- Email: [email protected]
Point-of-care blood analysis of hypotensive patients in the emergency department
American Journal of Emergency Medicine xxx (xxxx) xxx
Contents lists available at ScienceDirect
American Journal of Emergency Medicine journal homepage: www.elsevier.com/locate/ajem
Point-of-care blood analysis of hypotensive patients in the emergency department☆ Hyungoo Shin, MD, PhD 1, Inhye Lee, MD 1, Changsun Kim, MD, PhD ⁎, Hyuk Joong Choi, MD, PhD Department of Emergency Medicine, College of Medicine, Hanyang University Guri Hospital, Guri, Republic of Korea
a r t i c l e
i n f o
Article history: Received 26 April 2019 Received in revised form 15 July 2019 Accepted 21 July 2019 Available online xxxx Keywords: Point-of-care systems Capillary blood Emergency medical services Hypotension
a b s t r a c t Objective: The aim of this study is to compare a point-of-care (POC) analysis, Enterprise POC (epoc), using the capillary blood obtained from skin puncture with conventional laboratory tests using arterial and venous blood in hypotensive patients. Methods: This study was conducted at the emergency department of a tertiary care hospital between June and November 2018. 231 hypotensive patients were enrolled. Three types of blood samples (capillary blood from skin puncture and arterial and venous blood from blood vessel puncture) were collected and analyzed. We compared a total of 13 parameters (pH, pCO2, pO2, HCO3−, Ca2+, lactate, Na+, K+, Cl−, glucose, Hb, Hct, and creatinine) between the POC analysis and reference analyzers by performing the equivalence test and Bland-Altman plot analysis. Results: In hypotensive patients, with the exception of two parameters (pCO2, pO2), the pH, HCO3−, Ca2+, lactate, Na+, K+, Cl−, glucose, Hb, Hct, and creatinine parameters measured by the POC analysis were equivalent to or correlated with the reference values. In the patients with cardiac arrest group, nine parameters (pH, HCO3−, Ca2+, Na+, K+, glucose, Hb, Hct, and creatinine) analyzed by the epoc system were equivalent to the reference values. Conclusion: Most parameters, except pO2, measured by the epoc system using the capillary blood in hypotensive patients were equivalent to or correlated with those measured by the reference analyzers. © 2019 Published by Elsevier Inc.
1. Introduction In hemodynamically unstable patients, especially those with cardiac arrest, serum biomarkers, such as electrolytes (e.g. hyperkalemia), acidbase imbalance (metabolic acidosis), and point-of-care (POC) blood sampling can help recognize reversible causes at an early stage [1,2]. Thus, the interest in these analyses in the pre-hospital stage has increased. With the evolution in technology over the last few decades, POC analysis has become increasingly available in many clinical settings, including the prehospital environment [3-9]. We used a skin punctured capillary blood-based POC analysis in a previous study and showed 2+ that parameters including pH, pCO2, HCO− , lactate, Na+, K+, 3 , Ca Cl−, glucose, hemoglobin (Hb), and hematocrit (Hct), except for pO2, measured by POC analysis were equivalent to or correlated with those
measured by conventional laboratory tests [3]. Although a few critically ill patients were enrolled, most were hemodynamically stable in that study. Thus, we could not confirm that the POC analysis using the capillary blood of hypotensive patients were equivalent to the conventional laboratory tests, because of reduced peripheral blood circulation which could affect the results. The aim of this study was to investigate whether the results of the POC analysis using capillary blood samples were equivalent to the conventional laboratory tests in case of hypotensive patients. We conducted this study at an urban, tertiary care, teaching hospital as a primary step to test the POC analysis in prehospital settings anticipating that eventually, the reversible symptoms in the hemodynamically unstable patients can be corrected ahead of arrival at the emergency department (ED). 2. Methods and materials
☆ This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. ⁎ Corresponding author at: Department of Emergency Medicine, College of Medicine, Hanyang University Guri Hospital, 153, Gyeongchun-ro, Guri-si, Gyeonggi-do 11923, Republic of Korea. E-mail address: fl[email protected] (C. Kim). 1 These authors contributed equally to this work.
https://doi.org/10.1016/j.ajem.2019.158363 0735-6757/© 2019 Published by Elsevier Inc.
2.1. Study design This prospective experimental study was performed from June to October 2018. Evaluation of the POC analysis was conducted at an ED of an urban, tertiary care, teaching hospital. The local ethics committee
2
H. Shin et al. / American Journal of Emergency Medicine xxx (xxxx) xxx
2.3. Equipment and material
Table 1 Characteristics of the study population Characteristics
Values
Male, n (%) Age, year (SD) Diagnosis, n (%) Cardiac arrest Cardiogenic Acute coronary syndrome Congestive heart failure Cardiomyopathy Hypovolemic Gastrointestinal bleeding Trauma related blood loss Burn Septic Pneumonia Abdominal infection Kidney infection etc Distributive Anaphylaxis Salicylate toxicity Neurogenic Total
149 (64.5) 68.2 (16.6)
2.3.1. Blood collection We obtained three types of blood samples simultaneously after the hypotensive patients were presented at the ED. 1) The EMTs working in the ED obtained the capillary blood samples from the fingertip of the patients upon arrival to the ED. 2) The arterial blood samples from the radial artery of the patients were obtained with a heparinized syringe (BD Preset™, Becton Dickinson and Company, UK) by the emergency physician and analyzed. 3) The venous blood samples were obtained by trained ED nursing staff from the vein of the forearm and were collected in a collection tube (BD Vacutainer) for laboratory measurement. We collected each blood sample in accordance with the World Health Organization guidelines [11] and analyzed them in the devices described below.
30 (13.0) 46 (19.9) 27 (11.7) 17 (7.4) 2 (0.9) 67 (29.0) 52 (22.5) 14 (6.1) 1 (0.4) 76 (32.9) 41 (17.7) 23 (10.0) 9 (3.9) 3 (1.3) 12 (5.2) 4 (1.7) 1 (0.4) 7 (3.0) 231 (100)
approved this study in May 2018 (GURI 2018–01–004-004) and the study protocol was registered with the Clinical Research Information Service (cris.nih.go.kr: KCT0003159). 2.2. Participants We included hypotensive patients who were presented at the ED; both venous and arterial blood samples had been taken. Patients with a systolic blood pressure of less than 90 mmHg or mean arterial pressure of less than 65 mmHg were included [10]. Patients with cardiac arrest at ED arrival were also included. We excluded those patients who were already receiving crystalloid with electrolytes since it would affect the results. Patients under the age of 18 years were also excluded. The sample size was calculated based on a preliminary study and the sample size calculation (G-power 3.1.9; Heine Heinrich University, Düsseldorf, German) revealed a required sample of 242 participants. A total of 250 patients were enrolled in this study, taking into account the potential for dropouts and test failure rate. All participants signed a written consent form before being included. In case of cardiac arrest patients, if the consent could not be obtained from the patients, it was obtained from the next-of-kin.
