If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Sodium bicarbonate (“bicarb”) administration in out-of-hospital cardiac arrest (OHCA) is intended to counteract acidosis, although there is limited clinical evidence to support its routine administration. We sought to analyze the association of bicarb with resuscitation outcomes in non-traumatic OHCA.
Methods
Records were obtained from the 2019–2020 ESO Data Collaborative prehospital electronic health record database, spanning 1,322 agencies in 50 states. OHCAs with resuscitations lasting 5–40 minutes were stratified by presenting ECG rhythm (VF/VT, pulseless electrical activity (PEA), asystole) for analysis. The outcomes of any prehospital ROSC and survival to discharge were compared by bicarb status using propensity score matching and logistic regressions with/without adjustment.
Results
We analyzed 23,567 records, 6,663 (28.3 %) of which included bicarb administration. Most patients presented in asystole (67.4 %), followed by PEA (16.6 %), and VF/VT (15.1 %). In the propensity-matched cohort, ROSC was higher in the bicarb group for the asystole group (bicarb 10.6 % vs control 8.8 %; p = 0.013), without differences in the PEA or VF/VT groups. Survival was higher in the bicarb group for asystole (bicarb 3.3 % vs control 2.4 %; p = 0.020) and for PEA (bicarb 8.1 % vs control 5.4 %; p = 0.034), without differences in the VF/VT group. These results were consistent across adjusted/unadjusted logistic regression analyses: bicarb was associated with ROSC and survival in asystole [uOR (95 % CI): ROSC 1.23 (1.04–1.44), survival 1.40 (1.05–1.87)] and with survival in PEA (1.54 (1.03–2.31).
Conclusions
Bicarb was associated with survival in non-shockable rhythms and ROSC in asystole. Findings from this observational study should be corroborated with prospective randomized work.
Though out-of-hospital cardiac arrest (OHCA) recommendations have included the use of sodium bicarbonate (NaHCO3, hereafter “bicarb”) for nearly-four decades, the effectiveness of this medication has limited evidence. Bicarb is thought to combat acidosis arising during cardiac arrest by buffering acid-base imbalances.
However, several studies from recent years have failed to support its use in OHCA, citing either no difference in outcomes or an association with poorer outcomes.
Sodium bicarbonate on severe metabolic acidosis during prolonged cardiopulmonary resuscitation: A double-blind, randomized, placebo-controlled pilot study.
Sodium bicarbonate use and the risk of hypernatremia in thoracic aortic surgical patients with metabolic acidosis following deep hypothermic circulatory arrest.
Part 12: Cardiac arrest in special situations: 2010 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care.
Proposed explanations for adverse effects are hypernatremia, inhibited oxygen release from hemoglobin, reduction of systemic vascular resistance, and inactivation of simultaneously-administered catecholamines. Nevertheless, in many of these observational studies, bicarb was often given late in the resuscitation after earlier resuscitative attempts are not successful. Thus, the association of bicarb with poor outcomes is confounded by long resuscitation attempts which have inherently poor prognoses. More recent updates to the American Heart Association guidelines have removed support for bicarb administration and recommend use only in cases of cyclic antidepressant overdose, ventricular arrhythmias, and hyperkalemic cardiac arrests.
Part 12: Cardiac arrest in special situations: 2010 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care.
Prehospital sodium bicarbonate use could worsen long term survival with favorable neurological recovery among patients with out-of-hospital cardiac arrest.
There is a need to investigate the impact of bicarb on cardiac arrest outcomes when controlling for known predictors of OHCA outcomes, including the length of the resuscitation attempt. In this work, we sought to investigate the use of bicarb in OHCA and associated outcomes through analysis of a nationwide database of EMS-treated patients with cardiac arrest.
Methods
Study design and setting
We conducted a nationwide retrospective cross-sectional study of EMS-treated OHCA using fully anonymized prehospital electronic health records maintained by the ESO Data Collaborative (Austin, TX). As of 2019, the ESO database contains de-identified EMS records from 1,322 participating agencies across all fifty states and provides comprehensive clinical information including demographics, treatments, and pertinent timestamps for each medical event. Data collected within the ESO database are National EMS Information System compliant.
