Advertisement

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

Open AccessPublished:May 04, 2021DOI:https://doi.org/10.1016/j.resuscitation.2021.04.028

      Abstract

      Aims

      We aimed to describe the characteristics and outcomes of adults admitted to hospital with suspected COVID-19 according to their DNACPR decisions, and identify factors associated with DNACPR decisions.

      Methods

      We undertook a secondary analysis of 13,977 adults admitted to hospital with suspected COVID-19 and included in the Pandemic Respiratory Infection Emergency System Triage (PRIEST) study. We recorded presenting characteristics and outcomes (death or organ support) up to 30 days. We categorised patients as early DNACPR (before or on the day of admission) or late/no DNACPR (no DNACPR or occurring after the day of admission). We undertook descriptive analysis comparing these groups and multivariable analysis to identify independent predictors of early DNACPR.

      Results

      We excluded 1249 with missing DNACPR data, and identified 3929/12748 (31%) with an early DNACPR decision. They had higher mortality (40.7% v 13.1%) and lower use of any organ support (11.6% v 15.7%), but received a range of organ support interventions, with some being used at rates comparable to those with late or no DNACPR (e.g. non-invasive ventilation 4.4% v 3.5%). On multivariable analysis, older age (p < 0.001), active malignancy (p < 0.001), chronic lung disease (p < 0.001), limited performance status (p < 0.001), and abnormal physiological variables were associated with increased recording of early DNACPR. Asian ethnicity was associated with reduced recording of early DNACPR (p = 0.001).

      Conclusions

      Early DNACPR decisions were associated with recognised predictors of adverse outcome, and were inversely associated with Asian ethnicity. Most people with an early DNACPR decision survived to 30 days and many received potentially life-saving interventions.

      Registration

      ISRCTN registry, ISRCTN28342533, http://www.isrctn.com/ISRCTN28342533

      Keywords

      Introduction

      In-hospital cardiac arrest is relatively common in patients with COVID-19 and often results in poor outcome. A multicentre cohort study from the United States
      • Hayek Salim S.
      • Brenner Samantha K.
      • Azam Tariq U.
      • et al.
      In-hospital cardiac arrest in critically ill patients with covid-19: multicenter cohort study.
      reported that 701/5019 (14.0%) critically ill patients with COVID-19 had in-hospital cardiac arrest, with 400/701 (57.1%) receiving CPR, and only 7% of these surviving to hospital discharge with normal or mildly impaired neurological status. Management of cardiac arrest in COVID-19 is further complicated by concerns about infection risk associated with aerosol-generating procedures and consequent risks to staff.
      • Perkins G.D.
      • Morley P.T.
      • Nolan J.P.
      • et al.
      International Liaison Committee on Resuscitation: COVID-19 consensus on science, treatment recommendations and task force insights.
      These concerns have raised awareness about the need to consider do not attempt cardiopulmonary resuscitation (DNACPR) decisions when patients are admitted to hospital with suspected COVID-19. An appropriately implemented DNACPR decision can ensure that the patient’s wishes and best interests are addressed, while avoiding futile medical intervention.
      • Curtis J.R.
      • Kross E.K.
      • Stapleton R.D.
      The importance of addressing advance care planning and decisions about do-not-resuscitate orders during novel coronavirus 2019 (COVID-19).
      However, concerns have been raised about inappropriate use of DNACPR decisions during the pandemic,
      • Clare Dyer
      Covid-19: Campaigner calls for national guidance to stop DNR orders being made without discussion with patients and families.
      leading to the Care Quality Commission being asked to review their use in the United Kingdom (UK).

      Care Quality Commission. Reviewing the use of do not resuscitate decisions during COVID-19, 17/11/2020. https://www.cqc.org.uk/news/stories/reviewing-use-do-not-resuscitate-decisions-during-covid-19 (Accessed 11/12/2020).

      Previous studies have estimated the prevalence of DNACPR decision in patients admitted to hospital with community-acquired pneumonia
      • Egelund G.B.
      • Jensen A.V.
      • Petersen P.T.
      • et al.
      Do-not-resuscitate orders in patients with community-acquired pneumonia: a retrospective study.
      • Marrie T.J.
      • Fine M.J.
      • Kapoor W.N.
      • Coley C.M.
      • Singer D.E.
      • Obrosky D.S.
      Community-acquired pneumonia and do not resuscitate orders.
      • Oshitani Y.
      • Nagai H.
      • Matsui H.
      Rationale for physicians to propose do-not-resuscitate orders in elderly community-acquired pneumonia cases.
      • Walkey A.J.
      • Weinberg J.
      • Wiener R.S.
      • Cooke C.R.
      • Lindenauer P.K.
      Association of do-not-resuscitate orders and hospital mortality rate among patients with pneumonia.
      • Mulder T.
      • van Werkhoven C.H.
      • Huijts S.M.
      • Bonten M.J.
      • Postma D.F.
      • Oosterheert J.J.
      Treatment restrictions and empirical antibiotic treatment of community-acquired pneumonia in elderly patients.
      and sepsis,
      • Sarkari N.N.
      • Perman S.M.
      • Ginde A.A.
      Impact of early do-not-attempt-resuscitation orders on procedures and outcomes of severe sepsis.
      • Bradford M.A.
      • Lindenauer P.K.
      • Wiener R.S.
      • Walkey A.J.
      Do-not-resuscitate status and observational comparative effectiveness research in patients with septic shock.
      • Powell E.
      • Sauser K.
      • Cheema N.
      • et al.
      Severe sepsis in do-not-resuscitate patients: intervention and mortality rates.
      • Huang C.T.
      • Chuang Y.C.
      • Tsai Y.J.
      • Ko W.J.
      • Yu C.J.
      High mortality in severe sepsis and septic shock patients with do-not-resuscitate orders in East Asia.
      and have attempted to identify factors associated with DNACPR use, but we currently know very little about how DNACPR decisions have been used in people admitted with suspected COVID-19. The Pandemic Respiratory Infection Emergency System Triage (PRIEST) study was established to develop and evaluate triage tools for people presenting to hospital emergency departments with suspected COVID-19.
      • Goodacre S.
      • Thomas B.
      • Lee E.
      • et al.
      Characterisation of 22445 patients attending UK emergency departments with suspected COVID-19 infection: Observational cohort study.
      DNACPR decisions were recorded to facilitate evaluation of triage tools in pre-specified subgroups. We present a post hoc secondary analysis of patients admitted with suspected COVID-19 that aims to describe their characteristics and outcomes according to their DNACPR decision and identify factors associated with recording of a DNACPR decision.

