- Email: [email protected]
First or Second Drop of Blood in Capillary Glucose Monitoring: Findings from a Quantitative Study
RESEARCH ARTICLE
FIRST OR SECOND DROP OF BLOOD IN CAPILLARY GLUCOSE MONITORING: FINDINGS FROM A QUANTITATIVE STUDY Authors: Alvisa Palese, MNS, BNS, RN, Elisa Fabbro, BNS, RN, Anica Casetta, MS, and Irene Mansutti, MNS, BNS, RN, Udine, Italy Introduction: For clinical nurses, especially those working in
emergency departments, it is crucial to measure blood glucose (BG) in an accurate, timely, and safe manner. Many differences in practice exist with regard to use of the first or second drop of blood for testing, and no consistent guidelines are available for capillary BG testing at home or in ED settings. The purpose of this study is to evaluate the BG differences between the first and second drop of capillary blood collected from the same site in patients with type 1 diabetes. Methods: A consecutive sample of 195 persons with type 1 diabetes who had washed their hands and were not suspected of having hypoglycemia were included in the study. Descriptive and inferential statistical analysis for non-normal distributed variables was performed. Results: A strong correlation emerged between the BG reported in the first and the second drops (Spearman’s rho test [rs] 0.979, P b .001; Pearson r 0.978, P b .001). The average BG values obtained from the first and second drops were 184.30 mg/dL (median, 166) and 187.6 mg/dL (median, 172), respectively, and thus the second drop showed higher glucose values compared with the first drop. However, BG values of the
ccording to the International Diabetes Federation, the incidence of diabetes mellitus has increased considerably in all countries. 1 Health care professionals regularly encounter patients with altered blood
A
Alvisa Palese is Associate Professor in Nursing Science, Udine University, Italy. Elisa Fabbro is Registered Nurse, Udine University, Italy. Anica Casetta is Statistician, Epidemiological Unit, Udine University, Italy. Irene Mansutti is Registered Nurse, Udine University, Italy. For correspondence, write: Alvisa Palese, MNS, BNS, RN, Udine University, School of Nursing, Viale Ungheria 20, 33100 Udine, Italy; E-mail: [email protected]. J Emerg Nurs ■. 0099-1767 Copyright © 2016 Emergency Nurses Association. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jen.2016.03.027
■
second drop were not higher in all occasions: whereas some evaluations reported higher BG values in the second drop capillary sample (n = 123), others reported higher values in the first drop (n = 65), and still others reported identical measurements in the first and second drops (n = 7). Five outliers were present with a BG difference from − 39 to − 53 mg/dL in the first drop compared with the second drop, and 3 outliers were present with a BG difference from + 46 to + 57 mg/dL in the first drop compared with the second drop. However, the differences that emerged were not affected by glucose concentration (P = .221). Discussion: Using the first drop of blood in a patient with
clean hands allows emergency nurses to perform the test more quickly, resulting in immediate information. Findings indicate that the first drop of blood is adequate for clinical decision making, but the clinician should use judgment if using protocols in which small values (eg, 6 mg/dL) are important, because the first drop is more likely to have a slightly lower value. Key words: Accuracy; Blood glucose evaluation; Capillary; Drop of blood; Emergency nurses; First drop; Second drop; Nursing
glucose (BG) levels in a variety of outpatient clinics, emergency departments, and hospitals. 2 The purpose of this study is to evaluate the BG differences between the first and second drop of capillary blood collected from the same site in patients with type 1 diabetes. Most patients admitted to the hospitals are first evaluated by ED teams, and hyperglycemia is one of the most common problems detected. 3 In critically ill patients, hyperglycemia has been associated with poor clinical outcomes, 4 especially in persons affected by trauma, in whom increased infections, multiple organ failure, and mortality have been documented. 5–8 However, hypoglycemia is also reported as a common problem in emergency departments 9; it has been estimated that from 2% to 4% of deaths in persons with type 1 diabetes are attributable to hypoglycemia. 10 In addition, detecting altered BG levels is also important to identify persons with undiagnosed
■ • ■
WWW.JENONLINE.ORG
Master Proof ymen2954.pdf
1
RESEARCH ARTICLE/Palese et al
prediabetic conditions. Early identification of slightly elevated BG levels allows health care professionals to initiate an educational intervention, giving patients the opportunity to make choices to prevent or delay complications, 11 thus avoiding negative outcomes. 12 BG measurement accuracy is crucial to appropriate and timely clinical decisions, especially during emergency care such as that provided in ambulances, where the most common procedure adopted entails sticking a finger with a lancet device to obtain a small blood sample. Although this procedure is easy, fast, and safe, there is still no agreement regarding whether the first or second drop of capillary blood should be used for testing. 13 The first drop could contain more interstitial fluid, which differs slightly in glucose concentration; therefore, dilution of the first drop of blood with interstitial fluid may lead to lower BG values. 13 However, in highly perfused skin such as fingers, the glucose concentration in blood and interstitial fluid should be almost the same and thus should not affect the BG measurement accuracy. 14 For clinical nurses, especially those working in emergency departments, it is crucial to measure the BG in an accurate, timely, and safe manner. Advancing knowledge of the accuracy of BG measurements, aimed at improving the accuracy and quality of the measurements and consequent decision making, was the general intent of this study. Several previous studies examined the differences in BG levels in a different context, such as with soiled hands or different sites—for example, capillary to venous. In one study, in 53 healthy volunteers, no differences between the first and second drops of capillary blood from finger sites were found. 14 Another study in healthy volunteers evaluated the effect of fingers soiled with products containing sugar (eg, dextrose, fruit, jam, honey, and chocolate paste) and the effect of disinfection with chlorhexidine in removing traces of glucose. 15 Sequential drops of blood were obtained in different conditions: (1) after “intervention 1,” which was based on sequential evaluation of the first, second, and third drops from soiled fingers; and (2) after “intervention 2,” in which the BG measurement was performed using the first drop from soiled fingers, after which finger disinfection was performed and sequential BG measurement was performed using the second and third drops. The BG concentration in the third drop was 10% higher than the control measurement obtained in washed hands with neutral soap. In both interventions the highest glucose concentrations were found in the first drop of blood, with a significant decrease in sequential BG concentrations from the first drop of blood to the control. 15 Further studies are needed in persons who have diseases or disorders that alter BG, in addition to persons who are healthy.
