Midline intravenous catheters (MCs) have been in clinical use since the 1950s and are commonly used as an alternative for intravenous (IV) access to short peripheral IV catheters and central venous access devices.1–5 Use of MCs is associated with lower phlebitis rates and infections than for central venous catheters (CVCs) and can be inserted without the need for radiologic verification.2,6 Indications for using MCs are to administer intravenous medications and infusions for up to 30 days and are recommended when treatment is anticipated for 5–6 days or for long-term therapy in patients with limited IV access.2,4–6 In addition to MC use for infusions and medication administration, using the MC for blood sampling has been included as an indication for those who have difficult IV access.6 This option may reduce need for direct venipuncture in patients who have MCs with ongoing or frequent blood sampling needs for testing.6
While studies have been reported on the outcomes of use with MCs, such as dwell time, complications, completion of therapy, and recommended processes for maintenance and administration of fluids and medications,2,4–6 little is known or reported about procedures for blood sampling or outcomes from performance for this function. The Infusion Nursing Society standards7 includes guidance for blood sampling from central venous and short peripheral catheters (SPCs). However, when referencing MCs for blood sampling, the statement in the standards is “no evidence regarding obtaining blood samples from MLCs [midline catheters].” Frey8 also reported methods for blood sampling from various vascular access devices, but MCs were not included in this review and recommendations for practice.
Generally, intravenous catheters are designed to administer fluids and medication in a forward, antegrade motion into the venous system rather than to withdraw venous blood in a retrograde fashion, except for aspiration of blood return to verify patency. Little is known about the dynamics and physical effects on vasculature, catheter function, and clinical outcomes when retrograde maneuvers are used to withdraw samples of blood from MCs. Furthermore, no standardized procedures exist for withdrawing blood from MCs for sampling, though anecdotally, this practice is done in various clinical settings.
Hemolysis in vitro is the injury or breakdown of red blood cells that can occur during blood sampling.9 Reports in the literature show that using SPCs for blood sampling results in higher rates of hemolysis,9–12 mostly owing to mechanical shearing (Reynolds stress) and turbulence that occurs at the rough surface edges or kinks in tubing when withdrawing blood from the catheter.13 Hemolysis of samples that occurs during the preanalytic phase of blood collection can result in rejection of those samples for a myriad of blood tests. It is estimated about 3%–32% hemolysis occurs in acute care settings,14 accounts for 40%–70% of specimens unacceptable for processing,15 has a significant financial impact,16 places an additional burden on the patient for repeated venipunctures and vein injury, and results in delays in diagnosis or treatment.15 The American Society for Clinical Pathology benchmark for hemolysis rate is at 2%, thus reported hemolysis for blood sampling generally exceeds that mark.17
More recently, anecdotal feedback from practitioners to a company conducting clinical assessment of one of their MCs was the perception of increased hemolysis of samples when withdrawing blood from the catheter. To evaluate this further, the company conducted labora tory studies of fluid dynamics through the catheter structure and concluded that when performing fluid withdrawal maneuvers with a syringe, the distal tip of the catheter would collapse, potentially contributing to hemolysis during blood sampling. To address these challenges, the company made design changes to reinforce the tip of the MC to overcome distal catheter tip collapse with retrograde flow during blood sampling. The goal of this design change of the MC was to enhance success in collecting blood samples with reductions in hemolysis. In August 2016, the blood draw improvement with the reinforced tip was submitted to the U.S. Food and Drug Administration (FDA) as a special 510(k) application and received clearance from the FDA in September 2016.18 This MC (PowerGlide ProTM; BD, Salt Lake City, UT) demonstrated improved aspiration flow rates and reports of less hemolysis, but clinical testing of the MC for blood-sampling performance had not been previously done and was needed.