      Methods

      PRIEST was an observational cohort study of patients attending an emergency department (ED) in the UK with suspected COVID-19 infection during the first wave of the pandemic. We included patients if the assessing clinician recorded that the patient had suspected COVID-19 in the ED records or completed a standardised assessment form for suspected COVID-19 patients. The clinical diagnostic criteria for COVID-19 during the study were of fever (≥37.8 °C) and at least one of the following respiratory symptoms, which must be of acute onset: persistent cough (with or without sputum), hoarseness, nasal discharge or congestion, shortness of breath, sore throat, wheezing, sneezing. We did not seek consent to collect data but information about the study was provided in the ED and patients could withdraw their data at their request. Patients with multiple presentations to hospital were only included once, using data from the first presentation identified by research staff.
      We only included patients who were admitted to hospital after ED assessment because DNACPR planning was considered unlikely to be routinely undertaken for discharged patients and would be limited to a minority of highly selected cases. We also only included adults (age ≥ 16 years) because previous analysis
      • Goodacre S.
      • Thomas B.
      • Lee E.
      • et al.
      Characterisation of 22445 patients attending UK emergency departments with suspected COVID-19 infection: Observational cohort study.
      showed that children with suspected COVID-19 had very low rates of confirmed COVID-19 or adverse outcome.
      Baseline characteristics at presentation to the ED were recorded prospectively, using a standardised assessment form that doubled as a clinical record (Appendix 1: Standardised Data Collection Form), or retrospectively, through research staff extracting data onto the standardised form using the clinical records. Research staff collected follow-up data onto a standardised follow-up form (Appendix 2: Follow-up Form) using clinical records up to 30 days after presentation. This included recording whether research staff identified a DNACPR decision made at any time between initial presentation and follow-up, and if so, the date of the decision. We only recorded the first decision and did not record whether it was subsequently changed.
      Patients who died or required respiratory, cardiovascular, or renal support were classified as having an adverse outcome. Patients who survived to 30 days without requiring respiratory, cardiovascular, or renal support were classified as having no adverse outcome. Respiratory support was defined as any intervention to protect the patient’s airway or assist their ventilation, including non-invasive ventilation, or acute administration of continuous positive airway pressure. It did not include supplemental oxygen alone or nebulised bronchodilators. Cardiovascular support was defined as any intervention to maintain organ perfusion, such as inotropic drugs, or invasively monitor cardiovascular status, such as central venous pressure or pulmonary artery pressure monitoring, or arterial blood pressure monitoring. We included invasive monitoring as these imply an intention to use cardiovascular support even if it not actually required. It did not include peripheral intravenous cannulation, or fluid administration. Renal support was defined as any intervention to assist renal function, such as haemofiltration, haemodialysis, or peritoneal dialysis. It did not include intravenous fluid administration.
      We compared the characteristics and outcomes of those with a DNACPR decision recorded on or before the day of ED assessment (early DNACPR) to those with no DNACPR recorded or a DNACPR decision recorded at a later date (late/no DNACPR). We categorised patients in this way on the assumption that patients with a late DNACPR decision were likely to be systematically different from those with an early decision, with implementation of a late DNACPR decision reflecting the response to intervention. Our categorisation was therefore based on a theoretical framework in which patient characteristics at admission could determine whether a DNACPR decision was recorded at hospital admission, and recording of a DNACPR decision at admission could then determine subsequent use of interventions.
      We calculated a National Early Warning Score (2nd version, NEWS2) for adults,
      • Royal College of Physicians
      National early warning score (NEWS) 2: Standardising the assessment of acute-illness severity in the NHS. Updated report of a working party.
      to provide an overall assessment of acute illness severity on a score from zero to 20, based on the first recorded respiratory rate, oxygen saturation, systolic blood pressure, heart rate, level of consciousness and temperature. We calculated a PRIEST COVID-19 clinical severity score, to provide an overall prediction of the risk of adverse outcome on a score from zero to 29, based on NEWS2, age, sex, and performance status.
      • Goodacre S.
      • Thomas B.
      • Sutton L.
      • et al.
      Derivation and validation of a clinical severity score for acutely ill adults with suspected COVID-19: the PRIEST observational cohort study.
      We also undertook multivariable logistic regression modelling to identify independent predictors of DNACPR decisions. Variables were selected on the basis of clinical interest. Collinearity was observed between Glasgow Coma Scale (GCS) scores and consciousness, recorded as alert, responsive to verbal stimuli, responsive to pain or unconscious (AVPU). Missing AVPU data were imputed using GCS as follows, and GCS dropped from the list of predictors: GCS 15 = Alert, GCS 9–14 = Verbal, GCS 7–8 = Pain, and GCS 3–6 = Unresponsive. Continuous physiological predictors were categorised in accordance with NEWS2 risk categories
      • Royal College of Physicians
      National early warning score (NEWS) 2: Standardising the assessment of acute-illness severity in the NHS. Updated report of a working party.
      where the reference category denotes normal range and increasing category levels indicate increasing deviation from the norm. Data were analysed using SAS v9.4.