2
In a study involving 123 patients with diabetes, BG levels using the first and second drops of blood in different circumstances were measured: (1) without washing hands, (2) after handling fruit, (3) after washing the fruit-exposed hands, and (4) during the application of different amounts of pressure around the finger. 16 The effects of different circumstances were evaluated; according to the findings, the first drop of blood was adequate for self-monitored glucose testing, but only after having washed the hands. When hand washing is not possible, or in case of visible soiled fingers, using the second drop of blood after wiping away the first drop was suggested. 16 In a recent study of 526 hospitalized patients, the accuracy of the first or second drop of blood obtained from different finger sites was compared with venous BG, which was considered the gold standard. 17 No significant differences in the BG concentration were obtained from the first drop, the second drop, and the venous sample of blood. However, after stratifying the evaluations into 6 groups according to glucose concentration, differences were found in the groups reporting values of glucose b 9.9 mmol/ L and between 20 to 30 mmol/L. According to the findings, the first drop was more accurate in the group reporting a b 9.9 mmol/L glucose concentration, whereas the second drop was more accurate in the group reporting N 20 mmol/L with respect to the venous measurement. As previously reported, studies regarding differences, if any, between the first and second drops have mainly regarded BG measurements performed with healthy subjects, patients whose hands were exposed to dirt (soiled fingers and unwashed hands), or comparing hospitalized patients’ BG accuracy to venous blood sampled. Our evaluation focused on the differences of first and second drops of capillary blood in patients who had type 1 diabetes with cleanly washed hands in the outpatient setting. Methods AIM AND STUDY DESIGN
This quantitative study evaluated the differences between BG measured in the first and the second drops of capillary blood samples consecutively obtained from the same site fingertip at an Italian diabetic outpatient clinic. The data were collected in 2013 after obtaining authorization from the Hospital Internal Review Board. The BG data from the first and the second drops from the same subject have been compared, thus reducing the effects of potential confounding factors. 18 The hypotheses were as follows: 1.There
is no difference between the repeated measurements (first and second) obtained for a single subject.
■ ■ •
JOURNAL OF EMERGENCY NURSING
Master Proof ymen2954.pdf
■
Palese et al/RESEARCH ARTICLE
2.Different concentrations of glucose will not affect differences between the first and second drop of blood.
SAMPLING AND SAMPLE SIZE
A consecutive sample of 195 persons with type 1 diabetes who were attending their periodic follow-up in an Italian outpatient clinic, in the morning, after having consumed breakfast, was obtained in 2013. Patients with the following criteria were included: (1) age ≥ 18 years; (2) normal body and peripheral temperature to avoid vasoconstriction 19; (3) hands were washed; (4) hypoglycemia was not suspected; and (5) willing to participate in the study after having received information regarding its aims and after having provided a written consent. No patients refused to participate in the study, and all those eligible according to the inclusion criteria were enrolled. VARIABLES AND DATA COLLECTION PROCESS
Information regarding age, gender, and BG procedures adopted in daily practice concerning the drop used in the glucose concentration measurement were collected in a face-to-face interview. In addition, data regarding the education received by nurses, physicians, or other health care professionals regarding which drop should be considered for BG evaluation were also collected. Thereafter, participants were asked to wash their hands with warm water and neutral soap. 15 The blood sample was collected using standard procedures. 20 To avoid any possible bias, researchers applied the procedure considering the same hand (left) and finger (third) in all patients. It was agreed that only gentle pressure would be applied in case of difficulty in obtaining the blood sample. After the first drop was obtained, the BG was evaluated. The drop was then wiped away with clean gauze. The second drop of blood was obtained within a few seconds (b 60 seconds). 16 The same portable glucose meter (Verio IQ, OneTouch, LifeScan Inc, Milpitas, CA) 21 and same type of finger stick (Accu-Chek, Roche Diagnostics, Mannheim, Germany) 22 were used. The BG measurements obtained were immediately registered by the researcher performing the data collection process. Interview data and BG data had been pilot tested in 10 patients; these data are not included in this final report. The researcher was first trained both in the procedure and in the data collection process, aiming at ensuring fidelity with regard to the technique and the accuracy of the data collected. In cases of clinical significant difference between the measurements obtained at patient level requiring some clinical decisions according to the registered nurse and physician evaluation, a third evaluation
■
would be offered to patients measuring the venous blood. No cases emerged, and thus no further venous blood evaluation was needed. Concentrations of glucose were defined in accordance with Li and colleagues 17 and American Diabetes Association statements. 23 The criteria were BG b 179 mg/dL; from 180 to 358 mg/dL; and N 359 mg/dL. These criteria were applied to the BG list of values obtained from the first drop and then to the list of BG values obtained from the second drop. DATA ANALYSIS
The data were analyzed with SPSS software, version 22 (IBM Corp, Armonk, NY). Averages, standard deviations (SDs), medians, frequencies, and proportions were evaluated. Then, according to the non-normal distributed BG values obtained through the first and the second drops, Spearman’s rho test (rs) was performed to evaluate their correlation. To test the first hypothesis that there were no differences between the first and second drops of blood, data were visualized using a Bland-Altman plot, and the nonparametric Wilcoxon rank and sign test was used. To test the second hypothesis that the differences were unrelated to the concentration of BG, the KruskalWallis test was conducted. Statistical significance was fixed at P b .05.
Results
The average age of the patients involved was 57.5 years (SD 4.6), and the majority were male (65.6%; n = 128). In their daily self-monitoring BG measurement, a large proportion of patients were used to considering the first blood drop (88.7%; n = 173), and a limited proportion were used to considering the second drop (11.3%; n = 22). The majority (86.7%; n = 169) had never received recommendations from health care professionals regarding which drop to consider in order to obtain an accurate BG measurements, and the decision of which one to consider was made autonomously. BG concentrations were as follows: 109 patients had a BG concentration b 178 mg/dL; 78 had a BG concentration from 179 to 358 mg/dL; and 8 had a BG concentration of 359 mg/dL or higher. No case crossed groups between the first and second measurement. Table 1 relays the descriptive statistics for the BG measures: the differences, both in the averages and in the medians, were limited to around 3 mg/dL and 6 mg/dL, respectively, indicating that the second drop contained higher glucose values that were highly correlated, statistically different, but
■ • ■
WWW.JENONLINE.ORG
Master Proof ymen2954.pdf
3
RESEARCH ARTICLE/Palese et al
TABLE 1
TABLE 2
Blood glucose reported in the first and second drop of blood: descriptive statistics
Blood glucose reported in the first and second drop of blood
Blood glucose measures
First drop
BG measures
Mean (mg/dL) Median (mg/dL) Minimum-maximum (mg/dL) Percentile 25% (mg/dL) 50% (mg/dL) 75% (mg/dL)
184.3 187.6 + 3.3 166 172 + 6.0 77-455 77-467 0-10 135 166 222
Second Difference drop
140 172 222
5 6 0
clinically comparable with the first drop measured in the same finger a few seconds earlier. The results for the tests of the first hypothesis, that there are no differences between the first and second drop of blood, are depicted in Figure 1 and Table 2. As reported in Figure 1, a strong correlation has emerged between the BG level reported in the first and the second
FIGURE 1 Blood glucose reported in the first and second drop of blood: correlation.
