- Email: [email protected]
The Effects of Heparin Versus Normal Saline for Maintenance of Peripheral Intravenous Locks in Pregnant Women
PDFlib PLOP: PDF Linearization, Optimization, Protection Page inserted by evaluation version www.pdflib.com – [email protected]
CLINICAL RESEARCH
The Effects of Heparin Versus Normal Saline for Maintenance of Peripheral Intravenous Locks in Pregnant Women Kathryn M. Niesen, Denise Y. Harris, Linda S. Parkin, Lynn T. Henn
Objective: To compare the efficacy of two available preparations (heparin, 10 U/mL, 1 mL, vs. normal saline, 1 mL) used for maintaining patency in peripheral intravenous (IV) locks during pregnancy. Design: Prospective, randomized, and doubleblind. Eligible patients who were to receive a peripheral intermittent IV lock were randomly assigned to receive either heparin flushes or normal saline flushes for IV lock maintenance. IV locks were flushed after each medication administration, or at least every 24 hours, with the assigned blinded flush solution. Intermittent IV lock sites were also evaluated every 12 hours for the development of phlebitis. Setting: A large academic medical center in the Midwest that has both community-based and regional-referral obstetric practices with more than 2,000 deliveries per year. Participants: A convenience sample included 73 hospitalized pregnant women who were between 24 and 42 weeks gestation. Exclusions from the study were women with significant abnormalities in the fetal heart tracing on admission, cervical dilation > 4 cm, presence of hypersensitivity to heparin, presence of clotting abnormalities, and anticoagulation therapy (including low-dose aspirin). Results: Data indicate there were no statistically significant differences in IV lock patency nor in phlebitis between heparin or normal saline flushes. Conclusions: This study provides support that both normal saline and heparin in the doses studied may be equally effective in the maintenance of peripheral IV locks. Due to small sample size, additional studies are needed to determine optimal therapy over
July/August 2003
time. JOGNN, 32, 503–508; 2003. DOI: 10.1177/ 0884217503255203 Keywords: Heparin/saline lock—IV lock maintenance—Peripheral intermittent intravenous devices— Peripheral IV lock—Pregnancy Accepted: July 2002 Peripheral intermittent intravenous (IV) locks are commonly used as a convenient, readily accessible route for fluid or medication administration. They provide the patient with increased mobility and decrease costs in patients who do not need continuous IV infusion. Among the potential problems that could be encountered when using IV locks are loss of patency, decreased duration of use of the catheter, and phlebitis. The Intravenous Nurses Society (INS) has stated that the accepted phlebitis rate is 5% or less in any given population (INS, 2000). To prevent thrombosis at the IV site, catheters are commonly flushed after each administration of IV medication and every 8 to 24 hours when the device is not in use (Dunn & Lennihan, 1987; Tuten & Gueldner, 1991). Traditionally, dilute solutions of heparin in 0.9% sodium chloride injection (e.g., 10-100 U/mL) have been used for the periodic flush. Ongoing controversy exists, however, as to the optimal type and amount of flush solution. In fact, staff nurses on the obstetric units at this Midwestern teaching tertiary referral medical center perceived heparin to be more effective clinically than normal saline for IV lock maintenance in pregnant women, but were unable to locate supporting literature.
JOGNN 503
Thus, the purpose of this study was to compare the efficacy of two commonly used peripheral IV lock maintenance solutions, heparin, 10 U/mL, 1 mL, versus normal saline, 1 mL, in pregnant women who were between 24 and 42 weeks gestation. In 1990, the American Association of Critical Care Nurses conducted a nationwide random sample survey of members (N = 1,072) to examine the practices for maintaining IV lock patency. This survey revealed that 40% of respondents used 100 U/mL of heparin, 37% used 10
P
regnancy is an acquired hypercoagulable state.