      Ethical approval

      The North West - Haydock Research Ethics Committee (REC) gave a favourable opinion on the PAINTED study on 25 June 2012 (reference 12/NW/0303) and on the updated PRIEST study on 23rd March 2020. The Confidentiality Advisory Group of the Health Research Authority granted approval to collect data without patient consent in line with Section 251 of the National Health Service Act 2006. The REC approved a substantial amendment to undertake this secondary analysis on 7 January 2021.

      Results

      The PRIEST study recruited 22,484 patients from 70 EDs across 53 sites between 26 March 2020 and 28 May 2020 (22,445 after 39 withdrew their data), including 13,997 adults admitted following ED assessment. Research staff did not complete the DNACPR data item for 1249 patients, so they were excluded from the analysis. The remaining 12,748 were grouped into those with DNACPR decisions made on or before the day of initial ED assessment (N = 3929, 31%) and those with no DNACPR or DNACPR decisions made at a later date (N = 8819, including 7109 with no DNACPR and 1710 late DNACPR). Fig. 1 shows the flow of patients into this analysis.
      Table 1 shows presenting characteristics for both groups. Patients with a DNACPR decision recorded on or before their day of attendance tended to be older and have a higher prevalence of comorbidities, limited activity or self-care.
      Table 1Presenting characteristics for admitted adults with DNACPR decisions in place by the end of ED assessment (N = 3929) and adults with no DNACPR or DNACPR decision made later (N = 8819).
      CharacteristicStatistic/levelEarly DNACPRNo/late DNACPR
      Age (years)N39298819
      Mean (SD)79.4 (10.9)63.9 (17.6)
      Median (IQR)81 (73, 87)65 (52, 78)
      SexMissing4278
      Male2027 (52.1%)4704 (53.8%)
      Female1860 (47.9%)4037 (46.2%)
      EthnicityMissing/prefer not to say6011746
      UK/Irish/other white3102 (93.2%)6027 (85.2%)
      Asian83 (2.5%)468 (6.6%)
      Black/African/Caribbean75 (2.3%)289 (4.1%)
      Mixed/multiple ethnic groups27 (0.8%)102 (1.4%)
      Other41 (1.2%)187 (2.6%)
      Presenting featuresCough2117 (53.9%)5514 (62.5%)
      Shortness of breath2955 (75.2%)6678 (75.7%)
      Fever1777 (45.2%)4688 (53.2%)
      ComorbiditiesNo Chronic disease323 (8.2%)2012 (22.8%)
      Heart Disease1620 (41.2%)2036 (23.1%)
      Renal impairment739 (18.8%)843 (9.6%)
      Steroid therapy160 (4.1%)258 (2.9%)
      Asthma450 (11.5%)1446 (16.4%)
      Diabetes1123 (28.6%)2034 (23.1%)
      Active malignancy370 (9.4%)508 (5.8%)
      Immunosuppression126 (3.2%)326 (3.7%)
      Other chronic lung disease1219 (31%)1685 (19.1%)
      Hypertension1721 (43.8%)3039 (34.5%)
      Symptom duration (days)N33967986
      Mean (SD)5.7 (6.8)7.5 (8.4)
      Median (IQR)3 (1, 7)5 (2, 10)
      Heart rate (beats/min)N38678664
      Mean (SD)95.5 (23)97.4 (22)
      Median (IQR)93 (80, 109)96 (82, 111)
      Respiratory rate (breaths/min)N38678625
      Mean (SD)25.5 (7.5)23.9 (6.9)
      Median (IQR)24 (20, 29)22 (19, 28)
      Systolic BP (mmHg)N38348629
      Mean (SD)131.8 (27.8)134.1 (25.6)
      Median (IQR)130 (112, 149)132 (117, 149)
      Diastolic BP (mmHg)N38138587
      Mean (SD)73.5 (17.1)77.4 (16.3)
      Median (IQR)72 (62, 84)77 (67, 87)
      Temperature (°C)N38038576
      Mean (SD)37.1 (1.2)37.3 (1.2)
      Median (IQR)37 (36.4, 37.9)37.2 (36.5, 38.1)
      Oxygen saturation (%)N38938741
      Mean (SD)92.6 (8)94.1 (6.8)
      Median (IQR)95 (91, 97)96 (93, 98)
      Glasgow Coma ScaleN29856643
      Mean (SD)13.9 (2.1)14.7 (1.3)
      Median (IQR)15 (14, 15)15 (15, 15)
      AVPUMissing6101009
      Alert2877 (86.7%)7486 (95.9%)
      Verbal291 (8.8%)229 (2.9%)
      Pain97 (2.9%)50 (0.6%)
      Unresponsive54 (1.6%)45 (0.6%)
      Performance statusMissing222455
      Unrestricted normal activity453 (12.2%)4341 (51.9%)
      Limited strenuous activity, can do light activity448 (12.1%)1160 (13.9%)
      Limited activity, can self care923 (24.9%)1351 (16.2%)
      Limited self care1153 (31.1%)993 (11.9%)
      Bed/chair bound, no self care730 (19.7%)519 (6.2%)
      NEWS2 scoreN39118723
      Mean (SD)6.4 (3.3)5.1 (3.1)
      Median (IQR)6 (4, 9)5 (3, 7)
      PRIEST scoreN38708645
      Mean (SD)12.5 (3.9)8.9 (4.1)
      Median (IQR)12 (10, 15)9 (6, 12)
      Fig. 2 compares the NEWS2 scores for the two groups and shows that those with early DNACPR decisions tended to be more acutely unwell (median score 6 versus 5). Fig. 3 compares the PRIEST COVID-19 clinical severity scores for the two groups and shows that those with early DNACPR decisions were at a higher risk of adverse outcome (median score 12 versus 9, respectively indicating 38% versus 26% expected risk of a 30-day adverse outcome).
      • Goodacre S.
      • Thomas B.
      • Sutton L.
      • et al.
      Derivation and validation of a clinical severity score for acutely ill adults with suspected COVID-19: the PRIEST observational cohort study.
      Fig. 2
      Fig. 2NEWS2 score distribution by DNACPR status.
      Fig. 3
      Fig. 3PRIEST score distribution by DNACPR status.
