, p < .001). Emergent - ED Care Needed - Not Preventable/Avoidable declined by 19% (from 17 to 13, p < .001) having identical proportional reduction with the Emergent/Primary Care Treatable group which also declined by 19% (from 18 to 15, p < .001). Emergent – ED Care Needed – Preventable/Avoidable were reduced by 28% (from 7.7 to 5.6, p < .001). Drug related visits were reduced by 29% (from 0.10 to 0.07, p = 0.01) whereas alcohol related presentation had a non-significant trend to increase by 23% (from 1.7 to 2.2, p = .09. (Table 1, Fig. 2)

Chi-square showed a statistically significant difference in distributions of Non-Emergent (from 16–12%, p < 001), alcohol-related (from 1.6–2.5%, p < .001), Psychiatric (from 1.7–2.2%, p = .01) and Unclassified (18.4–20.4%, p < .001) visits. Injuries constituted the majority of visits but the difference in proportions was not statistically significant (from 24.2–23.7%, p = .45), which was the case for the other NYU-EDA groups as well (Table 2).

We observed significant reduction in the total number of arrivals and in almost every NYU-ED urgency class. The proportional reduction is in line with reports from SARS and MERS pandemics as well as early reports from COVID-19 pandemic. Somewhat expectedly non emergent visits were almost halved. However, we also observed a significant reduction of 23% in emergent, non-preventable cases. However, we also observed a significant reduction of 19% in emergent, non-preventable cases. This suggests that acute health concerns have been left untreated due to people’s unwillingness to seek medical advice even if that may have been necessary. This may be explained by the widespread fear of the virus or by the perception by the public that limited healthcare resources should be reserved for COVID-19 patients, causing them to underestimate the importance of their medical problems.

There was almost a 30% reduction in number of preventable emergency department visits in the study period. These visits are characterized by exacerbation of chronic conditions such as diabetes, chronic heart failure and asthma that could be prevented by close monitoring and non-emergent management. This may suggest that the speculated negative effect of the lockdown on the management of chronic conditions did not materialize in increased number of visits. Our setting does not shed light on the definite cause of this phenomenon, but we speculate that it is partly explained by elevated threshold to seek medical attention and successful primary care despite and because of the pandemic. One might also speculate, that the closing of the society by government restrictions might have reduced respiratory tract infections in general and thus exacerbations of some chronic diseases. Respiratory tract infections are known to be associated e.g. with asthma exacerbations and acute cardiac events (18, 19). It must also be acknowledged that the time period of three months may be insufficient for significant deterioration of chronic conditions to translate into a detectable number of ED visits.

Injury -related presentations decreased by 25%. This falls between the previously reported 50% reduction in a British trauma center and a 16% reduction in emergency surgery in Finnish hospitals (20, 21). There was not a statistically significant difference in the number of Alcohol related presentations but it was the only group that was increased during the study period. If the observation is indeed real, it. This may be explained by the increased domestic alcohol use as nightlife establishments were closed during the pandemic (22).

The NYU-EDA ICD-10CM chart conversion to ICD-10 was straightforward as the United States version is essentially a more detailed version of the original ICD-10. The utilization of NYU-EDA has clear advantages over ICD-class based -stratification of emergency department visits as many diagnostic classes contain drastically different conditions and without case review -based chart, ICD-diagnosis is difficult to reliably translate into urgency. Although NYU-EDA has not been validated in European datasets, it has been validated in nationwide studies in the United States (23).

The weaknesses of this study are its retrospective nature and limited historical data of two years, the latter of which is caused by a reform in emergency department organization at the institution, preventing direct comparisons with patient flow before year 2018. Additionally, certain patient groups such as high-energy polytrauma and ST-elevation myocardial infarctions are treated in a tertiary center in the neighboring hospital district. However, the proportion of these patients is low and we believe the potential bias from that course to be limited. Additionally, the conversion of NYU-EDA ICD-10-CM to ICD-10 chart was limited to 95% of presented codes. However, we consider this acceptable as total conversion of 70,000 ICD-10CM codes would have provided very little additional information on overall demographics of the ED presentations.

4.1 Conclusions

COVID-19 and the state of emergency significantly affected the number of visits in a Finnish emergency department. Both emergent and non-emergent visits were significantly reduced. Whether this will materialize as a delayed increase in overall morbidity will be an interesting subject for later investigation. The lessons from earlier SARS and MERS pandemics seem to apply to ED patient flow dynamics during COVID-19 pandemic.

