key: cord-1042146-utoa1d7o
authors: Wendt, Christine; Mobus, Kristi; Wiener, Dan; Eskin, Barnet; Allegra, John R.
title: Prone Positioning on Non-intubated COVID 19 Patients in Hypoxic Respiratory Distress: Single Site Retrospective Health Records Review
date: 2020-12-26
journal: J Emerg Nurs
DOI: 10.1016/j.jen.2020.12.006
sha: 90ac0ae1fd6649745821f2f53f169d87510236db
doc_id: 1042146
cord_uid: utoa1d7o

INTRODUCTION: In March and April 2020 of the COVID-19 pandemic, site clinical practice guidelines were implemented for prone positioning of awake, alert, spontaneously breathing suspected COVID-19 patients in hypoxic respiratory distress. The purpose of this pandemic disaster practice improvement project was to measure changes in pulse oximetry associated with prone positioning on awake, alert, spontaneously breathing non-intubated adult acute respiratory distress, or ARDS, patients with COVID-19 infection. METHODS: A retrospective chart review of ED COVID-19 positive patients from 3/30/2020 to 4/30/2020 was conducted for patients with a room air pulse oximetry < 90% and a pre-prone position pulse oximetry ≤ 94% who tolerated prone positioning for at least 30 minutes. The primary outcome was change in pulse oximetry associated with prone positioning, measured on room air, with supplemental oxygen, and approximately 30 minutes after initiating prone positioning. Median differences were compared with the Wilcoxon signed-rank test. RESULTS: Of the 440 COVID-19 patients, 31 met inclusion criteria. Median pulse oximetry increased as 83% (IQR= 75%-86%) on room air, 90% (IQR=89%-93%) with supplemental oxygen, and 96% (IQR=94%-98%) with prone positioning (x.xx, p<.001). 45% (N=14) were intubated during their hospital stay and 26% (N=8) of the included patients died. CONCLUSION: In awake, alert, and spontaneously breathing patients with COVID-19, an initially low pulse oximetry reading improved with prone positioning. Future studies are needed to determine the association of prone positioning with subsequent endotracheal intubation and mortality.

Introduction : In March and April 2020 of the COVID-19 pandemic, site clinical practice guidelines were 6

implemented for prone positioning of awake, alert, spontaneously breathing suspected COVID-19 7 patients in hypoxic respiratory distress. The purpose of this pandemic disaster practice improvement 8 project was to measure changes in pulse oximetry associated with prone positioning on awake, alert, 9

spontaneously breathing non-intubated adult acute respiratory distress, or ARDS, patients with COVID-10 19 infection. 11

Methods : A retrospective chart review of ED COVID-19 positive patients from 3/30/2020 to 4/30/2020 12 was conducted for patients with a room air pulse oximetry < 90% and a pre-prone position pulse 13 oximetry ≤ 94% who tolerated prone positioning for at least 30 minutes. The primary outcome was 14 change in pulse oximetry associated with prone positioning, measured on room air, with supplemental 15 oxygen, and approximately 30 minutes after initiating prone positioning. Median differences were 16 compared with the Wilcoxon signed-rank test. 17

Results : Of the 440 COVID-19 patients, 31 met inclusion criteria. Median pulse oximetry increased as 18 83% (IQR= 75%-86%) on room air, 90% (IQR=89%-93%) with supplemental oxygen, and 96% (IQR=94%-19 98%) with prone positioning (x.xx, p<.001). 45% (N=14) were intubated during their hospital stay and 20 26% (N=8) of the included patients died. 21

The 2019 Novel Coronavirus (COVID-19) emerged out of the Hubei Provence of China in November 28 2019. The first United States (US) case of COVID-19 was confirmed on 1/20/2020 in Seattle, WA. In the 29 subsequent weeks, the virus spread globally and was declared a pandemic by the World Health 30

Organization on 3/11/2020. In our emergency department in northern New Jersey, the first COVID-19 31 patient arrived on 3/11/2020. 32 J o u r n a l P r e -p r o o f Emergency services worldwide were tasked with responding to a crisis with presentations ranging from 33 asymptomatic patients to those in hypoxic respiratory distress. Testing centers appeared across the US in 34 the form of tents and drive-throughs, thus accommodating the patients' requests for testing, however this 35 met only the need of the "walking well". 36

EDs in areas experiencing clustered high incidences of the spread of COVID-19 were inundated with 37 symptomatic patients, many of whom arrived critically ill. As this crisis unfolded, little was known about RR of 0.77 (95% CI 0.61 to 0.99); and participants with more severe hypoxemia at trial entry (six trials; 84 1108 participants showed an RR of 0.77 (95% CI 0.65 to 0.92)). The study also showed an improvement 85 in oxygenation. The mean difference in the PaO2/FIO2 between prone and supine positioning was 24.6 86 mmHg (95% CI 13.9 to 35.2 mmHg). PP reduced ventilator-associated pneumonia and days on the 87 ventilator but appears to have increased length of ICU and hospital stay. 88