      Table 2 gives location of admission, pathogen confirmation and adverse outcome data for the two groups. Patients with an early DNACPR decision had a higher mortality rate but most (59.4%) survived to 30 days. They also had lower use of critical care and organ support, but a significant proportion (11.6%) received organ support. Table 2 shows the highest level of organ support received, according to a predefined hierarchy (corresponding to the order presented in the table). Patients with early DNACPR decisions received a wide range of interventions, some at comparable rates to those with no or a late DNACPR decision (e.g. non-invasive ventilation and high flow nasal oxygen).
      Table 2Outcome data for admitted adults with DNACPR decisions in place by the end of ED assessment (N = 3929) and adults with no DNACPR or DNACPR decision made later (N = 8819).
      OutcomeLevelEarly DNACPRNo/late DNACPR
      Location of first admissionMissing98194
      Ward3717 (97%)7802 (90.5%)
      ITU57 (1.5%)667 (7.7%)
      HDU57 (1.5%)156 (1.8%)
      Respiratory pathogenCOVID-191791 (45.6%)3367 (38.2%)
      Influenza6 (0.2%)14 (0.2%)
      Other235 (6%)701 (7.9%)
      None identified1897 (48.3%)4737 (53.7%)
      Mortality statusMissing00
      Alive2328 (59.3%)7668 (86.9%)
      Dead1601 (40.7%)1151 (13.1%)
       Death with organ support251 (15.7%)373 (32.4%)
       Death without organ support1350 (84.3%)778 (67.6%)
      Organ supportRespiratory423 (10.8%)1313 (14.9%)
       Mechanical ventilation51 (1.3%)509 (5.8%)
       Non-invasive ventilation173 (4.4%)292 (3.3%)
       Continuous positive airway pressure125 (3.2%)386 (4.4%)
       High-flow nasal oxygen74 (1.9%)126 (1.4%)
      Cardiovascular47 (1.2%)426 (4.8%)
       Extracorporeal membrane oxygenation0 (0%)13 (0.1%)
       Inotropic/vasopressor drugs29 (0.7%)283 (3.2%)
       Central venous pressure measurement2 (0.1%)25 (0.3%)
       Intra-arterial BP measurement16 (0.4%)105 (1.2%)
      Renal29 (0.7%)172 (2%)
       Haemofiltration7 (0.2%)93 (1.1%)
       Haemodialysis22 (0.6%)75 (0.9%)
       Peritoneal dialysis0 (0%)4 (0%)
      Any455 (11.6%)1386 (15.7%)
      Table 3 shows the results of the multivariable logistic regression model. Older age, active malignancy, chronic lung disease, limited performance status, abnormal respiratory rate, lower oxygen saturation, and lower alertness were all associated with increased use of early DNACPR. Asian ethnicity was associated with a lower use of early DNACPR.
      Table 3Multivariable analysis of predictors of early DNACPR use.
      EffectOdds ratio95% CIp-Value
      Age1.054(1.049, 1.060)<0.001
      Male sex1.010(0.905, 1.128)0.859
      Ethnicity (ref = UK/Irish/other white)
       Asian0.571(0.416, 0.783)0.001
       Black/African/Caribbean0.730(0.524, 1.017)0.063
       Mixed/multiple ethnic groups0.993(0.568, 1.737)0.982
       Other0.647(0.410, 1.020)0.061
      Shortness of breath1.150(1.005, 1.316)0.042
      Cough1.012(0.903, 1.133)0.841
      Fever0.954(0.851, 1.069)0.415
      No chronic disease0.753(0.609, 0.931)0.009
      Heart disease1.182(1.048, 1.333)0.006
      Renal impairment1.241(1.067, 1.445)0.005
      Steroid therapy1.268(0.971, 1.657)0.082
      Asthma0.900(0.765, 1.060)0.209
      Diabetes1.120(0.987, 1.271)0.080
      Active malignancy1.604(1.319, 1.951)<0.001
      Immunosuppression1.117(0.835, 1.494)0.455
      Other chronic lung disease1.456(1.280, 1.656)<0.001
      Hypertension0.883(0.786, 0.993)0.038
      Symptom duration (days)0.993(0.986, 1.001)0.076
      Pulse rate (beats/min; ref = 51−90)
       41−50 or 91−1101.121(0.987, 1.274)0.079
       111−1301.095(0.929, 1.290)0.280
       ≤40 or ≥1311.030(0.816, 1.300)0.802
      Respiratory rate (breaths/min; ref = 12−20)
       9−110.772(0.115, 5.173)0.790
       21−241.175(1.019, 1.355)0.027
       ≤8 or ≥251.361(1.187, 1.560)<0.001
      Systolic BP (mmHg; ref = 111−219)
       101−1101.164(0.968, 1.400)0.107
       91−1001.347(1.066, 1.703)0.013
       ≤90 or ≥2201.177(0.891, 1.555)0.251
      Temperature (°C; ref = 36.1−38.0)
       35.1−36.0 or 38.1−39.00.916(0.807, 1.039)0.173
       ≥39.10.885(0.693, 1.131)0.330
       ≤35.01.085(0.764, 1.540)0.650
      Oxygen saturation (%; ref=≥96)
       94−951.041(0.898, 1.206)0.595
       92−931.127(0.938, 1.354)0.203
       ≤911.329(1.153, 1.532)<0.001
      AVPU (ref = Alert)
       Verbal1.905(1.563, 2.323)<0.001
       Pain2.651(1.644, 4.272)<0.001
       Unresponsive2.620(1.429, 4.803)0.002
      Performance status (ref = Unrestricted normal activity)
       Limited strenuous activity, can do light activity1.885(1.565, 2.269)<0.001
       Limited activity, can self care2.629(2.222, 3.110)<0.001
       Limited self care4.100(3.448, 4.876)<0.001
       Bed/chair bound, no self care5.437(4.438, 6.660)<0.001
      Finally, we undertook a post hoc sensitivity analysis to determine whether the multivariable analysis was robust to using a more liberal definition of early DNACPR, in which 870 patients with a DNACPR decision on the day after attendance were reclassified as having an early decision. This is presented in the supplementary table and shows that the most significant predictors were unchanged.