While some study findings in the literature address outcomes from MC use in general, no evidence was found about processes or clinical outcomes from using MCs for blood sampling, such as impact on dwell time, hemolysis of samples, and functional aspects of the MC to obtain samples. No clear procedures exist to guide techniques to perform blood sampling or ways to avoid hemolysis with MCs. Thus, variation in practice likely exists across many settings and caregivers with regard to blood specimen collection from MCs. Given the paucity of published data or studies related to blood collection from MCs, the investigators aimed to use exploratory and observational methods to summarize practices and outcomes of using the MC for blood collection.
The primary purpose of this study was to evaluate the processes, uses, and outcomes of using MCs for the purpose of blood collection for laboratory analysis. The aims of this study were to (1) evaluate the rate of hemolysis when the MC was used for blood collection, (2) evaluate outcomes from using the MCs for blood collection in the acute care setting (such as catheter performance during blood sampling, dwell time), and (3) evaluate nurse perceptions and practices used for withdrawing blood specimens from the MC.
Methods
This investigator-led study used a prospective, observational, and mixed methods design at a single-site hospital to evaluate the function and outcomes of MCs for blood specimen collection in two medical and two surgical units at the study facility. The study was approved by the organizational institutional review board through expedited review and waived informed consent from patients for data collection on MC use and outcomes since there were no interventions and only observation of usual care was done (IRB #19.216.07). No private health information was retained. However, informed consent was required by nurses to participate in the qualitative component of the study.
Subjects and settings
The study was conducted in a large, tertiary care center in southeastern United States. The units selected for the study were based on volumes of MC use from medical and surgical populations. As there were no interventions in the study and no comparison, we used historical MC use to drive sample estimates for the study and used a convenience sample of patients over a 3-month period to generate a reasonable sample size for this observational study. The aim was to achieve 300 total MCs inserted and to capture at least 150 cases where blood was drawn from the MC at any time during its dwell time. The MC was considered the study “subject,” and it was possible for patients to have more than one MC during their hospital stay. Criteria for inclusion were adult patients (≥18 years) whose initial MC was inserted within 24 hours during their stay on 1 of 4 study units. The study units were 2 medical progressive care units and 2 surgical units with similar patient characteristics. Excluded were those MCs on the study units that were inserted >24 hours prior to enrollment.
As a standard of practice at the study organization, all MCs are placed by specialized and trained nurses on the vascular access team (VAT) using sterile technique. Nurses used ultrasound-guided placement using an all-in-one, no-touch method with an integrated guidewire technique used with the type of MC, which is also described in the literature.4Figure 1 illustrates the type of MC with a reinforced tip used for all insertions in this study and exclusively at the study organization. In discussions with the VAT, they indicated that their practice was to assess each patient for the most appropriate vascular access prior to insertion of an MC.
Study measures
The investigator collected data on all MCs on the study units meeting inclusion of cases whose MCs were used for blood collection as well as those that were used only for clinical indications, such as for IV medications, fluids, difficult intravenous access, and blood administration. Difficult intravenous access was defined as those with a significant history of venous access issues requiring MC or other venous access devices or unsuccessful achievement of venous access with an SPC after two reasonable attempts.
To meet study aims 1 and 2, outcomes from catheter performance in all MCs, including both those used or not used for blood sampling, included complications, such as phlebitis, extravasation, infiltration, and catheter performance (dwell time, completion of therapy). Rates of hemolysis of blood samples sent to the laboratory were determined by direct observation of data obtained from the electronic healthcare record (EHR). Hemolysis rates were determined by the number of tests that were hemolyzed divided by the total number of blood tests done in those patients who had their MC used for blood collection. At the study facility, the laboratory result reports in the EHR customarily and reliably note when samples for blood tests are hemolyzed with instructions for resampling for testing. No other methods of reporting hemolysis were used by the laboratory system, thus manual observation and EHR reviews were needed.
Timing metrics were defined as follows:
Other measures included reasons for insertion (medications, fluid administration, blood transfusions, blood specimen collection), removal (complications, completion of therapy/discharge, no longer indicated, inadvertent removal), and for stopping use of the MC for blood collection. Patient demographics of age, gender, and reason for admission were obtained from the EHR. Types of medications administered either continuously or intermittently via the MC were collected and categorized for the study using the hospital medication list for irritating or nonirritating drugs.