U/mL of heparin, 18% used normal saline only, and the remaining 5% used other dilutions or solutions. Goode et al. (1991) conducted a meta-analysis to estimate the effects of heparin flush versus saline flush solutions on IV site patency, incidence of phlebitis, and duration of peripheral heparin locks. The total sample sizes of the meta-analysis were 3,490 participants across 15 studies for patency; 2,356 participants across 13 studies for phlebitis; and 1,960 participants across 6 studies for duration. Their results indicated that heparin is no more effective than saline in preventing clotting, in decreasing the incidence of phlebitis, or in increasing the duration (number of hours) that an IV lock lasts (p < .05). All patients included in the meta-analysis were from the adult medical, surgical, or critical care populations, except for one study with 74 pediatric patients. Goode et al. (1991) recommended their results not be generalized to children or neonates, as this one small study does not provide enough information to warrant a change in practice. None of these studies included pregnant women. The American Society of Hospital Pharmacists (1994) issued a position statement that 0.9% sodium chloride (normal saline) injection is a safe and effective means for maintaining catheter patency of peripheral, indwelling, intermittent infusion devices (e.g., IV locks) in all adult patients in institutional settings. Obstetric patients were not specifically addressed, however. This statement also recommended that until further controlled studies are conducted, heparin should be administered to pediatric patients in the smallest amount necessary to maintain patency of the device. Pregnancy is an acquired hypercoagulable state (Cerneca et al., 1997; Comeglio et al., 1996; Cunningham et al., 2001; Peric-Knowlton, 1992). The term hypercoagulable implies an increased tendency for arterial or venous thrombosis. Pregnancy normally induces appreciable increases in the concentrations of coagulation factors I 504 JOGNN
(fibrinogen), VII (proconvertin), VIII (antihemophilic factor), IX (plasma thromboplastin component), and X (Stuart factor). These hemostatic changes during pregnancy are thought to result in an ongoing low-grade activation of the coagulation system beginning as early as 10 weeks gestation and lasting through the first 8 weeks after delivery (Cerneca et al., 1997; Sipes & Weiner, 1990). It has been suggested by Blackburn and Loper (1992) and in Gerbasi, Bottoms, Farag, and Mammen’s study (as cited in Cunningham et al., 2001) that this hypercoagulability is normal, protecting the placental implantation site throughout pregnancy and the woman from excessive blood loss at delivery by providing rapid hemostasis following placental removal. Meyer, Little, Thorp, Cohen, and Yeast (1995) questioned whether pregnant women were more likely to experience the failure of intermittent IV locks because of the normal physiological changes of pregnancy that result in a hypercoagulable state. They hypothesized that in pregnancy use of heparin flushes would improve intermittent IV lock patency, compared with normal saline flushes. In a double-blind study including 44 pregnant women who were between 26 and 34 weeks gestation and required serial phlebotomy, Meyer et al. compared heparin sodium (100 U/mL) with normal saline for use in the maintenance of IV locks. Catheter sites were examined and flushed with the study solution at least once every 6 hours. Partial thromboplastin times were also measured at catheter insertion and 48 hours later. Although no differences were found in partial thromboplastin times, there was a significant increase in catheter patency rate at 48 and 72 hours in the heparin group (84% vs. 52% and 68% vs. 27%, respectively; p < .01). There also was a significantly lower rate of catheter complications in the heparin group (13% vs. 31%, p < .01). Only one study (Meyer et al., 1995) was found that compared the efficacy of heparin with normal saline flushes to maintain peripheral IV lock patency in pregnant patients. Therefore, a prospective, randomized clinical trial to compare the efficacy of heparin, 10 U/mL, 1 mL, and normal saline, 1 mL, flushes to maintain peripheral IV locks in pregnant women was conducted. This investigation adds to the scarce amount of literature covering this topic for pregnant women and provides the necessary data comparing a lower dosage of heparin (10 U) with normal saline for maintenance of peripheral IV locks in pregnant women who are between 24 and 42 weeks gestation.
Design and Methods A prospective, randomized, double-blind design was used to compare the efficacy of using heparin, 10 U/mL, 1 mL, versus normal saline, 1 mL, for maintaining patency in peripheral IV locks in pregnant women. The null hypotheses identified by the investigators for this study Volume 32, Number 4
were (a) there is no difference in patency when peripheral IV locks in pregnant women are flushed with heparin, 10 U/mL, 1 mL, or when they are flushed with normal saline, 1 mL, and (b) there is no difference in the incidence of phlebitis when peripheral IV locks in pregnant women are flushed with heparin, 10 U/mL, 1 mL, or when they are flushed with normal saline, 1 mL.
Sample Participants were 73 pregnant women who were between 24 and 42 weeks gestation, required peripheral intermittent IVs, and were hospitalized in the obstetric units at a large Midwestern teaching hospital. Exclusion criteria included cervical dilation greater than 4 cm; presence of any significant abnormalities in the fetal heart tracing on admission; presence of hypersensitivity to heparin; presence of clotting abnormalities; and treatment with anticoagulant medication, including aspirin. The medical center’s institutional review board approved the study protocol.
Procedure After their initial assessment and provision of informed consent, eligible women who were to receive a peripheral intermittent IV lock were entered into the study per a random allocation list generated by the statistician. They were assigned treatment with either normal saline or heparin for IV lock maintenance. Each woman received placement of a 16- or 18-gauge IV catheter, with an extension set (IV Catheter Extension Set, InterLink System, Baxter Healthcare Corp, Deerfield, IL) attached. After placement, the extension set and catheter (IV lock) were flushed with the study solution. The participants were assigned to have IV locks flushed with one of two solutions: preservative-free normal saline, 1 mL; or heparin, 10 Units/mL, 1 mL. The hospital pharmacy prepared batches of the two treatment solutions, using identical syringes labeled as study flush solution. Only the pharmacy staff had access to the randomization list and the code. When a pregnant woman was entered into the study, the dispensing pharmacist assigned her the next available number and dispensed treatment flushes labeled with the patient study number. A heparin/saline studyalert sticker with the coded study number was placed on the patient’s chart and room door. Once enrolled in the study, participants received only the assigned solution throughout the duration of that peripheral IV lock. If their IV lock needed to be restarted, those participants were out of the study. Data collected before the IV restart were included in the analysis, however. All IV locks were assessed by registered nurses for complications at least every 12 hours (more often sooner, which was related to hanging fluids), and findings were documented. The IV locks were flushed at least every 24 hours or after each medication administration, whichever was more frequent. July/August 2003
Staff RNs providing care to the patients on the unit collected the data. The nurses were educated about the study at meetings conducted by the principal investigator. Because there were multiple data collectors, the RNs received information weekly at follow-up question-andanswer sessions conducted by members of the research team. The research team frequently offered food and beverages at these meetings, which served the secondary purpose of keeping data collectors enthused about the study. Content included the purpose of the study, study questions, inclusion/exclusion criteria, how to obtain consent, and review of the data collection tool. A copy of the study protocol was placed on each participating nursing unit, along with a step-by-step instruction list and a study flow diagram for nurses to follow. A member of the research team made rounds almost daily to assist staff nurses who were data collectors with questions, and to motivate them to complete the data collection tools. Specific information collected included the patient’s age, race, gravidity and parity, gestational age in weeks, time and date of IV insertion and discontinuation, IV location, number of flushes given, and IV complications. Complications of IV therapy were defined as resistance with flush or fluid infusion and phlebitis. Phlebitis was further defined as presence of pain, erythema, palpable venous cord, induration, and ecchymosis, as adapted from the INS Phlebitis Rating Scale (INS, 2000). Patients were enrolled and data were collected for 15 months. Because assessments were made at least every 12 hours or when initiating fluid infusion or administering
T
his study provides support that both normal saline, 1 mL, and heparin, 10 U/mL, 1 mL, may be effective in the maintenance of peripheral IV locks.