4
First drop Second drop
N
Average Rank sum rank
Negative 123 a 97.83 ranks 88.21 Positive ranks 65 b c Equal values 7 Total 195 Z score − 4.218, P b .001
12,032.50 5,733.50
BG, Blood glucose. a Number of cases first drop BG measure was less than second drop BG measure. b Number of cases first drop BG measure was greater than second drop BG measure. c Number of cases first drop BG measure was equal to second drop BG measure.
drops (rs = 0.979, P b .001; Pearson r = 0.978, P b .001). The average BG level obtained from the first drop was 184.30 mg/dL (median, 166), and from the second drop the average BG level was 187.6 mg/dL (median, 172), as reported in Table 2. The differences were statistically significant. As reported in Table 2, in 123 measurements (63%), the BG reported from the second drop was higher (from + 1 to + 57 mg/dL) compared with the values reported from the first drop. In particular, in 81.3% of the cases (n = 100), the values ranged from + 1 to + 15 mg/dL. However, in 65 cases (33.3%), the second drop reported lower BG values (range − 1 to − 53 mg/dL) compared with those in the first drop; in the majority of cases (78.5%; n = 51), these values ranged from − 1 to − 15 mg/dL. Finally, only 7 cases (3.6%) had the same BG values in both measurements. The Bland-Altman plot in Figure 2 shows that some evaluations reported higher BG values in the second drop capillary sample, others reported higher values in the first drop, and still others reported identical measurements in the first and the second drops. Eight outliers were also present in which a BG difference from − 39 to − 53 mg/dL was reported in the first drop compared with the second drop (n = 5, − 39, − 40, − 40, − 42, and − 53 mg/dL, respectively) and from + 46 to + 57 mg/dL in the first drop as compared with the second drop (n = 3, + 46, + 48, and + 57, respectively). Therefore, the BG measurement performed with the second drop was not higher in all occasions. The differences that emerged in the BG measurements between the first and second drops were not significantly affected by glucose concentration, supporting our second hypothesis. In fact, when the cases were classified in groups with BG concentration ranging from b 178 mg/dL, from
■ ■ •
JOURNAL OF EMERGENCY NURSING
Master Proof ymen2954.pdf
■
Palese et al/RESEARCH ARTICLE
mg/dl
60
1st drop - 2nd drop
40 +1.96 SD 24,5
20
Mean -3,4
0 -20
-1.96 SD -31,2
-40 -60 0
100
200
300
400
500
600
mg/dl
Mean of 1st drop and 2nd drop Legend: SD, Standard Deaviation
FIGURE 2 Findings from the Bland-Altman plot.
179 to 358 mg/dl, and 359 mg/dL or higher, 17 the same number of evaluations from the first and the second blood samples were classified in each group. Nevertheless, when evaluating statistically the BG differences between the first and second drops for each group created on the basis of the glucose concentration, no significant differences have emerged (P = .221), as reported in Table 3.
Discussion
The purpose of this study was to evaluate differences in BG level between the first and second drop of capillary blood collected from the same site. Although the 2 samples were
TABLE 3
Differences in blood glucose reported by the first and second drop of blood at different glucose concentration Glucose concentration (mg/dL) a
N, first and second drop
Average rank
b 179 mg/dL 180-358 mg/dL N 359 mg/dL Total χ 2 (df) P value
109 78 8 195 3.015 (2) b .221
96.82 96.17 131.88
df, Degrees of freedom. a According to Li et al.17 b Kruskal-Wallis test.
■
strongly correlated, we found small, statistically significant differences between the first and second drop. The first drop was more likely to have a lower value. However, these differences of 3 to 6 mg/dL may not be large enough to have clinical relevance in most cases. Glucose concentration did not significantly affect the differences between the first and second drop, but our descriptive plot (Figure 1) indicates that caution should be used in persons with higher BG levels. Thus, our first hypothesis that there would be no difference in the first or second drop was not supported statistically but may still be appropriate clinically. Our second hypothesis that glucose concentration would not affect the difference between the first and second drop was supported by our results. No consistent guidelines were available about whether to use the first or second drop in capillary BG testing in the home or ED settings. 17,20,24 Consequently, in their daily practice, patients and practitioners still appear to be unsure and have inconsistencies regarding the procedure. For example, approximately 72% of European nurse educators advise patients to wash their hands with water and soap and to use the first drop of blood. 13 When it is not possible to wash the hands, 52% of European educators suggest that the second drop is considered more accurate. The clinical question regarding the use of the first or the second drops in BG evaluations has been evaluated in several previous studies. 15–17,23,25 Most of these previous studies focused on avoiding false hyperglycemia results in patients with soiled fingers and unwashed hands, as well as evaluating the accuracy of these measures compared with the venous blood evaluation. Our study is unique in the evaluation of BG values from the first and second drops taken in cleanly washed hands for routine care for type 1 diabetes. Most of the patients in our study had not been given any previous education on which drop of blood to use in home testing, possibly reflecting the inconsistencies in the evidence available. In addition, although a large proportion of patients routinely used the first drop, a few routinely used the second drop. Using the second drop may increase the duration of the procedure, the risk of soiling by blood, or the intensity of pain as a result of use of the milking procedure, which may affect the quality of care in general. In the case of emergency nurses, waiting for the second drop may increase the duration of the procedure and the associated biologic risk. The findings show that there is a statistically significant difference in the BG measured after a few seconds in a sequential of drops obtained from the same fingertip site in the same patient with type 1 diabetes. According to previous data, 14 BG measurements from the first drop tend to be
■ • ■
WWW.JENONLINE.ORG
Master Proof ymen2954.pdf
5
RESEARCH ARTICLE/Palese et al
lower, possibly because of dilution with interstitial fluid. Limited comparison can be made with previous studies available in the field 15–17,23,25 given that different endpoints, procedure, or patient inclusion (persons with diabetes or healthy volunteers) were adopted. Our findings showed that in the majority of the cases (63%), the second drop returned on average higher BG values of approximately 3 to 4 mg/dL compared with the first drop. In the remaining patients (around 33%), higher BG values were found in the first drop. BG values overlapped in only 7 patients. However, the differences observed, both in positive or negative direction, ranged from ± 1 to 15 mg/dL for 80% of our participants. Therefore, the clinical relevance of this difference is limited. However, a difference of + 57 and − 53 as reported by some outliers may be clinically significant and may change treatment decisions. Further study is needed for in-depth understanding of the factors affecting these relevant differences, both at the individual and at the procedural levels. In addition, the difference that emerged in BG as measured in the consequent drops is not significantly affected by BG concentration, in contrast to that reported by Li and colleagues, 17 who in comparing the accuracy of the BG measurements concluded that the first drop was more accurate in the group with BG b 179 mg/dL whereas the second drop was more accurate in the group reporting N 359 mg/dL with respect to the venous evaluations. The sample size was not defined a priori. However, the sample of 195 subjects guarantees, when confronting the difference between the averages of the first and second drop of 3.3 mg/dL, an SD of 14.2 mg/dL and a significance level of 5%, for a power sample equal to 90%.
Limitations
The study was affected by several limitations. The comparison between the first and second drops was made without considering a gold standard measure such as venous BG results, as reported by previous studies (eg, Li et al 17), thus without considering an appropriate comparison method, such as the measurement of the same capillary sample using an analytically robust method. The drops were consecutive and not obtained from a second site or measured with different glucose meters as previous researchers have documented (eg, Li et al 17); in addition, the evaluations were performed after having washed the hands, reflecting the best recommendations available. 20 The same procedure was adopted considering the same hand (left) and the same finger (third), which was aimed at standardizing the data collection process and limiting bias.
6
To our knowledge, evidence is limited to date with regard to the variability of the BG measurements between different hand or finger collection sites. Our findings are limited to persons with type 1 diabetes. Outpatients were tested in the morning, after having consumed breakfast and having washed their hands with soap and water, and they were not suspected of having hypoglycemia. All of these conditions may differ from those encountered in ED settings in which patients’ hands could be unclean, hand washing may be difficult, and patients arrive in different moments of the day, as well as with hypoglycemia. This may have affected the limited variance in the BG emerged, thus not reflecting the variability of the conditions encountered by clinical nurses in emergency departments. Furthermore, a relatively narrow clustering of glucose values around the median has emerged, with no participant showing evidence of significant biochemical hypoglycemia. Accurate detection of hypoglycemia is arguably the most important glucose measurement task undertaken in the emergency room. Therefore, future research should address the full range of conditions representing the complexity of care in emergency departments.
