key: cord-0868773-4b54fh18
authors: Piscator, Eva; Göransson, Katarina; Forsberg, Sune; Herlitz, Johan; Djärv, Therese
title: Do-Not-Attempt-Cardiopulmonary-Resuscitation (DNACPR) decisions in patients admitted through the emergency department in a Swedish University Hospital – An observational study of outcome, patient characteristics and changes in DNACPR decisions
date: 2022-02-04
journal: Resusc Plus
DOI: 10.1016/j.resplu.2022.100209
sha: 4ab549f5ee902b3072fc45fe48a346fdf95618e2
doc_id: 868773
cord_uid: 4b54fh18

AIMS: The aims were to examine patient and hospital characteristics associated with Do-Not-Attempt-Cardiopulmonary-Resuscitation (DNACPR) decisions for adult admissions through the emergency department (ED), for patients with DNACPR decisions to examine patient and hospital characteristics associated with hospital mortality, and to explore changes in CPR status. METHODS: This was a retrospective observational study of adult patients admitted through the ED at Karolinska University Hospital 1 January to 31 October 2015. RESULTS: The cohort included 25,646 ED admissions, frequency of DNACPR decisions was 11% during hospitalisation. Patients with DNACPR decisions were older, with an overall higher burden of chronic comorbidities, unstable triage scoring, hospital mortality and one-year mortality compared to those without. For patients with DNACPR decisions, 63% survived to discharge and one-year mortality was 77%. Age and comorbidities for patients with DNACPR decisions were similar regardless of hospital mortality, those who died showed signs of more severe acute illness on ED arrival. Change in CPR status during hospitalisation was 5% and upon subsequent admission 14%. For patients discharged with DNACPR decisions, reversal of DNACPR status upon subsequent admission was 32%, with uncertainty as to whether this reversal was active or a consequence of a lack of consideration. CONCLUSION: For a mixed population of adults admitted through the ED, frequency of DNACPR decisions was 11%. Two-thirds of patients with DNACPR decisions were discharged, but one-year mortality was high. For patients discharged with DNACPR decisions, reversal of DNACPR status was substantial and this should merit further attention.

A Do-Not-Attempt-Cardiopulmonary-Resuscitation (DNACPR) decision may be issued when cardiopulmonary resuscitation (CPR) is considered not to be beneficial for the patient, or when CPR is not aligned with the patient's values and goals of care. 1, 2 Incidence, outcome and hospital organisation for in-hospital cardiac arrests (IHCA) are dependent on the clinical practice of DNACPR decisions and the associated limitations of life-sustaining treatments 3 and prevalence, patient demographics, and patterns of decision-making have been published for different subgroups of patients with DNACPR decisions. [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] However, publications characterising a mixed patient population with regards to DNACPR decisions are more scarce. [23] [24] [25] [26] [27] Studies reflecting clinical practice have shown relatively low changes in life-sustaining treatment decisions during hospitalisation (8% changes), 26 but surprisingly high changes in DNACPR decisions upon readmission (45% reversal of DNACPR decisions). 21 Previous qualitative research has shown that patients' end-of-life preferences can be dynamic but are mostly stable over time. 28 To our knowledge, changes in CPR status have not previously been studied in the Swedish setting. We performed this study therefore with the following aims: to examine patient and hospital characteristics associated with DNACPR decisions for adult admissions through the emergency department (ED), for patients with DNACPR decisions to examine patient and hospital characteristics associated with hospital mortality and explore changes in CPR status during hospitalisation and upon subsequent admission.

This retrospective observational cohort study included patients 18 years with a Personal Identity Number admitted to Karolinska University Hospital through the ED between 1 January and 31 October 2015. Patients admitted for obstetric care were excluded. The cohort was based on a previous publication on DNACPR decisions 29 and was a sub-cohort of a larger cohort of adult ED admissions with complete pre-collected data. Karolinska University Hospital is a teaching hospital and serves as the trauma referral centre in Stockholm, Sweden, a region with 2.2 million inhabitants 2015. The health care structure in the region has been described in detail previously. 30 In 2015 the hospital had two sites, both with adult and paediatric ED (one site predominant regarding paediatric care) open to ambulatory patients and patients brought in by the emergency medical services according to prehospital emergency priority and the predefined health service area of the hospital. Elective admissions were not assessed in the ED, all emergency patients were admitted through the ED, except for patients with a suspicion of ST-elevation myocardial infarction on prehospital assessment that were admitted directly to the angiography laboratory, and there were no direct hospital admissions from e.g., primary care. There were no observational units or units for pure palliative management. The hospital provided care that included neurosurgery-, vascular surgery-, cardiothoracic surgery, trauma-and general surgical units as well as paediatric, obstetric, gynaecologic, otolaryngologic, neurologic, cardiologic, oncologic, geriatric, and internal medicine units. KUH received approximately 220,000 ED visits in 2015 and had approximately 1100 hospital beds.

