Clinical Pearl 77: Carbon Monoxide Poisoning: Is the RAD-57 Useful?

Currently, the standard for measuring Carbon Monoxide (CO) is CO-oximetry spectrophotometry via blood gas analysis. However the RAD-57 from Masimo, claims the ability to detect CO concentrations using a non-invasive instrument based on light spectrophotometry – a device similar to a pulse oximeter that measures CO. In fact, the device manufacturer claims the RAD-57 has the ability to measure functional O2-Hb as well as CO-Hb (SpCO). The question we ask is whether there is a role in the use of RAD-57 in the detection of CO in a prehospital setting or Emergency Department, and how reliable are these measurements? Can the RAD-57 readings be used clinical to truly risk stratify patients with CO poisoning?

Four prospective studies currently exist and the rest are case studies. These studies compared two variables; blood CO-Oximetry vs RAD-57 (SpCO) values, and were used to asses for % difference (bias), accuracy, and precision.

The first study, Barker et al, investigated the device in 10 healthy volunteers who were exposed to CO in a gas mixture until their CarboxyHb level reached 15%. The comparison revealed an uncertainty of +/- 2% from CO-Ox readings.

The prospective observational study by Suner et al used RAD-57 to measure CO levels for 10,856 patients from the ED. In the study 28 patients read positive with the RAD-57 of which 11 had no apparent clinical suspicion. Even though the study was able to show cases of occult CO poisoning, the data on the accuracy of those devices were not reported.

The third study Touger et al enrolled 120 patients with suspected CO toxicity based on history and was found to have an accuracy of 1.4% but had confidence intervals from -11.6 to 14.4 which demonstrates very poor accuracy of the device. Moreover the study revealed the RAD-57 only detected 11 out of the 23 patients that had COHb > 15%. Which means the preciseness of readings were also questionable due to the high false negative rate. This Annals of Emergency Paper concluded that Rad57 should not be used interchangeably with blood readings.

Roth et al. a prospective study actually found a positive outcome. This study measured the RAD-57 SpCO in 1278 ED pts in which 17 were positive, with a relatively low bias of 2.32% and precision of 4.01%. This study also revealed there as an increase in erroneous readings with increasing CO-Hb concentrations. The study might have suffered from selection bias, as not all patients that had SpCO readings had comparison CO-Hb results.

Nilson et al. demonstrated instances of elevated SpCO in the pre-hospital setting (n=1700). This initial elevated non-invasive reading did lead to faster blood CO-Hb measurement and time to hyperbaric therapy. It should be noted that that clinical outcomes were not measured. Case reports revealed when SpCO was used as a screening measure, all 5 cases that had a positive reading had a lower clinically acceptably level on CO-Ox (15% vs 10%). False positive results in such scenarios especially in the prehospital setting could lead to mismanagement of resources as well as add additional costs to the healthcare system.

Now how does SpCO stand up to concomitant hemoglobinopathies. Feiner et al. assessed the accuracy of SpCO reading with concomitant Methemglobenima, which revealed a linear increase in error with increasing MetHb levels, also questioning the reliability of SpCO readings.

From these limited studies we can infer that the use of the RAD-57 has significant limitations. Poor precision and sensitivity issues render that clinician unable to rule out CO poisoning in the field as well as the Emergency Department. Positive readings may guide triage however with the devices’ poor accuracy; the benefits might be overshadowed by the harm. Until improvements in device accuracy, we cannot recommend routine use of this device.