Implications for Emergency Nurses
Nurses working in the emergency department routinely care for patients with diabetes and/or patients who require BG measurement. BG measurement procedures using capillary blood are very common among nurses in the clinical setting and patients in the home setting, and many differences in practice exist with regard to use of the first or second drop of blood for testing. No guidelines are available because of limited studies in the field. Nurses should be aware that the first drop of blood is likely to have a lower BG level. In our study, this difference averaged 3 to 6 mg/dL, which is not likely to be clinically significant in most cases. These differences between the first and second drop were unrelated to the overall BG level (low, normal, or high), although practicing nurses should interpret the findings seen in Figure 1, where the correlations fall further away from the dotted line at higher BG levels, as a sign that continued caution and research on the topic is needed for high BG measurements. Our findings only have implications for measurements from patients’ cleanly washed hands, because our findings were different from previous ones in which BG measurements were performed in patients whose hands were exposed to dirt (soiled fingers and unwashed hands). We did not compare our results with venous blood measurements, which would be common in the ED setting.
■ ■ •
JOURNAL OF EMERGENCY NURSING
Master Proof ymen2954.pdf
■
Palese et al/RESEARCH ARTICLE
Therefore, emergency nurses may educate patients that the first drop of blood in home testing is likely to be adequate but also likely to demonstrate slightly lower values than the second drop. Emergency nurses also may reinforce that patients who already use the first drop should continue that practice. Our recommendations to practicing nurses are based on a limited amount of evidence, including our study, and nurses should continue to conduct and evaluate continued research with patients admitted in emergency departments to evaluate if the differences that emerged in this study are similar. Our findings did not apply to all participants. Patients with extreme differences, or outliers, between the first and second drop of ± 50 mg/dL require further study.
9. Harding AD. Hypoglycemia in the emergency department. J Emerg Nurs. 2012;38(4):362-363. 10. Cryer PE. Diverse causes of hypoglycemia-associated autonomic failure in diabetes. N Engl J Med. 2004;350(22):2272-2279. 11. Bergman M. Preface. Prediabetes and diabetes prevention. Med Clin North Am. 2011;95(2):xi-xiii. 12. Hoerger TJ, Hicks KA, Sorensen SW, et al. Cost-effectiveness of screening for pre-diabetes among overweight and obese U.S. adults. Diabetes Care. 2007;30(11):2874-2879. 13. Hortensius J, van der Bijl JJ, Kleefstra N, Houweling ST, Bilo HJ. Selfmonitoring of blood glucose: professional advice and daily practice of patients with diabetes. Diabetes Educ. 2012;38(1):101-107. 14. Fruhstorfer H, Quarder O. Blood glucose monitoring: milking the finger and using the first drop of blood give correct glucose values. Diabetes Res Clin Pract. 2009;85(1):14-15.
Conclusion
The findings from this study support that the first or second drop of BG is adequate for clinical decision making if the patient has cleanly washed hands, but the clinician should use judgment if using protocols in which small values (eg, 6 mg/dL) are important because the first drop is more likely to have a slightly lower value. The differences in BG are not related to the overall BG level (high, normal, or low). Using the first drop in a patient with cleanly washed hands allows emergency nurses to conduct testing more quickly, resulting in immediate information. REFERENCES 1. International Diabetes Federation. IDF Diabetes Atlas. 7th ed., Brussels, Belgium: International Diabetes Federation; 2015. http://www. diabetesatlas.org/. Accessed April 13, 2016. 2. Rouse MD, Shoukry CL. Elevated blood glucose levels in the emergency department: missed opportunities. J Emerg Nurs. 2014;40(4):311-316.
15. Hortensius J, Kleefstra N, Slingerland RJ, et al. The influence of a soiled finger in capillary blood glucose monitoring. Neth J Med. 2010;68(1): 330-331. 16. Hortensius J, Slingerland RJ, Kleefstra N, et al. Self-monitoring of blood glucose: the use of the first or the second drop of blood. Diabetes Care. 2011;34(3):556-560. 17. Li M, Wang X, Shan Z. Deciding between using the first or second drop of blood for the self monitoring of blood glucose. Prim Care Diabetes. 2014;8(4):365-369. 18. Maclure M, Mittleman MA. Should we use a case-crossover design?. Annu Rev Public Health. 2000;21:193-221. 19. Center for Devices and Radiological Health, US Food and Drug Administration. Draft guidance document: review criteria assessment of portable blood glucose monitoring in vitro diagnostic devices using glucose oxidase, dehydrogenase or hexokinase methodology. http:// www.fda.gov/ohrms/dockets/dockets/06p0313/06p-0313-cp00001-10Exhibit-I-CDRH-ODE-Gluc-vol1.pdf. Accessed April 13, 2016. 20. Clar C, Barnard K, Cummins E, Royle P, Waugh N. Aberdeen Health Technology Assessment Group. Self-monitoring of blood glucose in type 2 diabetes: systematic review. Health Technol Assess. 2010;14(12):1-140.
3. Bernard JB, Munoz C, Harper J, Muriello M, Rico E, Baldwin D. Treatment of inpatient hyperglycemia beginning in the emergency department: a randomized trial using insulin aspart and detemir compared with usual care. J Hosp Med. 2011;6(5):279-284.
21. LifeScan. OneTouch Verio®IQ. http://www.lifescan.it/ourproducts/ meter/one-touch-verio-iq. Accessed April 13, 2016.
4. McCowen KC, Malhotra A, Bistrian BR. Stress-induced hyperglycemia. Crit Care Clin. 2001;17(1):107-124.
22. Roche Diagnostics. Accu-Chek Mobile. https://www1.accu-chek.it/ documents/Accu-Chek_Mobile_Manuale.pdf. Accessed April 13, 2016.
5. Sperry JL, Frankel HL, Vanek SL, et al. Early hyperglycemia predicts multiple organ failure and mortality but not infection. J Trauma. 2007; 63(3):487-493.
23. American Diabetes Association. eAG/A1C conversion calculator. http:// professional.diabetes.org/eAG. Accessed April 16, 2016.
6. Wahl WL, Taddonio M, Maggio PM, Arbabi S, Hemmila MR. Mean glucose values predict trauma patient mortality. J Trauma. 2008;65(1):42-47. 7. Kreutziger J, Schlaepfer J, Wenzel V, Constantinescu MA. The role of admission blood glucose in outcome prediction of surviving patients with multiple injuries. J Trauma. 2009;67(4):704-708.
■
8. Bochicchio GV, Sung J, Joshi M, et al. Persistent hyperglycemia is predictive of outcome in critically ill trauma patients. J Trauma. 2005; 58(5):921-924.
24. American Diabetes Association. Standards of medical care in diabetes—2014. Diabetes Care. 2014;37(Suppl. 1):S14-S80. 25. Palese A, Chiandetti R, Mansutti I. The effect of 2% chlorhexidine and cleansing wipes on capillary blood glucose sampling accuracy of a fingertip soiled with apple: a case crossover study design. J Clin Nurs. 2014;23(17–18):2672-2674.