Patients were recruited from the hospital's central data warehouse that has collected data on patient demographics, hospital admission characteristics, and mortality since 2009. Comorbidities were extracted from the National Patient Register (NPR) from 1997 to admission in 2015. The NPR records comorbidities according to the International Statistical Classification of Diseases-10 (ICD-10). Comorbidities were reported as single comorbidities and according to Charlson Comorbidity Index (CCI) score (eTable 1). 31, 32 The Rapid Emergency Triage and Treatment System (RETTSÓ) 33 was used because it was the only variable available representing the severity of acute illness in the pre-collected data set. The RETTSÓ is a Swedish triage scale, with widespread routine use in Swedish ED. 34 It is used by nurses in EDs and weighs together vital signs, major complaints and comorbidities in a structured algorithm that results in a five-level triage scale, where level 1 represents patients in need of immediate medical attendance, levels 1 and 2 are classified as unstable, and levels 3-5 are classified as stable.

The admission ward was the first ward the patient was admitted to from the ED. It was categorised into the following: general ward, high dependency unit (HDU) and intensive care unit (ICU). HDU included wards with the possibility of continuous monitoring of cardiac rhythm and/or oxygen saturation and more frequent controls of vital signs as compared to general wards. ICU included intermediate care units with similar high monitoring possibilities, access to non-invasive ventilation and vasoactive drugs as intensive care units, but without invasive ventilation, dialysis, more advanced invasive monitoring, or multiple vasoactive drugs. Detailed data regarding in which ward and at what time the DNACPR forms were issued were not available.

According to Swedish ethical guidelines, 1 a conversation concerning DNACPR should take place with all patients with increased risk of inhospital cardiac arrest or where a DNACPR decision could be in line with the values and goals of the patient. There is no special routine for DNACPR decisions on admission to Karolinska University Hospital. When a DNACPR decision is made, it is mandatory to fill out a form. If there is no form, the standard procedure is to initiate CPR in case of a cardiac arrest event. Patients may have multiple DNACPR forms, as a change of ward requires a reassessment of the DNACPR status, and patient conditions may change during hospitalisation. 1 Besides DNACPR, the form specifies other limitations of life-sustaining treatments (LLST) such as intensive care, invasive ventilation or dialysis. It can also specify that there is no DNACPR decision, which in clinical practice is the same thing as having no form. To be able to describe changes, this is called "initiate CPR status" in the reporting of this study.

Univariable analyses were performed for associations between patient and hospital characteristics and DNACPR decisions, and for patients with DNACPR decisions for associations between patient and hospital characteristics and hospital mortality. For patients with DNACPR decisions, time variables were based on the first DNACPR decision placed after arrival to the ED.

Changes in CPR status during hospitalisation were analysed based on admissions with at least one form regarding CPR status. Changes in CPR status upon subsequent admissions were analysed based on cases with known CPR status in the previous hospitalisation during the study period. The CPR status upon subsequent admission (first DNACPR decision, initiate CPR, or no form) was compared to the last CPR status on previous hospitalisation.

Categorical variables were presented as numbers and percentages, binary variables were compared using chi-squared and ordinal/nominal variables were compared using the Wald test. Non-normally distributed data were described by median, interquartile range, and range and were compared using the Mann-Whitney test. Significance tests were two-sided with a significance level of 0.05. Analyses were performed using Stata 13 for Windows (Stata Corp, College Station, TX).

The study was approved by the Swedish Ethical Review Authority to be conducted without informed consent (2019-02142, 2020-05465).

During the study period, 25,646 patients were admitted through the ED, of which 10.9% were admissions where at least one DNACPR decision was issued during the hospital stay. A total of 4000 forms were issued, of which 3861 were DNACPR decisions and 139 were forms with a directive to initiate CPR in case of cardiac arrest. In 18.8% the DNACPR decision was only DNACPR, whereas the rest were associated with other forms of LLST. The most common associated LLST were invasive ventilation and intensive care, and 79% of the DNACPR decisions were combined with either of these.