A subset of agencies participates in a bi-directional data exchange allowing direct linkage of prehospital data with hospital outcome data including diagnoses and dispositions. The Institutional Review Board at the University of Pittsburgh approved this investigation with a letter of determination that this was not human subject research (STUDY21030059).
Study population
We limited the analysis to encounters with response by Advanced Life Support (ALS) capable units for non-traumatic cardiac arrests with linked hospital outcome data occurring between January 2019 and December 2020. We included cases with cardiac etiologies where resuscitation was attempted. Excluded were patients with traumatic arrest etiologies, responses from Basic Life Support (BLS) units only, encounters initiating within a healthcare setting, patients with do-not-resuscitate orders or treatment refusal documented, patients with known pregnancies, and children under 18 years old. We also excluded those with missing survival status, initial rhythm, demographics, or resuscitation times.
Measures
Prehospital bicarb use status was defined by the prehospital administration of bicarb in any dose and by any route by EMS. Outcomes of interest were survival to hospital discharge and return of spontaneous circulation (ROSC) during any point in the prehospital encounter. Patients whose resuscitations were terminated in the field were included in the analysis.
Statistical analysis
Statistical analyses were performed using STATA/SE version 17.0 (StataCorp, College Station, TX, USA). All analyses were stratified by initial presenting rhythm [ventricular fibrillation/pulseless ventricular tachycardia (VF/VT), asystole, or pulseless electrical activity (PEA)]. Descriptive statistics comparing bicarb use status were performed using Chi-square tests, student’s t-tests, and Mann-Whitney U tests. Descriptive statistics were reported either as frequency (percentage) or median (interquartile range, 25th to 75th percentile). We restricted analyses to cases with resuscitation lengths (start to end of EMS CPR) of forty minutes or less.
Association between Duration of Resuscitation and Favorable Outcome after Out-of-Hospital Cardiac Arrest: Implications for Prolonging or Terminating Resuscitation.
In addition, we limited analyses to cases with length of resuscitation over five minutes because IV access (route most commonly used to administer bicarb), in under 5 minutes was uncommon, is technically difficult to execute and rarely occurs. Thus, any arrests that resolved within five minutes likely were not eligible to receive bicarb and therefore were not included in analyses.
We then examined the association of bicarb use with survival to hospital discharge and any ROSC during the EMS encounter using propensity score matching and logistic regression analyses with and without adjustment. We created pairs of bicarb and control patients (1:1 matching) without replacement using the user-generated pstest and psmatch2 programs,
Evaluation of pre-hospital administration of adrenaline (epinephrine) by emergency medical services for patients with out of hospital cardiac arrest in Japan: Controlled propensity matched retrospective cohort study.
which balance covariates across the groups and within strata of the propensity score to create a matched sample. Potential confounding variables used to generate the propensity score were selected a priori. We included patient age, gender, race, witnessed status, bystander CPR, prearrival instructions, any defibrillation attempt, use of CPR feedback devices, impedance threshold device (ITD) use, any attempted ventilation, length of resuscitation, and number of adrenaline doses. The overall covariate imbalance and quality of match were assessed with the Stata pstest command balance diagnostics.
The matched pairs were then isolated for subsequent logistic regressions with and without adjustment for the variables included in the propensity score.
All statistical analyses utilized an alpha level of 0.05 as a threshold for determining statistical significance.
Results
Of the 148,110 cardiac arrests in the database between January 2019 and December 2020, 23,567 (15.9 %) met the inclusion criteria (Fig. 1). Median age was 68 years (IQR 57–78 years) and 38.2 % were female. Most patients had a presenting rhythm of asystole (67.4 %), followed by PEA (16.6 %), and VF/VT (15.1 %). Overall, bicarb was administered in 28.3 % of cases. Bicarb was administered most often to patients with a presenting rhythm of asystole (29.5 %), followed by PEA (27.3 %), and then VF/VT (24.2 %). The median time to bicarb administration (CPR start to first bicarb) was 17.7 minutes (IQR 12.9–22.9).