■ • ■
WWW.JENONLINE.ORG
Master Proof ymen2954.pdf
7
FIRST OR SECOND DROP OF BLOOD IN CAPILLARY GLUCOSE MONITORING: FINDINGS FROM A QUANTITATIVE STUDY Authors: Alvisa Palese, MNS, BNS, RN, Elisa Fabbro, BNS, RN, Anica Casetta, MS, and Irene Mansutti, MNS, BNS, RN, Udine, Italy Introduction: For clinical nurses, especially those working in
emergency departments, it is crucial to measure blood glucose (BG) in an accurate, timely, and safe manner. Many differences in practice exist with regard to use of the first or second drop of blood for testing, and no consistent guidelines are available for capillary BG testing at home or in ED settings. The purpose of this study is to evaluate the BG differences between the first and second drop of capillary blood collected from the same site in patients with type 1 diabetes. Methods: A consecutive sample of 195 persons with type 1 diabetes who had washed their hands and were not suspected of having hypoglycemia were included in the study. Descriptive and inferential statistical analysis for non-normal distributed variables was performed. Results: A strong correlation emerged between the BG reported in the first and the second drops (Spearman’s rho test [rs] 0.979, P b .001; Pearson r 0.978, P b .001). The average BG values obtained from the first and second drops were 184.30 mg/dL (median, 166) and 187.6 mg/dL (median, 172), respectively, and thus the second drop showed higher glucose values compared with the first drop. However, BG values of the
ccording to the International Diabetes Federation, the incidence of diabetes mellitus has increased considerably in all countries. 1 Health care professionals regularly encounter patients with altered blood
A
Alvisa Palese is Associate Professor in Nursing Science, Udine University, Italy. Elisa Fabbro is Registered Nurse, Udine University, Italy. Anica Casetta is Statistician, Epidemiological Unit, Udine University, Italy. Irene Mansutti is Registered Nurse, Udine University, Italy. For correspondence, write: Alvisa Palese, MNS, BNS, RN, Udine University, School of Nursing, Viale Ungheria 20, 33100 Udine, Italy; E-mail: [email protected]. J Emerg Nurs ■. 0099-1767 Copyright © 2016 Emergency Nurses Association. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jen.2016.03.027
■
second drop were not higher in all occasions: whereas some evaluations reported higher BG values in the second drop capillary sample (n = 123), others reported higher values in the first drop (n = 65), and still others reported identical measurements in the first and second drops (n = 7). Five outliers were present with a BG difference from − 39 to − 53 mg/dL in the first drop compared with the second drop, and 3 outliers were present with a BG difference from + 46 to + 57 mg/dL in the first drop compared with the second drop. However, the differences that emerged were not affected by glucose concentration (P = .221). Discussion: Using the first drop of blood in a patient with
clean hands allows emergency nurses to perform the test more quickly, resulting in immediate information. Findings indicate that the first drop of blood is adequate for clinical decision making, but the clinician should use judgment if using protocols in which small values (eg, 6 mg/dL) are important, because the first drop is more likely to have a slightly lower value. Key words: Accuracy; Blood glucose evaluation; Capillary; Drop of blood; Emergency nurses; First drop; Second drop; Nursing
glucose (BG) levels in a variety of outpatient clinics, emergency departments, and hospitals. 2 The purpose of this study is to evaluate the BG differences between the first and second drop of capillary blood collected from the same site in patients with type 1 diabetes. Most patients admitted to the hospitals are first evaluated by ED teams, and hyperglycemia is one of the most common problems detected. 3 In critically ill patients, hyperglycemia has been associated with poor clinical outcomes, 4 especially in persons affected by trauma, in whom increased infections, multiple organ failure, and mortality have been documented. 5–8 However, hypoglycemia is also reported as a common problem in emergency departments 9; it has been estimated that from 2% to 4% of deaths in persons with type 1 diabetes are attributable to hypoglycemia. 10 In addition, detecting altered BG levels is also important to identify persons with undiagnosed
■ • ■
WWW.JENONLINE.ORG
Master Proof ymen2954.pdf
1
RESEARCH ARTICLE/Palese et al
prediabetic conditions. Early identification of slightly elevated BG levels allows health care professionals to initiate an educational intervention, giving patients the opportunity to make choices to prevent or delay complications, 11 thus avoiding negative outcomes. 12 BG measurement accuracy is crucial to appropriate and timely clinical decisions, especially during emergency care such as that provided in ambulances, where the most common procedure adopted entails sticking a finger with a lancet device to obtain a small blood sample. Although this procedure is easy, fast, and safe, there is still no agreement regarding whether the first or second drop of capillary blood should be used for testing. 13 The first drop could contain more interstitial fluid, which differs slightly in glucose concentration; therefore, dilution of the first drop of blood with interstitial fluid may lead to lower BG values. 13 However, in highly perfused skin such as fingers, the glucose concentration in blood and interstitial fluid should be almost the same and thus should not affect the BG measurement accuracy. 14 For clinical nurses, especially those working in emergency departments, it is crucial to measure the BG in an accurate, timely, and safe manner. Advancing knowledge of the accuracy of BG measurements, aimed at improving the accuracy and quality of the measurements and consequent decision making, was the general intent of this study. Several previous studies examined the differences in BG levels in a different context, such as with soiled hands or different sites—for example, capillary to venous. In one study, in 53 healthy volunteers, no differences between the first and second drops of capillary blood from finger sites were found. 14 Another study in healthy volunteers evaluated the effect of fingers soiled with products containing sugar (eg, dextrose, fruit, jam, honey, and chocolate paste) and the effect of disinfection with chlorhexidine in removing traces of glucose. 15 Sequential drops of blood were obtained in different conditions: (1) after “intervention 1,” which was based on sequential evaluation of the first, second, and third drops from soiled fingers; and (2) after “intervention 2,” in which the BG measurement was performed using the first drop from soiled fingers, after which finger disinfection was performed and sequential BG measurement was performed using the second and third drops. The BG concentration in the third drop was 10% higher than the control measurement obtained in washed hands with neutral soap. In both interventions the highest glucose concentrations were found in the first drop of blood, with a significant decrease in sequential BG concentrations from the first drop of blood to the control. 15 Further studies are needed in persons who have diseases or disorders that alter BG, in addition to persons who are healthy.
2
In a study involving 123 patients with diabetes, BG levels using the first and second drops of blood in different circumstances were measured: (1) without washing hands, (2) after handling fruit, (3) after washing the fruit-exposed hands, and (4) during the application of different amounts of pressure around the finger. 16 The effects of different circumstances were evaluated; according to the findings, the first drop of blood was adequate for self-monitored glucose testing, but only after having washed the hands. When hand washing is not possible, or in case of visible soiled fingers, using the second drop of blood after wiping away the first drop was suggested. 16 In a recent study of 526 hospitalized patients, the accuracy of the first or second drop of blood obtained from different finger sites was compared with venous BG, which was considered the gold standard. 17 No significant differences in the BG concentration were obtained from the first drop, the second drop, and the venous sample of blood. However, after stratifying the evaluations into 6 groups according to glucose concentration, differences were found in the groups reporting values of glucose b 9.9 mmol/ L and between 20 to 30 mmol/L. According to the findings, the first drop was more accurate in the group reporting a b 9.9 mmol/L glucose concentration, whereas the second drop was more accurate in the group reporting N 20 mmol/L with respect to the venous measurement. As previously reported, studies regarding differences, if any, between the first and second drops have mainly regarded BG measurements performed with healthy subjects, patients whose hands were exposed to dirt (soiled fingers and unwashed hands), or comparing hospitalized patients’ BG accuracy to venous blood sampled. Our evaluation focused on the differences of first and second drops of capillary blood in patients who had type 1 diabetes with cleanly washed hands in the outpatient setting. Methods AIM AND STUDY DESIGN
This quantitative study evaluated the differences between BG measured in the first and the second drops of capillary blood samples consecutively obtained from the same site fingertip at an Italian diabetic outpatient clinic. The data were collected in 2013 after obtaining authorization from the Hospital Internal Review Board. The BG data from the first and the second drops from the same subject have been compared, thus reducing the effects of potential confounding factors. 18 The hypotheses were as follows: 1.There
is no difference between the repeated measurements (first and second) obtained for a single subject.