2.3.2. Blood sample analyzers The capillary blood samples were analyzed using Enterprise Pointof-care (epoc® Blood Analysis System, Alere, USA) system and 13 pa2+ rameters (pH, pCO2, pO2, HCO− , lactate, Na+, K+, Cl−, glucose, 3 , Ca Hb, Hct, and creatinine) were measured. The epoc system is a portable device that can examine blood gases, electrolytes, and metabolites using arterial, venous, and capillary blood samples. This system is a handheld device with a test cartridge containing the sensors. The parameters including pH, pCO2, Ca2+, Na+, K+, and Cl− are measured by potentiometry using a membrane-covered sensing electrode. The parameters including pO2, lactate, glucose, and creatinine are measured by amperometry. In addition, Hct is a measured parameter, whereas Hb and HCO–3 are calculated parameters. These values are shown in approximately 195 s (cartridge calibration time of 165 s and an analysis time of 30 s) on the display of the device. 2+ The reference parameters (pH, pCO2, pO2, HCO− , and lactate) 3 , Ca of the arterial blood samples were analyzed by an arterial blood gas analyzer (pHOx®Ultra, Nova Biomedical, USA). The reference central laboratory testing of the venous blood samples was performed on a clinical chemistry analyzer (AU5800, Beckman Coulter Inc., USA) for the electrolytes (Na+, K+, Cl−, glucose, and creatinine) and on a cellular analyzer (DxH2401, Beckman Coulter Inc., USA) for Hb and Hct. 2.4. Statistical analysis The parameters of the capillary blood samples obtained from the epoc system were compared with those obtained from the reference analyzers of a central laboratory and an equivalence test was performed. 2+ The parameters including pH, pCO2, pO2, HCO− , lactate, Na+, 3 , Ca
Table 2 Statistics for comparison of the epoc system and reference analyzers Parameters pH pCO2, mmHg pO2, mmHg HCO3, mmol/L Ionized calcium, mmol/L Lactate, mmol/L Sodium, mEq/L Potassium, mEq/L Chloride, mEq/L Glucose, mg/dL Hemoglobin, g/dL Hematocrit, % Creatinine, mg/dL
Reference value, mean (SD) b
7.378 (0.156) 33.27 (15.01)b 82.23 (26.08)b 19.27 (5.39)b 1.21 (0.08)b 3.27 (3.23) 136.21 (6.60)e 4.42 (1.18)e 102.94 (6.31)e 188.24 (156.34)e 11.98 (2.92) 35.96 (8.75)e 1.44 (1.57)e
epoc, mean (SD) 7.351 (0.171) 39.32 (16.18) 58.85 (12.49) 20.97 (5.85) 1.13 (0.02) 3.59 (3.23) 139.03 (6.73) 4.73 (1.19) 106.43 (6.30) 181.40 (148.62) 11.80 (2.96) 34.83 (8.74) 1.44 (1.53)
Mean difference (95% CI) 0.027 (0.018, 0.035) −6.06 (−6.75, −5.36) 23.38 (20.042, 26.722) −1.69 (−1.927, −1.461) 0.077 (0.064, 0.089) −0.324 (−0.406, −0.241) −2.82 (−3.104, −2.532) −0.31 (−0.365, −0.257) −3.50 (−4.062, −2.934) 6.84 (4.317, 9.354) 0.17 (0.093, 0.257) 1.13 (0.864, 1.391) 0.005 (−0.025, 0.036)
SD: standard deviation, CI: confidence interval, LOA: limit of agreement. a LOA of difference = ±1.96·SD. b Analyzed by using the arterial blood gas analyzer (pHOx®Ultra, Nova biomedical, USA). c Defined based on the Clinical Laboratory Improvement Amendments (CLIA) guidelines [12,13]. d Defined as 15% of the reference mean value based on alternative criteria of the CLIA guidelines [12,13]. e Analyzed by using the venous blood analyzer (AU5800 and DxH2401, Beckman Coulter Inc. USA).
LOA of differencea ±0.13 ±10.46 ±50.49 ±3.53 ±0.18 ±1.25 ±4.33 ±0.82 ±8.53 ±38.07 ±1.24 ±3.98 ±0.46
Agreement target c
±0.04 ±8%c ±15%d ±15%d ±15%d ±15%d ±4c ±0.5c ±5%c ±10%c ±7%c ±6%c ±0.3 or ±15%c
Results within target values 77.06% 16.90% 25.54% 82.30% 88.74% 75.76% 84.40% 81.82% 70.13% 80.09% 80.95% 60.60% 89.20%
H. Shin et al. / American Journal of Emergency Medicine xxx (xxxx) xxx
3
Fig. 1. Bland-Altman plot between epoc system and the arterial blood gas analyzer.
K+, Cl−, glucose, Hb, Hct, and creatinine were analyzed for agreement. We defined the agreement targets for pH, pCO2, Na+, K+, Cl−, glucose, Hb, Hct, and creatinine based on the Clinical Laboratory Improvement Amendments (CLIA) guidelines [12,13]. The agreement targets for 2+ pO2, HCO− , and lactate were defined as 15% of the mean value de3 , Ca termined by the reference analyzers based on the alternative criteria of the CLIA guidelines. The parameters analyzed by the epoc system were considered to be equivalent to those analyzed by the reference analyzers when the 95% confidence interval of the mean difference between epoc system and reference analyzers was included within the range of the agreement target. We also performed a Bland-Altman
plot analysis to identify the agreement between epoc system and reference analyzers. The Bland-Altman method is a simple, graphical method to analyze the agreement between two different measurement techniques by analyzing the bias between the mean difference and constructing limits of agreements. The graph displays three horizontal lines which are drawn at the mean difference and limits of agreement (±1.96 times the standard deviation of differences). The scatter diagram of differences between the two assays is plotted against the average of the two, which defines the interval of agreements. Further, it shows the trend of differences between these two methods according to the average value of the two measurements [14]. When the
4
H. Shin et al. / American Journal of Emergency Medicine xxx (xxxx) xxx
Fig. 2. Bland-Altman plot between epoc system and the venous blood analyzers.
H. Shin et al. / American Journal of Emergency Medicine xxx (xxxx) xxx
5
Table 3 Statistics for comparison of the epoc system and reference analyzers in cardiac arrest group Parameters
Reference value, mean (SD)
epoc, mean (SD)
Mean difference (95% CI)
LOA of differencea
Agreement target
Results within target values
pH pCO2, mmHg pO2, mmHg HCO3, mmol/L Ionized calcium, mmol/L Lactate, mmol/L Sodium, mEq/L Potassium, mEq/L Chloride, mEq/L Glucose, mg/dL Hemoglobin, g/dL Hematocrit, % Creatinine, mg/dL
7.012 (0.144)b 52.84 (25.82)b 74.67 (60.23)b 15.19 (7.58)b 1.24 (0.12)b 3.15 (3.82) 138.43 (10.93)e 5.40 (1.72)e 103.40 (8.19)e 249.07 (213.97)e 11.54 (3.44) 36.01 (10.17)e 1.65 (1.32)e
6.993 (0.148) 69.00 (34.73) 52.09 (25.05) 15.39 (6.90) 1.16 (0.20) 3.58 (3.53) 140.13 (11.21) 5.69 (1.71) 107.36 (8.29) 2354.93 (197.99) 11.55 (3.59) 34.87 (10.27) 1.78 (1.42)
0.02 (0.004, 0.036) −16.16 (−25.07, −7.26) 22.58 (2.708, 42.446) −0.2 (−1.037, 0.637) 0.08 (0.027, 0.136) −0.434 (−0.737, −0.131) −1.7 (−2.956, −0.444) −0.29 (−0.473, −0.107) −3.97 (−6.456, −1.477) 14.13 (4.072, 24.195) −0.01 (−0.28, 0.26) 1.15 (0.398, 1.895) −0.13 (−0.257, −0.006)
±0.09 ±46.74 ±104.29 ±4.39 ±0.29 ±1.59 ±6.59 ±0.96 ±13.07 ±52.81 ±1.42 ±3.93 ±0.66
±0.04c ±8%c ±15%d ±15%d ±15%d ±15%d ±4c ±0.5c ±5%c ±10%c ±7%c ±6%c ±0.3or ± 15%c
80.00% 20.00% 16.70% 70.00% 66.67% 70.00% 83.33% 83.33% 50.00% 73.33% 80.00% 63.33% 83.33%
SD: standard deviation, CI: confidence interval, LOA: limit of agreement. a LOA of difference = ±1.96·SD. b Analyzed by using the arterial blood gas analyzer (pHOx®Ultra, Nova biomedical, USA). c Defined based on the Clinical Laboratory Improvement Amendments (CLIA) guidelines [12,13]. d Defined as 15% of the reference mean value based on alternative criteria of the CLIA guidelines [12,13]. e Analyzed by using the venous blood analyzer (AU5800 and DxH2401, Beckman Coulter Inc. USA).
