a b s t r a c t

Objective: Comparing a point-of-care (POC) test using the capillary blood obtained from skin puncture with con- ventional laboratory tests.

Methods: In this study, which was conducted at the emergency department of a tertiary care hospital in April-July 2017, 232 patients were enrolled, and three types of blood samples (capillary blood from skin puncture, arterial and venous blood from blood vessel puncture) were simultaneously collected. Each blood sample was analyzed using a POC analyzer (epoc(R) system, USA), an arterial Blood gas analyzer (pHOx(R)Ultra, Nova biomedical, USA) and venous blood analyzers (AU5800, DxH2401, Beckman Coulter, USA). Twelve parameters were compared between the epoc and reference analyzers, with an equivalence test, Bland-Altman plot analysis and linear re- gression employed to show the agreement or correlation between the two methods.

Results: The pH, HCO₃, Ca²⁺, Na⁺, K⁺, Cl^-, glucose, Hb and Hct measured by the epoc were equivalent to the ref- erence values (95% confidence interval of mean difference within the range of the agreement target) with clin- ically inconsequential mean differences and narrow limits of agreement. All of them, except pH, had clinically acceptable agreements between the two methods (results within target value >= 80%). Of the remaining three parameters (pCO_2, pO₂ and lactate), the epoc pCO₂ and lactate values were highly correlated with the reference device values, whereas pO₂ was not. (pCO₂: R² = 0.824, y = -1.411 + 0.877.x; lactate: R² = 0.902,y = -0.544 + 0.966.x; pO₂: R² = 0.037, y = 61.6 + 0.431.x).

Conclusion: Most parameters, except only pO₂, measured by the epoc were equivalent to or correlated with those from the reference method.

(C) 2017

1. Introduction

Point-of-care (POC) analysis involves simple medical tests that can be conducted at or near the point of care. This approach has obvious benefits, including rapid critical decision making and early disease iden- tification, which can ensure that appropriate treatment is provided dur- ing acute care. As technology has developed over the past few decades, POC tests have been increasingly used in many clinical settings, includ- ing the prehospital setting [1-6]. Prehospital electrocardiogram for cardiac cath lab activation has already been applied, and blood analysis of selected parameters, such as Blood sugar, cardiac enzymes and lactate, has also been widely used prehospitally. Prehospital blood gas analysis and electrolyte measurement are also extremely valuable, especially for patients with cardiac arrest, as electrolyte (e.g., hyperkalemia) and acid-base imbalance (metabolic acidosis) are typical treatable (reversible) causes of sudden cardiac arrest. However, its

* Corresponding author at: Department of Emergency Medicine, Hanyang University Guri Hospital, 153, Gyeongchun-ro(st), Guri-si, GyeongGi-do 471-701, Republic of Korea.

E-mail address: [email protected] (C. Kim).

¹ CK & HK contributed equally to this paper.

Prehospital use has not been widely adopted because of several limita- tions, even though its feasibility for prehospital application has techni- cally been proven [7,8].

A prehospital healthcare provider must be able to use the results of a POC test to impact prehospital patient care. If a lab result cannot impact prehospital care or transport decisions, then blood need not be drawn during prehospital care. Although paramedics can interpret the lab re- sult and provide the proper management prehospitally in some coun- tries, most prehospital healthcare providers (usually EMTs) have no proper training to understand its significance. Recently, as telecommu- nication between an in-hospital physician and a prehospital provider has emerged, this limitation has ceased to be a problem [9-11]. Howev- er, prehospital blood sampling is still difficult and time consuming. This technique must be practiced regularly to maintain proficiency; howev- er, Prehospital providers usually do not have enough experience of puncturing blood vessels, although individual experiences may vary.