■ ■ •
JOURNAL OF EMERGENCY NURSING
Master Proof ymen2954.pdf
■
Palese et al/RESEARCH ARTICLE
2.Different concentrations of glucose will not affect differences between the first and second drop of blood.
SAMPLING AND SAMPLE SIZE
A consecutive sample of 195 persons with type 1 diabetes who were attending their periodic follow-up in an Italian outpatient clinic, in the morning, after having consumed breakfast, was obtained in 2013. Patients with the following criteria were included: (1) age ≥ 18 years; (2) normal body and peripheral temperature to avoid vasoconstriction 19; (3) hands were washed; (4) hypoglycemia was not suspected; and (5) willing to participate in the study after having received information regarding its aims and after having provided a written consent. No patients refused to participate in the study, and all those eligible according to the inclusion criteria were enrolled. VARIABLES AND DATA COLLECTION PROCESS
Information regarding age, gender, and BG procedures adopted in daily practice concerning the drop used in the glucose concentration measurement were collected in a face-to-face interview. In addition, data regarding the education received by nurses, physicians, or other health care professionals regarding which drop should be considered for BG evaluation were also collected. Thereafter, participants were asked to wash their hands with warm water and neutral soap. 15 The blood sample was collected using standard procedures. 20 To avoid any possible bias, researchers applied the procedure considering the same hand (left) and finger (third) in all patients. It was agreed that only gentle pressure would be applied in case of difficulty in obtaining the blood sample. After the first drop was obtained, the BG was evaluated. The drop was then wiped away with clean gauze. The second drop of blood was obtained within a few seconds (b 60 seconds). 16 The same portable glucose meter (Verio IQ, OneTouch, LifeScan Inc, Milpitas, CA) 21 and same type of finger stick (Accu-Chek, Roche Diagnostics, Mannheim, Germany) 22 were used. The BG measurements obtained were immediately registered by the researcher performing the data collection process. Interview data and BG data had been pilot tested in 10 patients; these data are not included in this final report. The researcher was first trained both in the procedure and in the data collection process, aiming at ensuring fidelity with regard to the technique and the accuracy of the data collected. In cases of clinical significant difference between the measurements obtained at patient level requiring some clinical decisions according to the registered nurse and physician evaluation, a third evaluation
■
would be offered to patients measuring the venous blood. No cases emerged, and thus no further venous blood evaluation was needed. Concentrations of glucose were defined in accordance with Li and colleagues 17 and American Diabetes Association statements. 23 The criteria were BG b 179 mg/dL; from 180 to 358 mg/dL; and N 359 mg/dL. These criteria were applied to the BG list of values obtained from the first drop and then to the list of BG values obtained from the second drop. DATA ANALYSIS
The data were analyzed with SPSS software, version 22 (IBM Corp, Armonk, NY). Averages, standard deviations (SDs), medians, frequencies, and proportions were evaluated. Then, according to the non-normal distributed BG values obtained through the first and the second drops, Spearman’s rho test (rs) was performed to evaluate their correlation. To test the first hypothesis that there were no differences between the first and second drops of blood, data were visualized using a Bland-Altman plot, and the nonparametric Wilcoxon rank and sign test was used. To test the second hypothesis that the differences were unrelated to the concentration of BG, the KruskalWallis test was conducted. Statistical significance was fixed at P b .05.
Results
The average age of the patients involved was 57.5 years (SD 4.6), and the majority were male (65.6%; n = 128). In their daily self-monitoring BG measurement, a large proportion of patients were used to considering the first blood drop (88.7%; n = 173), and a limited proportion were used to considering the second drop (11.3%; n = 22). The majority (86.7%; n = 169) had never received recommendations from health care professionals regarding which drop to consider in order to obtain an accurate BG measurements, and the decision of which one to consider was made autonomously. BG concentrations were as follows: 109 patients had a BG concentration b 178 mg/dL; 78 had a BG concentration from 179 to 358 mg/dL; and 8 had a BG concentration of 359 mg/dL or higher. No case crossed groups between the first and second measurement. Table 1 relays the descriptive statistics for the BG measures: the differences, both in the averages and in the medians, were limited to around 3 mg/dL and 6 mg/dL, respectively, indicating that the second drop contained higher glucose values that were highly correlated, statistically different, but
■ • ■
WWW.JENONLINE.ORG
Master Proof ymen2954.pdf
3
RESEARCH ARTICLE/Palese et al
TABLE 1
TABLE 2
Blood glucose reported in the first and second drop of blood: descriptive statistics
Blood glucose reported in the first and second drop of blood
Blood glucose measures
First drop
BG measures
Mean (mg/dL) Median (mg/dL) Minimum-maximum (mg/dL) Percentile 25% (mg/dL) 50% (mg/dL) 75% (mg/dL)
184.3 187.6 + 3.3 166 172 + 6.0 77-455 77-467 0-10 135 166 222
Second Difference drop
140 172 222
5 6 0
clinically comparable with the first drop measured in the same finger a few seconds earlier. The results for the tests of the first hypothesis, that there are no differences between the first and second drop of blood, are depicted in Figure 1 and Table 2. As reported in Figure 1, a strong correlation has emerged between the BG level reported in the first and the second
FIGURE 1 Blood glucose reported in the first and second drop of blood: correlation.