To meet aim 3, evaluation of current practice with regard to MC blood specimen, collection was done through qualitative methods to capture experiences to explore this unstudied practice. Nurses on the VAT and study units were given the opportunity to participate in 1 of 3 hour-long focus group discussions about the methods they used to withdraw blood from the MC, facilitators and barriers to performance, or any other information about blood sampling from the MC. All willing participants received an informed consent form describing the study and voluntary nature of participating in the focus groups. Sessions were audio-recorded and transcribed for qualitative analysis by the research team. Three of the investigators evaluated information from the transcripts using constant comparative analysis techniques to develop codes and generate themes about using MCs for blood sampling. The initial and exploratory questions used to prompt group discussion were as follows:
Data collection
Every weekday, the investigator obtained a patient list from the EHR on the study units and identified those patients whose MCs were inserted by the VAT less than 24 hours prior to enrollment. The investigator reviewed physician orders for laboratory tests requiring blood collection. The investigator was present on the study units every day during the week, usually starting at 4 am, when most blood sampling occurred, and then made rounds throughout the day for other laboratory orders. A second investigator would make rounds in the late afternoons to capture other orders or blood collections. Any new MC insertions beyond that time were usually found by the next morning on rounds. The investigator queried each nurse about the method used for blood collection: either from the MC, direct venipuncture (RN/phlebotomist), or other intravascular device, and which tools were used, that is, syringe/size, tourniquet, or vacutainer. In those cases in which the MC was used to withdraw blood, the investigator reviewed laboratory results to capture any cases of reported hemolysis from tests done.
In patients identified with MCs, the investigator directly observed the MC site daily for its appearance and any complications, such as phlebitis, using the Infusion Nurses Society 2016 Phlebitis Scale,7 signs of extravasation or infiltration (acute pain, burning, failure to flush), and thrombosis (failure to flush). The investigator also queried the nurse for any issues encountered when using the MC for blood sampling. Timing and other metrics as described above was collected and documented daily by the investigator as described earlier.
Analysis
All observational data were entered into an Excel spreadsheet (Microsoft Corp., Redmond, WA) on a password-protected computer and server. To analyze the rate of hemolysis, the total number of blood tests that were withdrawn from the MC divided by the number of reported hemolyzed samples was used. Simple, descriptive statistics were used to summarize data, such as mean, standard deviation, and median values. To detect differences between groups, nonparametric testing with independent samples (using the Mann-Whitney U test) was used since the data were skewed and not normally distributed.
Results
After 3 months of data collection, 397 MCs in 378 patients were included for study. Of those patients, 19 had more than 1 MC, 16 had 2, and 3 had 3 MCs during their hospitalization. Reasons for removal and replacement in those patients were phlebitis, infiltration, occlusion, accidental removal, one no longer indicated, and another at a patient's request. While the sample was heterogeneous, the most typical patient was female, 62 years old, and being treated for infection. The mean age of the sample was 61.53 ± 18 years (median 63 years; range, 19–98 years). A significant difference was seen in gender, with females composing 61.5% of the total sample, but age of subjects was not significantly different (P=0.11). A wide diversity of reasons for admission was seen in the sample. Demographics for patients in the sample are summarized in Table 1.
| Parameter | Value |
|---|---|
| Gender, n (%) | |
| Male | 153 (38.5) |
| Female | 244 (61.5) |
| Total | 397 (100) |
| Age, range, mean (SD), y | |
| Male | 19–98, 59.71 (17.55) |
| Female | 19–95; 62.66 (18.33) |
| Total | 19–98; 61.53 (18) |
| P | 0.11 |
| Reason for admission, n (%) | |
| Infections | 129 (32.5) |
| GI | 88 (22) |
| Pulmonary | 35 (9) |
| Surgical | 32 (8) |
| Other | 27 (7) |
| Cardiovascular | 24 (6) |
| Orthopedic | 20 (5) |
| Oncologic | 16 (4) |
| Neurological | 8 (2) |
| Hematologic | 7 (2) |
VAT nurses cannulated all patients with the MC with the reinforced tip as described earlier. Most MCs (89%) were placed in the cephalic vein using a 20-gauge MC. The reasons listed in the VAT consultation form for insertion was predominately for difficult venous access, either by patient history or actual attempts at venipuncture (61%) and venous access for medication administration (36%). Table 2 demonstrates the characteristics of MCs inserted and site location.