medication, most participants had only one assessment. For example, it was common for a pregnant woman to have an IV lock for a period of time before the onset of active labor, but this time period was usually less than 12 hours. When labor became active, the infusion of IV fluids was started. Outcomes measured and compared in the two groups included the patency of the peripheral IV lock for each patient within 12 hours and the incidence of phlebitis for each patient within 12 hours. Descriptive statistical analysis techniques were used for demographic information, and chi-square and Fisher’s exact tests were used for categorical data. Statistical significance was defined as p < .05.
JOGNN 505
TABLE 1
Demographic Characteristics of Participants (N = 73) Saline n = 35 Maternal age Mean SD Range Gestation Mean SD Range
pvalue* 0.41
28.8 5.8 18-41
29 5.6 20-40
39.5 2.2 32-42
37.7 4 25-42
.02
Saline n (%) Gravida 1 2 3+ Para 0 1 2+
Heparin n = 38
Heparin n (%)
pvalue* .07
15 (42.9) 6 (17.1) 14 (40.0)
15 (39.5) 15 (39.5) 8 (21.0)
17 (48.6) 12 (34.3) 6 (17.1)
20 (52.6) 14 (36.8) 4 (10.6)
.71
*Significant at p < .05.
Results The final sample consisted of 73 IV sites in 73 participants, 35 in the normal saline group and 38 in the heparin group. The sample was 97.1% White and 2.9% other (missing data = 4). This was consistent with the population seen at the time of the study. Groups were similar for maternal age, gravidity, and parity (see Table 1). The saline group did have a significantly higher gestational age (p = .02). The original intent of the study was to measure and compare outcomes in the two groups within 12, 24, 48, and 72 hours. However, assessments of IV lock status were rare beyond 12 hours, generally because of the initiation of fluid infusion. Because there were insufficient data to make comparisons in the 24-hour, 48-hour, and 72-hour groups, only the assessments within 12 hours for patency are reported here. IV lock patency was determined by the ability of the nurse to flush the catheter freely without obstruction, or if initiated, by IV fluids infusing freely. Of the 35 IV locks flushed with saline, 68.6% had no obstruction within 12 hours, as compared to no obstruction within 12 hours in 76.3% of the IV locks flushed with heparin (n = 38). This difference in
506 JOGNN
patency was not statistically significant, as shown by a chi-square test (p = .459). The following catheter complications related to phlebitis were reported: Three participants from the normal saline group (n = 35) and 2 participants from the heparin group (n = 38) reported pain at the catheter site. One participant from each group had erythema at the catheter site. One participant in the saline group had a palpable venous cord, and 1 participant had ecchymosis. Included in the aforementioned complications was 1 participant whose IV lock would not flush freely and who had pain, erythema, palpable venous cord, and ecchymosis. None of these complications demonstrated statistically significant differences between groups, using Fisher’s exact test.
Discussion In this randomized, double-blind trial, there were no statistically significant differences in IV lock patency, nor in phlebitis with normal saline, 1mL, or heparin 10 U/ml, 1mL, used for peripheral IV lock maintenance. Both of the null hypotheses identified by the investigators for this study were supported. Admittedly, the sample size for this study is small. For the first hypothesis, the heparin group had a patency rate of 76%, whereas the saline group had a patency rate of 69%. This study had only 11% power to detect a difference of this magnitude. However, this study does rule out large differences in patency rates between saline and heparin, as there was 80% power to detect odds ratios for patency of 40 or larger. The limitations in the statistical power are even more pronounced for the second hypothesis, because the complication rates were less than 10% in each study group.
A
dditional studies in this population are needed. Current evidence does not indicate clear direction for practice.
This study provides support that both normal saline, 1 mL, and heparin, 10 U/mL, 1 mL, may be effective in the maintenance of peripheral IV locks. However, normal saline is inherently safer than heparin. One of the main reasons the American Society of Hospital Pharmacists (1994) recommended switching to normal saline flush was to minimize patient exposure to heparin and thus possibly decrease a rare but significant adverse reaction that can develop in heparin-exposed patients, which is called heparin-induced thrombocytopenia.