Fig. 1Sample population statistics. Flow chart illustrating selection criteria from ESO database and basic descriptive statistics of the resulting sample population. OHCA = out-of-hospital cardiac arrest; EMS = emergency medical services; ALS = advanced life support; DNR = do-not-resuscitate, VF/VT = ventricular fibrillation or ventricular tachycardia; PEA = pulseless electrical activity.
EMS ROSC and survival to hospital discharge rates were 18.4 % and 7.6 % overall. ROSC and survival rates were highest among those with a presenting rhythm of VF/VT (40.5 % and 23.3 %), followed by PEA (28.8 % and 9.9 %) and then asystole (10.6 % and 3.3 %). Among those who received bicarb, 15.2 % experienced ROSC during the EMS encounter and 4.9 % survived to hospital discharge (Table 1).
Table 1Survival to hospital discharge and EMS ROSC rates for all patients.
Survival Rate % (n) N = 23,567
EMS ROSC % (n) N = 23,567
All cases
7.6 (1,783)
18.4 (4,337)
Received bicarb
4.9 (324)
15.2 (1,014)
VF/VT
23.3 (828)
40.5 (1,439)
Asystole
3.3 (525)
10.6 (1,687)
PEA
9.9 (387)
28.8 (1,126)
EMS-witnessed
12.4 (233)
31.5 (592)
Non-EMS witnessed
7.2 (1,550)
17.3 (3,745)
Unwitnessed
3.7 (573)
12.2 (1,865)
All variables are categorical and presented as a percentage.
EMS = emergency medical services; ROSC = return of spontaneous circulation; VF/VT = ventricular fibrillation or ventricular tachycardia; PEA = pulseless electrical activity.
In all presenting rhythm groups, patients who received bicarb were more often patients that had unwitnessed arrests, non-EMS-witnessed arrests, long resuscitations, and many adrenaline doses (Table 2). In asystole and PEA, patients receiving bicarb were less frequently Black, more frequently Hispanic, and had more use of CPR feedback devices, in addition to other differences seen in the distinct rhythm groups.
Table 2Demographics, characteristics, and treatments before and after PSM stratified by rhythm group.
2A. VF/VT
Before PSM,N = 3,553
After PSM,N = 1,328 (664 pairs)
VF/VT
Bicarb
Control
P value
Bicarb
Control
P value
% Bias*
Age (y)
66(57–76)
66(56–75)
0.522
66(57–76)
66(57–76)
0.640
−3.0
Female
26.5 %
28.5 %
0.254
26.3 %
27.7 %
0.601
−3.0
White
70.7 %
71.3 %
0.748
71.7 %
73.9 %
0.397
−4.9
Black
17.2 %
19.1 %
0.233
17.4 %
16.9 %
0.817
1.3
Hispanic
4.2 %
3.2 %
0.177
4.0 %
3.9 %
0.882
0.9
Other race
1.7 %
1.7 %
0.935
1.7 %
1.5 %
0.817
1.4
Unwitnessed
48.5 %
37.8 %
0.000*
46.0 %
46.4 %
0.907
−0.7
Prearrival instructions
64.1 %
66.0 %
0.359
62.7 %
66.1 %
0.225
−7.1
Bystander CPR
24.2 %
25.9 %
0.310
25.5 %
26.8 %
0.597
−3.0
EMS-witnessed
7.0 %
11.3 %
0.000*
7.3 %
7.3 %
1.000
0.0
Defibrillated
39.0 %
41.9 %
0.130
39.8 %
40.6 %
0.767
−1.7
CPR feedback
43.6 %
41.0 %
0.182
43.5 %
42.2 %
0.639
2.7
ITD
14.3 %
14.5 %
0.915
14.2 %
15.7 %
0.464
−4.3
Attempted ventilation
88.9 %
87.9 %
0.403
88.9 %
89.2 %
0.853
−1.0
Length of resus. (min.)