      Discussion

      We found that 31% of adults admitted to hospital with suspected COVID-19 during the first phase of the pandemic had a DNACPR decision recorded on or before their day of attendance, after excluding those who could not be classified. Most patients (59.4%) with an early DNACPR decision survived to 30 days and 11.6% received some form of organ support. These findings show that potentially life-saving treatments were provided to a significant proportion of people, potentially addressing concerns that DNACPR decisions may be conflated with ‘do not provide active treatment’.
      • Perkins G.D.
      • Griffiths F.
      • Slowther A.M.
      • et al.
      Do-not-attempt-cardiopulmonary-resuscitation decisions: an evidence synthesis.
      The use of invasive intervention, particularly mechanical ventilation, in people with a DNACPR decision was an unexpected finding. Contact with participating site investigators suggested that this could be explained by use of the ReSPECT process in discussions about resuscitation, in which the patient is encouraged to explicitly indicate which treatments they want in a future situation where they are unable to make or express choices.

      Resuscitation Council UK. Resuscitation Council UK Statement on the role of the ReSPECT Process during COVID-19, 23 April 2020. https://www.resus.org.uk/sites/default/files/2020-06/COVID%20ReSPECT%20Guidance%2023042020_0.pdf (Accessed 14/01/2021).

      The ReSPECT process therefore allows patients to consent to mechanical ventilation but decline cardiopulmonary resuscitation if it is subsequently required.
      Older age, active malignancy, chronic lung disease (excluding asthma), and lower performance status were associated with increased use of early DNACPR, whereas Asian ethnicity was associated with decreased use. Patients with early DNACPR decisions tended to be more acutely ill, with higher NEWS2 scores. Abnormal respiratory rate, lower oxygen saturation, and lower alertness were associated with increased use of early DNACPR.
      Our findings suggest a higher rate of DNACPR use (31%) than identified in previous studies of other respiratory conditions or sepsis. Studies of patients admitted with community-acquired pneumonia reported rates of DNACPR use ranging from 13% to 29%,
      • Egelund G.B.
      • Jensen A.V.
      • Petersen P.T.
      • et al.
      Do-not-resuscitate orders in patients with community-acquired pneumonia: a retrospective study.
      • Marrie T.J.
      • Fine M.J.
      • Kapoor W.N.
      • Coley C.M.
      • Singer D.E.
      • Obrosky D.S.
      Community-acquired pneumonia and do not resuscitate orders.
      • Oshitani Y.
      • Nagai H.
      • Matsui H.
      Rationale for physicians to propose do-not-resuscitate orders in elderly community-acquired pneumonia cases.
      • Walkey A.J.
      • Weinberg J.
      • Wiener R.S.
      • Cooke C.R.
      • Lindenauer P.K.
      Association of do-not-resuscitate orders and hospital mortality rate among patients with pneumonia.
      • Mulder T.
      • van Werkhoven C.H.
      • Huijts S.M.
      • Bonten M.J.
      • Postma D.F.
      • Oosterheert J.J.
      Treatment restrictions and empirical antibiotic treatment of community-acquired pneumonia in elderly patients.
      while studies in severe sepsis or septic shock reported rates ranging from 9% to 20%.
      • Sarkari N.N.
      • Perman S.M.
      • Ginde A.A.
      Impact of early do-not-attempt-resuscitation orders on procedures and outcomes of severe sepsis.
      • Bradford M.A.
      • Lindenauer P.K.
      • Wiener R.S.
      • Walkey A.J.
      Do-not-resuscitate status and observational comparative effectiveness research in patients with septic shock.
      • Powell E.
      • Sauser K.
      • Cheema N.
      • et al.
      Severe sepsis in do-not-resuscitate patients: intervention and mortality rates.
      • Huang C.T.
      • Chuang Y.C.
      • Tsai Y.J.
      • Ko W.J.
      • Yu C.J.
      High mortality in severe sepsis and septic shock patients with do-not-resuscitate orders in East Asia.
      DNACPR decisions in these studies were associated with older age, but conflicting findings were reported around the use of invasive procedures. Sakari et al.
      • Sarkari N.N.
      • Perman S.M.
      • Ginde A.A.
      Impact of early do-not-attempt-resuscitation orders on procedures and outcomes of severe sepsis.
      and Bradford et al.
      • Bradford M.A.
      • Lindenauer P.K.
      • Wiener R.S.
      • Walkey A.J.
      Do-not-resuscitate status and observational comparative effectiveness research in patients with septic shock.
      reported that DNACPR decision were associated with lower rates of invasive procedures, while Powell et al.
      • Powell E.
      • Sauser K.
      • Cheema N.
      • et al.
      Severe sepsis in do-not-resuscitate patients: intervention and mortality rates.
      reported no difference, and Huang et al.
      • Huang C.T.
      • Chuang Y.C.
      • Tsai Y.J.
      • Ko W.J.
      • Yu C.J.
      High mortality in severe sepsis and septic shock patients with do-not-resuscitate orders in East Asia.