Volume 32, Number 4
TABLE 2
Participants With Phlebitis Complications Within 12 Hours Participant # 13 53 69 47 17 20 p-value
Study Group Heparin Heparin Heparin Saline Saline Saline
Pain
Erythema
Yes Yes No Yes Yes Yesa 0.67
No No Yes No No Yesa 1.00
Palpable Cord No No No No No Yesa 0.48
Induration
Ecchymosis
No No No No No No 1.00
No No No No No Yesa 0.48
Note. No other participants experienced complications. Heparin 35/38 (92.1%); saline 32/35 (91.4%). Used Fisher’s exact test rather than chisquare owing to low complication rates. aSame
participant.
The findings of this study differ from those of Meyer et al. (1995) in their study of 44 pregnant women who were between 23 and 34 weeks gestation. Meyer et al. found that the use of heparin sodium, 100 U/mL, 1 mL, resulted in a significant increase in catheter patency and lower rates of catheter complications as compared with the use of normal saline. However, Meyer et al. used 10 times the strength of heparin (100 U/mL, 1 mL) compared with this study (10 U/mL, 1 mL), and their IV locks remained in place much longer.
Implications for Nursing Although the findings of this study provide support that the use of normal saline, 1 mL, to flush IV locks in obstetric patients may be as effective as heparin, 10 U/mL, 1 mL, the results must be interpreted with caution. As with all clinical studies using very small samples, the statistical power in this study was low. This means that it is possible that a difference between population groups might have been detected with a larger sample. In addition, because fluids may have been infused soon after the IV lock was initiated, only the within-12-hour data were used. Current evidence does not indicate clear direction for practice at this time. Additional studies comparing different strengths of heparin and normal saline, which address the precise dwell time of the catheter in this population of pregnant women, are needed to better determine optimal therapy over time. Also, future investigators’ use of the Intravenous Nurses Society Phlebitis Rating Scale would provide a uniform measurement standard across multiple studies (INS, 2000).
July/August 2003
REFERENCES American Association of Critical Care Nurses. (1990, August). Nationwide practice survey results announced: Survey studied heparin use. AACN News, p. 3. American Society of Hospital Pharmacists. (1994). ASHP therapeutic position statement on the institutional use of 0.9% sodium chloride injection to maintain patency of peripheral indwelling intermittent infusion devices. American Journal of Hospital Pharmacy, 51, 1572-1574. Blackburn, S. T., & Loper, D. L. (1992). Maternal, fetal, and neonatal physiology: A clinical perspective. Philadelphia: Saunders. Cerneca, F., Ricci, G., Simeone, R., Malisano, M., Alberico, S., & Guaschino, S. (1997). Coagulation and fibrinolysis changes in normal pregnancy. European Journal of Obstetrics, Gynecology, & Reproductive Biology, 73(1), 31-36. Comeglio, P., Fedi, S., Liotta, A. A., Cellai, A. P., Chiarantini, E., Prisco, D., et al. (1996). Blood clotting activation during normal pregnancy. Thrombosis Research, 84(3), 199-202. Cunningham, F. G., Gant, N. F., Leveno, K. J., Gilstrap, L. C., Hauth, J. C., & Wenstrom, K. D. (2001). Obstetrical hemorrhage: Consumptive coagulopathy. In Williams obstetrics (21st ed., pp. 619-669). New York: McGraw-Hill. Dunn, D., & Lennihan, S. (1987). The case for the saline flush. American Journal of Nursing, 87, 798-799. Goode, C. J., Titler, M., Rakel, B., Ones, D. S., Kleiber, C., Small, S., et al. (1991). A meta-analysis of effects of heparin flush and saline flush: Quality and cost implications. Nursing Research, 40(6), 324-330. Intravenous Nurses Society. (2000). Infusion nursing standards of practice. Journal of Intravenous Nursing, 23(Suppl. 6), S56. Meyer, B. A., Little, C. J., Thorp, J. A., Cohen, G. R., & Yeast, J. D. (1995). Heparin versus normal saline as a peripheral line flush in maintenance of intermittent intravenous
JOGNN 507
lines in obstetric patients. Obstetrics and Gynecology, 85(3), 433-436. Peric-Knowlton, W. (1992). Acquired hypercoagulable states. Nurse Practitioner Forum, 3(2), 72-81. Sipes, S. L., & Weiner, C. P. (1990). Venous thromboembolic disease in pregnancy. Seminars in Perinatology, 14, 103. Tuten, S. H., & Gueldner, S. H. (1991). Efficacy of sodium chloride versus dilute heparin for maintenance of peripheral intermittent intravenous devices. Applied Nursing Research, 2, 63-71.
Kathryn M. Niesen, MSN, RN, is director of clinical nursing, Department of Nursing, Mayo Clinic, Rochester, MN, and is an assistant professor of nursing, Mayo Medical School, Rochester, MN.
508 JOGNN
Denise Y. Harris, PharmD, is a pharmacist in Hospital Pharmacy Services, Mayo Clinic, Rochester, MN. Linda S. Parkin, RN, is a staff nurse, formerly in the Department of Nursing, Mayo Clinic, Rochester, MN, and now in the Department of Family Medicine. Lynn T. Henn, RN, is a staff nurse, Department of Nursing, Mayo Clinic, Rochester, MN. Address for correspondence: Kathryn M. Niesen, Department of Nursing, Mayo Clinic, 200 First St. SW, Rochester, MN, 55905. E-mail: [email protected].
