Original Contribution

Relationship between laboratory findings and the outcome of cardiopulmonary arrest

Youichi Yanagawa PhD, MD^a,?, Toshihisa Sakamoto PhD, MD^a, Hiroki Sato MHSc^b

^aDepartment of Traumatology and Critical Care Medicine, National Defense Medical College (NDMC),

Saitama 359-8513, Japan

^bMedical Information Division, NDMC, Saitama 359-8513, Japan

Received 16 January 2008; accepted 1 March 2008

Abstract

Purpose: The aim of this study was to investigate whether laboratory findings on arrival may be useful in predicting the outcome of out-of-hospital cardiopulmonary arrest (CPA).

Methods: Between January 2005 and September 2007, a medical chart review was retrospectively performed for CPA. The individual medical records were reviewed for the following data: background of CPA, arterial blood gas, cell blood count, serum biochemical, and Cerebral Performance Category 1 month after the CPA. The subjects were divided into patients with a CPC ranging from 3 to 5 (CPC 3-5 group) and from 1 to 2 (CPC 1-2 group).

Findings: The total protein, platelets, pH, and PO2 in the CPC 1-2 group tended be higher than those in the CPC 3-5 group. The PCO2, potassium, phosphorus, and ammonia in the CPC 1-2 group tended be lower than those in the CPC 3-5 group. However, there were no factors independently associated with the outcome by multivariate analysis.

Conclusion: Some of the biochemical-hematologic parameters demonstrate significant changes concerning the outcome. However, initial blood work cannot be used to make Clinical decisions because there are no factors independently associated with the outcome.

(C) 2009

Introduction

Based on the results of biochemical analyses, the serum S100 protein level, which is a calcium-binding protein astroglial protein; the serum neuron-specific enolase level, which is located in the neurons and neuroectodermal cells; creatine kinase BB in the cerebrospinal fluid, which is located throughout the entire brain; and the neurofilament, which is neuronal protein in the cerebrospinal fluid, have been reported to be useful markers for predicting the neurological outcome of cardiopulmonary arrest (CPA) [1-8]. These markers directly reflect the damage to the central nervous

* Corresponding author. Tel.: +81 4 2995 1888; fax: +81 4 2996 5221.

E-mail address: [email protected] (Y. Yanagawa).

system. However, these markers are not commonly used in general hospitals, and thus, they tend to not be immediately examined. Cardiopulmonary arrest and the consequent interruption of blood flow to metabolically active tissues cause intense hypercapnic and metabolic acidosis at the tissue level, thus resulting in the accumulation of end products such as CO₂, lactate, and hydrogen ions. As a result, base excess , which is the amount of base in millimoles required to titrate 1 L of whole arterial blood to a pH of 7.40 and can be easily determined in blood samples even in general hospitals, thus decreases in the course of cardiac arrest because of the promotion of acidosis. Takasu et al [9] reported that the duration of cardiac arrest correlated with the BE, and the BE indirectly predicted the outcome of CPA because the severity of anoxia correlated with the duration of

0735-6757/$ – see front matter (C) 2009 doi:10.1016/j.ajem.2008.03.001

the CPA. The CPA also induces Ischemic injury on whole organs. The ischemic organ injury may affect the laboratory findings dependent on the severity of the ischemic insult. However, the laboratory findings on arrival of CPA have not yet been sufficiently evaluated. Therefore, this study retro- spectively investigated the relationship between bioChemical and hematologic parameters on admission and outcome of out-of-hospital CPA.

Material and methods

The National Defense Medical College is an 800-bed hospital in the suburban area of Tokyo, with a neighboring population of approximately 800 000. There are approxi- mately 3000 CPAs per year in this area, and the emergency department (ED) treats more than 150 cases of sudden-onset, unexpected out-of-hospital CPA every year.

Between January 2005 and September 2007, medical charts were retrospectively reviewed for all out-of-hospital CPA patients who were treated in the ED. Patients meeting the following criteria were excluded: (1) died of trauma; (2) nonwitnessed CPA; (3) chronic renal failure or creatinine level greater than 2.0 mg/dL on arrival; (4) liver cirrhosis or choline esterase level less than 177 IU/L; (5) arterial blood sample was not obtained on arrival; (6) induced hypothermic therapy was applied into anoxic encephalopathy; and (7) prehospital infusion was performed by emergency medical services teams.

