key: cord-0834246-vg115zqp
authors: Lo Presti, G.; Biggiogero, M.; Glotta, A.; Biondi, C.; Horvath, Z.; Leo, R.; Franzetti-Pellanda, A.; Saporito, A.; Ceruti, S.; Capdevila, X.
title: DNR order in SARS-CoV-2 patients: preliminary guidelines validation and quality of palliative care in a single Swiss COVID-19 Center.
date: 2021-07-15
journal: nan
DOI: 10.1101/2021.07.12.21260359
sha: 7fb7f2d398c0faf63e31ea5345bab54a9c4edb12
doc_id: 834246
cord_uid: vg115zqp

Background: The worldwide pandemic situation forced many hospitals to improve COVID-19 management. In this scenario, the Swiss Academy of Medical Sciences (SAMW/ASSM) organized guidelines based on expert opinion to identify Do-Not-Resuscitate (DNR) patients, which often need palliative care (PC) due to a large symptoms burden. In our specialized COVID-19 Center, we investigated characteristics and mortality of DNR patients according to SAMW/ASSM guidelines, comparing to non-DNR patients. Methods: Pilot retrospective validation study, evaluating consecutive COVID-19 patients admitted to Internal Medicine Department. A sub-analysis of the deceased patients care was performed, using both Richmond Agitation-Sedation Scale-Palliative care (RASS-PAL) for sedation and agitation (+4/-5) and modified Borg Scale for dyspnea (1-10). Primary outcome was a 30-days survival of DNR patients comparing to non-DNR patients. Secondary outcomes reported quality of PC, especially on dyspnea and restlessness. Results: From March 16 to April 1 2020, 213 consecutive patients were triaged; of 40 DNR patients 25 (64%) were males with a mean age of 79.9 years. At 30-days follow-up, 9 (22.5%) of DNR patients died; 4 patients (2.2%) died in the control group. The higher mortality rate in DNR group was further confirmed by Log-Rank Mantel-Cox (23.104, p < 0.0001). PC was applied using oxygen (100%), opioids (100%) and sedatives (89%); the mean RASS-PAL improved from 2.2 to -1.8 (p < 0.0001) and Borg scale improved from 5.7 to 4.7 (p = 0.581). Conclusion: The SAMW/ASSM guidelines identified patients at higher risk of short-term death. In the end-of-life management, an early access to PC becomes crucial, both improving patients' end-of-life treatment and reducing psychological families' post-traumatic disorders.

During the beginning of 2020, the world witnessed the first COVID-19 pandemic. To date, around 115 million COVID-19 cases have been diagnosed, with more than 2 million deaths worldwide; in Switzerland around 560`000 cases are reported, with more than 10`000 deaths 1 . The pandemic forced many hospitals to improve their competence, with subsequent development of COVID-19 centers 2 dedicated to improve patients' specific management.

The Swiss Academy of Medical Sciences (SAMW/ASSM) implemented and applied a national standardized protocol of care based on expert opinion, with the aim to guarantee a better treatment of COVID-19 patients in an emergency situation with limited resources 3, 4 The protocol determined the indications for a correct evaluation and identification of patients with a Do-Not-Resuscitate (DNR) order. This tool supported doctors' decision in the triage and at hospital admission, with the aim to identify patients at high risk of death despite a potential ICU support, thus avoiding futile treatments, ICU overload and health costs 3, 5 .

Nevertheless, until today, the effects of the SAMW/ASSM guidelines are still unknown.

In this scenario, it has become essential to guarantee a tool able to identify with high accuracy DNR patients;

according to other authors, in comparison with not-DNR patients, patients with a DNR order usually presented a high fatality rate, a rapid disease progression and the frequent need of an early Palliative Care (PC) approach 6,7 . PC was the forefront of this pandemic, as most DNR patients with SARS-CoV-2 need a palliative approach due to the large symptoms burden, including dyspnea, restlessness, anxiety and delirium 8 . PC contributed to a better symptomatic care in all dying COVID-19 patients, implementing a context of multidisciplinary discussion between intensive care physicians, internist physicians and general practitioners.