Volume 32, Number 4
CLINICAL RESEARCH
The Effects of Heparin Versus Normal Saline for Maintenance of Peripheral Intravenous Locks in Pregnant Women Kathryn M. Niesen, Denise Y. Harris, Linda S. Parkin, Lynn T. Henn
Objective: To compare the efficacy of two available preparations (heparin, 10 U/mL, 1 mL, vs. normal saline, 1 mL) used for maintaining patency in peripheral intravenous (IV) locks during pregnancy. Design: Prospective, randomized, and doubleblind. Eligible patients who were to receive a peripheral intermittent IV lock were randomly assigned to receive either heparin flushes or normal saline flushes for IV lock maintenance. IV locks were flushed after each medication administration, or at least every 24 hours, with the assigned blinded flush solution. Intermittent IV lock sites were also evaluated every 12 hours for the development of phlebitis. Setting: A large academic medical center in the Midwest that has both community-based and regional-referral obstetric practices with more than 2,000 deliveries per year. Participants: A convenience sample included 73 hospitalized pregnant women who were between 24 and 42 weeks gestation. Exclusions from the study were women with significant abnormalities in the fetal heart tracing on admission, cervical dilation > 4 cm, presence of hypersensitivity to heparin, presence of clotting abnormalities, and anticoagulation therapy (including low-dose aspirin). Results: Data indicate there were no statistically significant differences in IV lock patency nor in phlebitis between heparin or normal saline flushes. Conclusions: This study provides support that both normal saline and heparin in the doses studied may be equally effective in the maintenance of peripheral IV locks. Due to small sample size, additional studies are needed to determine optimal therapy over
July/August 2003
time. JOGNN, 32, 503–508; 2003. DOI: 10.1177/ 0884217503255203 Keywords: Heparin/saline lock—IV lock maintenance—Peripheral intermittent intravenous devices— Peripheral IV lock—Pregnancy Accepted: July 2002 Peripheral intermittent intravenous (IV) locks are commonly used as a convenient, readily accessible route for fluid or medication administration. They provide the patient with increased mobility and decrease costs in patients who do not need continuous IV infusion. Among the potential problems that could be encountered when using IV locks are loss of patency, decreased duration of use of the catheter, and phlebitis. The Intravenous Nurses Society (INS) has stated that the accepted phlebitis rate is 5% or less in any given population (INS, 2000). To prevent thrombosis at the IV site, catheters are commonly flushed after each administration of IV medication and every 8 to 24 hours when the device is not in use (Dunn & Lennihan, 1987; Tuten & Gueldner, 1991). Traditionally, dilute solutions of heparin in 0.9% sodium chloride injection (e.g., 10-100 U/mL) have been used for the periodic flush. Ongoing controversy exists, however, as to the optimal type and amount of flush solution. In fact, staff nurses on the obstetric units at this Midwestern teaching tertiary referral medical center perceived heparin to be more effective clinically than normal saline for IV lock maintenance in pregnant women, but were unable to locate supporting literature.
JOGNN 503
Thus, the purpose of this study was to compare the efficacy of two commonly used peripheral IV lock maintenance solutions, heparin, 10 U/mL, 1 mL, versus normal saline, 1 mL, in pregnant women who were between 24 and 42 weeks gestation. In 1990, the American Association of Critical Care Nurses conducted a nationwide random sample survey of members (N = 1,072) to examine the practices for maintaining IV lock patency. This survey revealed that 40% of respondents used 100 U/mL of heparin, 37% used 10
P
regnancy is an acquired hypercoagulable state.
U/mL of heparin, 18% used normal saline only, and the remaining 5% used other dilutions or solutions. Goode et al. (1991) conducted a meta-analysis to estimate the effects of heparin flush versus saline flush solutions on IV site patency, incidence of phlebitis, and duration of peripheral heparin locks. The total sample sizes of the meta-analysis were 3,490 participants across 15 studies for patency; 2,356 participants across 13 studies for phlebitis; and 1,960 participants across 6 studies for duration. Their results indicated that heparin is no more effective than saline in preventing clotting, in decreasing the incidence of phlebitis, or in increasing the duration (number of hours) that an IV lock lasts (p < .05). All patients included in the meta-analysis were from the adult medical, surgical, or critical care populations, except for one study with 74 pediatric patients. Goode et al. (1991) recommended their results not be generalized to children or neonates, as this one small study does not provide enough information to warrant a change in practice. None of these studies included pregnant women. The American Society of Hospital Pharmacists (1994) issued a position statement that 0.9% sodium chloride (normal saline) injection is a safe and effective means for maintaining catheter patency of peripheral, indwelling, intermittent infusion devices (e.g., IV locks) in all adult patients in institutional settings. Obstetric patients were not specifically addressed, however. This statement also recommended that until further controlled studies are conducted, heparin should be administered to pediatric patients in the smallest amount necessary to maintain patency of the device. Pregnancy is an acquired hypercoagulable state (Cerneca et al., 1997; Comeglio et al., 1996; Cunningham et al., 2001; Peric-Knowlton, 1992). The term hypercoagulable implies an increased tendency for arterial or venous thrombosis. Pregnancy normally induces appreciable increases in the concentrations of coagulation factors I 504 JOGNN
