Comparison of Laboratory Values Obtained by Phlebotomy versus Saline Lock Devices Jill Corbo, MD, RDMS, Lequan Fu, MD, Matthew Silver, MD, Hany Atallah, MD, Polly Bijur, PhD, MPH
Abstract Objectives: To assess the utility of a peripheral saline lock device (SLD) as an alternative to a second venipuncture for obtaining selected blood samples. Methods: This prospective study used a comparative design and was conducted in an urban emergency department (ED). Adult patients with an existing SLD in place who required serial phlebotomy were eligible for inclusion in the study. Each subject had blood samples obtained by venipuncture (control) with a Vacutainer adapter according to standard protocols. Within 5 minutes of obtaining the control samples, a sample was obtained from the patients’ SLDs; a tourniquet was applied proximal to the intravenous line, a 5-mL waste portion was obtained, and a Vacutainer adapter was placed to draw specimens for testing. Each of the paired samples was analyzed for hematocrit, electrolytes, and cardiac enzymes. The Bland-Altman method was used to analyze the concordance between each pair of measurements. Paired t-tests for each of the eight laboratory tests were used to assess whether the values were statistically different from each other. The 95% limits of agreement around the mean differences were calculated. Differences between SLD aspirates and venipuncture aspirates also were compared with the federal regulatory standards that ensure reliable and accurate laboratory testing. Results: Eighty-one patients were eligible for the study; in 73 (90.1%; 95% confidence interval [CI] = 81.5% to 95.6%) of the patients, the SLD could be aspirated for testing. The paired t-tests indicated that there were no statistically significant differences between the mean values of the two methods of testing. Of the 584 paired values analyzed, 35 (6.0%; 95% CI = 4.3% to 8.2%) exceeded the Bland-Altman limits of agreement, and 43 (7.4%; 95% CI = 5.4% to 9.8%) fell outside the acceptable range determined by the federal regulation of clinical laboratories. Of those values that exceeded the acceptable Bland-Altman limits of agreement, none would have resulted in clinical intervention. Conclusions: Aspirating blood via an SLD is an acceptable method of obtaining serial laboratory values in a group of stable, consenting adult ED patients. ACADEMIC EMERGENCY MEDICINE 2007; 14:23–28 ª 2007 by the Society for Academic Emergency Medicine Keywords: phlebotomy, saline lock devices
P
atients in the emergency department (ED) frequently are subjected to multiple blood draws for the purpose of obtaining new clinical data or to monitor serial values over time. With the trend of ED overcrowding leading to longer ED stays,1 one can
From the Department of Emergency Medicine, Jacobi Medical Center, Albert Einstein College of Medicine (JC, LF, MS, HA, PB), Bronx, NY. Dr. Fu currently is at Bronx-Lebanon Hospital Center, Bronx, NY; Dr. Atallah currently is at Emory University School of Medicine, Atlanta, GA. Presented at the Society for Academic Emergency Medicine Annual Meeting, Orlando, Florida, May 2004. Received January 16, 2006; revisions received May 16, 2006, and June 22, 2006; accepted June 23, 2006. Contact for correspondence and reprints: Jill Corbo, MD, RDMS; e-mail:
[email protected].
ª 2007 by the Society for Academic Emergency Medicine doi: 10.1197/j.aem.2006.06.053
assume that the frequency of obtaining serial phlebotomies from some patients will be increasing. Common practice, which is not evidence based, dictates that laboratory samples are obtained by means of a second venipuncture, despite the presence of an existing peripheral catheter. The use of an existing catheter as a sampling site offers potential benefits to patients and staff.2 Serial blood draws are painful to patients, potentially could increase infection rate, and may also increase the risk of needle sticks to health care providers.3 Central venous lines have been accepted widely as an alternative route for obtaining serial blood tests.4 Infusing intravenous lines also have been shown to be a reliable tool for blood specimen collection.5–7 Recent research has examined the use of saline lock devices (SLDs) as a mean of obtaining blood samples.8–11 Past studies have used SLDs with running infusate, and many have looked specifically at isolated hematology, chemistry, or coagulation panels.10 To the best of our
ISSN 1069-6563 PII ISSN 1069-6563583
23
24
knowledge, no studies have used needleless adapters as a means for obtaining samples from the existing SLD. The use of a needleless method of obtaining serial phlebotomy would decrease the potential risk of blood exposure to the health care providers. The objective of this study was to evaluate equivalence of eight commonly performed laboratory tests when comparing blood specimens obtained by means of standard venipuncture with those obtained through SLDs. Secondary objectives included evaluation of the success rate of aspiration from the SLD with a needleless Vacutainer (Becton, Dickinson and Company, Franklin Lakes, NJ) and to determine whether there was an increased rate of hemolysis in blood samples that are obtained by this needleless method.