difference between the values of epoc system and reference analyzers was at least 80% within the range of agreement target, we regarded the parameters analyzed by the epoc system as being clinically acceptable. We performed a linear regression to identify the correlation between the results of the epoc system and reference analyzers if the parameters were not equivalent or clinically acceptable between the two methods. Likewise, we conducted a subgroup analysis for the cardiac arrest group. 3. Results Of the 250 participants, 19 were excluded due to test error (test failure rate: 7.6%) and 231 were enrolled. Out of the 19, 17 test failures were due to test card failure (mainly internal quality control error) and only two were due to the sample injection errors (insufficient blood volume and fast sample injection). The characteristics of patients enrolled are shown in Table 1. A total of 13 parameters (pH, pCO2, pO2, 2+ HCO− , lactate, Na+, K+, Cl−, glucose, Hb, Hct, and creatinine) 3 , Ca were compared. 3.1. Comparison of the epoc system and reference analyzers for hypotensive patients
becomes wider as the average increased when the average value is more than approximately 1.5 mg/dL) (Figs. 1 and 2). 3.2. Subgroup analysis: comparison of the epoc system and reference analyzers for the cardiac arrest group In the subgroup analysis for the cardiac arrest group, two additional parameters (lactate and Cl−) measured by the epoc system were not equivalent to the reference values (Table 3). As a result, a total of nine parameters by POC analysis were equivalent to the reference values for this group. The mean differences between the two methods were also inconsequential and the limit of agreement was narrow along with total patients' group. Of these nine parameters, five (pH, Na+, K+, Hb, and creatinine) had a clinically acceptable agreement between the two methods. Of the four parameters (pCO2, pO2, lactate, Cl−) that were not equivalent, only lactate measured by epoc system was highly correlated with the reference values. (lactate: R2 = 0.958, y = −0.653 + 1.061·x; pCO2: R2 = 0.529, y = 15.535 + 0.541·x; pO2: R2 = 0.223, y = 15.565 + 1.135·x; Cl: R2 = 0.453, y = 32.0 + 0.665·x) The regression model accounts for 95.8% of the variance for lactate. The trends in the Bland-Altman plot for the cardiac arrest group are similar to those for total patients' group (Figs. 3 and 4). 4. Discussion
The mean values of each parameter measured by each method (epoc system and reference analyzers) and the mean difference with 95% confidence interval (CI) are shown in Table 2. With the exception of two parameters (pCO2, pO2), the 95% CI of the mean difference between the epoc system and the reference analyzers were included within the range of the agreement target. The 11 parameters measured by the epoc system were equivalent to the reference values. The mean differences between the two methods were inconsequential and the limit of agreement was narrow. Moreover, of these 11 parameters, seven 2+ (HCO− , Na+, K+, glucose, Hb, and creatinine) showed that the re3 , Ca sults were within the target values for at least 80% (185 of 231 tests) of the patients. Thus, these seven parameters were regarded as having a clinically acceptable agreement between the two methods. Of the remaining two parameters (pCO2 and pO2), which neither had equivalence nor acceptability with the reference analyzers, the pCO2 measured by epoc system was highly correlated with the reference values, whereas pO2 was not. (pCO2: R2 = 0.891, y = −1.163 + 0.876·x; pO2: R2 = 0.07, y = 49.7 + 0.552·x). The regression model accounts for 89.1% of the variance for pCO2. The scatter plots of some parameters showed trends (negative trend: Ca2+, positive trend: pO2 and glucose) and inconsistent variability (the scatter plot for creatinine
2+ Most of the parameters (pH, pCO2, HCO− , lactate, Na+, K+, 3 , Ca Cl−, glucose, Hb, and Hct), except for pO2, measured by the epoc system were equivalent to or correlated with those measured by the reference methods in our previous study. The capillary blood-based epoc system might have some clinical utility, however, the capillary blood sample might be less accurate due to ischemia and vasoconstriction in patients who are in cardiac arrest or hypotension. Given that these hemodynamically unstable patients were not sufficiently represented in our previous study, further study on these patients' group was needed [3]. In this study, the result was similar to the previous study; except for pO2, 12 of the 13 parameters (we compared an additional parameter of creatinine in this study) were equivalent to or correlated with the reference values. Further, except for two parameters (lactate and creatinine), a similar result was shown in the cardiac arrest group, even though the sample size was small (n = 30). Previous studies have reported that various confounding factors might interfere with the analysis of POC glucometry in critically ill patients [15,16]. With regard to shock, several possible mechanisms may explain the discrepancies in blood glucose measurements obtained by different methods [17]. Peripheral vasoconstriction in hypoperfusion
6
H. Shin et al. / American Journal of Emergency Medicine xxx (xxxx) xxx
Fig. 3. Bland-Altman plot between epoc system and the arterial blood gas analyzer in the cardiac-arrested group.
states can lead to increased glucose extraction by tissues due to low capillary flow and increased glucose transit time, resulting in altered glucose measurements in the capillary blood [18]. Moreover, Garingarao et al. concluded that POC blood glucose measurements were significantly less accurate in the hypotensive subgroup of ICU patients as compared to the normotensive group. However, most patients included in those studies received a vasopressor such as dopamine or norepinephrine, while this study included patients who had not yet received those drugs. In addition, we believe that the peripheral vasoconstriction of the patients in our study group was less prominent than those in previous study groups because the period of exposure to the stress situation in our patients was markedly shorter than in previous studies. All patients included in
previous studies had been in the ICU for a certain period of time, while our study design was in the ED setting. There have been several previous studies comparing POC blood gas analysis and the central laboratory analysis [19,20]. However, to the best of our knowledge, there has been no study comparing the POC analysis using the peripheral capillary blood and reference analysis in hypotensive patients. This is the first study to compare peripheral capillary blood and venous blood in hemodynamically unstable patients, including not only glucose but also other significant parameters, such as pH, K+, lactate, Hb, creatinine, etc. that may be needed in an emergency situation. Therefore, we do not think that the conclusion of previous studies, in which the peripheral blood tests (skin puncture) were inaccurate in hypotensive patients, is applicable
H. Shin et al. / American Journal of Emergency Medicine xxx (xxxx) xxx
Fig. 4. Bland-Altman plot between epoc system and the venous blood analyzers in the cardiac-arrested group.
7
8
H. Shin et al. / American Journal of Emergency Medicine xxx (xxxx) xxx
to the acute patients in ED. Most parameters of the capillary blood measured by the epoc system were equivalent to or correlated with that measured by the reference analyzers in hypotensive patients in ED. Elevated serum lactate is an early indicator of shock and is associated with increased mortality [21]. The lactate assay remains a clinically useful test that can aid a clinician to identify underlying hypoperfusion in need of immediate treatment or an etiology not readily apparent on initial evaluation [22]. Hyperkalemia, one of the reversible causes of outof-hospital cardiac arrest (OHCA), can be identified in the field, so the patients can immediately receive a reversible drug [23]. Although hemoglobin is not an indicator of acute blood loss, patients with hypovolemic shock should be carefully monitored by checking base excess or lactate together [24,25]. Laboratory tests may help early evaluation of reversible causes in hemodynamically unstable patients with undifferentiated hypotension or shock to identify the cause. In this study, we have confirmed that these parameters measured by epoc system using the capillary blood could be useful and applicable even in the hemodynamically unstable patients. We added one parameter (creatinine) in this study. Although the incidence of contrast-induced nephropathy without confirming the creatinine level was low, when computed tomography with intravenous contrast was used to assess time-sensitive emergency patients, such as acute stroke and trauma patients [26,27], creatinine could be a valuable parameter for hemodynamically unstable patients [2830]. You et al. suggested that the results of POC creatinine analysis correlate well with the values obtained from reference laboratory tests [31]. However, they compared the POC creatinine levels using whole venous blood and central laboratory serum creatinine levels and did not use the capillary blood for POC analyzer. In addition, it was unclear whether the participants included in that study were hemodynamically unstable or hypotensive. In our study, creatinine is regarded as having a clinically acceptable agreement between the epoc system using capillary blood and the reference analyzers in hypotensive patients, including the cardiac arrest group. Thus, POC creatinine analysis using capillary blood can be identified in a hemodynamically unstable patient, which may be beneficial in the management of critically ill patients. The epoc system is an automated, self-monitoring system and may show some types of malfunction. The analyzer or test card can generate errors associated with failure of calibration and quality control (test card failure). An examiner can also contribute to test failure (sample injection error: insufficient volume of sample or fast sample injection). According to a previous study, the complete card failure rate was 13.0% for the epoc system [32]. In our previous study [3], the total failure rate was 10.8% (test card failure rate: 7.3%, sample injection error 3.5%), which was lower than that in Stotler's study. We further reduced the failure rate (7.6%) as compared to our previous failure rate (10.8%). The failure rate originated from the analyzer or test card themselves (test card failure rate) was not significantly different between the two studies (7.3% in our previous study vs. 6.8% in this study). However, the failure rate due to the examiner was markedly reduced in this study (3.5% vs. 0.8%). It can be a little difficult for an inexperienced examiner to obtain a sufficient volume of sample from the peripheral skin (90 μL) and inject the sample to the card at an appropriate rate. Our two examiners (emergency medical technician) had an experience of conducting more than 100 examinations prior to this study, which seems to have contributed to minimizing sample injection errors. Additionally, the cost-effectiveness of the epoc system should be considered. The cost per test is usually high for the epoc system due to high reagent/cartridge costs. It may be a barrier to the usage of the epoc system for patients. A prospective study with a large number of patients will be needed to compare the diagnostic utility and costeffectiveness of the epoc system. Further study in the prehospital environment may be warranted.