(fibrinogen), VII (proconvertin), VIII (antihemophilic factor), IX (plasma thromboplastin component), and X (Stuart factor). These hemostatic changes during pregnancy are thought to result in an ongoing low-grade activation of the coagulation system beginning as early as 10 weeks gestation and lasting through the first 8 weeks after delivery (Cerneca et al., 1997; Sipes & Weiner, 1990). It has been suggested by Blackburn and Loper (1992) and in Gerbasi, Bottoms, Farag, and Mammen’s study (as cited in Cunningham et al., 2001) that this hypercoagulability is normal, protecting the placental implantation site throughout pregnancy and the woman from excessive blood loss at delivery by providing rapid hemostasis following placental removal. Meyer, Little, Thorp, Cohen, and Yeast (1995) questioned whether pregnant women were more likely to experience the failure of intermittent IV locks because of the normal physiological changes of pregnancy that result in a hypercoagulable state. They hypothesized that in pregnancy use of heparin flushes would improve intermittent IV lock patency, compared with normal saline flushes. In a double-blind study including 44 pregnant women who were between 26 and 34 weeks gestation and required serial phlebotomy, Meyer et al. compared heparin sodium (100 U/mL) with normal saline for use in the maintenance of IV locks. Catheter sites were examined and flushed with the study solution at least once every 6 hours. Partial thromboplastin times were also measured at catheter insertion and 48 hours later. Although no differences were found in partial thromboplastin times, there was a significant increase in catheter patency rate at 48 and 72 hours in the heparin group (84% vs. 52% and 68% vs. 27%, respectively; p < .01). There also was a significantly lower rate of catheter complications in the heparin group (13% vs. 31%, p < .01). Only one study (Meyer et al., 1995) was found that compared the efficacy of heparin with normal saline flushes to maintain peripheral IV lock patency in pregnant patients. Therefore, a prospective, randomized clinical trial to compare the efficacy of heparin, 10 U/mL, 1 mL, and normal saline, 1 mL, flushes to maintain peripheral IV locks in pregnant women was conducted. This investigation adds to the scarce amount of literature covering this topic for pregnant women and provides the necessary data comparing a lower dosage of heparin (10 U) with normal saline for maintenance of peripheral IV locks in pregnant women who are between 24 and 42 weeks gestation.
Design and Methods A prospective, randomized, double-blind design was used to compare the efficacy of using heparin, 10 U/mL, 1 mL, versus normal saline, 1 mL, for maintaining patency in peripheral IV locks in pregnant women. The null hypotheses identified by the investigators for this study Volume 32, Number 4
were (a) there is no difference in patency when peripheral IV locks in pregnant women are flushed with heparin, 10 U/mL, 1 mL, or when they are flushed with normal saline, 1 mL, and (b) there is no difference in the incidence of phlebitis when peripheral IV locks in pregnant women are flushed with heparin, 10 U/mL, 1 mL, or when they are flushed with normal saline, 1 mL.
Sample Participants were 73 pregnant women who were between 24 and 42 weeks gestation, required peripheral intermittent IVs, and were hospitalized in the obstetric units at a large Midwestern teaching hospital. Exclusion criteria included cervical dilation greater than 4 cm; presence of any significant abnormalities in the fetal heart tracing on admission; presence of hypersensitivity to heparin; presence of clotting abnormalities; and treatment with anticoagulant medication, including aspirin. The medical center’s institutional review board approved the study protocol.
Procedure After their initial assessment and provision of informed consent, eligible women who were to receive a peripheral intermittent IV lock were entered into the study per a random allocation list generated by the statistician. They were assigned treatment with either normal saline or heparin for IV lock maintenance. Each woman received placement of a 16- or 18-gauge IV catheter, with an extension set (IV Catheter Extension Set, InterLink System, Baxter Healthcare Corp, Deerfield, IL) attached. After placement, the extension set and catheter (IV lock) were flushed with the study solution. The participants were assigned to have IV locks flushed with one of two solutions: preservative-free normal saline, 1 mL; or heparin, 10 Units/mL, 1 mL. The hospital pharmacy prepared batches of the two treatment solutions, using identical syringes labeled as study flush solution. Only the pharmacy staff had access to the randomization list and the code. When a pregnant woman was entered into the study, the dispensing pharmacist assigned her the next available number and dispensed treatment flushes labeled with the patient study number. A heparin/saline studyalert sticker with the coded study number was placed on the patient’s chart and room door. Once enrolled in the study, participants received only the assigned solution throughout the duration of that peripheral IV lock. If their IV lock needed to be restarted, those participants were out of the study. Data collected before the IV restart were included in the analysis, however. All IV locks were assessed by registered nurses for complications at least every 12 hours (more often sooner, which was related to hanging fluids), and findings were documented. The IV locks were flushed at least every 24 hours or after each medication administration, whichever was more frequent. July/August 2003
Staff RNs providing care to the patients on the unit collected the data. The nurses were educated about the study at meetings conducted by the principal investigator. Because there were multiple data collectors, the RNs received information weekly at follow-up question-andanswer sessions conducted by members of the research team. The research team frequently offered food and beverages at these meetings, which served the secondary purpose of keeping data collectors enthused about the study. Content included the purpose of the study, study questions, inclusion/exclusion criteria, how to obtain consent, and review of the data collection tool. A copy of the study protocol was placed on each participating nursing unit, along with a step-by-step instruction list and a study flow diagram for nurses to follow. A member of the research team made rounds almost daily to assist staff nurses who were data collectors with questions, and to motivate them to complete the data collection tools. Specific information collected included the patient’s age, race, gravidity and parity, gestational age in weeks, time and date of IV insertion and discontinuation, IV location, number of flushes given, and IV complications. Complications of IV therapy were defined as resistance with flush or fluid infusion and phlebitis. Phlebitis was further defined as presence of pain, erythema, palpable venous cord, induration, and ecchymosis, as adapted from the INS Phlebitis Rating Scale (INS, 2000). Patients were enrolled and data were collected for 15 months. Because assessments were made at least every 12 hours or when initiating fluid infusion or administering
T
his study provides support that both normal saline, 1 mL, and heparin, 10 U/mL, 1 mL, may be effective in the maintenance of peripheral IV locks.