The Enterprise Point-of-care (epoc(R) Blood Analysis System, Alere, USA) blood analysis system is a handheld device that can test blood gas, electrolytes, and metabolites using capillary blood samples as well as arterial and venous blood (Fig. 1). Compared to puncturing a blood vessel, the capillary blood can be easily obtained by puncturing the

https://doi.org/10.1016/j.ajem.2017.12.025

0735-6757/(C) 2017

Fig. 1. Point-of-care analyzer (epoc(R)) using the test cartridge containing the sensors.

skin of a fingertip or ear lobe. Although several previous studies have shown the correlation between venous blood and skin puncture blood samples using a conventional device, no studies evaluating the effec- tiveness of a POC test using skin puncture blood samples exist [12-15]. We hypothesized that the results of a POC test using skin puncture blood samples for electrolytes and hydrogen ions, which are regarded as life-threatening yet Reversible causes of cardiac arrest, would be equivalent to those obtain from conventional laboratory tests (refer- ence tests) using samples obtained from blood vessel puncture. We wished to apply the skin puncture blood sample based-POC test to the prehospital setting, especially to cases of out-of-hospital cardiac arrest (OHCA), by verifying this hypothesis. In this study, as a first step, we in- vestigated whether results from the skin punctured POC blood analysis were equivalent to those from the conventional Laboratory analysis systems.

1. Methods and materials
   1. Study design and setting

This prospective experimental study was conducted at an emergen- cy department (ED) of an urban tertiary care teaching hospital from April to July 2017. The local ethics committee approved the study. We registered the study protocol in Clinical Trials before study initiation (Clinicaltrials.gov: NCT03096665).

Study participants

Participants included cases needed to be taken both venous and ar- terial blood samples during the study period. It was calculated that at least 200 samples were required for this study, and considering the po- tential for dropouts and the test failure rate, 250 patients were enrolled in this study. Patients who were already receiving oxygen therapy or fluid with electrolytes, which would affect the result, were also excluded.

Blood collection

After the participants signed a written consent form, three types of blood samples were obtained almost simultaneously. The skin puncture blood sample of 90 uL was obtained by an emergency medical

technician (EMT) from the fingertip (Fig. 2). The arterial blood samples (1 mL) were drawn into a heparinized syringe (BD Preset(TM), Becton Dickinson and Company, UK) by the emergency physician from the ra- dial artery, and the venous blood samples (3 mL) were acquired by punctures made by trained ED nursing staff from the vein of forearm and were collected in a collection tube (BD vacutainer) for laboratory measurement. The collection of each blood sample was performed in ac- cordance with the WHO guidelines [16] and analyzed by the below devices.

Blood sample analyzers

The skin puncture blood sample was analyzed using a POC blood analysis system, namely, the epoc(R) Blood Analysis System (Alere, USA), which is a handheld device, via the test cartridge containing the sensors (Fig. 1). The epoc can measure the parameters of pH, pCO₂, Na⁺, K⁺ and Ca²⁺ using selective electrode potentiometry. Metabolites (glucose and lactate) and pO₂ are also measured amperometrically. Bi- carbonate (HCO^-) is calculated. These values can be shown in approxi- mately 195 s (cartridge calibration time of 165 s and analysis time of 30 s) on the display of the device. In addition, the results can be sent to the physician via Wi-Fi.

3

The arterial blood samples for blood gas parameters (pH, pCO₂, pO₂, HCO^-) as well as lactate and ionized calcium were analyzed using an ar- terial blood gas analyzer (pHOx(R)Ultra, Nova biomedical, USA). The ve- nous blood samples were sent to the laboratory, and the laboratory testing was performed on a clinical chemistry analyzer (AU5800, Beckman Coulter Inc., USA) for electrolytes and metabolites and on a cellular analyzer (DxH2401, Beckman Coulter Inc., USA) for hemoglobin (Hb) and hematocrit (Hct).

3

Statistical analysis

We compared the parameters of the skin punctured blood sample obtained from the epoc system and those obtained from the reference blood analysis systems (arterial blood gas analyzer: pHOx(R)Ultra, ve- nous blood analyzers: AU5800 and DxH2401). An equivalence test was conducted to analyze whether the results from the skin punctured POC blood analysis were equivalent to those from the above reference analysis systems conducted in a central laboratory. The agreement tar- gets for pH, pCO₂, Na⁺, K⁺, Cl^-, glucose, Hb and Hct, which are shown in Table 2, were defined based on the Clinical Laboratory Improvement Amendments (CLIA) guidelines [17,18]. The agreement targets for the remaining parameters (pO₂, HCO^-, Ca²⁺, and lactate) were defined as 15% of the mean value determined by the reference analyzers based on alternative criteria of the CLIA guidelines. When the 95% confidence interval of mean difference between reference device and epoc was in- cluded within the range of the agreement target, the parameters mea- sured by the epoc were equivalent to those measured by the reference device. Bland-Altman plot analysis was also conducted to show the agreement between the two methods (skin punctured epoc analysis vs. reference analysis). We regarded the parameters measured by the epoc as having a clinically acceptable agreement when at least 80% of the results (difference between the reference device and epoc) were within the range of agreement target. If the parameters did not have a clinically acceptable agreement (results within target value b 80%), lin- ear regression was conducted to evaluate the correlation between the two methods.