METHODS Study Design This was a prospective, comparison study conducted in an urban ED to evaluate whether a series of eight common laboratory tests are equivalent when obtained either through an SLD or venipuncture. The institutional review board of our hospital approved the study, and informed consent was obtained before enrollment. Study Setting and Population The study was conducted over a two-month period in the adult ED of an urban, tertiary-care, Level 1 trauma center. The ED has an estimated volume of 70,000 patient visits per year. Adult patients with an existing SLD in place who required repeat laboratory tests were eligible for inclusion in the study. Patients were excluded if they refused consent or were incapable of giving consent. Study Protocol A convenience sample of patients was enrolled once eligibility and exclusion criteria were confirmed and informed consent was obtained. Each subject had blood samples obtained by venipuncture (control) with a 21-G butterfly needle attached to a Vacutainer adapter, with direct aspiration of blood into three vacuum-prepared specimen tubes. Within 5 minutes of obtaining the control samples, specimens were obtained from the patient’s SLD. The study specimens were obtained in the following manner: if infusion of solution was occurring through the SLD, it was halted 2 minutes before placement of the tourniquet. The tourniquet was applied proximal to the SLD, the SLD was wiped with an alcohol pad, and a 5mL syringe was attached to the SLD. Five milliliters of blood were drawn and discarded as waste, and a needleless Vacutainer adapter then was inserted into the SLD. Three vacuum-prepared specimen tubes were filled for testing. On the basis of prior research, a 5-mL discarded sample, drawn within 2 minutes of discontinuation of infusate, is acceptable to ensure no contamination of blood specimens by the infusate.6,7 Venipuncture and SLD samples were analyzed for hematocrit with a GenS 33425 analyzer (Beckman Coulter, Inc., Fullerton, CA); serum sodium, potassium, bicarbonate, chloride, glucose, and creatine phosphokinase (CPK) were analyzed with the Bayer-Technicon 1650 (Bayer HealthCare Diag-
Corbo et al.
COMPARISON OF LABORATORY VALUES
nostics Division, Tarrytown, NY); and troponin I was analyzed with an Access 2 analyzer (Beckman Coulter). At our institution, the troponin I reference range of 97.5 percentile for healthy individuals is less than 0.03 mg/L, and the laboratory cutoff for positive test is R0.10 mg/L. Venipuncture and SLD samples were analyzed with the same laboratory machines. Also analyzed were the success rates of aspiration relative to intravenous line (IV) cannula size, as well as rates of adverse events such as hemolysis, clotting, or IV infiltration. Federal guidelines set the maximum allowable analytical error for laboratories in the Clinical Laboratory Improvement Amendment (CLIA) of 1988.12 CLIA specifies national standards to ensure that laboratories are performing these tests with acceptable accuracy and reliability. Thus, for each test, there is a set value or percentage that is the acceptable variance between test results. For example, the laboratory test for serum sodium has a permissible analytical error of a set value of 4 mEq/L, so a sodium level of 140 mEq/L could have a value from 136 to 144 mEq/L because of laboratory error alone. The laboratory test for serum glucose has a percentage permissible analytical error of 10%, so a serum glucose of 100% could have a value from 90% to 110% because of laboratory error alone. Results from two different techniques of obtaining blood samples can be a result of the analytical error of the process of testing of the laboratory, or from the techniques used (venipuncture vs. SLD). By using the CLIA standards, the variation of the process of analysis is removed. Data Analysis We used the Bland-Altman13 method to analyze the concordance between each pair of measurements. Paired t-tests of each of the eight laboratory tests were used to assess whether the values were statistically different from each other. Differences between values were inferred to be statistically significant at a p-value of