medication, most participants had only one assessment. For example, it was common for a pregnant woman to have an IV lock for a period of time before the onset of active labor, but this time period was usually less than 12 hours. When labor became active, the infusion of IV fluids was started. Outcomes measured and compared in the two groups included the patency of the peripheral IV lock for each patient within 12 hours and the incidence of phlebitis for each patient within 12 hours. Descriptive statistical analysis techniques were used for demographic information, and chi-square and Fisher’s exact tests were used for categorical data. Statistical significance was defined as p < .05.
JOGNN 505
TABLE 1
Demographic Characteristics of Participants (N = 73) Saline n = 35 Maternal age Mean SD Range Gestation Mean SD Range
pvalue* 0.41
28.8 5.8 18-41
29 5.6 20-40
39.5 2.2 32-42
37.7 4 25-42
.02
Saline n (%) Gravida 1 2 3+ Para 0 1 2+
Heparin n = 38
Heparin n (%)
pvalue* .07
15 (42.9) 6 (17.1) 14 (40.0)
15 (39.5) 15 (39.5) 8 (21.0)
17 (48.6) 12 (34.3) 6 (17.1)
20 (52.6) 14 (36.8) 4 (10.6)
.71
*Significant at p < .05.
Results The final sample consisted of 73 IV sites in 73 participants, 35 in the normal saline group and 38 in the heparin group. The sample was 97.1% White and 2.9% other (missing data = 4). This was consistent with the population seen at the time of the study. Groups were similar for maternal age, gravidity, and parity (see Table 1). The saline group did have a significantly higher gestational age (p = .02). The original intent of the study was to measure and compare outcomes in the two groups within 12, 24, 48, and 72 hours. However, assessments of IV lock status were rare beyond 12 hours, generally because of the initiation of fluid infusion. Because there were insufficient data to make comparisons in the 24-hour, 48-hour, and 72-hour groups, only the assessments within 12 hours for patency are reported here. IV lock patency was determined by the ability of the nurse to flush the catheter freely without obstruction, or if initiated, by IV fluids infusing freely. Of the 35 IV locks flushed with saline, 68.6% had no obstruction within 12 hours, as compared to no obstruction within 12 hours in 76.3% of the IV locks flushed with heparin (n = 38). This difference in
506 JOGNN
patency was not statistically significant, as shown by a chi-square test (p = .459). The following catheter complications related to phlebitis were reported: Three participants from the normal saline group (n = 35) and 2 participants from the heparin group (n = 38) reported pain at the catheter site. One participant from each group had erythema at the catheter site. One participant in the saline group had a palpable venous cord, and 1 participant had ecchymosis. Included in the aforementioned complications was 1 participant whose IV lock would not flush freely and who had pain, erythema, palpable venous cord, and ecchymosis. None of these complications demonstrated statistically significant differences between groups, using Fisher’s exact test.
Discussion In this randomized, double-blind trial, there were no statistically significant differences in IV lock patency, nor in phlebitis with normal saline, 1mL, or heparin 10 U/ml, 1mL, used for peripheral IV lock maintenance. Both of the null hypotheses identified by the investigators for this study were supported. Admittedly, the sample size for this study is small. For the first hypothesis, the heparin group had a patency rate of 76%, whereas the saline group had a patency rate of 69%. This study had only 11% power to detect a difference of this magnitude. However, this study does rule out large differences in patency rates between saline and heparin, as there was 80% power to detect odds ratios for patency of 40 or larger. The limitations in the statistical power are even more pronounced for the second hypothesis, because the complication rates were less than 10% in each study group.
A
dditional studies in this population are needed. Current evidence does not indicate clear direction for practice.
This study provides support that both normal saline, 1 mL, and heparin, 10 U/mL, 1 mL, may be effective in the maintenance of peripheral IV locks. However, normal saline is inherently safer than heparin. One of the main reasons the American Society of Hospital Pharmacists (1994) recommended switching to normal saline flush was to minimize patient exposure to heparin and thus possibly decrease a rare but significant adverse reaction that can develop in heparin-exposed patients, which is called heparin-induced thrombocytopenia.