SARS-CoV-2 Severe Acute Respiratory Syndrome

Virus 2

SARS Serious Acute Respiratory Syndrome

MERS Middle East respiratory syndrome virus

NYU-EDA New York University Emergency Department Algorithm

ICD-10-CM International Statistical Classification of Diseases and Related Health Problems Clinical Modification

ED Emergency Department

EDCN Emergency Department Care Needed

PCT Primary Care Treatable

Ethics approval and consent to participate

Institutional approval for the study was obtained.

Consent for publication

Not applicable

Availability of data and materials

Not available

Competing interests

We declare no competing interests

Funding

No external funding

Authors’ contributions

Jalmari Tuominen performed statistical analysis. Antti Roine, Ari Palomäki, Niku Oksala and Ville Hällberg designed the study protocol. Timo Lukkarinen acquired the data. All participated in writing the manuscript.

Acknowledgements

The authors thank Kanta-Häme Central hospital research coordinator Anu Pakkanen for enabling this study.

Wang L, Wang Y, Ye D, Liu Q. Review of the 2019 novel coronavirus (SARS-CoV-2) based on current evidence. Int J Antimicrob Agents [Internet]. 2020;55(6):105948. Available from: https://doi.org/10.1016/j.ijantimicag.2020.105948
Verity R, Okell LC, Dorigatti I, Winskill P, Whittaker C, Imai N, et al. Estimates of the severity of coronavirus disease 2019: a model-based analysis. Lancet Infect Dis. 2020;20(6):669–77.
Aziz S, Arabi YM, Alhazzani W, Evans L, Citerio G, Fischkoff K, et al. Managing ICU surge during the COVID-19 crisis: rapid guidelines. Intensive Care Med [Internet]. 2020;46(7):1303–25. Available from: https://doi.org/10.1007/s00134-020-06092-5
Li R, Rivers C, Tan Q, Murray MB, Toner E, Lipsitch M. Estimated Demand for US Hospital Inpatient and Intensive Care Unit Beds for Patients With COVID-19 Based on Comparisons With Wuhan and Guangzhou, China. JAMA Netw open. 2020;3(5):e208297.
De Wit E, Van Doremalen N, Falzarano D, Munster VJ. SARS and MERS: Recent insights into emerging coronaviruses. Nat Rev Microbiol [Internet]. 2016;14(8):523–34. Available from: http://dx.doi.org/10.1038/nrmicro.2016.81
Petrosillo N, Viceconte G, Ergonul O, Ippolito G, Petersen E. COVID-19, SARS and MERS: are they closely related? Clin Microbiol Infect [Internet]. 2020;26(6):729–34. Available from: https://doi.org/10.1016/j.cmi.2020.03.026
Chen WK, Cheng YC, Chung YT, Lin CC. The impact of the SARS outbreak on an urban emergency department in Taiwan. Med Care. 2005;43(2):168–72.
Man CY, Yeung RSD, Chung JYM, Cameron PA. Impact of SARS on an emergency department in Hong Kong. Emerg Med. 2003;15(5–6):418–22.
Heiber M, Lou WYW. Effect of the SARS outbreak on visits to a community hospital emergency department. Can J Emerg Med. 2006;8(5):323–8.
Lateef F. SARS changes the ED paradigm. Am J Emerg Med. 2004;22(6):483–7.
Bellan M, Gavelli F, Hayden E, Patrucco F, Soddu D, Pedrinelli AR, et al. Pattern of emergency department referral during the Covid-19 outbreak in Italy. Panminerva Med. 2020;
Santana R, Sousa JS, Soares P, Lopes S, Boto P, Rocha JV. The Demand for Hospital Emergency Services: Trends during the First Month of COVID-19 Response. Port J Public Heal. 2020;
Anteby R, Zager Y, Barash Y, Nadler R, Cordoba M, Klang E, et al. The Impact of the Coronavirus Disease 2019 Outbreak on the Attendance of Patients with Surgical Complaints at a Tertiary Hospital Emergency Department. J Laparoendosc Adv Surg Tech. 2020;00(00):1–7.
Hartnett KP, Kite-Powell A DJ. Impact of the COVID-19 Pandemic on Emergency Department Visits — United States, January1, 2019–May 30, 2020. MMWR Morb Mortal Wkly Rep. 2020;69:699–704.
Tam HL, Chung SF, Lou CK. A review of triage accuracy and future direction 11 Medical and Health Sciences 1117 Public Health and Health Services. BMC Emerg Med. 2018;18(1):1–7.
ICD-10 : international statistical classification of diseases and related health problems : tenth revision, 2nd ed [Internet]. Available from: https://apps.who.int/iris/handle/10665/42980
Billings J, Parikh N, Mijanovich T. Emergency department use: the New York Story. Issue Brief (Commonw Fund). 2000 Nov;(434):1–12.
Tomita K, Sano H, Iwanaga T, Ishihara K, Ichinose M, Kawase I, et al. Association between episodes of upper respiratory infection and exacerbations in adult patients with asthma. J Asthma. 2012;49(3):253–9.
Cilli A, Cakin O, Aksoy E, Kargin F, Adiguzel N, Karakurt Z, et al. Acute cardiac events in severe community-acquired pneumonia: A multicenter study. Clin Respir J. 2018;12(7):2212–9.
Park C, Sugand K, Nathwani D, Bhattacharya R, Sarraf KM. Impact of the COVID-19 pandemic on orthopedic trauma workload in a London level 1 trauma center: the “golden month”: The COVid Emergency Related Trauma and orthopaedics (COVERT) Collaborative. Acta Orthop. 2020;3674.
Ponkilainen V, Kuitunen I, Hevonkorpi TP, Paloneva J, Reito A, Launonen AP, et al. The effect of nationwide lockdown and societal restrictions due to COVID-19 on emergency and urgent surgeries. Br J Surg. 2020;2019–20.
Kim JU, Majid A, Judge R, Crook P, Nathwani R, Selvapatt N, et al. Effect of COVID-19 lockdown on alcohol consumption in patients with pre-existing alcohol use disorder. Lancet Gastroenterol Hepatol [Internet]. 2020;1253(20):19–20. Available from: http://dx.doi.org/10.1016/S2468-1253(20)30251-X
Gandhi SO, Sabik L. Emergency department visit classification using the NYU algorithm. Am J Manag Care. 2014;20(4):315–20.