The ED covers an urban and suburban area in which emergency medical services teams are based at fire stations and provide prehospital basic life support, including automatic external defibrillation, Bag-mask ventilation with 10 L/min oxygen, and airway control using laryngeal mask airways, esophageal-tracheal combitubes, or tracheal intuba- tion. The mean call-to-response time is approximately 6 minutes. After arrival at the ED, CPA patients were provided advanced cardiac life support. Arterial blood sampling was routinely performed to diagnose the cause and grasp the severity of the CPA of the patients. If the CPA patients obtained spontaneous return of circulation (ROSC) and

Table 1 Background of subjects

showed convulsions or required aggressive ventilation support, then intravenous barbiturates (2-6 mg kg^-1 h^-1) or midazolam (0.2-0.5 mg kg^-1 h^-1) was continuously administered until the convulsions or Respiratory function improved. If the Blood sugar exceeded 250 mg/dL, titrated regular insulin was administered. If the body temperature was more than 38.0?C, then Surface cooling and antipyretics were applied.

Individual medical records were reviewed regarding the following data: demographic characteristics; cause of CPA; electrocardiogram rhythm at the scene (categorized ventri- cular fibrillation [VF] or non-VF); whether bystander cardiopulmonary resuscitation (CPR) was performed; esti- mated duration of CPA (in case of prehospital ROSC: the time from first call to ROSC; in cases without prehospital ROSC: the time from first call to arrival at our ED); arterial blood gas data, including pH, PCO2, PO2, bicarbonate, and BE; cell blood count; serum biochemical data, including total protein, albumin, glucose, total bilirubin, aspartate aminotransferase, alanine aminotransferase, gamma-gluta- myl transpeptidase, alkaline phosphatase, choline esterase, amylase, blood urea nitrogen, creatinine, creatine phospho- kinase, sodium, potassium, chloride, calcium, phosphorus, and ammonia; and cerebral performance category after 1 month of CPA (outcome). CPC 1 is normal or mild cerebral disability, CPC 2 is moderate cerebral disability, CPC 3 is severe cerebral disability, CPC 4 is coma/vegetative state, and CPC 5 is death.

The subjects were divided into 2 groups. The first group

included the patients with CPC from 3 to 5 (CPC 3-5 group). The second group included those patients with CPC from 1 to 2 (CPC 1-2 group). The variables listed above were investigated between the 2 groups. The relationship between the estimated duration of CPA and the significant factors was evaluated using a univariate analysis.

Both the ?2 test and Student t test were used for the statistical analyses. A P value of less than .05 was considered to be significant.

This study was not invasive or time consuming and did not include randomization; moreover, on account of its retrospective nature, approval by the institutional review board was not needed.

CPC 3-5 (n = 119)		CPC 1-2 (n = 16)	P value
Age, mean +- SE, y	63.1 +- 1.4	60.3 +- 2.5	n.s.
Sex, n (male/female)	69/50	10/6	n.s.
Bystander CPR performed, n (%)	44 (37)	7 (44)	n.s.
VF at scene, n (%)	22 (18)	10 (63)	.0005
Cause of arrest, n (%)
Cardiac disease	94 (79)	15 (94)
Respiratory disease	13 (11)	1 (6)
Cerebral vascular disease	12 (10)	0
Estimated duration of CPA, mean +- SE	31.8 +- 0.8	14.3 +- 2.4	b.0001

Results

Table 3 Results of blood cell count on arrival

White blood cells (/uL)

Hemoglobin (g/dL) Platelets (x10⁴/uL)

CPC 3-5

(n = 110)

10 790 +- 671

CPC 1-2

(n = 16)

12 246 +- 1399

P value

n.s.

12.1 +- 0.2

16.6 +- 0.6

13.3 +- 0.5 n.s.

22.6 +- 1.7 .001

Values are presented as mean +- SE.