Volume 32, Number 4
TABLE 2
Participants With Phlebitis Complications Within 12 Hours Participant # 13 53 69 47 17 20 p-value
Study Group Heparin Heparin Heparin Saline Saline Saline
Pain
Erythema
Yes Yes No Yes Yes Yesa 0.67
No No Yes No No Yesa 1.00
Palpable Cord No No No No No Yesa 0.48
Induration
Ecchymosis
No No No No No No 1.00
No No No No No Yesa 0.48
Note. No other participants experienced complications. Heparin 35/38 (92.1%); saline 32/35 (91.4%). Used Fisher’s exact test rather than chisquare owing to low complication rates. aSame
participant.
The findings of this study differ from those of Meyer et al. (1995) in their study of 44 pregnant women who were between 23 and 34 weeks gestation. Meyer et al. found that the use of heparin sodium, 100 U/mL, 1 mL, resulted in a significant increase in catheter patency and lower rates of catheter complications as compared with the use of normal saline. However, Meyer et al. used 10 times the strength of heparin (100 U/mL, 1 mL) compared with this study (10 U/mL, 1 mL), and their IV locks remained in place much longer.
Implications for Nursing Although the findings of this study provide support that the use of normal saline, 1 mL, to flush IV locks in obstetric patients may be as effective as heparin, 10 U/mL, 1 mL, the results must be interpreted with caution. As with all clinical studies using very small samples, the statistical power in this study was low. This means that it is possible that a difference between population groups might have been detected with a larger sample. In addition, because fluids may have been infused soon after the IV lock was initiated, only the within-12-hour data were used. Current evidence does not indicate clear direction for practice at this time. Additional studies comparing different strengths of heparin and normal saline, which address the precise dwell time of the catheter in this population of pregnant women, are needed to better determine optimal therapy over time. Also, future investigators’ use of the Intravenous Nurses Society Phlebitis Rating Scale would provide a uniform measurement standard across multiple studies (INS, 2000).
July/August 2003
REFERENCES American Association of Critical Care Nurses. (1990, August). Nationwide practice survey results announced: Survey studied heparin use. AACN News, p. 3. American Society of Hospital Pharmacists. (1994). ASHP therapeutic position statement on the institutional use of 0.9% sodium chloride injection to maintain patency of peripheral indwelling intermittent infusion devices. American Journal of Hospital Pharmacy, 51, 1572-1574. Blackburn, S. T., & Loper, D. L. (1992). Maternal, fetal, and neonatal physiology: A clinical perspective. Philadelphia: Saunders. Cerneca, F., Ricci, G., Simeone, R., Malisano, M., Alberico, S., & Guaschino, S. (1997). Coagulation and fibrinolysis changes in normal pregnancy. European Journal of Obstetrics, Gynecology, & Reproductive Biology, 73(1), 31-36. Comeglio, P., Fedi, S., Liotta, A. A., Cellai, A. P., Chiarantini, E., Prisco, D., et al. (1996). Blood clotting activation during normal pregnancy. Thrombosis Research, 84(3), 199-202. Cunningham, F. G., Gant, N. F., Leveno, K. J., Gilstrap, L. C., Hauth, J. C., & Wenstrom, K. D. (2001). Obstetrical hemorrhage: Consumptive coagulopathy. In Williams obstetrics (21st ed., pp. 619-669). New York: McGraw-Hill. Dunn, D., & Lennihan, S. (1987). The case for the saline flush. American Journal of Nursing, 87, 798-799. Goode, C. J., Titler, M., Rakel, B., Ones, D. S., Kleiber, C., Small, S., et al. (1991). A meta-analysis of effects of heparin flush and saline flush: Quality and cost implications. Nursing Research, 40(6), 324-330. Intravenous Nurses Society. (2000). Infusion nursing standards of practice. Journal of Intravenous Nursing, 23(Suppl. 6), S56. Meyer, B. A., Little, C. J., Thorp, J. A., Cohen, G. R., & Yeast, J. D. (1995). Heparin versus normal saline as a peripheral line flush in maintenance of intermittent intravenous
JOGNN 507
lines in obstetric patients. Obstetrics and Gynecology, 85(3), 433-436. Peric-Knowlton, W. (1992). Acquired hypercoagulable states. Nurse Practitioner Forum, 3(2), 72-81. Sipes, S. L., & Weiner, C. P. (1990). Venous thromboembolic disease in pregnancy. Seminars in Perinatology, 14, 103. Tuten, S. H., & Gueldner, S. H. (1991). Efficacy of sodium chloride versus dilute heparin for maintenance of peripheral intermittent intravenous devices. Applied Nursing Research, 2, 63-71.
Kathryn M. Niesen, MSN, RN, is director of clinical nursing, Department of Nursing, Mayo Clinic, Rochester, MN, and is an assistant professor of nursing, Mayo Medical School, Rochester, MN.
508 JOGNN
Denise Y. Harris, PharmD, is a pharmacist in Hospital Pharmacy Services, Mayo Clinic, Rochester, MN. Linda S. Parkin, RN, is a staff nurse, formerly in the Department of Nursing, Mayo Clinic, Rochester, MN, and now in the Department of Family Medicine. Lynn T. Henn, RN, is a staff nurse, Department of Nursing, Mayo Clinic, Rochester, MN. Address for correspondence: Kathryn M. Niesen, Department of Nursing, Mayo Clinic, 200 First St. SW, Rochester, MN, 55905. E-mail: [email protected].
Volume 32, Number 4