4
First drop Second drop
N
Average Rank sum rank
Negative 123 a 97.83 ranks 88.21 Positive ranks 65 b c Equal values 7 Total 195 Z score − 4.218, P b .001
12,032.50 5,733.50
BG, Blood glucose. a Number of cases first drop BG measure was less than second drop BG measure. b Number of cases first drop BG measure was greater than second drop BG measure. c Number of cases first drop BG measure was equal to second drop BG measure.
drops (rs = 0.979, P b .001; Pearson r = 0.978, P b .001). The average BG level obtained from the first drop was 184.30 mg/dL (median, 166), and from the second drop the average BG level was 187.6 mg/dL (median, 172), as reported in Table 2. The differences were statistically significant. As reported in Table 2, in 123 measurements (63%), the BG reported from the second drop was higher (from + 1 to + 57 mg/dL) compared with the values reported from the first drop. In particular, in 81.3% of the cases (n = 100), the values ranged from + 1 to + 15 mg/dL. However, in 65 cases (33.3%), the second drop reported lower BG values (range − 1 to − 53 mg/dL) compared with those in the first drop; in the majority of cases (78.5%; n = 51), these values ranged from − 1 to − 15 mg/dL. Finally, only 7 cases (3.6%) had the same BG values in both measurements. The Bland-Altman plot in Figure 2 shows that some evaluations reported higher BG values in the second drop capillary sample, others reported higher values in the first drop, and still others reported identical measurements in the first and the second drops. Eight outliers were also present in which a BG difference from − 39 to − 53 mg/dL was reported in the first drop compared with the second drop (n = 5, − 39, − 40, − 40, − 42, and − 53 mg/dL, respectively) and from + 46 to + 57 mg/dL in the first drop as compared with the second drop (n = 3, + 46, + 48, and + 57, respectively). Therefore, the BG measurement performed with the second drop was not higher in all occasions. The differences that emerged in the BG measurements between the first and second drops were not significantly affected by glucose concentration, supporting our second hypothesis. In fact, when the cases were classified in groups with BG concentration ranging from b 178 mg/dL, from
■ ■ •
JOURNAL OF EMERGENCY NURSING
Master Proof ymen2954.pdf
■
Palese et al/RESEARCH ARTICLE
mg/dl
60
1st drop - 2nd drop
40 +1.96 SD 24,5
20
Mean -3,4
0 -20
-1.96 SD -31,2
-40 -60 0
100
200
300
400
500
600
mg/dl
Mean of 1st drop and 2nd drop Legend: SD, Standard Deaviation
FIGURE 2 Findings from the Bland-Altman plot.
179 to 358 mg/dl, and 359 mg/dL or higher, 17 the same number of evaluations from the first and the second blood samples were classified in each group. Nevertheless, when evaluating statistically the BG differences between the first and second drops for each group created on the basis of the glucose concentration, no significant differences have emerged (P = .221), as reported in Table 3.
Discussion
The purpose of this study was to evaluate differences in BG level between the first and second drop of capillary blood collected from the same site. Although the 2 samples were
TABLE 3
Differences in blood glucose reported by the first and second drop of blood at different glucose concentration Glucose concentration (mg/dL) a
N, first and second drop
Average rank
b 179 mg/dL 180-358 mg/dL N 359 mg/dL Total χ 2 (df) P value
109 78 8 195 3.015 (2) b .221
96.82 96.17 131.88
df, Degrees of freedom. a According to Li et al.17 b Kruskal-Wallis test.
■
strongly correlated, we found small, statistically significant differences between the first and second drop. The first drop was more likely to have a lower value. However, these differences of 3 to 6 mg/dL may not be large enough to have clinical relevance in most cases. Glucose concentration did not significantly affect the differences between the first and second drop, but our descriptive plot (Figure 1) indicates that caution should be used in persons with higher BG levels. Thus, our first hypothesis that there would be no difference in the first or second drop was not supported statistically but may still be appropriate clinically. Our second hypothesis that glucose concentration would not affect the difference between the first and second drop was supported by our results. No consistent guidelines were available about whether to use the first or second drop in capillary BG testing in the home or ED settings. 17,20,24 Consequently, in their daily practice, patients and practitioners still appear to be unsure and have inconsistencies regarding the procedure. For example, approximately 72% of European nurse educators advise patients to wash their hands with water and soap and to use the first drop of blood. 13 When it is not possible to wash the hands, 52% of European educators suggest that the second drop is considered more accurate. The clinical question regarding the use of the first or the second drops in BG evaluations has been evaluated in several previous studies. 15–17,23,25 Most of these previous studies focused on avoiding false hyperglycemia results in patients with soiled fingers and unwashed hands, as well as evaluating the accuracy of these measures compared with the venous blood evaluation. Our study is unique in the evaluation of BG values from the first and second drops taken in cleanly washed hands for routine care for type 1 diabetes. Most of the patients in our study had not been given any previous education on which drop of blood to use in home testing, possibly reflecting the inconsistencies in the evidence available. In addition, although a large proportion of patients routinely used the first drop, a few routinely used the second drop. Using the second drop may increase the duration of the procedure, the risk of soiling by blood, or the intensity of pain as a result of use of the milking procedure, which may affect the quality of care in general. In the case of emergency nurses, waiting for the second drop may increase the duration of the procedure and the associated biologic risk. The findings show that there is a statistically significant difference in the BG measured after a few seconds in a sequential of drops obtained from the same fingertip site in the same patient with type 1 diabetes. According to previous data, 14 BG measurements from the first drop tend to be
■ • ■
WWW.JENONLINE.ORG
Master Proof ymen2954.pdf
5
RESEARCH ARTICLE/Palese et al
lower, possibly because of dilution with interstitial fluid. Limited comparison can be made with previous studies available in the field 15–17,23,25 given that different endpoints, procedure, or patient inclusion (persons with diabetes or healthy volunteers) were adopted. Our findings showed that in the majority of the cases (63%), the second drop returned on average higher BG values of approximately 3 to 4 mg/dL compared with the first drop. In the remaining patients (around 33%), higher BG values were found in the first drop. BG values overlapped in only 7 patients. However, the differences observed, both in positive or negative direction, ranged from ± 1 to 15 mg/dL for 80% of our participants. Therefore, the clinical relevance of this difference is limited. However, a difference of + 57 and − 53 as reported by some outliers may be clinically significant and may change treatment decisions. Further study is needed for in-depth understanding of the factors affecting these relevant differences, both at the individual and at the procedural levels. In addition, the difference that emerged in BG as measured in the consequent drops is not significantly affected by BG concentration, in contrast to that reported by Li and colleagues, 17 who in comparing the accuracy of the BG measurements concluded that the first drop was more accurate in the group with BG b 179 mg/dL whereas the second drop was more accurate in the group reporting N 359 mg/dL with respect to the venous evaluations. The sample size was not defined a priori. However, the sample of 195 subjects guarantees, when confronting the difference between the averages of the first and second drop of 3.3 mg/dL, an SD of 14.2 mg/dL and a significance level of 5%, for a power sample equal to 90%.
Limitations
The study was affected by several limitations. The comparison between the first and second drops was made without considering a gold standard measure such as venous BG results, as reported by previous studies (eg, Li et al 17), thus without considering an appropriate comparison method, such as the measurement of the same capillary sample using an analytically robust method. The drops were consecutive and not obtained from a second site or measured with different glucose meters as previous researchers have documented (eg, Li et al 17); in addition, the evaluations were performed after having washed the hands, reflecting the best recommendations available. 20 The same procedure was adopted considering the same hand (left) and the same finger (third), which was aimed at standardizing the data collection process and limiting bias.
6
To our knowledge, evidence is limited to date with regard to the variability of the BG measurements between different hand or finger collection sites. Our findings are limited to persons with type 1 diabetes. Outpatients were tested in the morning, after having consumed breakfast and having washed their hands with soap and water, and they were not suspected of having hypoglycemia. All of these conditions may differ from those encountered in ED settings in which patients’ hands could be unclean, hand washing may be difficult, and patients arrive in different moments of the day, as well as with hypoglycemia. This may have affected the limited variance in the BG emerged, thus not reflecting the variability of the conditions encountered by clinical nurses in emergency departments. Furthermore, a relatively narrow clustering of glucose values around the median has emerged, with no participant showing evidence of significant biochemical hypoglycemia. Accurate detection of hypoglycemia is arguably the most important glucose measurement task undertaken in the emergency room. Therefore, future research should address the full range of conditions representing the complexity of care in emergency departments.