Accurate PC also provided an adequate integration of families, in order to improve patients' and their families' care in the last moments of life 9 .

Aim of this study was to report and analyze the survival rate of DNR COVID-19 patients, identified and managed according to expert opinion like SAMW/ASSM guidelines, hospitalized in our Center, a COVID-19 center in Southern Switzerland, in an emergency situation with limited resources. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 

This was a pilot retrospective validation study, evaluating consecutive DNR patients hospitalized with a diagnosis of COVID-19 during the first two weeks of the first pandemic wave. DNR order was defined according to the first version of SAMW/ASSM national guidelines (Table 1) 3 , defining patients ineligible for ICU admission (especially in case of severe cerebral deficits after a stroke, NYHA class III or IV heart failure, COPD GOLD 4, liver cirrhosis with refractory ascites or encephalopathy at stage > I, stage V chronic kidney disease, a Frailty index higher than 4, age greater than 80 years and/or an estimated survival of less than 24 months) 3 .

Through the health medical records, a dataset containing biological, clinical and medical information of DNR patients hospitalized during the study period was created. Biological data concerned age, sex, frailty status 10 , comorbidities (oncological disease, cardiac failure, COPD, cirrhosis and neurodegenerative disease) and all characteristics regarding DNR application were collected. Finally, the 30-day mortality rate was also reported.

Clinical symptoms were evaluated using standard score systems. Dyspnea was evaluated according to modified Borg scale 11 , a semiquantitative numerical scale frequently used as measure of perceived exertion during physical activity. Sedation and agitation were evaluated according to Richmond Agitation Sedation Scale modified for palliative care inpatients (RASS-PAL) 12 . The Borg scale and RASS-PAL were regularly evaluated twice a day by nursing staff; moreover, patients were further reassessed and revaluated with the above-mentioned scales whenever a variation in dyspnea or agitation level was registered. All medical relevant information concerning most important PC pharmacological therapies (use of opioids, sedative, neuroleptics), fluid therapy, oxygen delivery and other characteristics were also reported ( Table 2) . As pandemic is a social health problem, family visit restriction was implemented for all patients due to Public Law 13 , the only exception was a single 15-minute visit performed by one family member for patients who entered an end-of-life care approach. Data concerning family visits were also collected and reported. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 15, 2021. ; https://doi.org/10.1101/2021.07.12.21260359 doi: medRxiv preprint Outcomes Primary outcome was the 30-day overall survival in patients defined as DNR according to the SAMW/ASSM guidelines, analyzing with validation tests whether this group of patients was at greater risk of death than patients defined as not-DNR, intended as control group. Secondary outcomes were to report the quality of PC according to the patients' clinical status, reporting their clinical characteristics and with a special focus on dyspnea and degree of agitation/sedation.

A descriptive statistic of frequency was reported. Categorical variables were expressed as number (percentage); continuous variables were expressed as average (standard deviation, SD). Data distribution was verified with Kolmogorov-Smirnov test and with Shapiro-Wilk test. To identify significant difference between continuous variables, t-tests were performed. A Chi-Square statistic with 1 degree of freedom was carried out to study differences between categorical variables. Time to death was portrayed by Kaplan-Meier plot and compared with a log-rank test. Significant results were intended with p value < 0.01. Statistical data analysis was performed using the SPSS 25.0 package (SPSS Inc, USA).

The project has been approved by the local Ethics Committees according to the local Federal rules. All the participants provided a written informed consent form.