5. Limitation Although we concluded that the parameters, including pH, HCO− 3 , Ca2+, lactate, Na+, K+, Cl−, glucose, Hb, Hct, and creatinine, measured by the epoc system were equivalent to those measured by the reference analyzers, the reliability of these results can be compromised because the test was conducted by an emergency physician whose primary expertise is not laboratory work. 6. Conclusion The parameters measured by the epoc system using the capillary 2+ blood, which include pH, pCO2, pO2, HCO− , lactate, Na+, K+, 3 , Ca − Cl , glucose, Hb, Hct, and creatinine, were equivalent to or correlated with those measured by the reference analyzers in hypotensive patients. Thus, POC analysis using capillary blood may be useful in identifying the reversible causes for the life-threatening conditions in an emergency situation. References [1] Prause G, Ratzenhofer-Komenda B, Offner A, Lauda P, Voit H, Pojer H, et al. Prehospital point of care testing of blood gases and electrolytes - an evaluation of IRMA. Crit Care. 1997;1:79–83. https://doi.org/10.1186/cc108. [2] 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. Scand J Trauma Resusc Emerg Med. 2015;23:54. https://doi.org/10.1186/s13049-015-0134-y. [3] Kim C, Kim H. Emergency medical technician-performed point-of-care blood analysis using the capillary blood obtained from skin puncture. Am J Emerg Med. 2018; 36:1215–21. https://doi.org/10.1016/j.ajem.2017.12.025. [4] Gilbert HC, Szokol JW. Point of care technologies. Int Anesthesiol Clin. 2004;42: 73–94. [5] Peterson MC, Syndergaard T, Bowler J, Doxey R. A systematic review of factors predicting door to balloon time in ST-segment elevation myocardial infarction treated with percutaneous intervention. Int J Cardiol. 2012;157:8–23. https://doi. org/10.1016/j.ijcard.2011.06.042. [6] Foss KT, Subhi Y, Aagaard R, Bessmann EL, Bøtker MT, Graumann O, et al. Developing an emergency ultrasound app – a collaborative project between clinicians from different universities. Scand J Trauma Resusc and Emerg Med. 2015;23:47. https://doi. org/10.1186/s13049-015-0130-2. [7] Nelson BP, Sanghvi A. Out of hospital point of care ultrasound: current use models and future directions. Eur J Trauma Emerg Surg. 2016;42:139–50. https://doi.org/ 10.1007/s00068-015-0494-z. [8] Stengaard C, Sørensen JT, Ladefoged SA, Christensen EF, Lassen JF, Bøtker 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: 1361–6. https://doi.org/10.1016/j.amjcard.2013.06.026. [9] Guyette FX, Meier EN, Newgard C, McKnight B, Daya M, Bulger EM, et al. A comparison of prehospital lactate and systolic blood pressure for predicting the need for resuscitative care in trauma transported by ground. J Trauma Acute Care Surg. 2015; 78:600–6. https://doi.org/10.1097/ta.0000000000000549. [10] Callaway CW, Donnino MW, Fink EL, Geocadin RG, Golan E, Kern KB, et al. Part 8: post-cardiac arrest care: 2015 American Heart Association guidelines update for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation. 2015; 132:S465–82. https://doi.org/10.1161/cir.0000000000000262. [11] WHO. Guidelines approved by the guidelines review committee. WHO guidelines on drawing blood: Best practices in phlebotomy. Geneva: World Health Organization; 2010. [12] Bachner P, Hamlin W. Federal regulation of clinical laboratories and the clinical laboratory improvement amendments of 1988–part I. Clin Lab Med. 1993;13:739–52. https://doi.org/10.1016/s0272-2712(18)30438-4. [13] 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. . https://www.westgard.com/ biodatabase1.htm. Accessed date: 25 April 2019. [14] Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1:307–10. https://doi.org/10.1016/ s0140-6736(86)90837-8. [15] Dungan K, Chapman J, Braithwaite SS, Buse J. Glucose measurement: confounding issues in setting targets for inpatient management. Diabetes Care. 2007;30:403–9. https://doi.org/10.2337/dc06-1679. [16] Heinemann L. Quality of glucose measurement with blood glucose meters at the point-of-care: relevance of interfering factors. Diabetes Technol Ther. 2010;12: 847–57. https://doi.org/10.1089/dia.2010.0076. [17] Pati-An Garingarao CJ, Buenaluz-Sedurante M, Jimeno CA. Accuracy of point-of-care blood glucose measurements in critically ill patients in shock. J Diabetes Sci Technol. 2014;8:937–44. https://doi.org/10.1177/1932296814538608. [18] Juneja D, Pandey R, Singh O. Comparison between arterial and capillary blood glucose monitoring in patients with shock. Eur J Intern Med. 2011;22:241–4. https:// doi.org/10.1016/j.ejim.2011.01.004.
H. Shin et al. / American Journal of Emergency Medicine xxx (xxxx) xxx [19] Dashevsky M, Bernstein SL, Barsky CL, Taylor RA. Agreement between serum assays performed in ED point-of-care and hospital central laboratories. West J Emerg Med. 2017;18:403–9. https://doi.org/10.5811/westjem.2017.1.30532. [20] Bloom BM, Connor H, Benton S, Harris T. A comparison of measurements of sodium, potassium, haemoglobin and creatinine between an emergency department-based point-of-care machine and the hospital laboratory. Eur J Emerg Med. 2014;21: 310–3. https://doi.org/10.1097/mej.0000000000000076 Aug. [21] Bou Chebl R, El Khuri C, Shami A, Rajha E, Faris N, Bachir R, et al. Serum lactate is an independent predictor of hospital mortality in critically ill patients in the emergency department: a retrospective study. Scand J Trauma Resusc Emerg Med. 2017;25:69. https://doi.org/10.1186/s13049-017-0415-8. [22] Andersen LW, Mackenhauer J, Roberts JC, Berg KM, Cocchi MN, Donnino MW. Etiology and therapeutic approach to elevated lactate levels. Mayo Clin Proc. 2013;88: 1127–40. https://doi.org/10.1016/j.mayocp.2013.06.012. [23] Truhlář A, Deakin CD, Soar J, Khalifa GEA, Alfonzo A, Bierens JJLM, et al. Cardiac arrest in special circumstances: section 4 of European resuscitation council guidelines for resuscitation 2015. Resuscitation. 2015;95:148–201. https://doi.org/10.1016/j.resuscitation.2010.08.015. [24] EH1 Kincaid, Miller PR, Meredith JW, Rahman N, Chang MC. Elevated arterial base deficit in trauma patients: a marker of impaired oxygen utilization. J Am Coll Surg. 1998;187:384–92. https://doi.org/10.1016/s1072-7515(98)00202-6. [25] Moomey CB, Melton SM, Croce MA, Fabian TC, Proctor KG. Prognostic value of blood lactate, base deficit, and oxygen-derived variables in an LD50 model of penetrating trauma. Crit Care Med. 1999;27:154–61. https://doi.org/10.1097/00003246199901000-00044.
9
[26] Ang TE, Bivard A, Levi C, Ma H, Hsu CY, Campbell B, et al. Multi-modal CT in acute stroke: wait for a serum creatinine before giving intravenous contrast? No! Int J Stroke. 2015;10:1014–7. https://doi.org/10.1111/ijs.12605 Oct. [27] Colling KP, Irwin ED, Byrnes MC, Reicks P, Dellich WA, Reicks K, et al. Computed tomography scans with intravenous contrast: low incidence of contrast-induced nephropathy in blunt trauma patients. J Trauma Acute Care Surg. 2014;77:226–30. https://doi.org/10.1097/ta.0000000000000336 Aug. [28] Hoste EA, Clermont G, Kersten A, Venkataraman R, Angus DC, De Bacquer D, et al. RIFLE criteria for acute kidney injury are associated with hospital mortality in critically ill patients: a cohort analysis. Crit Care. 2006;10:R73. https://doi.org/10.1186/ cc4915. [29] Hinson JS, Al Jalbout N, Ehmann MR, Klein EY. Acute kidney injury following contrast media administration in the septic patient: a retrospective propensity-matched analysis. J Crit Care. 2019;51:111–6. https://doi.org/10.1016/j.jcrc.2019.02.003. [30] Shah K, Kilian B, Hsieh WJ, Kyrillou E, Hedge V, Newman DH. Can urine dipstick predict an elevated serum creatinine? Am J Emerg Med. 2010;28:613–6. https://doi. org/10.1016/j.ajem.2009.02.020. [31] You JS, Chung YE, Park JW, Lee W, Lee HJ, Chung TN, et al. The usefulness of rapid point-of-care creatinine testing for the prevention of contrast-induced nephropathy in the emergency department. Emerg Med J. 2013;30:555–8. https://doi.org/10. 1136/emermed-2012-201285. [32] 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;140:715–20. https://doi.org/10.1309/AJCP7QB3QQIBZPEK.