31(25–36)
25(16–32)
0.000*
30(25–35)
31(25–36)
0.581
−3.9
Total doses adrenaline
4(3–5)
3(2–4)
0.000*
4(3–5)
4(3–5)
0.937
−2.5
VF
82.6 %
82.4 %
0.901
83.1 %
86.5 %
0.106
−8.9
VT
9.2 %
8.0 %
0.270
8.6 %
7.1 %
0.331
5.5
2B. Asystole
Before PSM,N=15,886
After PSM,N=7,472 (3,736 pairs)
Asystole
Bicarb
Control
P value
Bicarb
Control
P value
% Bias*
Age (y)
68(57–79)
67(57–77)
0.002*
67(57–77)
67(56–78)
0.737
0
Female
39.2%
40.8%
0.060
39.4
40.0%
0.639
−1.1
White
64.3%
64.0%
0.699
65.8%
65.7%
0.899
0.3
Black
21.8%
24.5%
0.000*
21.6%
21.8%
0.837
−0.5
Hispanic
6.5%
4.4%
0.000*
6.0%
5.8%
0.608
1.3
Other race
2.0%
1.9%
0.406
2.0%
2.2%
0.736
−0.8
Unwitnessed
76.0%
74.3%
0.019*
74.5%
74.2%
0.803
0.6
Prearrival instructions
67.6%
66.0%
0.096
67.6%
67.2%
0.757
0.7
Bystander CPR
17.0%
17.3%
0.708
20.7%
20.6%
0.858
0.4
EMS-witnessed
3.5%
4.8%
0.000*
3.8%
3.9%
0.753
−0.7
Defibrillated
3.4%
3.1%
0.253
3.3%
3.7%
0.394
−2.2
CPR feedback
41.0%
38.6%
0.005*
40.4%
39.7%
0.554
1.4
ITD
17.2%
15.5%
0.007*
16.8%
16.9%
0.923
−0.2
Attempted ventilation
85.6%
86.1%
0.434
84.9%
84.2%
0.403
2.1
Length of resus. (min.)
30(24–34)
28(22–33)
0.000*
30(24–34)
30(24–34)
0.351
−1.9
Total doses adrenaline
4(3–5)
3(2–4)
0.000*
4(3–5)
4(3–5)
0.4652
0.5
2C. PEA
Before PSM,N=3,910
After PSM,N=1,702 (851 pairs)
PEA
Bicarb
Control
P value
Bicarb
Control
P value
% Bias†
Age (y)
72(59–81)
71(60–81)
0.833
72(60–81)
71(60–80)
0.737
0.5
Female
39.2%
38.9%
0.899
39.5%
38.0%
0.567
2.9
White
66.2%
64.7%
0.382
66.5%
65.9%
0.830
1.1
Black
23.9%
26.5%
0.001*
23.0%
23.5%
0.810
−1.2
Hispanic
6.0%
3.2%
0.000*
3.9%
4.1%
0.796
−1.3
Other race
1.2%
1.4%
0.596
1.5%
1.7%
0.840
−1.1
Unwitnessed
53.6%
45.9%
0.000*
51.5%
50.9%
0.800
1.3
Prearrival instructions
53.7%
46.9%
0.001*
52.7%
50.3%
0.335
4.9
Bystander CPR
15.7%
14.4%
0.297
17.6%
15.0%
0.170
6.8
EMS-witnessed
14.6%
22.1%
0.000*
17.1%
19.3%
0.264
−5.4
Defibrillated
3.6%
5.2%
0.032*
3.5%
4.0%
0.593
−2.5
CPR feedback
42.3%
38.0%
0.014*
41.9%
41.8%
0.959
0.3
ITD
14.9%
15.3%
0.778
13.4%
14.7%
0.465
−3.6
Attempted ventilation
87.7%
87.4%
0.801
87.5%
88.4%
0.586
−2.7
Length of resus. (min.)