References:

Barker SJ, Curry J, Redford D. Measurement of carboxyhemo- globin and methemoglobin by pulseoximetry: a human vol- unteer study. Anesthesiology 2006;105:892–7

Suner S, Partridge R, Sucov A, Valente J, Chee K, Hughes A, Jay G. Non-invasive pulse CO-oximetry screening in the emergency department identifies occult carbon monoxide toxicity. J Emerg Med 2008;34:441–50

Touger M, Birnbaum A, Wang J, Chou K, Pearson D, Bijur P. Performance of the Rad-57 pulse COoximeter compared with standard laboratory carboxyhemoglobin measurement. Ann Emerg Med 2010;56:382– 8

Feiner JR, Bickler PE, Mannheimer PD. Accuracy of methemoglobin detection by pulse CO-oximetry during hypoxia. Anesth Analg 2010;111:143– 8

Roth D, Herkner H, Schreiber W, Hubmann N, Gamper G, Laggner AN, Havel C. Accuracy of noninvasive multiwave pulse oximetry compared with carboxyhemoglobin from blood gas analysis in unselected emergency department patients. Ann Emerg Med 2011;58:74 –9

Nilson D, Partridge R, Suner S, Jay G. Non-invasive carboxy- hemoglobin monitoring: screening emergency medical services patients for carbon monoxide exposure. Prehosp Disaster Med 2010;25:253– 6

O’Malley GF. Non-invasive carbon monoxide measurement is not accurate. Ann Emerg Med 2006;48:477– 8

UpToDate

Clinical Pearl 76: Ventilation Strategies in Cardiac Arrest

You have a patient in cardiac arrest who needs some sort of ventilation or oxygenation strategy.  Three choices exist:

1. 1Passive Oxygenation
2. Asynchronous Ventilation
3. Ventilation with interposed intermittent compression

Which do you choose?  First we should acknowledge a few things before we look at the science.

1.      For the noncardiac arrest patient, passive oxygenation is a great thing for patients who are breathing at least 4 times per minute and saturation is at least 93 percent.  If the patient is below these numbers you should bag until you get to these numbers and then go back to passive oxygenation (NRB with High-Flow Nasal Cannula or BVM with Peep Valve and High Flow Nasal Cannula Underneath WITHOUT squeezing the BVM or CPAP). For three minutes if the decision to intubate has been made

2.      CC-CPR (chest compression CPR is a great thing) for “Untrained Rescuers.”  When a layperson arrives at the scene the benefit of good compressions outweighs stopping compressions to ventilate.  This practice is proven in multiple studies.

3.      Bagging people can cause harm.  It decreases myocardial blood flow,  decreases cerebral blood flow, causes shunting by distending the alveoli and opens the Lower Esophageal Sphincter which results in gastric distention and aspiration from vomiting.  Additionally this makes it harder to intubate patients.

So now the science. .. First we should acknowledge that not breathing is a bad thing so if we are not going to ventilate cardiac arrest patients based on the science we better make absolutely sure the science is very very overwhelming for not ventilating.   This is because not ventilating/breathing results in increased PCO2, decreased oxygenation and leads to anoxic brain injury/brain death.  In lieu of great science (published papers) we need many expert societies to back us up that we are doing the correct thing. 

Luckily, this is one of the less controversial topics for us because we really only have one paper that the current recommendation is based on. ..that is correct one paper on what to do when EMTs or higher medical personnel oxygenate/ventilation in cardiac arrest.   The 2015 AHA guideline states “for witnessed cardiac arrest in a shockable rhythm, it may be reasonable for EMS systems with priority-based multi-tiered response to delay positive-pressure ventilation by using a strategy of up to three cycles of 200 continuous compressions with passive oxygenation insufflation and airway adjuncts”  was based on an older paper in 2009 by Dr. Bobrow which appeared in Annals of Emergency Medicine.  No further research research supports this position. 

Now guidelines are somewhat silly because by the time they come out more science exists or they are already outdateed because of a lack of other published articles.  Before we look at the sole paper this was based on, we should be clear what it states:

1. This practice may be reasonable which is hardly an endorsement
2. It is only for Witnessed cardiac arrests.  If you are doing this based on unwitnessed cardiac arrests…it is dangerous and not based on any science or recommendation furthermore and you are risking harm.
3.  It is only for initial shockable rhythms.  So if the initial rhythm is not VFIB or Pulseless VT, (or the AED fires)..there is no science or recommendation for any society and you are risking doing harm.