The primary aim of this pandemic disaster practice improvement project was to measure changes in 90 pulse oximetry associated with prone positioning on awake, alert, spontaneously breathing non-91 intubated adult ARDS patients with COVID-19 infection. The secondary aim was to analyze changes in 92 respiratory rate and heart rate associated with proning in these patients. 93 94

This study was a pandemic disaster practice improvement initiative using retrospective chart review. 97

The practice site was a suburban hospital ED in northern New Jersey with an annual volume of 90,000. 99

The hospital was a Level 1 trauma center and has several residencies, including one in emergency 100

Site guidelines were implemented for PP of awake, alert, spontaneously breathing suspected COVID-19 103 infected patients with hypoxic respiratory distress (Supplemental Online Appendix). Included were 104 positioning recommendations and contraindications consistent with those described in the PROSEVA 105 trial. 7 Beginning March 30, 2020 emergency nurses and physicians were encouraged to prone position 106 J o u r n a l P r e -p r o o f this patient population. Guidelines were communicated to nurses staffing the emergency department at the 107 multiple shift change huddles and via email. Staff assisted the patient in assuming a prone position, with 108 the stretcher positioned in mild reverse Trendelenburg. The patient was asked to remain prone for at least 109 2 hours or as long as tolerated and no clinical deterioration was noted. Patients were provided with 110 pillows and/or blankets to position comfortably and to cushion bony prominences and were encouraged to 111 move frequently while maintaining PP. Providers were asked to enter a nursing communication "Keep 112

Prone" in the electronic medical record (EMR) for ease of data extraction. 113

Data source and inclusion criteria: A report was run in the EPIC (Epic Systems Corporation, Verona, 114 WI) EMR system to identify all COVID-19 adult patients admitted through the emergency department 115 between 3/30/2020 and 4/30/2020. Patients who met the following criteria were included: assuming PP 116 by themselves and tolerating it for at least 30 minutes, documented room air pulse oximetry (SpO2) < 117 90% and pre-PP SpO2 ≤ 94% despite supplemental oxygen. 118

The following data from the EMRs of the patients meeting inclusion criteria were extracted: 120 length of time from arrival to PP, SpO2 on room air, heart and respiratory rate, SpO2 pre and post PP, and 121 length of time proned. While these data are repeated measures on the clinical record, only 1 measure for 122 each variable was extracted. The post PP measure closest to 30 minutes after the onset of proning was 123 recorded. Demographic data, level of care on admission, intubation during hospitalization (including the 124 length of time from ED arrival to intubation), and the existence of a "do not intubate order," length of 125 hospital stay, and disposition upon discharge were collected. The primary outcome was the pre-to post-126 proning change in SpO2. Pre-to post-proning changes in respiratory and heart rate were analyzed as 127 secondary measures. Two of the authors abstracted data from the EMR. The two data abstracters 128 examined ten charts together, with excellent agreement. The rare circumstance of uncertainty was 129 resolved by consensus. Caucasian, 9% Asian, 6% African American and 6% unspecified. These demographic parameters are 147 different from the typical patient population in this emergency department (Table 1) . 148

Median time from patient arrival to PP was 85 minutes (IQR=46-174). For the 13 (42%) of 150 patients for whom the times were recorded, the duration of PP was 140 (SD=47) min. For the 31 patients 151 included in the study, the least recorded duration of PP was 51 minutes and for that individual patient, the 152 SpO2 rose from 93% to 96% during PP. All but 4 patients (13%) were given supplemental oxygen (from 153 J o u r n a l P r e -p r o o f 2 to 21 l/min by nasal cannula and/or non-rebreather mask) and then were proned. Median SpO2's were 154 83% (IQR= 75-86%) on room air, 90% (IQR=89%-93%) with supplemental oxygen, and 96% (IQR=94-155 98%) with PP. (see Table 2 and Figure 1 ). The 5% (IQR 4%, 9%) median change from before to with PP 156 was statistically significant (X.XX, p<0.001). 157

Supplemental oxygen was increased for 7 (23%) patients when placed in the prone position. 158

Considering only the 24 patients for whom supplemental oxygen was not increased, the median SpO2's 159 before and with PP were 92% (IQR=89-93%) and 96% (IQR=94-98%), respectively. For these 24 160 patients, the 4% (IQR=3-6%) change from before to with PP was statistically significant (X.XX, 161