30(23–35)
26(19–33)
0.000*
29(23–34)
30(23–35)
0.339
−3.6
Total doses adrenaline
4(3–5)
3(2–4)
0.000*
4(3–5)
4(3–5)
0.918
−2.1
Categorical variables were compared between groups using chi-square tests, and frequencies are presented as percentages. Continuous variables were compared using Mann-Whitney U tests, and values are presented as median(interquartile range, 25th to 75th percentile). PSM model adjusts for age, sex, race, witnessed status, prearrival instructions, bystander CPR, any defibrillation attempt, CPR feedback, any attempted ventilation, length of resuscitation, and total doses of adrenaline.
†% Bias: percent standardized bias is the percent difference of sample means in bicarb and control (unmatched-matched) sub-samples as a percentage of the square root of the average of the sample variances in the bicarb and control groups. Displayed are the p values of the likelihood-ratio test.
The propensity score matching balanced the distribution of all variables included in the propensity score calculation (imbalance < 10 % and p > 0.05 across all rhythm groups). In the matched sample, those who received bicarb had increased rates of survival compared to those not receiving bicarb (bicarb 5.3 % vs control 4.3 %; p = 0.019) overall (Table 3). There were no differences in ROSC overall. ROSC was not different in cases with VF/VT as the presenting rhythm. ROSC was higher in the bicarb group when asystole was the presenting rhythm (bicarb 10.6 % vs control 8.8 %; p = 0.013), but not when the presenting rhythm was PEA. There was no difference in survival in cases presenting with VF/VT. The bicarb group showed higher rates of survival than the control group when asystole was the presenting rhythm (bicarb 3.3 % vs control 2.4 %; p = 0.020) and when PEA was the presenting rhythm (bicarb 8.1 % vs control 5.4 %; p = 0.034).
Table 3Outcomes for propensity-matched cohort of cardiac arrest patients by presenting rhythm and bicarb status.
Initial Rhythm
Frequencies
Odds ratio (95 % CI)
Bicarb
Control
P value
Unadjusted†
Adjusted ‡
VF/VT,N = 1,186
Survival
12.8 %
13.5 %
0.731
0.94(0.67–1.32)
0.94(0.64–1.37)
EMS ROSC
28.8 %
30.5 %
0.525
0.92(0.72–1.18)
0.88(0.67–1.15)
Asystole,N = 6,848
Survival
3.33 %
2.4 %
0.020*
1.40(1.05–1.87)
1.46(1.08–1.96)
EMS ROSC
10.6 %
8.8 %
0.013*
1.23(1.04–1.44)
1.24(1.05–1.47)
PEA,N = 1,556
Survival
8.10 %
5.4 %
0.034*
1.54(1.03–2.31)
1.60(1.04–2.47)
EMS ROSC
25.2 %
24.3 %
0.681
1.05(0.83–1.32)
1.01(0.78–1.30)
All cases,N = 9,590
Survival
5.28 %
4.3 %
0.019*
1.25(1.04–1.51)
1.30(1.06–1.59)
EMS ROSC
15.2 %
14.0 %
0.094
1.10(0.98–1.23)
1.11(0.98–1.25)
Categorical variables were compared between groups using chi-square tests, and frequencies are presented as percentages. Odds ratios from logistic regressions are presented with‡ and without† adjustment for variables included in propensity score matching (age, sex, race, witnessed status, prearrival instructions, bystander CPR, any defibrillation attempt, CPR feedback, any attempted ventilation, length of resuscitation, and total doses of adrenaline) as odds ratio(95% confidence interval).
VF/VT = ventricular fibrillation or ventricular tachycardia; EMS = emergency medical systems; ROSC = return of spontaneous circulation; PEA = pulseless electrical activity.
*Value is statistically significant (p < 0.05).
†Univariate regression with propensity score as the predictor.
‡ Multivariate regression adjusting for all variables included in propensity score matching.
These results were consistent with the logistic regression analyses. Overall, bicarb use was associated with improved survival [unadjusted/adjusted odds ratio (95 % CI): OR 1.25 (1.04–1.51), aOR 1.30 (1.06–1.59)], but there was no difference in ROSC. In the asystole group, bicarb administration was significantly associated with survival [OR 1.40 (1.05–1.87), aOR 1.46 (1.08–1.96)] and ROSC during the EMS encounter [OR 1.23 (1.04–1.44), aOR 1.24 (1.05–1.47)]. In the PEA group, bicarb was associated with increased survival [OR 1.54 (1.03–2.31), aOR 1.60 (1.04–2.47)], but had no association with ROSC. There were no significant differences for either outcome between bicarb and control groups in VF/VT.