There were 475 patients with an out-of-hospital CPA directly transported to this department. Of these, 112 patients with traumatic CPA, 149 patients without witnessed collapse, 27 patients who underwent prehospital infusion, 42 patients with renal or hepatic failure, 3 patients without blood sampling on arrival, and 5 patients who underwent induced hypothermic therapy were all excluded. The remaining 135 patients were selected as subjects. Of these, 20 patients obtained a CPC score of 1 or 2 (CPC 1-2 group).

Table 1 shows the demographic characteristics, bystander CPR performance, electrocardiogram rhythm on the scene, etiology of CPA, and estimated duration of CPA in all 135 patients. The ratio of VF at the scene in the CPC 3-5 group was smaller than that in the CPC 1-2 group. The estimated duration of CPA in the CPC 3-5 group was longer than that in the CPC 1-2 group.

Table 2 Results of the serum biochemical analysis on arrival

Table 2 shows the results of a serum biochemical analysis on arrival. The average of total protein in the CPC 3-5 group was smaller than that in the CPC 1-2 group. The average potassium, inorganic phosphorus, and ammonia in the CPC 3-5 group were greater than those in the CPC 1-2 group.

Table 3 shows the results of cell blood count on arrival. The average platelet count in the CPC 3-5 group was smaller than that in the CPC 1-2 group.

Table 4 shows the results of ABG on arrival. The average pH, PO2, and BE in the CPC 3-5 group were smaller than those in the CPC 1-2 group. The average PCO2 in the CPC 3- 5 group was greater than that in the CPC 1-2 group.

CPC 3-5 n		CPC 1-2	n P value
Total protein (g/dL)	6.2 +- 0.5 112	6.7 +- 1.6	16 .0007
Albumin (g/dL)	3.5 +- 0.4 106	3.7 +- 1.3	15 n.s.
Glucose (mg/dL)	302 +- 13 110	253 +- 24	16 n.s.
Total bilirubin	0.5 +- 0.0 112	0.6 +- 0.1	16 n.s.
(mg/dL)
Aspartate	116 +- 12 112	85 +- 14	16 n.s.
aminotransferase
(IU/L)
Alanine	91 +- 9 112	61 +- 11	16 n.s.
aminotransferase
(IU/L)
Gamma-glutamyl	65 +- 9 102	90 +- 24	15 n.s.
transpeptidase (IU/L)
Alkaline phosphatase	292 +- 21 103	249 +- 18	14 n.s.
(IU/L)
Choline esterase	295 +- 6 97	327 +- 16	13 n.s.
(reference range:
178-483) (U/L)
Amylase (IU/L)	136 +- 11 107	107 +- 12	12 n.s.
Blood urea nitrogen	17 +- 1 113	17 +- 1	16 n.s.
(mg/dL)
Creatinine (mg/dL)	1.0 +- 0.0 113	0.9 +- 0.1	16 n.s.
Creatine	220 +- 27 112	136 +- 13	15 n.s.
phosphokinase
(IU/L)
Sodium (mEq/L)	140 +- 0 112	139 +- 1	16 n.s.
Potassium (mEq/L)	5.6 +- 1.6 112	4.2 +- 3.0	16 .003
Chloride (mEq/L)	102 +- 0 112	102 +- 1	16 n.s.
Calcium (mg/dL)	9.4 +- 0 106	9.1 +- 0.1	14 n.s.
Inorganic phosphorus	8.0 +- 0.2 93	5.7 +- 0.7	16 .006
(mg/dL)
Ammonia (mg/dL)	327 +- 27 66	124 +- 27	16 .001
Values are presented as mean +- SE. I indicates international; U, unit.

Multivariate analysis using a logistic regression analysis was used to evaluate the independent factors regarding the outcome. There were no factors independently associated with the outcome.

Discussion

This study demonstrated that patients with a favorable outcome from CPA tended to have higher total protein, platelets, pH, and PO2 and lower PCO2, potassium, phos- phorus, and ammonia in comparison with those with a poor outcome. However, there were no factors independently associated with the outcome by multivariate analysis.