For all 31 patients, both HR and respiratory rate showed small decreases after being placed in the 163 prone position. Mean HR and respiratory rate before PP were 93 (SD=18) and 31 (SD=9) beats/min, 164 respectively. With PP the rates were 88 (SD=15) and 26 (SD=8) beats/min, respectively. These changes 165 were statistically significant (HR change: 5, SD=11, beats/min, X.XX, p < 0.001 and respiratory rate 166 change: 5, SD=17 breaths/min, p = 0.007). We studied the association of PP and SpO2 on non-intubated ED patients with COVID19. This work was 178 a single-site, pragmatic pandemic process implementation in a "real world" clinical setting that 179 demonstrated feasibility and initial effectiveness of the intervention for the included patients, and should 180 not be interpreted as testing the efficacy of PP as a controlled clinical trial. Only two studies on this 181 patient population, with a total of 56 patients, have been previously published when we began this work. 182

Our results confirm most of the findings of these studies, discussed in more detail below, which increases 183 confidence in the reproducibility of these findings. Our study is unique for reporting the ethnicity of the 184 patients (the majority were Hispanic), in-hospital disposition (32% were admitted to the ICU) and 185 mortality (26% died). To contextualize our findings, we found two previous reports on the effectiveness 186 of PP on mortality and intubation rate, with conflicting results. 187

We found that for the 31 patients that were proned in the ED, SpO2 increased by a median of 5% (IQR 189 4%, 9%) with PP, from a borderline oxygenation level of 90% (IQR 89%, 93%) before PP to a more non-intubated COVID-19 patients but only one was done entirely on emergency department patients. 8-17 201 In this study of 50 patients, the median age of 59 (IQR 50,68) was similar to our median age of 62, but a 202 larger proportion were female (40%, compared to our 13%). 18 The median SpO2 on ED arrival of 80% 203 (IQR 69, 85) increased to 84% (IQR 75,90) after supplemental oxygen and then 94% (IQR 90,95) with 204 PP. This 10% increase from before to with PP was statistically significant (p=0.001). This change was 205 greater than the 5% change we found, although the SpO2 with PP was similar to our 96% finding. 206

Eighteen (36%) patients were intubated, with the median time until intubation in the 1-24 hour period 207 after ED arrival. This is slightly smaller than our 45% intubation rate, although the median time to 208 intubation was shorter than in our study (35 hours). Mortality statistics were not reported in these other 209 studies. The overall outcomes of the 13 previous studies mentioned above (including the ED study just 210 described) were reported as PaO2 in 2 studies (33 patients), PaO2/ FIO2 in 5 studies (78 patients), SpO2 211 in 6 studies (118 patients) and "oxygenation" in one study (10 patients). Mean changes in PaO2, 212

PaO2/FIO2 and SpO2 with PP were 30 ± 13 mmHg, 80 ± 87 mmHg and 8 ± 2%, respectively. Again, the 213 latter change was somewhat larger than what we found (5%). In the 13 studies, 59 (26%) patients were 214 intubated. Calculating the median rate for the individual studies yields a median intubation rate of 21% 215 (IQR 7,33). Only 8 studies (139 patients) reported mortality results and in those studies 11 (8%) died. 216

Calculating the median rate for individual studies yields a median mortality rate of 3% (IQR 0%,10%). 217

Both these intubation and mortality rates are less than what we found. Although not directly comparable 218 to our study, we did find one other study on PP in intubated COVID-19 patients. Carsetti et al 219 retrospectively reviewed 10 intubated COVID-19 patients, whose median PaO2/FIO2 before PP was 126 220 mmHg. 19 With PP for either 16-or 36-hour cycles, PaO2/FIO2 increased significantly to 177 and 394 221 mmHg, respectively, and remained elevated after subsequent supine repositioning (166 and 290 mmHg, 222 respectively). 223

When our findings are contextualized in the published literature, we interpret that our results 224 corroborate the association of PP with increased pulse oximetry outcomes. Effectiveness of PP on longer-225 term outcomes of mortality and intubation rates are conflicting. Future study is needed to determine the 226 required duration of PP to improve outcomes and the effect of PP on intubation and mortality in COVID-227 19 patients. 228

Limitations include a small sample size, demographics that may limit generalizability (87% male, 55% particularly for the duration of the PP intervention. Although SPO2 is less accurate than other invasive 235 measures, it is the standard method to monitor oxygenation in the emergency department. 236

As a retrospective review, there is no assurance that all patients who met the inclusion criteria 237 were place in PP by the emergency staff, nor that all PP intervention was accurately recorded in the EMR 238 for inclusion in the study. Retrospective data abstraction has innate problems and shortcomings. 20 239