Discussion
Using a large national sample of more than 23,000 EMS records with linked hospital outcome data for patients with non-traumatic cardiac arrest, we observed significant improvements in ROSC and survival for patients receiving prehospital bicarb when presenting with non-shockable rhythms, after adjustment for confounding variables. There were no associations between bicarb use status and outcomes for patients presenting with shockable rhythms (VF/VT). Collectively, these findings suggest that prehospital bicarb delivery may improve outcomes for patients presenting with non-shockable rhythms in OHCA.
Despite limited recommended use per guidelines, greater than 1-in-4 resuscitations in this study included bicarb administration, consistent with other work.
Prehospital sodium bicarbonate use could worsen long term survival with favorable neurological recovery among patients with out-of-hospital cardiac arrest.
though are in agreement with other works. A 2005 study by Bar-Joseph et al. found that agencies giving bicarb early and often had improved chance of ROSC and better long-term outcomes than those giving it late and infrequently.
One randomized, double-blinded pre-hospital trial by Vukmir and Katz found that survival in the bicarb group was double that of the control group in patients with resuscitations lasting over 15 minutes. Another study from a single emergency department in Korea conducted a matched case control and found an increase in ROSC with bicarb use.
Sodium bicarbonate on severe metabolic acidosis during prolonged cardiopulmonary resuscitation: A double-blind, randomized, placebo-controlled pilot study.
Our results suggest that stratifying OHCA cases by rhythm could clarify the aforementioned contrasting literature; bicarb administration appears beneficial in asystole and PEA, but no difference was seen in shockable rhythm outcomes. The findings of this work have direct implications for OHCA practices since there is no treatment indicated for patients in non-shockable rhythms other than continued CPR and repeat doses of adrenaline.
The improved outcomes in the non-shockable groups may be due to differences in underlying arrest characteristics, which can influence how the patient presents to EMS. A 2015 study by Hara et al. showed that the proportion of patients presenting with asystole increases as time from collapse to CPR increases.
Since acid-base status largely depends on the duration of the arrest, it is likely that those in asystole with long down time and/or long resuscitation attempts have a large base deficit. Thus, bicarb is a reasonable therapy to limit the complications of acidosis and potentially improve outcomes in patients presenting in non-shockable rhythms.
We must also consider how the timing of bicarb administration may influence these results. Bicarb use late in arrests likely explains the many observational studies associating bicarb use with poor outcomes. However, we were able to match bicarb and control patients on many arrest characteristics, including length of resuscitation, in this work. After propensity matching, we found that patients presenting in asystole who received bicarb had higher survival and ROSC rates than those not receiving bicarb. Additionally, patients who presented in PEA and received bicarb had higher survival than their control counterparts. There were no differences in survival or ROSC between bicarb and control groups for patients presenting in shockable rhythms.
Limitations
This work was limited by its retrospective nature. There are likely unmeasured variables underlying our findings which we cannot account for with this dataset. Agencies that deliver bicarb may have other characteristics that yield favorable outcomes. Particularly, given the limited support for bicarb in current AHA guidelines, it would be of interest to know why bicarb was used at the patient level. However, we are limited by the retrospective nature of these data, and that level of detail is not available here. Most states have statewide EMS protocols which are highly variable. Additionally, local city, county, and regional EMS Medical Directors have the license to modify the statewide protocols (by addition) as they see fit. As a result, some agencies may have chosen to retain bicarb use during OHCA, while others have not. There is also variability of the actual care given by EMS personnel at the scene. Individual providers may choose to give bicarb routinely, while others may use it selectively or not at all. Therefore, it is not possible to understand why bicarb was or was not given at the case level in this dataset.