      reported a higher rate of arterial or central venous cannulation in those with a DNACPR decision. Our data cannot determine why there was a relatively high rate of DNACPR use, so further research could be helpful to identify whether this reflects concerns around personal risks to staff or the lack of effective treatments in the first phase of the pandemic.
      We found an association between Asian ethnicity and decreased use of early DNACPR decisions compared to White ethnicity. The odds ratio for Black/African/Caribbean ethnicity also suggested decreased use but was not significant. Previous studies from the United States have shown less use of DNACPR decisions among African-American, Asian, and Hispanic patients,
      • Bailey F.A.
      • Allen R.S.
      • Williams B.R.
      • et al.
      Do-not-resuscitate orders in the last days of life.
      • Burgio K.L.
      • Williams B.R.
      • Dionne-Odom J.N.
      • et al.
      Racial differences in processes of care at end of life in VA medical centers: planned secondary analysis of data from the BEACON trial.
      • Richardson D.K.
      • Zive D.M.
      • Newgard C.D.
      End-of-life decision-making for patients admitted through the emergency department: hospital variability, patient demographics, and changes over time.
      and Black patients tend to receive more life-prolonging treatment at end of life care.
      • Mack J.W.
      • Paulk M.E.
      • Viswanath K.
      • Prigerson H.G.
      Racial disparities in the outcomes of communication on medical care received near death.
      A systematic review of end of life decisions for people from ethnic minority groups suggested that Hispanic and African American people had advance care plans documented less often, citing religious coping and spirituality as factors.
      • LoPresti M.A.
      • Dement F.
      • Gold H.T.
      End-of-life care for people with cancer from ethnic minority groups: a systematic review.
      A scoping review of culturally- and spiritually-sensitive end-of-life care highlighted a multitude of factors influencing end-of-life care and subsequent experiences by culturally- and spiritually-diverse groups.
      • Fang M.L.
      • Sixsmith J.
      • Sinclair S.
      • Horst G.
      A knowledge synthesis of culturally- and spiritually-sensitive end-of-life care: findings from a scoping review.
      Further research of DNACPR decisions in relation to ethnicity is clearly required.
      Studies of DNACPR use in COVID-19 are currently limited. Alhatem et al.
      • Alhatem A.
      • Spruijt O.
      • Heller D.S.
      • Chokshi R.J.
      • Schwartz R.A.
      • Lambert W.C.
      “Do-not-resuscitate (DNR)” status determines mortality in patients with COVID-19.
      analysed 1270 patients admitted across two hospitals with COVID-19, of whom 750/1270 (59%) had a DNACPR decision at admission, and 570/750 (76%) of these died. Age over 60, male sex, and comorbidities were associated with DNACPR at admission. Coleman et al.
      • Coleman J.J.
      • Botkai A.
      • Marson E.J.
      • et al.
      Bringing into focus treatment limitation and DNACPR decisions: How COVID-19 has changed practice.
      examined records of DNACPR decisions at a single centre from 2017 to 2020 and showed an increased rate of DNACPR use during pandemic, with patients tending to be younger and have fewer comorbidities. It is unclear whether these findings reflect an increased overall need for DNACPR decisions during the pandemic or increased willingness to use DNACPR decisions in COVID-19.
      This study was based on a large representative sample of adults admitted with suspected COVID-19, but has a number of limitations. DNACPR decisions were recorded to facilitate subgroup analyses addressing the primary purpose of the study rather than addressing the aims of this secondary analysis. We were unable to include 1249/13997 (9%) cases because data were missing or uncertain regarding the use or timing for DNACPR. Recording of DNACPR decision in the notes was inconsistent between sites and some recorded decisions may have been missed. Our categorisation on the basis of timing of DNACPR decision and assumption that later DNACPR decisions are qualitatively different to early decisions could be challenged, and the categorisation may have been too early to capture the impact of senior or specialist review. We did not collect any detailed data to allow us to explore the reasons behind DNACPR decisions, so we are unable to offer explanations for the associations identified in our analysis. We also did not collect data on whether sites used ReSPECT or other similar processes, or isolated DNACPR decisions. Our suggestion that the ReSPECT process could explain the use of invasive interventions in people with a DNACPR decision is based on informal contacts and requires further research. The use of the ReSPECT process could also undermine our rationale for categorising DNACPR decisions as early versus late or no decision, and suggests a complex relationship between DNACPR decisions and subsequent interventions.