Patient and hospital characteristics associated with DNACPR decisions are shown in Table 1 . Patients with DNACPR decisions were significantly older (median age 79 years versus. 64 years, p < 0.01) with an overall higher burden of chronic comorbidities as compared to those without. Further, a larger proportion of patients had unstable triage-scoring according to RETTSÓ and were admitted to wards with higher levels of care than patients without DNACPR decisions. Hospital mortality for patients with DNACPR decisions was 36.9%, 30-day mortality was 37.4% and one-year mortality was 76.9% compared to 1%, 1.8% and 12.9% respectively for patients without (p < 0.01 for all).

Out of 2,797 ED admissions with DNACPR decisions, 63.1% were discharged from hospital. When comparing these patients to those with hospital mortality, we found the two groups to be similar in terms of age, sex, and chronic comorbidities except for congestive heart failure (31.2%, versus 27.4%, p = 0.04) and dementia (16.3% versus 11.2%, p < 0.01) which were more prevalent in those discharged, and malignancy (42% versus 46%, p = 0.04) which was less prevalent in those discharged ( Table 2 ). The proportion of unstable RETTSÓ triage scorings on arrival to ED was higher for patients with hospital mortality than for those discharged. The time from the day of ED arrival to the first DNACPR decision did not differ (median 1 day, p > 0.99) in the two groups. For patients with DNACPR decisions and hospital mortality, the median time until death was 10 days [IQR 3;22] , and the median time from the first DNACPR order until death was 6 days [IQR 2 ;16] . Hospital length of stay for patients with DNACPR decisions that were discharged was in median 10 days [IQR 5;20] ( Table 2) .

During the study period, 2798 admissions received at least one form regarding CPR status (one admission had one decision to initiate CPR that was unchanged), see 

Out of the 25,646 admissions through the ED, we excluded 16,285 cases that were admitted only once, and 3709 cases with unknown previous admissions outside of the study period. For the remaining 5652 admissions, discharge CPR status in the previous hospitalisation was known (Table 4 ).

In 86.1% of cases, CPR status was unchanged upon subsequent admission.

Of 577 cases discharged with DNACPR decisions, a reversal of DNACPR status upon subsequent admission occurred in 32.2% of the cases. In 97.8% (182/186) of these cases this was an effect of no form being issued during subsequent admission, and thus there was uncertainty as to whether this reversal was active or a consequence of a lack of consideration. For 67.2% (388/577) of those discharged with DNACPR decisions, DNACPR status was unchanged upon subsequent admission, with an iteration of the DNACPR decision. In nine cases it could not be determined whether CPR status was changed, due to a lack of access to the exact time of documentation.

Out of 983 cases where a DNACPR decision was issued upon subsequent admission, CPR status was changed from initiate CPR (n = 2) or no form in the previous hospitalisation (n = 591) to DNACPR decisions in 60.3% of the cases. For 90.9% of these cases, there was no previous documentation regarding CPR status in previous hospitalisations during the study period.

A sensitivity analysis of the 577 cases discharged with DNACPR status showed that upon subsequent admission they were admitted from the ED to a general ward in 47.5% of cases, HDU in 48%, and ICU in 4.5%.