| Characteristic | Size or site | n(%) |
|---|---|---|
| Gauge | 20 | 354 (89) |
| 18 | 43 (11) | |
| Length | 8 cm | 260 (65.5) |
| 10 cm | 137 (35.5) | |
| Side inserted | Left upper arm | 192 (48.4) |
| Right upper arm | 205 (51.6) | |
| Site MC Placed | Cephalic | 184 (46.3) |
| Basilic | 129 (32.5) | |
| Brachial | 84 (21.2) | |
| Reasons listed for requests for VAT insertion | Venous access, history of difficult venous access | 242 (61) |
| Medication administration | 143 (36) | |
| For procedure | 12 (3) |
MC = midline catheter; VAT = vascular access team.
A wide variety of medications and intravenous fluids were administered in 338 patients. Timing for administration of medication was intermittent, secondary infusions, and continuous infusions. Continuous infusions included mostly isotonic or dextrose fluids without additives (n=135; 84.4%), and 22 had potassium chloride additives (13.8%). Twenty-eight patients had continuous medication infusions, including heparin (n=7), analgesics/sedation (n=14), cardiac drugs (n=2), and 5 others had octreotide and bumetanide infusions. The hospital pharmacy list of drug characteristics was used to classify irritating versus nonirritating drugs received by patients through their MC. A wide array of nonirritating antibiotic drugs, such as cephalosporins, clindamycin, and daptomycin, were given in 187 patients, 154 patients received at least 1 irritating medication, and 53 patients had 2 or more secondary irritating medications. Intermittent irritating medication infusion drugs included vancomycin, acyclovir, azithromycin, carboplatin, calcium gluconate, and iron sucrose. Seventeen patients in the sample received intermittent irritating drugs, including nonopiate analgesics to support minimizing opiate use (methocarbamol, ketorolac), promethazine, lorazepam, and potassium chloride. Patients received a wide variety of drugs with variable characteristics over the course of the MC dwell, including both irritating and nonirritating medications. Thus, the investigators could not determine individual medicinal contributions to MC function for blood sample collection.
To answer research aim 1, all blood sample tests collected through the MC were evaluated during its dwell time for laboratory alerts for hemolysis and prompts to redraw specimens. Blood was sampled from the MC for 1021 laboratory tests, and of those, only 7 tests from 4 blood samples (0.69%) had hemolyzed specimens when the MC was used for blood sampling.
To answer research aim 2, we evaluated outcomes from all MCs in the sample and then compared outcomes in those who did or did not have MCs used for the purpose of blood collection.
Catheter performance outcomes
Mean dwell time for all MCs was 108.5 ± 98 hours (4.52 days). Over the time span of the study period, the average patient length of stay on the 4 study units was approximately 5 days. Mean dwell time of the MC by gender was significantly different (P=0.038). Dwell time was significantly longer in those patients who had 18-gauge MCs (P=0.01), as were those who had any treatment (medications, intravenous fluids) through their MC (P<0.001) or received at least 1 irritating medication (P<0.001). When blood was withdrawn from the MC any time after an hour of insertion (n=204), mean dwell time was 127.14 ± 109.18 hours (5.17 days); median was 98.55 hours. If the MC was never used for blood draws or only withdrawn upon insertion (n=193), mean dwell time was 88.34 ± 79.86 hours (3.68 days); median was 64.63 hours. This was statistically significantly different between the groups (P<0.001) in favor of those who had blood withdrawn periodically over the MC dwell time. Table 3 summarizes these results.