During the first two weeks of pandemics, from March 16 th to April 1 st 2020, 242 consecutive patients were admitted to our Center; 29 patients were excluded as they had already been admitted as critically ill patients to the ICU. Of 213 not-critically ill patients, 40 (18.8%) were identified as DNR according to SAMW/ASSM guidelines, resulting ineligible for ICU admission; 173 (81.2%) patients were admitted to the Internal Medicine Department, without a DNR order (Figure 1 ). Among DNR group, 27 (67.5%) patients presented a Frailty Index greater than 4, 21 (52.5%) had at least one chronic disease, 7 (17.5%) had a diagnosis of severe All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 15, 2021. ; https://doi.org/10.1101/2021.07.12.21260359 doi: medRxiv preprint circulatory failure and 6 (15%) were affected by oncological active or metastatic disease. Among DNR patients, mean age was 79.9 years (62 -97, SD 7.9) while among control group it was 63.2 years (24 -86, SD 12.1, p = 0.011); in the DNR group 25 patients (64%) were men, while among control group 103 patients (59.5%) were men, p=0.257 (Table 2) .

At 30-day follow-up of DNR group, 9 (22.5%) of them were death due to cardiovascular arrest (Table 3) ;

instead, regarding patients identified as not-DNR, a mortality rate of 2.2% has been registered (Figure 1 and   2 ). Referring to deceased patients in DNR group, mean age was 82.4 years (72 -93, SD 6.9); all of them were resident at home before the hospitalization, 7 (78%) presented a Frailty index more than 4, 3 (33%) a pre-existing severe heart disease, 2 (22%) a metastatic oncological disease and finally 1 (11%) a chronic brain damage (Figure 1 ). In order to perform a preliminary guidelines validation 3 , a contingency table was structured according to the outcome stratification and DNR group allocation. A Chi-square test comparing mortality in DNR group with control group, revealed a higher mortality rate in DNR population, p < 0.0001 (Table 4SM) ; similarly, the Kaplan-Meier distribution ( Figure 2 ) reported a significant difference over time (Log-Rank Mantel-Cox 23.104, p < 0.0001).

Regarding deceased patients in DNR group, average hospital length-of-stay (LOS) was 6.2 days (3 -11, SD 2.4); for 8 (89%) of them it was possible to meet with a family member before the end of life. All 9 deceased patients (100%) received oxygen therapy until death; 4 (44%) of them had a continuous intravenous fluids infusion and 8 (89%) received at least one antibiotics or anti-retroviral therapy during the entire hospitalization (Table 3) . Concerning PC administered in these patients, all of them (100%) received opioids (89% morphine, 11% fentanyl) according to own clinical status; 8 (89%) received sedatives as needed, especially midazolam; no neuroleptics were administered. Consequently, after PC was started, median RASS-PAL value significantly improved from 2.2 (SD 1.2) to -1.8 (SD 2), p < 0.0001 ( Figure 3) ; dyspnea improved from a median of 5.7 (SD 3.3) to 4.7 (SD 3.7), p = 0.581 ( Figure 1SM ). All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 

COVID-19 is a disease characterized by important clinical and social impact, with a high mortality rate especially in patients over 60 years 6 . Identifying with high accuracy patients at high risk of death at shortterm represents a key point both in patients' triage and in-hospital management, in an emergency situation with limited resources such as the COVID-19 pandemic. Application of SAMW/ASSM guidelines made it possible to identify a group of COVID-19 patients defined as DNR, consequently not admitted to the ICU, with a short-term mortality rate significantly increased compared to not-DNR COVID-19 patients (22.5% vs 2.2%) 14, 15 . This evidence suggested that the SAMW/ASSM guidelines, even if based on expert opinion, pragmatically identify population at higher risk of death at short-term (30 days); therefore, they characterize a population requiring a different specialist treatments, namely PC, rather than an intensive specialist care.

COVID-19 is a disease with a significant social impact and the Public Authority required a limited family members' access to hospitalized patients in order to preserve Public Health 16 . In this context, physicians' competences to identify adequately patients at high risk of death allows the nosocomial internal organization to perform a selective family members' visit permit for patients at the end-of-life. This aspect becomes essential for Public Health, but also in a long-term social benefit for relatives, decreasing risks of posttraumatic-stress-disorder, of unprocessed mourning and of other psychological diseases 17 . Moreover, it may prevent future increase in health costs due to a better management of these important psychological aspects.