More importantly…lets look at this paper that this recommendation is based on……..a subgroup analysis of 200 patients.  In this subgroup analysis, patients who received up to 3 cycles of 200 compressions had better neurologically intact survival only in the witnessed group.  30 compressions is 18 seconds so 200 compressions is 120 seconds.  Therefore if you are doing passive oxygenation for more than 200 compressions (120 seconds), this is against the guideline as well.  So why was the recommendation only in Witnessed VF/Pulseless VT?  It is because all the other patients did WORSE when they got passive oxygenation.  Specifically:

1. Neurolocally Intact Survival in the Unwitnessed Cardiac Arrest group was  better in BVM group 13.8% vs 7.3%.  That is correct the chance of Neurologically intact survival was two times greater in the group who were bagged if they had an unwitnessed cardiac arrest.  If you are doing passive oxygenation in this group you are decreasing someones chance of neurologically intact survival.
2. Neurologically Intact Survival in nonshockable rhythms was better with BVM 3.7 vs. 1.3.  So you in this study three times more patients had neurologically intact survival if they had a BVM used vs. Passive oxygenation.  Again if you are doing passive oxygenation in this group you are decreasing someone change of survival!!!

So why did the one group do better when they had passive oxygenation…the answer may be very simple.  The study did not keep track of how fast EMS ventilated.  In studies you can show differences in two study groups by either the treatment group doing very good or the control group doing very bad.  It is quite possible since we know that excessive bagging is bad that patients in this study  were bagged too aggressively and that is why there was a difference between the intervention and control group.  The other major explanation is that patients who had witness cardiac arrests  required minimal need for ventilation since they were just breathing.  Remember they only did passive oxygenation for 2 minutes then they bagged them too!

Since there is really no evidence for this practice…are there any studies which demonstrate we should not be doing this?   The Annals of Emergency Medicine paper and current recommendation say we should not be doing this in unwittnessed arrests or nonshockable rhythms or greater 200 compressions.  Moreover all of the papers that show benefit and make recommendations for CC-CPR are for “untrained rescuers.” However all of the studies done with this were observational studies.  Finally a large EMS prehospital cardiac arrest trial was published in New England Journal of Medicine by Nichol in December 2015.  This major trial of 114 EMS agencies, 23,711 adult patients was done by the Resusciation Outcomes Consortium and actually supported complete ventilation!!!  It was statistically very superior to everything done and reinforced that some complete ventilations are necessary.  While its primary endpoint evaluation continuous compressions vs interrupted and found better outcomes with interrupted it showed that complete ventilation was better than asynchronous.  Although this study had a few problems…. it is the overwhelming the most well done study ever on the topic…  Finally in the last major review published by M. Chang April 2017 titled “The Past, Present and Future of Ventilation During CPR” in Current  Opinion in Critical Care Medicine discusses that ventilations of 8-10 are the best methods for cardiac arrest patients.   This author agrees!!!

Conclusions:

1. Based on one older and small paper, passive oxygenation for Witness Cardiac Arrests with Ventricular Fibrillation/Pulseless Ventricular Tachycardia (AED shockable rhythm) is not unreasonable however better literature suggests that even this practice is not best evidence.
2. Prolonged passive oxygenation by medical personal (greater than 200 compressions) has no evidence to support it and potentially can cause harm.
3. Passive oxygenation by medical personal for unwitnessed cardiac arrests or nonshockable rhythms is not supported in the literature and potentially can cause harm.

Clinical Pearl 73: Should We Pretreat with IV Calcium When giving CCBs for Stable Afib with RVR

Calcium channel blockers have long been an accepted treatment of hemodynamically stable Afib with RVR.  For those patients with borderline blood pressures or for those patients who are especially responsive to the antihypertensive effects of calcium channel blockers, there are few options available to prevent a hypotensive response. 