Contents lists available at ScienceDirect
American Journal of Emergency Medicine journal homepage: www.elsevier.com/locate/ajem
Point-of-care blood analysis of hypotensive patients in the emergency department☆ Hyungoo Shin, MD, PhD 1, Inhye Lee, MD 1, Changsun Kim, MD, PhD ⁎, Hyuk Joong Choi, MD, PhD Department of Emergency Medicine, College of Medicine, Hanyang University Guri Hospital, Guri, Republic of Korea
a r t i c l e
i n f o
Article history: Received 26 April 2019 Received in revised form 15 July 2019 Accepted 21 July 2019 Available online xxxx Keywords: Point-of-care systems Capillary blood Emergency medical services Hypotension
a b s t r a c t Objective: The aim of this study is to compare a point-of-care (POC) analysis, Enterprise POC (epoc), using the capillary blood obtained from skin puncture with conventional laboratory tests using arterial and venous blood in hypotensive patients. Methods: This study was conducted at the emergency department of a tertiary care hospital between June and November 2018. 231 hypotensive patients were enrolled. Three types of blood samples (capillary blood from skin puncture and arterial and venous blood from blood vessel puncture) were collected and analyzed. We compared a total of 13 parameters (pH, pCO2, pO2, HCO3−, Ca2+, lactate, Na+, K+, Cl−, glucose, Hb, Hct, and creatinine) between the POC analysis and reference analyzers by performing the equivalence test and Bland-Altman plot analysis. Results: In hypotensive patients, with the exception of two parameters (pCO2, pO2), the pH, HCO3−, Ca2+, lactate, Na+, K+, Cl−, glucose, Hb, Hct, and creatinine parameters measured by the POC analysis were equivalent to or correlated with the reference values. In the patients with cardiac arrest group, nine parameters (pH, HCO3−, Ca2+, Na+, K+, glucose, Hb, Hct, and creatinine) analyzed by the epoc system were equivalent to the reference values. Conclusion: Most parameters, except pO2, measured by the epoc system using the capillary blood in hypotensive patients were equivalent to or correlated with those measured by the reference analyzers. © 2019 Published by Elsevier Inc.
1. Introduction In hemodynamically unstable patients, especially those with cardiac arrest, serum biomarkers, such as electrolytes (e.g. hyperkalemia), acidbase imbalance (metabolic acidosis), and point-of-care (POC) blood sampling can help recognize reversible causes at an early stage [1,2]. Thus, the interest in these analyses in the pre-hospital stage has increased. With the evolution in technology over the last few decades, POC analysis has become increasingly available in many clinical settings, including the prehospital environment [3-9]. We used a skin punctured capillary blood-based POC analysis in a previous study and showed 2+ that parameters including pH, pCO2, HCO− , lactate, Na+, K+, 3 , Ca Cl−, glucose, hemoglobin (Hb), and hematocrit (Hct), except for pO2, measured by POC analysis were equivalent to or correlated with those
measured by conventional laboratory tests [3]. Although a few critically ill patients were enrolled, most were hemodynamically stable in that study. Thus, we could not confirm that the POC analysis using the capillary blood of hypotensive patients were equivalent to the conventional laboratory tests, because of reduced peripheral blood circulation which could affect the results. The aim of this study was to investigate whether the results of the POC analysis using capillary blood samples were equivalent to the conventional laboratory tests in case of hypotensive patients. We conducted this study at an urban, tertiary care, teaching hospital as a primary step to test the POC analysis in prehospital settings anticipating that eventually, the reversible symptoms in the hemodynamically unstable patients can be corrected ahead of arrival at the emergency department (ED). 2. Methods and materials
☆ This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. ⁎ Corresponding author at: Department of Emergency Medicine, College of Medicine, Hanyang University Guri Hospital, 153, Gyeongchun-ro, Guri-si, Gyeonggi-do 11923, Republic of Korea. E-mail address: fl[email protected] (C. Kim). 1 These authors contributed equally to this work.
https://doi.org/10.1016/j.ajem.2019.158363 0735-6757/© 2019 Published by Elsevier Inc.
2.1. Study design This prospective experimental study was performed from June to October 2018. Evaluation of the POC analysis was conducted at an ED of an urban, tertiary care, teaching hospital. The local ethics committee
2
H. Shin et al. / American Journal of Emergency Medicine xxx (xxxx) xxx
2.3. Equipment and material
Table 1 Characteristics of the study population Characteristics
Values
Male, n (%) Age, year (SD) Diagnosis, n (%) Cardiac arrest Cardiogenic Acute coronary syndrome Congestive heart failure Cardiomyopathy Hypovolemic Gastrointestinal bleeding Trauma related blood loss Burn Septic Pneumonia Abdominal infection Kidney infection etc Distributive Anaphylaxis Salicylate toxicity Neurogenic Total
149 (64.5) 68.2 (16.6)
2.3.1. Blood collection We obtained three types of blood samples simultaneously after the hypotensive patients were presented at the ED. 1) The EMTs working in the ED obtained the capillary blood samples from the fingertip of the patients upon arrival to the ED. 2) The arterial blood samples from the radial artery of the patients were obtained with a heparinized syringe (BD Preset™, Becton Dickinson and Company, UK) by the emergency physician and analyzed. 3) The venous blood samples were obtained by trained ED nursing staff from the vein of the forearm and were collected in a collection tube (BD Vacutainer) for laboratory measurement. We collected each blood sample in accordance with the World Health Organization guidelines [11] and analyzed them in the devices described below.
30 (13.0) 46 (19.9) 27 (11.7) 17 (7.4) 2 (0.9) 67 (29.0) 52 (22.5) 14 (6.1) 1 (0.4) 76 (32.9) 41 (17.7) 23 (10.0) 9 (3.9) 3 (1.3) 12 (5.2) 4 (1.7) 1 (0.4) 7 (3.0) 231 (100)
approved this study in May 2018 (GURI 2018–01–004-004) and the study protocol was registered with the Clinical Research Information Service (cris.nih.go.kr: KCT0003159). 2.2. Participants We included hypotensive patients who were presented at the ED; both venous and arterial blood samples had been taken. Patients with a systolic blood pressure of less than 90 mmHg or mean arterial pressure of less than 65 mmHg were included [10]. Patients with cardiac arrest at ED arrival were also included. We excluded those patients who were already receiving crystalloid with electrolytes since it would affect the results. Patients under the age of 18 years were also excluded. The sample size was calculated based on a preliminary study and the sample size calculation (G-power 3.1.9; Heine Heinrich University, Düsseldorf, German) revealed a required sample of 242 participants. A total of 250 patients were enrolled in this study, taking into account the potential for dropouts and test failure rate. All participants signed a written consent form before being included. In case of cardiac arrest patients, if the consent could not be obtained from the patients, it was obtained from the next-of-kin.