There have been no reports that described that ischemic stress or post-cardiac arrest caused a decrease in total protein. Hypoproteinemia might indicate the existence of malnutrition before CPA. Malnutrition is another risk factor of death [10,11]. Accordingly, the existence of malnutrition before CPA might affect the outcome of the CPA patients.

Table 4 Results of ABG analysis on arrival

CPC 3-5		CPC 1-2	P value
(n = 102)		(n = 16)
pH	7.029 +- 0.018	7.259 +- 0.036	b.0001
PCO2 (mm Hg)	63.0 +- 2.9	36.5 +- 4.6	.0008
PO2 (mm Hg)	103.2 +- 13.2	286.7 +- 38.7	b.0001
Bicarbonate (mmol/L)	15.3 +- 0.7	15.2 +- 1.0	n.s.
BE (mmol/L)	-13.4 +- 0.6	-9.5 +- 1.2	.02
Values are presented as mean +- SE.

During the early CPA interval, the platelet count decreases in a time-dependent manner because of aggregation or breakage after an initial increase due to the hypostatic phenomena [12]. The estimated duration of CPA in the CPC 3-5 group was longer than that in the CPC 1-2 group. Accordingly, the longer the duration of CPA may cause a lower platelet count. Hypoxia causes a decrease in the platelets even if the patient were not in CPA because the platelets may aggregate in pulmonary peripheral vascular tissues [13]. Such a mechanism might explain why longer hypoxia might cause a lower platelet count.

Geddes et al investigated the relationship between CPA and both ABG data and serum potassium, using a pig model with VF [14]. According to the findings of their paper, PO2 decreases whereas PCO2 and potassium increase in a time- dependent manner during CPA, even when chest compres- sion and artificial ventilation were applied. They hypothe- sized that (a) the cause of decreasing PO2 was the formation of pulmonary edema because the pulmonary artery pressure was equal to mean arterial pressure, (b) the cause of increasing PCO2 was alveolar hypoventilation, and (c) the cause of increasing potassium was the leakage of potassium from cells because of Na/K ion pump failure. The poor outcome has longer estimated duration of CPA than that in the favorable outcome. Accordingly, a lower value of PO2 and a higher value of PCO2 and potassium in the poor outcome in comparison with the favorable were observed in this study. Takasu et al [9] reported that the BE has a significant negative relationship with the duration of cardiac arrest because the BE has a close relationship with the peripheral tissue demand for oxygen. Geddes et al observed no trends with regard to pH during CPA [14]. However, they performed blood sampling for only 14 minutes after CPA. Even in their study, the value of PCO2 significantly increased and the value of bicarbonate tended to increase in a time-dependent manner during CPA. Therefore, if they performed a longer observation, the pH should increase, consistent with the observations of Takasu et al [9] and those in the current study.

The relationship between ammonia and CPA has been investigated by Japanese researchers [15,16]. The develop- ment of metabolic and respiratory acidosis during CPA induces the release of ammonia from red blood cells [16]. The longer duration of CPA tends to increase the value of ammonia. Hyperammonemia is reported to be one of the signs of poor neurological outcome [15,16]. In the current study, ammonia also had a strong relationship with pH (R =

-0.610, P b .0001). Therefore, the value of ammonia on the

poor outcome was higher than on the favorable outcome.

The relationship between serum inorganic phosphorus and the outcome of CPA patients or ischemic stress has not been studied so far. In contrast, the serum inorganic phosphorus level has been shown to correlate with the Serum lactate level because phosphorus may be released from damaged cells [17] The value of lactate was not investigated in this study. However, ischemia induces an

increase in serum lactate because of anaerobic glycolysis [18]. Accordingly, the longer duration of CPA and of anoxia might thus lead to an increase in the serum lactate and inorganic phosphorus levels. Makino et al [19] reported that the lactate and phosphorus levels of CPA patients were higher than those of mild traumatized patients. Their results may support the current hypothesis concerning the serum inorganic phosphorus level in patients presenting with CPA.