      Conclusion

      We found that many patients with an early DNACPR decision went on to receive life-saving interventions and most survived to 30 days. Early DNACPR decisions were associated with older age, lower performance status, active malignancy, chronic lung disease, and severe illness, as indicated by physiological parameters. We found some evidence of an association between ethnicity and DNACPR decisions that requires further research.

      Role of the funding source

      The PRIEST study was funded by the U nited Kingdom National Institute for Health Research Health Technology Assessment (HTA) programme (project reference 11/46/07). The funder played no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the article for publication. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health and Social Care.

      Credit author statement

      Steve Goodacre: conceived and designed the study. Ben Thomas: oversaw data acquisition. Laura Sutton: analysed the data. Sarah Connelly: undertook literature searches. Steve Goodacre, Laura Sutton, Ben Thomas and Sarah Connelly interpreted the data. All authors contributed to drafting the manuscript. Steve Goodacre is the guarantor of the paper. The corresponding author attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted.

      Conflict of interest statement

      All authors have completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf and declare: grant funding to their employing institutions from the National Institute for Health Research; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.

      Data sharing

      Anonymised data are available from the corresponding author upon reasonable request (contact details on first page).

      Acknowledgements

      We thank Katie Ridsdale for clerical assistance with the study, Erica Wallis (Sponsor representative), all members of the PRIEST Research Team, (Appendix 3), all members of the Study Steering Committee (Appendix 4) and the site research teams who delivered the data for the study (Appendix 5), and the research team at the University of Sheffield past and present (Appendix 6).

      References

        • Hayek Salim S.
        • Brenner Samantha K.
        • Azam Tariq U.
        • et al.
        In-hospital cardiac arrest in critically ill patients with covid-19: multicenter cohort study.
        BMJ. 2020; 371: m3513
        • Perkins G.D.
        • Morley P.T.
        • Nolan J.P.
        • et al.
        International Liaison Committee on Resuscitation: COVID-19 consensus on science, treatment recommendations and task force insights.
        Resuscitation. 2020; 151: 145-147
        • Curtis J.R.
        • Kross E.K.
        • Stapleton R.D.
        The importance of addressing advance care planning and decisions about do-not-resuscitate orders during novel coronavirus 2019 (COVID-19).
        JAMA. 2020; 323: 1771-1772
        • Clare Dyer
        Covid-19: Campaigner calls for national guidance to stop DNR orders being made without discussion with patients and families.
        BMJ. 2020; 369: m1856
      1. Care Quality Commission. Reviewing the use of do not resuscitate decisions during COVID-19, 17/11/2020. https://www.cqc.org.uk/news/stories/reviewing-use-do-not-resuscitate-decisions-during-covid-19 (Accessed 11/12/2020).