This retrospective observational study is the first to characterise a larger cohort of patients with DNACPR decisions in Sweden. It showed that 11% of patients admitted through the ED at a Swedish University Hospital received a DNACPR decision during hospitalisation. This is a high figure which may suggest that there is an ongoing increase in DNACPR decisions in Swedish hospitals, with a potential impact on the epidemiology of IHCA in terms of incidence and out-come. Patients with DNACPR decisions were older with more acute and chronic comorbidities, they were admitted to higher levels of care and had to stay in hospital for longer periods compared to those without. Although 63% of patients with DNACPR decisions survived to discharge, and short-term survival was high, one-year mortality was significant (77%). Age and comorbidities for patients with DNACPR decisions were similar regardless of hospital mortality. Patients with hospital mortality showed signs of more severe acute illness on arrival to the ED. The overall change in CPR status during hospitalisation and upon subsequent admission was low, but for patients discharged with DNACPR decisions, reversal of DNACPR status was substantial upon subsequent admission (32%) with uncertainty whether this reversal was active or a consequence of a lack of consideration. Frequency of DNACPR decisions, age, burden of chronic comorbidities, and hospital mortality in our study is in line with previous studies of mixed patient populations admitted through an ED in the US. 23, 26 However, direct comparison is difficult because there was a substantial difference in admission procedures, with CPR directives being an obligation upon admission in those studies. Although not directly comparable, our findings contrast with a point-prevalence study from 2006 from one of the two sites at KUH, which excluded patients in the ICU and showed a prevalence of DNACPR decisions of 4%. 17 A contributing factor to this discrepancy might be increased use of DNACPR decisions over time. 23 In a recent study of a mixed patient population in the UK 2017-2020, hospital mortality was in line with our study, with 32% hospital mortality for all patients with a Treatment Escalation and Limitation (TEAL) form, out of which 89% had DNACPR decisions. 27 A high proportion of patients with DNACPR decisions presented with unstable triage parameters, which can be seen as a proxy for more severe acute illness. They were at least initially treated at a high level of care implying that they were given an opportunity for more intensive emergency care. However, the observational nature of this study did not allow for an analysis of the sequence of events in relation to DNACPR decision placement. The majority were discharged, and for those, underlying comorbidities, and time to DNACPR decision placement from admission did not differ noticeably from those with hospital mortality. Patients with DNACPR decisions and hospital mortality showed signs of more severe acute illness in the ED. This corroborates the notion of DNACPR decisions being heterogenous and a result of complex decision processes involving the assessment of the severity of underlying chronic comorbidities, general health status, severity and progress of acute illness in combination with patient preferences and goals of care in the present situation and with the perspective of the near future. 16, 18, 20, [35] [36] [37] Increased knowledge on timing of DNACPR decision placement in relation to admission and death or discharge could aid in formulating more detailed clinical guidelines for DNACPR decision placement in the Swedish setting. This study showed that the change in CPR status upon subsequent admission was 14%, to our knowledge there are no previous studies for comparison. For the group of patients discharged with DNACPR decisions in the previous hospitalisation, reversal of DNACPR status occurred in one-third of patients. For the majority, it was not certain whether the decision was an active process or simply represented a lack of consideration, because no document was issued on rehospitalisation. Although not completely transferable to our setting as previously mentioned, a study from the US 21 similarly showed high DNACPR reversal upon readmission (45%) that was hypothesised to be driven by patient preferences but was instead strongly associated with institutional factors. This could be true for our setting as well, and exploration of what lies behind the reversal of DNACPR decisions upon subsequent admission should merit further attention.

Limitations of this study include the observational nature of the study that enabled the identification of associations but without the possibility to establish causality. However, it can constitute grounds for hypothesis generation to be tested in future studies. Because this cohort constituted ED admissions and elective admissions (approximately one-third of admissions) were not included, it did not reflect upon the rate of DNACPR decisions for the whole hospitalised population. However, according to a previous study, 23 83% of all DNACPR decisions were placed for patients admitted through the ED, thus capturing the majority of DNACPR decisions. Although the time-period was restricted to ten months, the sample size was large. Regarding changes in CPR status upon subsequent admission, we do not know if patients were admitted to another institution or electively to KUH with decisions regarding CPR made in between hospitalisations through the ED. Administrative data and ICD-10 coding have biases and have no reflection on patient characteristics as a whole or consider severity of illness. Stratification into scoring systems such as CCI further increases loss of details in the data. For our cohort there was a misclassification bias with a risk of over-estimating CCI because data from the NPR did not fulfil the detailed classification of diseases that CCI requires, as shown in eTable 1. The study was based on a cohort of pre-collected data from 2015; as the use of DNACPR decisions has been shown to increase, 23 the contemporary prevalence of DNACPR decisions could differ. Generalisability outside of Karolinska University Hospital is limited, as the use of DNACPR decisions is influenced by cultural, religious, and legal factors, as well as regional, and institutional policies. 23, 24, 38, 39 Conclusion For a mixed population of adults admitted through the ED in a Swedish University hospital, frequency of DNACPR decisions was 11%. The majority of patients with DNACPR decisions were discharged alive but died within one year. Age and comorbidities for patients with DNACPR decisions were similar regardless of hospital mortality, although patients with hospital mortality showed signs of more severe acute illness on arrival to ED. For patients discharged with DNACPR decisions, reversal of DNACPR status upon subsequent admission was substantial. This should merit attention as it could imply a need for strengthening of clinical practice regarding previous DNACPR decisions on admission. Increased knowledge on timing of DNACPR decision placement in relation to admission and death or discharge could aid in more detailed clinical guidelines for DNACPR decision placement in the Swedish setting. Furthermore, we need to know whether there is an ongoing increase in DNACPR decisions in Sweden, since an eventual increase may affect the epidemiology of IHCA. Thus, a continuous monitoring of the rate DNACPR decisions may be required.