3

1. Results

Of the 250 participants, 27 tests were not completed due to test fail- ure, which corresponded to a mean test failure rate of 10.8%. There are several reasons for test failure. Our tests were mainly failed due to test card failure (mainly internal quality control error) and the sample injec- tion errors (insufficient blood sample volume and fast sample

Fig. 2. point-of-care analysis using the capillary blood obtained from skin puncture; 1) insert cartridge, 2) prick fingertip with a needle, 3) capillary blood sampling, 4) capillary tube (Care- Fill(TM)) with 90 uL of capillary blood, 5) inject blood sample into the test card, 6) check result.

injection). Of the 223 patients enrolled, 178 (61.9%) were male, and their mean age was 64.7 years old (SD 17.6) (Table 1). Their diagnoses are also shown in Table 1. Twelve parameters were analyzed for agree- ment (Table 2). The mean values of each parameter measured by each method, and the mean difference are shown in Table 2. The parameters of pH, HCO₃, Ca²⁺, Na⁺, K⁺, Cl^-, glucose, Hb and Hct measured by the epoc were equivalent to the reference values (95% CI of mean difference within the range of the agreement target) (Table 2). For these variables, there were clinically inconsequential mean differences between the epoc and Reference devices, and the limit of agreement was narrow (Figs. 3 and 4). For all of them, except pH, there was also a clinically ac- ceptable agreement between the two methods (results within the tar- get value >= 80%). The remaining three parameters (pCO_2, pO₂ and lactate) had neither equivalence nor clinically acceptable agreement

Table 1

Characteristics of the study participants

Characteristics values

Male, n (%) 178 (61.9)

Age, year (SD) 64.7 (17.6)

Diagnosis, n (%)

Cardiac arrest 8 (3.4)

Congestive heart failure 31 (13.3)

Acute coronary syndrome 19 (8.2)

Cerebrovascular accident 6 (2.6)

Chronic obstructive pulmonary disease 23 (9.9)

Pneumonia 35 (15.0)

Acute respiratory distress syndrome 9 (3.9)

Metabolic acidosis 8 (3.4)

Sepsis 28 (12.0)

Cancer 6 (2.6)

Trauma 21 (9.0)

electrolyte imbalance 15 (6.4)

Renal failure 11 (4.7)

Other 13 (5.6)

SD: Standard deviation.

with the reference device. Among these, the pCO₂ and lactate measured by epoc were highly correlated with those measure by the reference de- vice, whereas PO₂ was not. (PCO₂: R² = 0.824, y = -1.411 + 0.877.x; lactate: R² = 0.902,y= -0.544 + 0.966.x; PO₂: R² = 0.037, y = 61.6

+ 0.431.x). The regression model accounts for 82.4% of the variance for PCO₂ and 90.2% of the variance for lactate.

The Bland-Altman plots for PO₂, HCO^-, Ca²⁺ and glucose all showed trends: the scatter plots for PO₂ and Ca²⁺ exhibited a negative trend, whereas the scatter plot for HCO^- had a positive trend. (Figs. 3 and 4) The scatter plot for glucose is distributed around the bias line when the average value of glucose is less than approximately 200 mg/dL but becomes more distant from the bias line as the average gets higher (Fig. 4).