2.3.2. Blood sample analyzers The capillary blood samples were analyzed using Enterprise Pointof-care (epoc® Blood Analysis System, Alere, USA) system and 13 pa2+ rameters (pH, pCO2, pO2, HCO− , lactate, Na+, K+, Cl−, glucose, 3 , Ca Hb, Hct, and creatinine) were measured. The epoc system is a portable device that can examine blood gases, electrolytes, and metabolites using arterial, venous, and capillary blood samples. This system is a handheld device with a test cartridge containing the sensors. The parameters including pH, pCO2, Ca2+, Na+, K+, and Cl− are measured by potentiometry using a membrane-covered sensing electrode. The parameters including pO2, lactate, glucose, and creatinine are measured by amperometry. In addition, Hct is a measured parameter, whereas Hb and HCO–3 are calculated parameters. These values are shown in approximately 195 s (cartridge calibration time of 165 s and an analysis time of 30 s) on the display of the device. 2+ The reference parameters (pH, pCO2, pO2, HCO− , and lactate) 3 , Ca of the arterial blood samples were analyzed by an arterial blood gas analyzer (pHOx®Ultra, Nova Biomedical, USA). The reference central laboratory testing of the venous blood samples was performed on a clinical chemistry analyzer (AU5800, Beckman Coulter Inc., USA) for the electrolytes (Na+, K+, Cl−, glucose, and creatinine) and on a cellular analyzer (DxH2401, Beckman Coulter Inc., USA) for Hb and Hct. 2.4. Statistical analysis The parameters of the capillary blood samples obtained from the epoc system were compared with those obtained from the reference analyzers of a central laboratory and an equivalence test was performed. 2+ The parameters including pH, pCO2, pO2, HCO− , lactate, Na+, 3 , Ca
Table 2 Statistics for comparison of the epoc system and reference analyzers Parameters pH pCO2, mmHg pO2, mmHg HCO3, mmol/L Ionized calcium, mmol/L Lactate, mmol/L Sodium, mEq/L Potassium, mEq/L Chloride, mEq/L Glucose, mg/dL Hemoglobin, g/dL Hematocrit, % Creatinine, mg/dL
Reference value, mean (SD) b
7.378 (0.156) 33.27 (15.01)b 82.23 (26.08)b 19.27 (5.39)b 1.21 (0.08)b 3.27 (3.23) 136.21 (6.60)e 4.42 (1.18)e 102.94 (6.31)e 188.24 (156.34)e 11.98 (2.92) 35.96 (8.75)e 1.44 (1.57)e
epoc, mean (SD) 7.351 (0.171) 39.32 (16.18) 58.85 (12.49) 20.97 (5.85) 1.13 (0.02) 3.59 (3.23) 139.03 (6.73) 4.73 (1.19) 106.43 (6.30) 181.40 (148.62) 11.80 (2.96) 34.83 (8.74) 1.44 (1.53)
Mean difference (95% CI) 0.027 (0.018, 0.035) −6.06 (−6.75, −5.36) 23.38 (20.042, 26.722) −1.69 (−1.927, −1.461) 0.077 (0.064, 0.089) −0.324 (−0.406, −0.241) −2.82 (−3.104, −2.532) −0.31 (−0.365, −0.257) −3.50 (−4.062, −2.934) 6.84 (4.317, 9.354) 0.17 (0.093, 0.257) 1.13 (0.864, 1.391) 0.005 (−0.025, 0.036)
SD: standard deviation, CI: confidence interval, LOA: limit of agreement. a LOA of difference = ±1.96·SD. b Analyzed by using the arterial blood gas analyzer (pHOx®Ultra, Nova biomedical, USA). c Defined based on the Clinical Laboratory Improvement Amendments (CLIA) guidelines [12,13]. d Defined as 15% of the reference mean value based on alternative criteria of the CLIA guidelines [12,13]. e Analyzed by using the venous blood analyzer (AU5800 and DxH2401, Beckman Coulter Inc. USA).
LOA of differencea ±0.13 ±10.46 ±50.49 ±3.53 ±0.18 ±1.25 ±4.33 ±0.82 ±8.53 ±38.07 ±1.24 ±3.98 ±0.46
Agreement target c
±0.04 ±8%c ±15%d ±15%d ±15%d ±15%d ±4c ±0.5c ±5%c ±10%c ±7%c ±6%c ±0.3 or ±15%c
Results within target values 77.06% 16.90% 25.54% 82.30% 88.74% 75.76% 84.40% 81.82% 70.13% 80.09% 80.95% 60.60% 89.20%
H. Shin et al. / American Journal of Emergency Medicine xxx (xxxx) xxx
3
Fig. 1. Bland-Altman plot between epoc system and the arterial blood gas analyzer.
K+, Cl−, glucose, Hb, Hct, and creatinine were analyzed for agreement. We defined the agreement targets for pH, pCO2, Na+, K+, Cl−, glucose, Hb, Hct, and creatinine based on the Clinical Laboratory Improvement Amendments (CLIA) guidelines [12,13]. The agreement targets for 2+ pO2, HCO− , and lactate were defined as 15% of the mean value de3 , Ca termined by the reference analyzers based on the alternative criteria of the CLIA guidelines. The parameters analyzed by the epoc system were considered to be equivalent to those analyzed by the reference analyzers when the 95% confidence interval of the mean difference between epoc system and reference analyzers was included within the range of the agreement target. We also performed a Bland-Altman
plot analysis to identify the agreement between epoc system and reference analyzers. The Bland-Altman method is a simple, graphical method to analyze the agreement between two different measurement techniques by analyzing the bias between the mean difference and constructing limits of agreements. The graph displays three horizontal lines which are drawn at the mean difference and limits of agreement (±1.96 times the standard deviation of differences). The scatter diagram of differences between the two assays is plotted against the average of the two, which defines the interval of agreements. Further, it shows the trend of differences between these two methods according to the average value of the two measurements [14]. When the
4
H. Shin et al. / American Journal of Emergency Medicine xxx (xxxx) xxx
Fig. 2. Bland-Altman plot between epoc system and the venous blood analyzers.
H. Shin et al. / American Journal of Emergency Medicine xxx (xxxx) xxx
5
Table 3 Statistics for comparison of the epoc system and reference analyzers in cardiac arrest group Parameters
Reference value, mean (SD)
epoc, mean (SD)
Mean difference (95% CI)
LOA of differencea
Agreement target
Results within target values
pH pCO2, mmHg pO2, mmHg HCO3, mmol/L Ionized calcium, mmol/L Lactate, mmol/L Sodium, mEq/L Potassium, mEq/L Chloride, mEq/L Glucose, mg/dL Hemoglobin, g/dL Hematocrit, % Creatinine, mg/dL
7.012 (0.144)b 52.84 (25.82)b 74.67 (60.23)b 15.19 (7.58)b 1.24 (0.12)b 3.15 (3.82) 138.43 (10.93)e 5.40 (1.72)e 103.40 (8.19)e 249.07 (213.97)e 11.54 (3.44) 36.01 (10.17)e 1.65 (1.32)e
6.993 (0.148) 69.00 (34.73) 52.09 (25.05) 15.39 (6.90) 1.16 (0.20) 3.58 (3.53) 140.13 (11.21) 5.69 (1.71) 107.36 (8.29) 2354.93 (197.99) 11.55 (3.59) 34.87 (10.27) 1.78 (1.42)
0.02 (0.004, 0.036) −16.16 (−25.07, −7.26) 22.58 (2.708, 42.446) −0.2 (−1.037, 0.637) 0.08 (0.027, 0.136) −0.434 (−0.737, −0.131) −1.7 (−2.956, −0.444) −0.29 (−0.473, −0.107) −3.97 (−6.456, −1.477) 14.13 (4.072, 24.195) −0.01 (−0.28, 0.26) 1.15 (0.398, 1.895) −0.13 (−0.257, −0.006)
±0.09 ±46.74 ±104.29 ±4.39 ±0.29 ±1.59 ±6.59 ±0.96 ±13.07 ±52.81 ±1.42 ±3.93 ±0.66
±0.04c ±8%c ±15%d ±15%d ±15%d ±15%d ±4c ±0.5c ±5%c ±10%c ±7%c ±6%c ±0.3or ± 15%c
80.00% 20.00% 16.70% 70.00% 66.67% 70.00% 83.33% 83.33% 50.00% 73.33% 80.00% 63.33% 83.33%
SD: standard deviation, CI: confidence interval, LOA: limit of agreement. a LOA of difference = ±1.96·SD. b Analyzed by using the arterial blood gas analyzer (pHOx®Ultra, Nova biomedical, USA). c Defined based on the Clinical Laboratory Improvement Amendments (CLIA) guidelines [12,13]. d Defined as 15% of the reference mean value based on alternative criteria of the CLIA guidelines [12,13]. e Analyzed by using the venous blood analyzer (AU5800 and DxH2401, Beckman Coulter Inc. USA).