Conclusion

This study demonstrated that a favorable outcome of CPA cases tended to show higher total protein, platelets, pH, and PO2 and lower PCO2, potassium, phosphorus, and ammonia in comparison with a poor outcome. However, because there were no factors independently associated with the outcome based on a multivariate analysis, unfortunately, the initial blood work cannot be used to make clinical decisions.

References

1. Korfias S, Stranjalis G, Papadimitriou A, et al. Serum S-100B protein as a biochemical marker of brain injury: a review of current concepts. Curr Med Chem 2006;13:3719-31.
2. Kaiser E, Kuzmits R, Pregant P, et al. Clinical biochemistry of neuron specific enolase. Clin Chim Acta 1989;183:13-31.
3. Prohl J, Rother J, Kluge S, et al. Prediction of short-term and long-term outcomes after cardiac arrest: a prospective multivariate approach combining biochemical, clinical, electrophysiological, and neuropsy- chological investigations. Crit Care Med 2007;35:1230-7.
4. Hachimi-Idrissi S, Van der Auwera M, Schiettecatte J, et al. S-100 protein as early predictor of regaining consciousness after out of hospital cardiac arrest. Resuscitation 2002;53:251-7.
5. Rosen H, Sunnerhagen KS, Herlitz J, et al. Serum levels of the brain- derived proteins S-100 and NSE predict long-term outcome after cardiac arrest. Resuscitation 2001;49:183-91.
6. Rosen H, Rosengren L, Herlitz J, et al. Increased serum levels of the S-100 protein are associated with Hypoxic brain damage after cardiac arrest. Stroke 1998;29:473-7.
7. Sherman AL, Tirschwell DL, Micklesen PJ, et al. Somatosensory potentials, CSF creatine kinase BB activity, and awakening after cardiac arrest. Neurology 2000;54:889-94.
8. Rosen H, Karlsson JE, Rosengren L. CSF levels of neurofilament is a valuable predictor of long-term outcome after cardiac arrest. J Neurol Sci 2004;221:19-24.
9. Takasu A, Sakamoto T, Okada Y. Arterial base excess after CPR: the relationship to CPR duration and the characteristics related to outcome. Resuscitation 2007;73:394-9.
10. Fenley J, Powers PS, Miller J, et al. Untreated Anorexia nervosa. A case study of the medical consequences. Gen Hosp Psych 1990;12:264-70.
11. Ravasco P, Camilo ME, Gouveia-Oliveira A, et al. A critical approach to nutritional assessment in critically ill patients. Clin Nutr 2002;21:73-7.
12. Thomsen H, Kaatsch HJ, Krisch B. How and why does the platelet count in postmortem blood change during the early postmortem interval? Forensic Sci Int 1999;101:185-94.
13. Lehmann T, Mairbaurl H, Pleisch B, et al. Platelet count and function at High altitude and in high-altitude pulmonary edema. J Appl Physiol 2006;100:690-4.
14. Geddes LA, Roeder RA, Rundell AE, et al. The natural biochemical changes during ventricular fibrillation with cardiopulmonary resusci- tation and the onset of postdefibrillation pulseless electrical activity. Am J Emerg Med 2006;24:577-81.
15. Nagamine K. Does blood ammonia level at time of initial treatment predict the outcome of patients in cardiopulmonary arrest on arrival? J Jpn Assoc Acute Med 2005;16:283-8.
16. Ishida H, Mastuoka T, Yokota J, et al. The clinical significance of an increase in the Blood concentration of ammonia in cases with cardiopulmonary arrest on arrival. J Jpn Soc Emerg Med 2002;5:490-4 [in Japanese].
17. Oster JR, Alpert HC, Vaamonde CA. Effect of acid-base status on plasma phosphorus response to lactate. Can J Physiol Pharmacol 1984;62:939-42.
18. Mullner M, Sterz F, Domanovits H, et al. The association between blood lactate concentration on admission, duration of cardiac arrest, and functional neurological recovery in patients resuscitated from ventricular fibrillation. Intensive Care Med 1997; 23:1138-43.
19. Makino J, Uchino S, Morimatsu H, et al. A quantitative analysis of the acidosis of cardiac arrest: a prospective observational study. Crit Care 2005;9:R357-62.