No disclosures for all authors.

Etiska riktlinjer för hjä rt-lungrä ddning (HLR) 2:a upplagan

European Resuscitation Council Guidelines 2021: Ethics of resuscitation and end of life decisions

Resuscitation Council Guidelines 2021: Epidemiology of cardiac arrest in Europe

Alignment of Do-Not-Resuscitate Status With Patients' Likelihood of Favorable Neurological Survival After In-Hospital Cardiac Arrest

Do Sex Differences Exist in the Establishment of "Do Not Attempt Resuscitation" Orders and Survival in Patients Successfully Resuscitated From In-Hospital Cardiac Arrest?

Patient and hospital-level characteristics associated with the use of do-not-resuscitate orders in patients hospitalized for sepsis

Do-not-attempt resuscitation independently predict in-hospital mortality in septic patients

Discharge patterns, survival outcomes, and changes in clinical management of hospitalized adult patients with cancer with a do-not-resuscitate order

Provider and patient gender influence on timing of do-not-resuscitate orders in hospitalized patients with cancer

The survival of patients with not-for-resuscitation orders

Do not attempt resuscitation' and 'cardiopulmonary resuscitation' in an inpatient setting: factors influencing physicians' decisions in Switzerland

Patients' characteristics associated with the decision of "do not attempt cardiopulmonary resuscitation" order in a Swiss hospital

Variation in Do-Not-Resuscitate Orders and Implications for Heart Failure Risk-Adjusted Hospital Mortality Metrics

Differences and Trends in DNR Among California Inpatients With Heart Failure

Early Do-Not-Resuscitate Orders and Outcome After Intracerebral Hemorrhage

Do-not-resuscitate orders in patients with community-acquired pneumonia: a retrospective study

Prevalence and sensitivity of MET-criteria in a Scandinavian University Hospital

Usage of do-not-attempt-to-resuscitate orders in a Swedish community hospital -patient involvement, documentation and compliance

A cross-sectional investigation of communication in Do-Not-Resuscitate orders in Dutch hospitals

Community-acquired pneumonia and do not resuscitate orders

Stability of do-not-resuscitate orders in hospitalized adults: a population-based cohort study

Do not attempt cardiopulmonary resuscitation (DNACPR) decisions in people admitted with suspected COVID-19: Secondary analysis of the PRIEST observational cohort study

End-of-life decision-making for patients admitted through the emergency department: hospital variability, patient demographics, and changes over time

Regional and institutional variation in the initiation of early do-not-resuscitate orders

Characteristics and outcome of patients with DNACPR orders in an acute hospital; an observational study

The natural history of changes in preferences for life-sustaining treatments and implications for inpatient mortality in younger and older hospitalized adults

Bringing into focus treatment limitation and DNACPR decisions: How COVID-19 has changed practice

Stability of end-of-life preferences: a systematic review of the evidence

Low adherence to legislation regarding Do-Not-Attempt-Cardiopulmonary-Resuscitation orders in a Swedish University Hospital

Association Between Hospital Bed Occupancy and Outcomes in Emergency Care: A Cohort Study in Stockholm Region

A new method of classifying prognostic comorbidity in longitudinal studies: development and validation

Updating and validating the Charlson comorbidity index and score for risk adjustment in hospital discharge abstracts using data from 6 countries

Medical Emergency Triage and Treatment System (METTS): a new protocol in primary triage and secondary priority decision in emergency medicine

An updated national survey of triage and triage related work in Sweden: a cross-sectional descriptive and comparative study

Do-not-attemptcardiopulmonary-resuscitation decisions: an evidence synthesis. Donot-attempt-cardiopulmonary-resuscitation decisions: an evidence synthesis

Perspectives on the DNR decision process: A survey of nurses and physicians in hematology and oncology

Do not attempt cardiopulmonary resuscitation (DNACPR) orders: a systematic review of the barriers and facilitators of decision-making and implementation

A survey of key opinion leaders on ethical resuscitation practices in 31 European Countries

Hospital Variation in Do-Not-Resuscitate Orders and End-of-Life Healthcare Use in the United States

We would like to offer special thanks to Martin J. Holzmann for his contribution to the study design and acquisition of data. Although no longer with us, he continues to inspire by his example and dedication to research throughout his career. TD was supported by the Stockholm County Council (clinical research appointment).

Supplementary data to this article can be found online at https://doi. org/10.1016/j.resplu.2022.100209.