Although the data abstracters were not blinded to the purpose of the study, there were well defined 240 objective data present in the same place in the EMR to limit bias. Despite these limitations, our study 241 design demonstrates an initial feasibility and effectiveness in achieving the intended clinical results at our 242 site to raise SpO2 by implementing a proning guideline for COVID-19 patients in the emergency 243 department. 244

Emergency nurses must implement practice changes to meet the needs of patients presenting with ARDS, 246

including those with COVID-19. As COVID-19 cases continue to occur in the US it is essential to 247 provide early intervention for patients presenting in respiratory failure. Management of this patient 248 population has been challenging from a logistical as well as clinical standpoint. Considering limitations in 249 use of non-invasive respiratory support devices, CPAP and HFNC, the application of PP is a potential 250 alternative to improve patient SpO2. The currently published evidence supports the early use of PP for 251 intubated patients. Implementation of PP guidelines for alert suspected COVID-19 patients arriving to the 252 emergency department is nurse driven and can be accomplished quickly and with little additional 253 expense. Although some patients did not tolerate PP, this intervention appears to be safe and feasible in 254 this patient population. Emergency nurses are pivotal in expanding the use of PP to awake, alert 255 spontaneously breathing patients. 256

We demonstrated a single site, pandemic practice guideline implementation of PP was feasible and 258 associated with improved Sp02 approximately 30 minutes after the initiation of PP for the included 259 awake, alert, and non-intubated COVID-19 patients. PP of awake and alert COVID-19 patients, not 260 receiving non-invasive or invasive respiratory support, presenting to the emergency department with low 261 pulse oximetry was associated with a 5% improvement in pulse oximetry readings. Future studies are 262 needed to determine the required duration of PP to improve outcomes, and the effect of PP on rates of 263 endotracheal intubation and long-term survival. • The main finding of this paper is significant improvement in oxygen saturation during prone positioning of emergency department awake, alert, non-intubated COVID-19 patients in hypoxic respiratory distress.

• Recommendations for translating the findings of this paper into clinical practice include: emergency nurse initiated prone positioning guidelines for awake, alert, non-intubated ED COVID-19 patients are easily implemented with positive patient impact.

Efficacy and safety of early prone positioning combined with 270

HFNC or NIV in moderate to severe ARDS: a multi-center prospective cohort study

A comprehensive review of prone position in ARDS. Resp Care

Improved oxygenation in patients with acute 275 respiratory failure: the prone position

Prone position for acute respiratory failure in adults

Response to the prone position in spontaneously 279 breathing patients with hypoxemic respiratory failure

Prone positioning improves oxygenation in 282 spontaneously breathing nonintubated patients with hypoxemic acute respiratory failure: A 283 retrospective study

PROSEVA Study Group. Prone positioning in severe 285 acute respiratory distress syndrome

Prone positioning in conscious patients on medical wards: A review of the 287 evidence and its relevance to patients with COVID-19 infection

Feasibility and physiological effects of prone 290 positioning in non-intubated patients with acute respiratory failure due to

COVID): a prospective cohort study

Prone positioning of non-Intubated patients with 294 COVID-19

Prone positioning combined with high-flow nasal or 296 conventional oxygen therapy in severe Covid-19 patients

Early application of prone position for management of 299

Covid-19 patients

Rationale and significance of patient selection in awake 301 prone positioning for COVID-19 pneumonia

Prone positioning in management of COVID-19 304 hospitalized patients

Prone positioning in awake, nonintubated 306 patients with COVID-19 hypoxemic respiratory failure

Prone positioning in high-flow nasal cannula for COVID-19 patients 309 with severe hypoxemia: a pilot study

Early awake prone position combined with high-flow nasal oxygen 312 therapy in severe COVID-19: a case series

Early self-proning in awake, non-intubated patients in 315 the emergency department: A single ED's experience during the COVID-19 pandemic

Prolonged prone position ventilation for SARS

Chart reviews in emergency medicine 320 research: Where are the methods?

Awake prone positioning does not reduce the risk 322 of intubation in COVID-19 treated with high-flow nasal oxygen therapy: a multicenter, adjusted 323 cohort study. Crit Care

Conflict of interest statement: The authors have no commercial associations or sources of 326

• Identify awake, alert, non-intubated COVID positive or COVID patients under 349 investigation experiencing respiratory distress and low SpO2 (>90%)

• Patient must be capable of repositioning with or without assistance at least every 2 hours 351

• Collaborate with ED provider regarding the appropriateness of PP the patient and 352

Acute bleeding (eg, hemorrhagic shock

Multiple fractures or trauma (eg, unstable fractures of femur, pelvis, face

O2 Δ): 24 patients with no change in supplemental oxygen when proned