PC played an important role in taking care of COVID-19 DNR patients, both in terms of psychological and physical symptomatic relief. 16, 18, 19 . The effectiveness that was shown in alleviating dyspnea and agitation, even in a relatively small group of DNR patients, underlined the importance of PC in this scenario. The results we obtained on sedation, besides being significant from a statistical point of view, showed us how the clinical handling of these patients has been satisfactory in guarantying them adequate comfort, even in the last moments of life. In order to achieve this important result, and to adequately adapt therapy and management in relation to the clinical status, we realized that it was mandatory to evaluate patient's agitation/sedation level through the use of appropriate international scales (RASS-PAL in our case) 12, 20 . The All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 15, 2021. ; https://doi.org/10.1101/2021.07.12.21260359 doi: medRxiv preprint assessments carried out so far, in accordance with evidences from other groups 18 , highlighted the importance of the PC specialist's role in the better management of symptoms such as agitation, dyspnea, thirst, fear and solitude, which are unfortunately very frequent in the dying population 19 .

Our study presented some limitations. Firstly, it was a retrospective collection of cohort patients, with small sample of DNR patients. However, the significance of statistical tests remains high and should induce other groups to carry out future prospective validations in this direction. Secondly, the group of DNR patients presented a median age greater than the control group; this may reflect the presence of co-morbidities, a keyelements necessary to identify DNR patients according to SAMW/ASSM guidelines. The increased mortality rate in DNR group was at a short-term of 30 days, so that the only presence of a median age of 80 years was not sufficient to explain this increase in mortality. Thirdly, in our scenario, due to resource re-allocations, DNR patients have not been treated by certified Palliative Physicians; this may have negatively influenced the obtained results on secondary outcomes. Fourth, in our analysis of palliative care we decided to evaluate only the administration of opiates and sedatives, and their effects due to the retrospective design of the study, as a more specific information in this context resulted incomplete. Finally, we did not investigate the followup in terms of hospital discharge, transfer to other facilities or survival after 30 days, mainly because the primary outcome of our study was to investigate the 30-day overall survival in COVID-19 patients, comparing DNR to not DNR.

In conclusion, the high mortality rate in COVID-19 patients enhances how it is relevant to identify a population at a higher risk of short-term death in a limited resources emergency situation; in this scenario, the validation analysis we performed supported the role of the SAMW/ASSM guidelines in DNR patients' identification. As PC had a key role in patients at high risk of death, an early identification and access to PC should be mandatory, with the aim to improve patients' management and at the same time to reduce the social impact of this pandemic.

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 15, 2021. ; https://doi.org/10.1101/2021.07.12.21260359 doi: medRxiv preprint

We thank the patients' families and the health care workers of our Center who took care of patients admitted during this pandemic.

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 15, 2021. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 15, 2021. ; https://doi.org/10.1101/2021.07.12.21260359 doi: medRxiv preprint 

la Clinica Luganese Moncucco diventa ospedale Covid-19

COVID-19 pandemic: triage for intensive-care treatment under resource scarcity

Development of a triage protocol for critical care during 9

Patients-Emergency Palliative Care in Action

Comparison of alternate scoring of variables on the performance of the frailty index

Usefulness of the modified 0-10 Borg scale in assessing the degree of dyspnea in patients with COPD and asthma

The Richmond Agitation-Sedation Scale modified for palliative care inpatients (RASS-PAL): A pilot study exploring validity and feasibility in clinical practice

ulteriori misure organizzative e di prevenzione

Clinical characteristics of coronavirus disease 2019 in China

Clinical course and outcomes of critically ill patients with SARS-CoV-2 pneumonia in Wuhan, China: a single-centered, retrospective, observational study

Palliative Care in the Time of COVID-19