Due to technical limitations, table 1,2 is only available as a download in the Supplemental Files section.

table1.xlsx
Table 1. Average daily presentations between March 16 and June 11 stratified by NYU-EDA. Δ2020 describes the difference between 2020 and mean of 2019 and 2018. Arr = daily arrivals, Std = standard deviation, EDCN = Emergency Department Care Needed, ~ = Not, P/A = Preventable/Avoidable, PCT = Primary Care Treatable. Proportional change is provided in the parenthesis.
table1.xlsx
Table 1. Average daily presentations between March 16 and June 11 stratified by NYU-EDA. Δ2020 describes the difference between 2020 and mean of 2019 and 2018. Arr = daily arrivals, Std = standard deviation, EDCN = Emergency Department Care Needed, ~ = Not, P/A = Preventable/Avoidable, PCT = Primary Care Treatable. Proportional change is provided in the parenthesis.
table2.xlsx
Table 2. Total arrivals between March 16 and June 11 stratified by NYU-EDA groups. Δ2020 describes the difference in proportion when comparing year 2020 with pooled visits from 2018 and 2019. EDCN = Emergency Department Care Needed, ~ = Not, P/A = Preventable/Avoidable, PCT = Primary Care Treatable, p.p. = percentage point
table2.xlsx
Table 2. Total arrivals between March 16 and June 11 stratified by NYU-EDA groups. Δ2020 describes the difference in proportion when comparing year 2020 with pooled visits from 2018 and 2019. EDCN = Emergency Department Care Needed, ~ = Not, P/A = Preventable/Avoidable, PCT = Primary Care Treatable, p.p. = percentage point

Download PDF

Journal Publication

published 11 Dec, 2020

Read the published version in BMC Emergency Medicine →

Review #2 received at journal
10 Nov, 2020
Editorial decision: Major revision
10 Nov, 2020
Reviewer #2 agreed at journal
09 Nov, 2020
Review #1 received at journal
04 Nov, 2020
Reviewer #1 agreed at journal
03 Nov, 2020
Editor assigned by journal
01 Nov, 2020
Reviewers invited by journal
01 Nov, 2020
Submission checks completed at journal
01 Nov, 2020
Editor invited by journal
01 Nov, 2020
First submitted to journal
27 Oct, 2020

You are reading this latest preprint version

NYU-EDA in Modelling the Effect of COVID-19 on Patient Volumes in A Finnish Emergency Department

Status:

Journal Publication

Version 1

Abstract

Figures

1. Background

2. Materials And Methods

2.1 Data

2.2 NYU-EDA and Conversion

2.3 Statistical tools

3. Results

4. Discussion

4.1 Conclusions

List Of Abbreviations

Declarations

References

Tables

Supplementary Files

Status:

Journal Publication

Version 1