3

3

1. Discussion

Although several studies have reported small differences between capillary blood and venous blood analysis using the same laboratory an- alyzer [12-15], to the best of our knowledge, this is the first study veri- fying the effectiveness of a POC device using the capillary blood in a clinical setting. The POC analyzer provides increased accessibility and cost-effectiveness in terms of reducing maintenance costs compared with traditional laboratory analyzers, enabling their increased use in the prehospital care setting. The epoc, a handheld POC blood analysis system incorporating a wireless card reader and a personal data assis- tant for data analysis, was approved by US Food and Drug Administra- tion in 2006. Similar to the i-STAT, which was previously the leading device, (Abbott Point of Care, Princeton, NJ), the epoc can analyze blood gas, electrolytes and metabolites. However, this device also has several advantages over the i-STAT regarding its use in the prehospital setting. Moreover, the epoc has individually barcoded cards and can be stored at Room temperature, which is one of the most important ad- vantages [19], as the requirement for refrigerated storage of the car- tridge (2 to 8 ?C) has prevented the prehospital use of i-STAT. The cost of the epoc test card is somewhat expensive; however, compared to

Table 2

Statistics for comparison of the epoc and reference analyzers

Parameters	Reference value, mean (SD)	epoc, mean (SD)	Mean difference	95% CI		LOA of differencee	Agreement target	Results within target values
pH	7.420 (0.098)a	7.405 (0.098)	-0.016	0.010	0.021	+-0.08	+-0.04c	74.4%
pCO2, mmHg	31.45 (10.09)a	37.12 (10.44)	-5.97	-6.550	-5.385	+-8.66	+-8%c	12.6%
pO2, mmHg	87.28 (28.17)a	59.52 (12.52)	27.76	23.981	31.534	+-55.96	+-15%d	26.0%
HCO3, mmol/L	20.34 (4.70)a	22.49 (4.94)	-2.15	-2.325	-1.972	+-2.63	+-15%d	83.4%
Ionized calcium, mmol/L	1.19 (0.08)a	1.10 (0.10)	0.09	0.081	0.101	+-0.14	+-15%d	90.6%
Lactate, mmol/L	2.65 (2.96)a	3.30 (2.91)	-0.66	-0.781	-0.534	+-1.83	+-15%d	36.3%
Sodium, mEq/L	135.01 (6.08)b	138.37 (6.53)	-3.35	-3.648	-3.061	+-4.36	+-4c	83.0%
Potassium, mEq/L	4.29 (0.94)b	4.53 (0.96)	-0.24	-0.278	-0.209	+-0.52	+-0.5c	86.1%
Chloride, mEq/L	102.74 (6.78)b	105.84 (6.0.)	-3.10	-3.633	-2.565	+-7.93	+-5%c	81.6%
Glucose, mg/dL	158.69 (88.35)b	155.75 (87.02)	2.94	1.371	4.503	+-23.26	+-10%c	83.4%
Hemoglobin, g/dL	12.02 (2.59)b	12.01 (2.61)	-0.002	-0.110	0.115	+-1.67	+-7%c	82.5%
Hematocrit, %	36.26 (7.61)b	35.64 (7.56)	0.625	0.378	0.873	+-3.67	+-6%c	80.7%

SD: standard deviation, CI: confidence interval, LOA: limit of agreement.

^a Analyzed by using the arterial blood gas analyzer (pHOx(R)Ultra, Nova biomedical, USA).

^b Analyzed by using the venous blood analyzer (AU5800 and DxH2401, Beckman Coulter Inc., USA).

^c Defined based on the Clinical Laboratory Improvement Amendments (CLIA) guidelines [17,18].

^d Defined as 15% of the reference mean value based on alternative criteria of the CLIA guidelines.

^e LOA of difference = +-1.96.SD.

laboratory devices, it rarely generates maintenance expenditure. Thus, the epoc may be more beneficial for use in the prehospital setting where its usage would be relatively low than in an in-hospital laborato- ry. In this study, we verified that the epoc system using capillary blood may be useful for analyzing most parameters, except pO₂. It could be used at the prehospital level.