Implications for Emergency Nurses
Nurses working in the emergency department routinely care for patients with diabetes and/or patients who require BG measurement. BG measurement procedures using capillary blood are very common among nurses in the clinical setting and patients in the home setting, and many differences in practice exist with regard to use of the first or second drop of blood for testing. No guidelines are available because of limited studies in the field. Nurses should be aware that the first drop of blood is likely to have a lower BG level. In our study, this difference averaged 3 to 6 mg/dL, which is not likely to be clinically significant in most cases. These differences between the first and second drop were unrelated to the overall BG level (low, normal, or high), although practicing nurses should interpret the findings seen in Figure 1, where the correlations fall further away from the dotted line at higher BG levels, as a sign that continued caution and research on the topic is needed for high BG measurements. Our findings only have implications for measurements from patients’ cleanly washed hands, because our findings were different from previous ones in which BG measurements were performed in patients whose hands were exposed to dirt (soiled fingers and unwashed hands). We did not compare our results with venous blood measurements, which would be common in the ED setting.
■ ■ •
JOURNAL OF EMERGENCY NURSING
Master Proof ymen2954.pdf
■
Palese et al/RESEARCH ARTICLE
Therefore, emergency nurses may educate patients that the first drop of blood in home testing is likely to be adequate but also likely to demonstrate slightly lower values than the second drop. Emergency nurses also may reinforce that patients who already use the first drop should continue that practice. Our recommendations to practicing nurses are based on a limited amount of evidence, including our study, and nurses should continue to conduct and evaluate continued research with patients admitted in emergency departments to evaluate if the differences that emerged in this study are similar. Our findings did not apply to all participants. Patients with extreme differences, or outliers, between the first and second drop of ± 50 mg/dL require further study.
9. Harding AD. Hypoglycemia in the emergency department. J Emerg Nurs. 2012;38(4):362-363. 10. Cryer PE. Diverse causes of hypoglycemia-associated autonomic failure in diabetes. N Engl J Med. 2004;350(22):2272-2279. 11. Bergman M. Preface. Prediabetes and diabetes prevention. Med Clin North Am. 2011;95(2):xi-xiii. 12. Hoerger TJ, Hicks KA, Sorensen SW, et al. Cost-effectiveness of screening for pre-diabetes among overweight and obese U.S. adults. Diabetes Care. 2007;30(11):2874-2879. 13. Hortensius J, van der Bijl JJ, Kleefstra N, Houweling ST, Bilo HJ. Selfmonitoring of blood glucose: professional advice and daily practice of patients with diabetes. Diabetes Educ. 2012;38(1):101-107. 14. Fruhstorfer H, Quarder O. Blood glucose monitoring: milking the finger and using the first drop of blood give correct glucose values. Diabetes Res Clin Pract. 2009;85(1):14-15.
Conclusion
The findings from this study support that the first or second drop of BG is adequate for clinical decision making if the patient has cleanly washed hands, but the clinician should use judgment if using protocols in which small values (eg, 6 mg/dL) are important because the first drop is more likely to have a slightly lower value. The differences in BG are not related to the overall BG level (high, normal, or low). Using the first drop in a patient with cleanly washed hands allows emergency nurses to conduct testing more quickly, resulting in immediate information. REFERENCES 1. International Diabetes Federation. IDF Diabetes Atlas. 7th ed., Brussels, Belgium: International Diabetes Federation; 2015. http://www. diabetesatlas.org/. Accessed April 13, 2016. 2. Rouse MD, Shoukry CL. Elevated blood glucose levels in the emergency department: missed opportunities. J Emerg Nurs. 2014;40(4):311-316.
15. Hortensius J, Kleefstra N, Slingerland RJ, et al. The influence of a soiled finger in capillary blood glucose monitoring. Neth J Med. 2010;68(1): 330-331. 16. Hortensius J, Slingerland RJ, Kleefstra N, et al. Self-monitoring of blood glucose: the use of the first or the second drop of blood. Diabetes Care. 2011;34(3):556-560. 17. Li M, Wang X, Shan Z. Deciding between using the first or second drop of blood for the self monitoring of blood glucose. Prim Care Diabetes. 2014;8(4):365-369. 18. Maclure M, Mittleman MA. Should we use a case-crossover design?. Annu Rev Public Health. 2000;21:193-221. 19. Center for Devices and Radiological Health, US Food and Drug Administration. Draft guidance document: review criteria assessment of portable blood glucose monitoring in vitro diagnostic devices using glucose oxidase, dehydrogenase or hexokinase methodology. http:// www.fda.gov/ohrms/dockets/dockets/06p0313/06p-0313-cp00001-10Exhibit-I-CDRH-ODE-Gluc-vol1.pdf. Accessed April 13, 2016. 20. Clar C, Barnard K, Cummins E, Royle P, Waugh N. Aberdeen Health Technology Assessment Group. Self-monitoring of blood glucose in type 2 diabetes: systematic review. Health Technol Assess. 2010;14(12):1-140.
3. Bernard JB, Munoz C, Harper J, Muriello M, Rico E, Baldwin D. Treatment of inpatient hyperglycemia beginning in the emergency department: a randomized trial using insulin aspart and detemir compared with usual care. J Hosp Med. 2011;6(5):279-284.
21. LifeScan. OneTouch Verio®IQ. http://www.lifescan.it/ourproducts/ meter/one-touch-verio-iq. Accessed April 13, 2016.
4. McCowen KC, Malhotra A, Bistrian BR. Stress-induced hyperglycemia. Crit Care Clin. 2001;17(1):107-124.
22. Roche Diagnostics. Accu-Chek Mobile. https://www1.accu-chek.it/ documents/Accu-Chek_Mobile_Manuale.pdf. Accessed April 13, 2016.
5. Sperry JL, Frankel HL, Vanek SL, et al. Early hyperglycemia predicts multiple organ failure and mortality but not infection. J Trauma. 2007; 63(3):487-493.
23. American Diabetes Association. eAG/A1C conversion calculator. http:// professional.diabetes.org/eAG. Accessed April 16, 2016.
6. Wahl WL, Taddonio M, Maggio PM, Arbabi S, Hemmila MR. Mean glucose values predict trauma patient mortality. J Trauma. 2008;65(1):42-47. 7. Kreutziger J, Schlaepfer J, Wenzel V, Constantinescu MA. The role of admission blood glucose in outcome prediction of surviving patients with multiple injuries. J Trauma. 2009;67(4):704-708.
■
8. Bochicchio GV, Sung J, Joshi M, et al. Persistent hyperglycemia is predictive of outcome in critically ill trauma patients. J Trauma. 2005; 58(5):921-924.
24. American Diabetes Association. Standards of medical care in diabetes—2014. Diabetes Care. 2014;37(Suppl. 1):S14-S80. 25. Palese A, Chiandetti R, Mansutti I. The effect of 2% chlorhexidine and cleansing wipes on capillary blood glucose sampling accuracy of a fingertip soiled with apple: a case crossover study design. J Clin Nurs. 2014;23(17–18):2672-2674.
■ • ■
WWW.JENONLINE.ORG
Master Proof ymen2954.pdf
7