L Type calcium channels are present in the vascular smooth muscle, mycardium, conducting system of the heart, and in the pacemaker cells of the SA and AV nodes.  Of the calcium channel blockers available for clinical use, dihydropyridines require a much higher serum concentration to achieve electrophysiological activity than the concentration needed to achieve potent vasodilation, so their application in cases requiring negative chronotropy are limited especially with their propensity for inducing a concomitant sympathetic response via the baroreceptor reflex pathway. The non-dihydropyridines, (e.g. verapamil and diltiazem) exert their effects on the L-type channels in the pacemaker cells and conducting system of the SA and AV nodes at much lower concentrations, and are therefore useful for any supraventricular arrhythmia requiring reduced frequency of conduction through the AV node.  This is the basis of choosing these drugs for the treatment of hemodynamically stable Afib with RVR.

Of the two non-dihydropyridines, both can result in hypotension via their action on the vascular smooth muscle, but diltiazem to a much lesser degree than verapamil.  When treating a patient with RVR, there is still a recognized risk to causing hypotension, and aside from the strategies of slowing the infusion rates of calcium channel blockers, or by administering fluid boluses, one thought is to pretreat the patients with intravenous calcium.  The thought arises from the treatment of calcium channel blocker toxicity which involves among other things, giving calcium chloride or calcium gluconate to overwhelm the receptor blocking effect of the drugs.   The question we explore here is whether pretreating with intravenous calcium when giving non-dihydropyridine calcium channel blockers actually works to prevent or mitigate hypotension.

The literature available on this topic is mainly from the 1980s and 1990s and focused completely on verapamil, except for one study in 2004 which looked at pretreatment with IV calcium for diltiazem.    All studies had a small N of 50 or less participants, and ranged from case series, to retrospective, prospective, and finally a prospective, randomized, double-blind, placebo-controlled study for diltiazem.  Each study and their findings are summarized in the references below. 

The summation of data support pretreatment with calcium when using verapamil to prevent hypotension without mitigating the desired rate control effect.  In contrast, the one study on diltiazem did not show a significant difference between either treatment arm.  The paper does not support routine use of intravenous calicum as pretreatment to prevent hypotension.  Until further research is performed with diltiazem, perhaps the prudent course would be to continue using IV fluid boluses for borderline blood pressure while keeping IV calcium ready for treatment in the case of diltiazem induced hypotension.  However this guideline is not supported by good evidence.  See below for a review of the literature

References and Results

1. Weiss AT, et al. Int J Cardiol 1983; 4:275-84.

Prospective study design,     N = 13, Verapamil, Ca gluconate 1 gm;                     Result: SBP ↑ 5 mmHg

2. Roguin N, et al. Clin Cardiol 1984; 7:613-6.

Case series                             N = 2,  Verapamil, Ca gluconate(peds);                   Result: No hypotension

3. Haft JI, et al. Arch Intern Med 1986; 146:1085-9.

Sequential study of                N = 50, Verapamil, CaCl 1 gm,                                   Result: SBP ↑ 2 mm Hg

2 treatment protocols

4. Salerno DM, et al. Ann Intern Med 1987; 107:623-8.

Sequential study of                N = 5,  Verapamil, Ca gluconate 1 gm                      Result: SBP ↓ 12 mmHg

2 treatment protocols

5. Stringer KA, et al. Drug Intell ClinPharm 1988; 22:575-6.

Case report                            N = 1,  Verapamil,  CaCl 1 gm,                                  Result: No hypotension

6.  Barnett JC, et al. Chest 1990; 97:1106-9.

Prospective report of protocol        N = 19, Verapamil, Ca gluconate 1 gm or CaCl 1 gm,           Result: SBP ↑ 4 mm Hg

7. Kuhn M, et al. Am Heart J 1992; 124:231-2.

Retrospective chart review  N = 18, Verapamil, Ca gluconate 3 gm or CaCl 1 gm,           Result: No hypotension

8. Miyagawa K, et al. J Cardiovasc Pharmacol 1993; 22:273-9.

Sequential study of                N = 7,  Verapamil, Ca gluconate 3.75 mg/kg,          Result: SBP no change

2 treatment protocols

9. Kolkebeck T, et al. J Emerg Med 2004; 26(4):395-400.

Prospective, randomized,     N = 34, Diltiazem, CaCl 0.333 gm                  Result: SBP ↓ 8  vs ↓ 14mmHg

double-blind, placebo-controlled