A paramedic or EMT frequently performed cardiopulmonary resus- citations of OHCA patients in the field in several countries. In South Korea, four to eight minutes of field CPR is generally conducted by an EMT. Even prehospital advanced field resuscitation, including the ad- ministration of epinephrine, has been tested as a pilot project in some model areas. The EMT usually stays at the scene for more than twenty minutes to conduct field Advanced cardiac life support with video communication-based medical direction from a remote emergen- cy physician. Reportedly, this pilot project led to three times more CPR survivals (unpublished data). If the appropriate life support could be provided in the field, then this long field stay may not be problematic to most OHCA patients because the most common cause of non-trau- matic OHCA is cardiogenic. However, we are worried that delayed transportation to the hospital could reduce the chance of treatment in some patients. For example, patients with a treatable cause, such as life-threatening electrolyte imbalance (hyperkalemia) or severe acido- sis, should have their blood analyzed rapidly and thus should be imme- diately transferred to the hospital instead of remaining in the field. However, these patients could be identified in the field if the blood anal- ysis could be performed on site. In this study, the blood level of potassi- um obtained from the capillary blood and measured by the POC device were highly equivalent to those measured by traditional laboratory analysis systems. Moreover, there is a clinically acceptable agreement between the POC and reference methods. Hyperkalemia, one of the re- versible causes of OHCA, can be identified by this POC analysis in the field, thereby allowing patients to receive a reversible drug, for instance, Calcium gluconate and sodium bicarbonate, in the field immediately if approved or to be transferred to the hospital without delay. Severe met- abolic acidosis could also be detected on the field given that the capillary blood-based epoc analysis for pH and HCO3^- had a good equivalence with the reference analysis, even though epoc analysis for pH did not reach a clinically acceptable agreement when the acceptable level was set at 80%. Although the latest American Heart Association (AHA) guide- lines for CPR and emergency cardiovascular care (ECC) do not recom- mend the routine use of sodium bicarbonate for patients in cardiac arrest, it can be beneficially used in some specific resuscitation situa- tions, such as preexisting metabolic acidosis, hyperkalemia and tricyclic antidepressant (TCA) overdose [20]. Adgey et al. also recommended the

use of bicarbonate for cardiac arrest patients in certain categories: 1) se- vere acidosis (pH b 7.1), 2) prolonged (N 10-20 min) CPR, 3) hyperkalemia, and 4) TCA overdose [21]. Thus, prehospital POC analysis for potassium, pH and HCO3^- would be useful for patients with cardiac arrest.

The 2005 AHA guidelines for CPR and ECC had defined hypoglycemia as the treatable causes of cardiac arrest (“H’s and T’s”) [22]. However, when the 2010 guidelines were established, hypoglycemia was deleted from the “H’s and T’s” in adult cardiac arrest [20,23]. Nevertheless, it has been still listed as a treatable cause of cardiac arrest in pediatric patients [24,25]. Considering that both hypoglycemia and hyperglycemia are true Medical emergencies, prehospital glucose analysis could also be helpful. In this study, glucose measurements from the epoc were highly equivalent to those from laboratory analyses and reached a clinically ac- ceptable agreement with the reference method. The scatter plot for glu- cose is distributed around the bias line when the average value of glucose is less than approximately 200 mg/dL but becomes more distant from the bias line as the average increases further, indicating that the glucose level measure by epoc may not be as clinically accepted when the measured value is rising. However, patients with a glucose level N 300 mg/dL were very few in this study; thus, further study would be needed.

Lactate acts as a surrogate marker for tissue hypoxia following sepsis or traumatic injury [26]. Prehospital POC lactate measurement for those suspected patients could enable rapid identification and prompt management, including fluid resuscitation, which could lead to im- proved outcomes [27]. Purcarea et al. reported that POC capillary lactate measurement was a better biomarker in the early stage of sepsis than classical lactate measurement [28]. Although there is no equivalence or agreement between the epoc and reference device in this study, the reference value can be predicted by using the epoc value according to the regression equation (y = -0.544 + 0.966.x), which accounts for 90.2% of the variance (R² = 0.902). Therefore, epoc lactate analysis using the capillary blood can be evaluated for use in a prehospital sepsis model.

We chose capillary measurement from the fingertip rather than the earlobe because EMTs in South Korea are familiar with performing a fin- gertip puncture. They usually perform a POC blood sugar test on a fin- gertip prehospitally. In addition, some researchers suggested that the laboratory values of the fingertip capillary blood are closer to those of whole blood than those of the earlobe [28]. Given that the capillary blood sampling from a fingertip is very simple and feasible, capillary blood sampling may be more suitable at the prehospital level than arte- rial or venous puncture.

Fig. 3. Bland-Altman plot between the reference analyzer (arterial blood gas analyzer) and epoc system.