Another limitation is that we only considered initial presenting rhythm because few cases had information about rhythm at hospital arrival. If patients in VF/VT later transitioned into asystole or PEA, we would expect to see an association between bicarb and survival in the VF/VT group as well. Likewise, we did not capture instances in which patients converted from non-shockable to shockable rhythms. Therefore, our findings must be interpreted cautiously regarding rhythm group. Additionally, we did not capture evidence of rearrests in this dataset. We also only considered whether a patient received any bicarb and did not consider total doses, so the impact of bicarb may be underestimated.
Despite controlling for the length of resuscitation, we were not able to control for down time, which could relate to acid-base imbalances. It is possible that those presenting with VF/VT had shorter down time before CPR began, and therefore were less acidotic than those presenting with asystole and PEA. It is also possible that outcomes in patients presenting with VF/VT rely on early defibrillation and are less influenced by acid-base status. Additionally, we only recorded whether ROSC occurred any time during the EMS phase of care. We did not measure the duration of ROSC prior to ED arrival. It is probable that many of these cases would have rearrested, which would have affected their ultimate survival outcome.
With regards to patient outcomes, while this study contained linked hospital discharge dispositions, we were not able to assess neurologic status at discharge. We were also unable to assess interventions and other outcomes such as myocardial dysfunction, which could be impacted by bicarb use.
Lastly, as with any study, the scope over our study population may limit generalization of our findings, although with data derived from more than 1,300 EMS agencies in all 50 states, we expect some degree of generalizability. That said, our analyses did not include granular geographic analysis, so it is impossible to know whether the associations we observed were concentrated or distributed geographically. The overall survival to hospital population was 7.6 %, which is on the lower end of expected rate.
The global survival rate among adult out-of-hospital cardiac arrest patients who received cardiopulmonary resuscitation: A systematic review and meta-analysis.
Randomized, prospective pre-clinical studies should be conducted in the future to confirm these findings.
Conclusions
In this large propensity score-matched analysis, prehospital bicarb was associated with increased survival to hospital discharge in patients experiencing OHCA. Specifically, patients presenting with asystole and PEA in out-of-hospital cardiac arrest had more favorable survival outcomes when administered prehospital bicarb than those not receiving bicarb, after accounting for important arrest characteristics. Additional prospective work is needed to confirm these findings and establish causality.
CRediT authorship contribution statement
Sara M. Niederberger: Investigation, Methodology, Software, Formal analysis, Resources, Data curation, Writing – original draft, Visualization, Funding acquisition. Remle P. Crowe: Software, Validation, Formal analysis, Data curation, Writing – review & editing. David D. Salcido: Software, Validation, Formal analysis, Writing – review & editing. James J. Menegazzi: Conceptualization, Methodology, Validation, Formal analysis, Resources, Supervision, Writing – review & editing, Visualization, Project administration.
Acknowledgements
The authors thank Chase Zikmund for the technical support.
Sources of funding
Research reported in this publication was supported in part by University of Pittsburgh School of Medicine T35 Grant Award (5T35DK065521-17) from the National Institutes of Health.
Disclosures
Dr. Crowe is employed by ESO. The authors have nothing else to disclose in relation to this work.
Sodium bicarbonate on severe metabolic acidosis during prolonged cardiopulmonary resuscitation: A double-blind, randomized, placebo-controlled pilot study.
Sodium bicarbonate use and the risk of hypernatremia in thoracic aortic surgical patients with metabolic acidosis following deep hypothermic circulatory arrest.
Part 12: Cardiac arrest in special situations: 2010 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care.
Prehospital sodium bicarbonate use could worsen long term survival with favorable neurological recovery among patients with out-of-hospital cardiac arrest.
Association between Duration of Resuscitation and Favorable Outcome after Out-of-Hospital Cardiac Arrest: Implications for Prolonging or Terminating Resuscitation.
Evaluation of pre-hospital administration of adrenaline (epinephrine) by emergency medical services for patients with out of hospital cardiac arrest in Japan: Controlled propensity matched retrospective cohort study.
The global survival rate among adult out-of-hospital cardiac arrest patients who received cardiopulmonary resuscitation: A systematic review and meta-analysis.
00713-4&id=gr1.jpg){kind=link}