difference between the values of epoc system and reference analyzers was at least 80% within the range of agreement target, we regarded the parameters analyzed by the epoc system as being clinically acceptable. We performed a linear regression to identify the correlation between the results of the epoc system and reference analyzers if the parameters were not equivalent or clinically acceptable between the two methods. Likewise, we conducted a subgroup analysis for the cardiac arrest group. 3. Results Of the 250 participants, 19 were excluded due to test error (test failure rate: 7.6%) and 231 were enrolled. Out of the 19, 17 test failures were due to test card failure (mainly internal quality control error) and only two were due to the sample injection errors (insufficient blood volume and fast sample injection). The characteristics of patients enrolled are shown in Table 1. A total of 13 parameters (pH, pCO2, pO2, 2+ HCO− , lactate, Na+, K+, Cl−, glucose, Hb, Hct, and creatinine) 3 , Ca were compared. 3.1. Comparison of the epoc system and reference analyzers for hypotensive patients
becomes wider as the average increased when the average value is more than approximately 1.5 mg/dL) (Figs. 1 and 2). 3.2. Subgroup analysis: comparison of the epoc system and reference analyzers for the cardiac arrest group In the subgroup analysis for the cardiac arrest group, two additional parameters (lactate and Cl−) measured by the epoc system were not equivalent to the reference values (Table 3). As a result, a total of nine parameters by POC analysis were equivalent to the reference values for this group. The mean differences between the two methods were also inconsequential and the limit of agreement was narrow along with total patients' group. Of these nine parameters, five (pH, Na+, K+, Hb, and creatinine) had a clinically acceptable agreement between the two methods. Of the four parameters (pCO2, pO2, lactate, Cl−) that were not equivalent, only lactate measured by epoc system was highly correlated with the reference values. (lactate: R2 = 0.958, y = −0.653 + 1.061·x; pCO2: R2 = 0.529, y = 15.535 + 0.541·x; pO2: R2 = 0.223, y = 15.565 + 1.135·x; Cl: R2 = 0.453, y = 32.0 + 0.665·x) The regression model accounts for 95.8% of the variance for lactate. The trends in the Bland-Altman plot for the cardiac arrest group are similar to those for total patients' group (Figs. 3 and 4). 4. Discussion
The mean values of each parameter measured by each method (epoc system and reference analyzers) and the mean difference with 95% confidence interval (CI) are shown in Table 2. With the exception of two parameters (pCO2, pO2), the 95% CI of the mean difference between the epoc system and the reference analyzers were included within the range of the agreement target. The 11 parameters measured by the epoc system were equivalent to the reference values. The mean differences between the two methods were inconsequential and the limit of agreement was narrow. Moreover, of these 11 parameters, seven 2+ (HCO− , Na+, K+, glucose, Hb, and creatinine) showed that the re3 , Ca sults were within the target values for at least 80% (185 of 231 tests) of the patients. Thus, these seven parameters were regarded as having a clinically acceptable agreement between the two methods. Of the remaining two parameters (pCO2 and pO2), which neither had equivalence nor acceptability with the reference analyzers, the pCO2 measured by epoc system was highly correlated with the reference values, whereas pO2 was not. (pCO2: R2 = 0.891, y = −1.163 + 0.876·x; pO2: R2 = 0.07, y = 49.7 + 0.552·x). The regression model accounts for 89.1% of the variance for pCO2. The scatter plots of some parameters showed trends (negative trend: Ca2+, positive trend: pO2 and glucose) and inconsistent variability (the scatter plot for creatinine
2+ Most of the parameters (pH, pCO2, HCO− , lactate, Na+, K+, 3 , Ca Cl−, glucose, Hb, and Hct), except for pO2, measured by the epoc system were equivalent to or correlated with those measured by the reference methods in our previous study. The capillary blood-based epoc system might have some clinical utility, however, the capillary blood sample might be less accurate due to ischemia and vasoconstriction in patients who are in cardiac arrest or hypotension. Given that these hemodynamically unstable patients were not sufficiently represented in our previous study, further study on these patients' group was needed [3]. In this study, the result was similar to the previous study; except for pO2, 12 of the 13 parameters (we compared an additional parameter of creatinine in this study) were equivalent to or correlated with the reference values. Further, except for two parameters (lactate and creatinine), a similar result was shown in the cardiac arrest group, even though the sample size was small (n = 30). Previous studies have reported that various confounding factors might interfere with the analysis of POC glucometry in critically ill patients [15,16]. With regard to shock, several possible mechanisms may explain the discrepancies in blood glucose measurements obtained by different methods [17]. Peripheral vasoconstriction in hypoperfusion
6
H. Shin et al. / American Journal of Emergency Medicine xxx (xxxx) xxx
Fig. 3. Bland-Altman plot between epoc system and the arterial blood gas analyzer in the cardiac-arrested group.
states can lead to increased glucose extraction by tissues due to low capillary flow and increased glucose transit time, resulting in altered glucose measurements in the capillary blood [18]. Moreover, Garingarao et al. concluded that POC blood glucose measurements were significantly less accurate in the hypotensive subgroup of ICU patients as compared to the normotensive group. However, most patients included in those studies received a vasopressor such as dopamine or norepinephrine, while this study included patients who had not yet received those drugs. In addition, we believe that the peripheral vasoconstriction of the patients in our study group was less prominent than those in previous study groups because the period of exposure to the stress situation in our patients was markedly shorter than in previous studies. All patients included in
previous studies had been in the ICU for a certain period of time, while our study design was in the ED setting. There have been several previous studies comparing POC blood gas analysis and the central laboratory analysis [19,20]. However, to the best of our knowledge, there has been no study comparing the POC analysis using the peripheral capillary blood and reference analysis in hypotensive patients. This is the first study to compare peripheral capillary blood and venous blood in hemodynamically unstable patients, including not only glucose but also other significant parameters, such as pH, K+, lactate, Hb, creatinine, etc. that may be needed in an emergency situation. Therefore, we do not think that the conclusion of previous studies, in which the peripheral blood tests (skin puncture) were inaccurate in hypotensive patients, is applicable
H. Shin et al. / American Journal of Emergency Medicine xxx (xxxx) xxx
Fig. 4. Bland-Altman plot between epoc system and the venous blood analyzers in the cardiac-arrested group.
7
8
H. Shin et al. / American Journal of Emergency Medicine xxx (xxxx) xxx
to the acute patients in ED. Most parameters of the capillary blood measured by the epoc system were equivalent to or correlated with that measured by the reference analyzers in hypotensive patients in ED. Elevated serum lactate is an early indicator of shock and is associated with increased mortality [21]. The lactate assay remains a clinically useful test that can aid a clinician to identify underlying hypoperfusion in need of immediate treatment or an etiology not readily apparent on initial evaluation [22]. Hyperkalemia, one of the reversible causes of outof-hospital cardiac arrest (OHCA), can be identified in the field, so the patients can immediately receive a reversible drug [23]. Although hemoglobin is not an indicator of acute blood loss, patients with hypovolemic shock should be carefully monitored by checking base excess or lactate together [24,25]. Laboratory tests may help early evaluation of reversible causes in hemodynamically unstable patients with undifferentiated hypotension or shock to identify the cause. In this study, we have confirmed that these parameters measured by epoc system using the capillary blood could be useful and applicable even in the hemodynamically unstable patients. We added one parameter (creatinine) in this study. Although the incidence of contrast-induced nephropathy without confirming the creatinine level was low, when computed tomography with intravenous contrast was used to assess time-sensitive emergency patients, such as acute stroke and trauma patients [26,27], creatinine could be a valuable parameter for hemodynamically unstable patients [2830]. You et al. suggested that the results of POC creatinine analysis correlate well with the values obtained from reference laboratory tests [31]. However, they compared the POC creatinine levels using whole venous blood and central laboratory serum creatinine levels and did not use the capillary blood for POC analyzer. In addition, it was unclear whether the participants included in that study were hemodynamically unstable or hypotensive. In our study, creatinine is regarded as having a clinically acceptable agreement between the epoc system using capillary blood and the reference analyzers in hypotensive patients, including the cardiac arrest group. Thus, POC creatinine analysis using capillary blood can be identified in a hemodynamically unstable patient, which may be beneficial in the management of critically ill patients. The epoc system is an automated, self-monitoring system and may show some types of malfunction. The analyzer or test card can generate errors associated with failure of calibration and quality control (test card failure). An examiner can also contribute to test failure (sample injection error: insufficient volume of sample or fast sample injection). According to a previous study, the complete card failure rate was 13.0% for the epoc system [32]. In our previous study [3], the total failure rate was 10.8% (test card failure rate: 7.3%, sample injection error 3.5%), which was lower than that in Stotler's study. We further reduced the failure rate (7.6%) as compared to our previous failure rate (10.8%). The failure rate originated from the analyzer or test card themselves (test card failure rate) was not significantly different between the two studies (7.3% in our previous study vs. 6.8% in this study). However, the failure rate due to the examiner was markedly reduced in this study (3.5% vs. 0.8%). It can be a little difficult for an inexperienced examiner to obtain a sufficient volume of sample from the peripheral skin (90 μL) and inject the sample to the card at an appropriate rate. Our two examiners (emergency medical technician) had an experience of conducting more than 100 examinations prior to this study, which seems to have contributed to minimizing sample injection errors. Additionally, the cost-effectiveness of the epoc system should be considered. The cost per test is usually high for the epoc system due to high reagent/cartridge costs. It may be a barrier to the usage of the epoc system for patients. A prospective study with a large number of patients will be needed to compare the diagnostic utility and costeffectiveness of the epoc system. Further study in the prehospital environment may be warranted.