1. Limitation

The test failure rate was 10.8% (27/250), which was lower than that in a previous study (16.1%) [19]. In this study, two EMTs, who

were trained to conduct the standard fingertip puncture and appro- priate manipulation method for the epoc device prior to the study, performed the capillary blood examination. In addition, all examina- tions were supervised by an emergency physician. This quality

Fig. 4. Bland-Altman plot between the reference analyzers (venous blood analyzers) and epoc system.

control might contribute to the relatively lower test failure rate than that in the previous study. However, compared to this study, prehospital use by non-trained paramedics without supervision could

lead to a higher test failure rate. Given that the epoc is used in emergen- cy situations, the time delay due to additional calibration and analysis (approximately 200 s) induced by the test failure could limit the active

use in the prehospital setting. Furthermore, the test card is relatively ex- pensive ($10-20).

Although several critically ill patients, including patients with cardi- ac arrest, were enrolled in this study, most enrolled patients were he- modynamically stable. In addition, this study was conducted in the ED setting. Further study on critically ill patients in the prehospital setting is needed because patients with peripheral hypoperfusion could reduce the accuracy of the capillary blood measurement.

1. Conclusion

The capillary blood-based epoc system may have some clinical utili- ty in the prehospital setting. Most parameters (pH, PCO₂, lactate, HCO₃, Ca²⁺, Na⁺, K⁺, Cl^-, glucose, Hb and Hct), except pO₂, were equivalent to or correlated with the reference method. Although this EMT-performed POC analysis may not be useful when the distance of transportation is short, it can be valuable when field CPR is provided or when transporta- tion time is long in an under-resourced area, such as a suburban or rural area.

Acknowledgements

We thank Professor Nam in the Biostatistical Consulting and Re- search Lab, Hanyang University, for assistance with statistical analysis.

Funding sources

We declare that we got grant money for investigator initiated re- search (conceived and written by author CK) from Alere health care (201700000000585). However, they had no involvement in this work.