        • Egelund G.B.
        • Jensen A.V.
        • Petersen P.T.
        • et al.
        Do-not-resuscitate orders in patients with community-acquired pneumonia: a retrospective study.
        BMC Pulm Med. 2020; 20: 201
        • Marrie T.J.
        • Fine M.J.
        • Kapoor W.N.
        • Coley C.M.
        • Singer D.E.
        • Obrosky D.S.
        Community-acquired pneumonia and do not resuscitate orders.
        J Am Geriatr Soc. 2002; 50: 290-299
        • Oshitani Y.
        • Nagai H.
        • Matsui H.
        Rationale for physicians to propose do-not-resuscitate orders in elderly community-acquired pneumonia cases.
        Geriatr Gerontol Int. 2014; 14: 54-61
        • Walkey A.J.
        • Weinberg J.
        • Wiener R.S.
        • Cooke C.R.
        • Lindenauer P.K.
        Association of do-not-resuscitate orders and hospital mortality rate among patients with pneumonia.
        JAMA Intern Med. 2016; 176: 97-104
        • Mulder T.
        • van Werkhoven C.H.
        • Huijts S.M.
        • Bonten M.J.
        • Postma D.F.
        • Oosterheert J.J.
        Treatment restrictions and empirical antibiotic treatment of community-acquired pneumonia in elderly patients.
        Neth J Med. 2016; 74: 56
        • Sarkari N.N.
        • Perman S.M.
        • Ginde A.A.
        Impact of early do-not-attempt-resuscitation orders on procedures and outcomes of severe sepsis.
        J Crit Care. 2016; 36: 134-139
        • Bradford M.A.
        • Lindenauer P.K.
        • Wiener R.S.
        • Walkey A.J.
        Do-not-resuscitate status and observational comparative effectiveness research in patients with septic shock.
        Crit Care Med. 2014; 42: 2042-2047
        • Powell E.
        • Sauser K.
        • Cheema N.
        • et al.
        Severe sepsis in do-not-resuscitate patients: intervention and mortality rates.
        J Emerg Med. 2013; 44: 742-749
        • Huang C.T.
        • Chuang Y.C.
        • Tsai Y.J.
        • Ko W.J.
        • Yu C.J.
        High mortality in severe sepsis and septic shock patients with do-not-resuscitate orders in East Asia.
        PLoS One. 2016; 11e0159501
        • Goodacre S.
        • Thomas B.
        • Lee E.
        • et al.
        Characterisation of 22445 patients attending UK emergency departments with suspected COVID-19 infection: Observational cohort study.
        PLoS One. 2020; 15e0240206
        • Royal College of Physicians
        National early warning score (NEWS) 2: Standardising the assessment of acute-illness severity in the NHS. Updated report of a working party.
        RCP, London2017
        • Goodacre S.
        • Thomas B.
        • Sutton L.
        • et al.
        Derivation and validation of a clinical severity score for acutely ill adults with suspected COVID-19: the PRIEST observational cohort study.
        PLoS ONE. 2021; 16 (In this issue)e0245840https://doi.org/10.1371/journal.pone.0245840
        • Perkins G.D.
        • Griffiths F.
        • Slowther A.M.
        • et al.
        Do-not-attempt-cardiopulmonary-resuscitation decisions: an evidence synthesis.
        Obtaining service providers’ perspectives on do-not-attempt-cardiopulmonary-resuscitation decisions. NIHR Journals Library (Health Services and Delivery Research, No. 4.11.), Southampton (UK)2016: 1-80 (Chapter 5)
      2. Resuscitation Council UK. Resuscitation Council UK Statement on the role of the ReSPECT Process during COVID-19, 23 April 2020. https://www.resus.org.uk/sites/default/files/2020-06/COVID%20ReSPECT%20Guidance%2023042020_0.pdf (Accessed 14/01/2021).

        • Bailey F.A.
        • Allen R.S.
        • Williams B.R.
        • et al.
        Do-not-resuscitate orders in the last days of life.
        J Palliat Med. 2012; 15: 751-759
        • Burgio K.L.
        • Williams B.R.
        • Dionne-Odom J.N.
        • et al.
        Racial differences in processes of care at end of life in VA medical centers: planned secondary analysis of data from the BEACON trial.
        J Palliat Med. 2016; 19: 157-163
        • Richardson D.K.
        • Zive D.M.
        • Newgard C.D.
        End-of-life decision-making for patients admitted through the emergency department: hospital variability, patient demographics, and changes over time.
        Acad Emerg Med. 2013; 20: 381-387
        • Mack J.W.
        • Paulk M.E.
        • Viswanath K.
        • Prigerson H.G.
        Racial disparities in the outcomes of communication on medical care received near death.
        Arch Intern Med. 2010; 170: 1533-1540
        • LoPresti M.A.
        • Dement F.
        • Gold H.T.
        End-of-life care for people with cancer from ethnic minority groups: a systematic review.
        Am J Hosp Palliat Care. 2016; 33: 291-305
        • Fang M.L.
        • Sixsmith J.
        • Sinclair S.
        • Horst G.
        A knowledge synthesis of culturally- and spiritually-sensitive end-of-life care: findings from a scoping review.
        BMC Geriatr. 2016; 16: 107
        • Alhatem A.
        • Spruijt O.
        • Heller D.S.
        • Chokshi R.J.
        • Schwartz R.A.
        • Lambert W.C.
        “Do-not-resuscitate (DNR)” status determines mortality in patients with COVID-19.
        Clin Dermatol. 2020; (Epub ahead of print)https://doi.org/10.1016/j.clindermatol.2020.11.013
        • Coleman J.J.
        • Botkai A.
        • Marson E.J.
        • et al.
        Bringing into focus treatment limitation and DNACPR decisions: How COVID-19 has changed practice.
        Resuscitation. 2020; 155: 172-179https://doi.org/10.1016/j.resuscitation.2020.08.006