5. Limitation Although we concluded that the parameters, including pH, HCO− 3 , Ca2+, lactate, Na+, K+, Cl−, glucose, Hb, Hct, and creatinine, measured by the epoc system were equivalent to those measured by the reference analyzers, the reliability of these results can be compromised because the test was conducted by an emergency physician whose primary expertise is not laboratory work. 6. Conclusion The parameters measured by the epoc system using the capillary 2+ blood, which include pH, pCO2, pO2, HCO− , lactate, Na+, K+, 3 , Ca − Cl , glucose, Hb, Hct, and creatinine, were equivalent to or correlated with those measured by the reference analyzers in hypotensive patients. Thus, POC analysis using capillary blood may be useful in identifying the reversible causes for the life-threatening conditions in an emergency situation. References [1] Prause G, Ratzenhofer-Komenda B, Offner A, Lauda P, Voit H, Pojer H, et al. Prehospital point of care testing of blood gases and electrolytes - an evaluation of IRMA. Crit Care. 1997;1:79–83. https://doi.org/10.1186/cc108. [2] 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. Scand J Trauma Resusc Emerg Med. 2015;23:54. https://doi.org/10.1186/s13049-015-0134-y. [3] Kim C, Kim H. Emergency medical technician-performed point-of-care blood analysis using the capillary blood obtained from skin puncture. Am J Emerg Med. 2018; 36:1215–21. https://doi.org/10.1016/j.ajem.2017.12.025. [4] Gilbert HC, Szokol JW. Point of care technologies. Int Anesthesiol Clin. 2004;42: 73–94. [5] Peterson MC, Syndergaard T, Bowler J, Doxey R. A systematic review of factors predicting door to balloon time in ST-segment elevation myocardial infarction treated with percutaneous intervention. Int J Cardiol. 2012;157:8–23. https://doi. org/10.1016/j.ijcard.2011.06.042. [6] Foss KT, Subhi Y, Aagaard R, Bessmann EL, Bøtker MT, Graumann O, et al. Developing an emergency ultrasound app – a collaborative project between clinicians from different universities. Scand J Trauma Resusc and Emerg Med. 2015;23:47. https://doi. org/10.1186/s13049-015-0130-2. [7] Nelson BP, Sanghvi A. Out of hospital point of care ultrasound: current use models and future directions. Eur J Trauma Emerg Surg. 2016;42:139–50. https://doi.org/ 10.1007/s00068-015-0494-z. [8] Stengaard C, Sørensen JT, Ladefoged SA, Christensen EF, Lassen JF, Bøtker 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: 1361–6. https://doi.org/10.1016/j.amjcard.2013.06.026. [9] Guyette FX, Meier EN, Newgard C, McKnight B, Daya M, Bulger EM, et al. A comparison of prehospital lactate and systolic blood pressure for predicting the need for resuscitative care in trauma transported by ground. J Trauma Acute Care Surg. 2015; 78:600–6. https://doi.org/10.1097/ta.0000000000000549. [10] Callaway CW, Donnino MW, Fink EL, Geocadin RG, Golan E, Kern KB, et al. Part 8: post-cardiac arrest care: 2015 American Heart Association guidelines update for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation. 2015; 132:S465–82. https://doi.org/10.1161/cir.0000000000000262. [11] WHO. Guidelines approved by the guidelines review committee. WHO guidelines on drawing blood: Best practices in phlebotomy. Geneva: World Health Organization; 2010. [12] Bachner P, Hamlin W. Federal regulation of clinical laboratories and the clinical laboratory improvement amendments of 1988–part I. Clin Lab Med. 1993;13:739–52. https://doi.org/10.1016/s0272-2712(18)30438-4. [13] 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. . https://www.westgard.com/ biodatabase1.htm. Accessed date: 25 April 2019. [14] Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1:307–10. https://doi.org/10.1016/ s0140-6736(86)90837-8. [15] Dungan K, Chapman J, Braithwaite SS, Buse J. Glucose measurement: confounding issues in setting targets for inpatient management. Diabetes Care. 2007;30:403–9. https://doi.org/10.2337/dc06-1679. [16] Heinemann L. Quality of glucose measurement with blood glucose meters at the point-of-care: relevance of interfering factors. Diabetes Technol Ther. 2010;12: 847–57. https://doi.org/10.1089/dia.2010.0076. [17] Pati-An Garingarao CJ, Buenaluz-Sedurante M, Jimeno CA. Accuracy of point-of-care blood glucose measurements in critically ill patients in shock. J Diabetes Sci Technol. 2014;8:937–44. https://doi.org/10.1177/1932296814538608. [18] Juneja D, Pandey R, Singh O. Comparison between arterial and capillary blood glucose monitoring in patients with shock. Eur J Intern Med. 2011;22:241–4. https:// doi.org/10.1016/j.ejim.2011.01.004.
H. Shin et al. / American Journal of Emergency Medicine xxx (xxxx) xxx [19] Dashevsky M, Bernstein SL, Barsky CL, Taylor RA. Agreement between serum assays performed in ED point-of-care and hospital central laboratories. West J Emerg Med. 2017;18:403–9. https://doi.org/10.5811/westjem.2017.1.30532. [20] Bloom BM, Connor H, Benton S, Harris T. A comparison of measurements of sodium, potassium, haemoglobin and creatinine between an emergency department-based point-of-care machine and the hospital laboratory. Eur J Emerg Med. 2014;21: 310–3. https://doi.org/10.1097/mej.0000000000000076 Aug. [21] Bou Chebl R, El Khuri C, Shami A, Rajha E, Faris N, Bachir R, et al. Serum lactate is an independent predictor of hospital mortality in critically ill patients in the emergency department: a retrospective study. Scand J Trauma Resusc Emerg Med. 2017;25:69. https://doi.org/10.1186/s13049-017-0415-8. [22] Andersen LW, Mackenhauer J, Roberts JC, Berg KM, Cocchi MN, Donnino MW. Etiology and therapeutic approach to elevated lactate levels. Mayo Clin Proc. 2013;88: 1127–40. https://doi.org/10.1016/j.mayocp.2013.06.012. [23] Truhlář A, Deakin CD, Soar J, Khalifa GEA, Alfonzo A, Bierens JJLM, et al. Cardiac arrest in special circumstances: section 4 of European resuscitation council guidelines for resuscitation 2015. Resuscitation. 2015;95:148–201. https://doi.org/10.1016/j.resuscitation.2010.08.015. [24] EH1 Kincaid, Miller PR, Meredith JW, Rahman N, Chang MC. Elevated arterial base deficit in trauma patients: a marker of impaired oxygen utilization. J Am Coll Surg. 1998;187:384–92. https://doi.org/10.1016/s1072-7515(98)00202-6. [25] Moomey CB, Melton SM, Croce MA, Fabian TC, Proctor KG. Prognostic value of blood lactate, base deficit, and oxygen-derived variables in an LD50 model of penetrating trauma. Crit Care Med. 1999;27:154–61. https://doi.org/10.1097/00003246199901000-00044.
9
[26] Ang TE, Bivard A, Levi C, Ma H, Hsu CY, Campbell B, et al. Multi-modal CT in acute stroke: wait for a serum creatinine before giving intravenous contrast? No! Int J Stroke. 2015;10:1014–7. https://doi.org/10.1111/ijs.12605 Oct. [27] Colling KP, Irwin ED, Byrnes MC, Reicks P, Dellich WA, Reicks K, et al. Computed tomography scans with intravenous contrast: low incidence of contrast-induced nephropathy in blunt trauma patients. J Trauma Acute Care Surg. 2014;77:226–30. https://doi.org/10.1097/ta.0000000000000336 Aug. [28] Hoste EA, Clermont G, Kersten A, Venkataraman R, Angus DC, De Bacquer D, et al. RIFLE criteria for acute kidney injury are associated with hospital mortality in critically ill patients: a cohort analysis. Crit Care. 2006;10:R73. https://doi.org/10.1186/ cc4915. [29] Hinson JS, Al Jalbout N, Ehmann MR, Klein EY. Acute kidney injury following contrast media administration in the septic patient: a retrospective propensity-matched analysis. J Crit Care. 2019;51:111–6. https://doi.org/10.1016/j.jcrc.2019.02.003. [30] Shah K, Kilian B, Hsieh WJ, Kyrillou E, Hedge V, Newman DH. Can urine dipstick predict an elevated serum creatinine? Am J Emerg Med. 2010;28:613–6. https://doi. org/10.1016/j.ajem.2009.02.020. [31] You JS, Chung YE, Park JW, Lee W, Lee HJ, Chung TN, et al. The usefulness of rapid point-of-care creatinine testing for the prevention of contrast-induced nephropathy in the emergency department. Emerg Med J. 2013;30:555–8. https://doi.org/10. 1136/emermed-2012-201285. [32] 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;140:715–20. https://doi.org/10.1309/AJCP7QB3QQIBZPEK.