References

1. Gilbert HC, Szokol JW. Point of care technologies. Int Anesthesiol Clin 2004;42: 73-94.
2. Peterson MC, Syndergaard T, Bowler J, Doxey R. A systematic review of factors predicting door to balloon time in ST-segment elevation myocardial infarction treat- ed with percutaneous intervention. Int J Cardiol 2012;157(1):8-23.
3. Foss KT, Subhi Y, Aagaard R, Bessmann EL, Botker MT, Graumann O, et al. Developing an emergency ultrasound app - a collaborative project between clinicians from dif- ferent universities. Scandinavian Journal of Trauma, Resuscitation and Emergency Medicine 2015;23:47.
4. Nelson BP, Sanghvi A. Out of hospital Point of Care Ultrasound: current use models and future directions. Eur J Trauma Emerg Surg 2016 Apr;42(2):139-50.
5. Stengaard C, Sorensen JT, Ladefoged SA, Christensen EF, Lassen JF, Botker HE, et al. Quantitative point-of-care troponin T measurement for diagnosis and prognosis in patients with a suspected acute myocardial infarction. Am J Cardiol 2013;112(9): 1361-6.
6. Guyette FX, Meier EN, Newgard C, McKnight B, Daya M, Bulger EM, et al. A compar- ison of prehospital lactate and systolic blood pressure for predicting the need for re- suscitative care in trauma transported by ground. J Trauma Acute Care Surg 2015; 78(3):600-6.
7. Mikkelsen S, Wolsing-Hansen J, Nybo M, Maegaard CU, Jepsen S. Implementation of the ABL-90 blood gas analyzer in a ground-based mobile emergency care unit. Scan- dinavian Journal of Trauma, Resuscitation and Emergency Medicine 2015 Jul 30;23: 54.
8. Prause G, Ratzenhofer-Komenda B, Offner A, Lauda P, Voit H, Pojer H. Prehospital point of care testing of blood gases and electrolytes - an evaluation of IRMA. Crit Care 1997;1(2):79-83.
9. Kim C, Hur J, Kang BS, Choi HJ, Shin JH, Kim TH, et al. Can an offsite expert remotely evaluate the visual estimation of ejection fraction via a social network Video call? J Digit Imaging 2017 Dec;30(6):718-25.
10. Lee Y, Kim C, Choi HJ, Kang B, Oh J, Lim TH. A feasibility study of telementoring for identifying the appendix using smartphone-based telesonography. J Digit Imaging 2017 Apr;30(2):148-55.
11. Kim C, Kang BS, Choi HJ, Lim TH, Oh J, Chee Y. Clinical application of real-time tele- ultrasonography in diagnosing pediatric acute appendicitis in the ED. Am J Emerg Med 2015 Oct;33(10):1354-9.
12. Kupke IR, Kather B, Zeugner S. On the composition of capillary and venous blood serum. Clin Chim Acta 1981 May 5;112(2):177-85.
13. Sobiech P, Pomianowski A, Stopyra A, Kuleta Z, Zbanyszek M, Procajlo A, et al. Acid- base balance parameters and ionic composition of arterial, venous and capillary blood in goats. Pol J Vet Sci 2004;7(2):117-21.
14. Blumenfeld TA, Hertelendy WG, Ford SH. Simultaneously obtained skin-puncture serum, skin-puncture plasma, and venous serum compared, and effects of warming the skin before puncture. Clin Chem 1977 Sep;23(9):1705-10.
15. Falch DK. Clinical chemical analyses of serum obtained from capillary versus venous blood, using microtainers and vacutainers. Scand J Clin Lab Invest 1981 Feb;41(1): 59-62.
16. WHO. Guidelines approved by the guidelines review committee. WHO Guidelines on Drawing Blood: Best Practices in Phlebotomy. Geneva: World Health Organiza- tion; 2010.
17. Bachner P, Hamlin W. Federal regulation of clinical laboratories and the Clinical Lab- oratory Improvement Amendments of 1988-Part I. Clin Lab Med 1993 Sep;13(3): 739-52 [discussion 737-8].
18. Ricos C, Alvarez V, Cava F, Garcia-Lario JV, Hernandez A, Jimenez CV, et al. Minimum, Optimal and Desirable Specifications for Total Error, Imprecision, and Bias, Derived from Intra- and Inter-individual Biologic Variation. Accessed 05 September 2017, at https://www.westgard.com/biodatabase1.htm.
19. Stotler BA, Kratz A. Analytical and Clinical performance of the epoc blood analysis system: experience at a large tertiary academic medical center. Am J Clin Pathol 2013 Nov;140(5):715-20.
20. Link MS, Berkow LC, Kudenchuk PJ, Halperin HR, Hess EP, Moitra VK, et al. Part 7: adult advanced cardiovascular life support: 2015 American Heart Association guide- lines update for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation 2015 Nov 3;132(18 Suppl 2):S444-64.
21. Adgey AA. Adrenaline dosage and buffers in cardiac arrest. Heart 1998;80(4):412-4.
22. American Heart Association. 2005 AHA guidelines for cardiopulmonary resuscitation and emergency cardiovascular care. Part 7.2: management of cardiac arrest. Circula- tion 2005;112 (IV-57-IV-66).
23. Neumar RW, Otto CW, Link MS, Kronick SL, Shuster M, Callaway CW, et al. Part 8: adult advanced cardiovascular life support: 2010 American Heart Association Guide- lines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circu- lation 2010 Nov 2;122(18 Suppl 3):S729-67.
24. Kleinman ME, de Caen AR, Chameides L, Atkins DL, Berg RA, Berg MD, et al. Pediatric basic and advanced life support chapter collaborators. Part 10: pediatric basic and advanced life support: 2010 international consensus on cardiopulmonary resuscita- tion and Emergency Cardiovascular Care Science with treatment recommendations. Circulation 2010 Oct 19;122(16 Suppl 2):S466-515.
25. de Caen AR, Berg MD, Chameides L, Gooden CK, Hickey RW, Scott HF, et al. Part 12: pediatric advanced life support: 2015 American Heart Association guidelines update for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation 2015 Nov 3;132(18 Suppl 2):S526-42.
26. Lewis CT, Naumann DN, Crombie N, Midwinter MJ. Prehospital point-of-care lactate following trauma: a systematic review. J Trauma Acute Care Surg 2016 Oct;81(4): 748-55.
27. Smyth MA, SJ Brace-McDonnell, Perkins GD. Identification of adults with sepsis in the prehospital environment: a systematic review. BMJ Open 2016 Aug 5; 6(8):e011218.
28. Purcarea A, Bourgarit A, Sovaila A, Ghiura C, Diemunsch P, Andres E. Brief report: se- rial capillary lactate measurement predict the evolution of early sepsis. J Med Life 2016 Jan-Mar;9(1):74-8.