| Dwell time category (n) | Dwell time hours | P value |
|---|---|---|
| For all patients (n=397) | ||
| Mean ± SD (days) | 108.5 ± 98 (4.52) | |
| Range (days) | 5–742.75 (0.21–30.94) | |
| Median (days) | 78.5 (3.27) | |
| When blood drawn from the MC during its dwell (n=204)a | ||
| Mean ± SD (days) | 127.19 ± 109.13 (5.17) | |
| Range (days) | 16.87–742.75 (0.7–30.94) | |
| Median (days) | 98.55 (4.1) | |
| When no blood was drawn from the MC after its initial insertion regardless of the reason (n=193) | <0.001 | |
| Mean ± SD (days) | 88.34 ± 79.86 (3.68) | |
| Range (days) | 5–455.17 (0.21–19) | |
| Median (days) | 64.63 (2.69) | |
| By sex, mean ± SD | 0.038 | |
| Men | 122 ± 112 | |
| Women | 100 ± 87 | |
| By MC gauge mean ± SD | 0.01 | |
| 20 | 105 ± 96.5 | |
| 18 | 136.1 ± 105 | |
| When receiving any irritating medication (n=154), mean ± SD | <0.001 | |
| Yes | 130.3 ± 102 | |
| No | 94.4 ± 93 | |
| With completion of therapy (continuous or intermittent IVs) (n=338), mean ± SD | 0.035 | |
| Yes (n=295) | 110 ± 96 | |
| No (n=43) | 144 ± 120.6 | |
| When any treatment through the MC, mean ± SD | <0.001 | |
| Yes (n=338) | 115.17 ± 100 | |
| No (n=59) | 69.17 ± 74.9 | |
IV = intravenous; MC = midline catheter; SD = standard deviation.
Sixteen cases were removed because the only blood draw was less than an hour after insertion.
Of those who had intravenous therapy through the MC (n=338), 294 (87%) completed their intended therapy. Reasons for not completing therapy were varied; some were related to the catheter or its function, and others were patient related. Catheter-related reasons for not completing therapy included catheter occlusion (n=15; 34%), pain during medication administration (n=5; 11.4%), infiltration (n=4; 9.1%), and phlebitis (n=4; 9.1%). Sixteen patients (36.4%) did not complete their therapy through the MC because of accidental removal by the patient or nurse. Those who had blood sampling during their MC dwell had an 88% completion rate as compared to 81% in the group in which the MC was not used for this purpose. However, dwell time was longer in the small sample of those who did not complete therapy (n=43).
Reasons for removal for the MC were predominately at discharge (n=304, 76.6%). Table 4 shows reasons for removal of catheters in the sample. Some patients were discharged with the MC in place for continued therapy, and others were discontinued for insertion of a CVC for medication infusions not amenable for use by the MC. Overall, after initial insertion, MCs remained in place for the duration of hospitalization.
| Reason | n (%) |
|---|---|
| Removed at discharge | 304 (76.6) |
| Occluded | 23 (5.6) |
| Accidental removal by patient or RN | 22 (5.5) |
| No longer needed or expired | 19 (4.8) |
| Discharged with the MC still in place | 7 (1.8) |
| Pain or burning at the site of med | 5 (1.3) |
| Infiltration | 5 (1.3) |
| Phlebitis | 5 (1.3) |
| Patient request | 4 (1) |
| Replaced by a PICC | 3 (0.8) |
MC = midline catheter; med = medication; PICC = peripherally inserted central venous catheter.
Outcomes from MC use for blood sampling
In this sample, 176 patients never had blood withdrawn from their MC, either because there were no orders for blood sampling, blood was collected from other existing lines or venipuncture, or it was never attempted. In 221 cases, MCs were used for blood sampling (55.7%). Since laboratory tests were not ordered regularly over the duration of the dwell time in most patients, the last known recorded time that the MC was functional for drawing blood was used for analysis. The average time recorded that the MC was known to be successfully used for blood sampling was 64 ± 85 hours (2.67 days), with a median of 34.7 hours. The longest time over which blood could be sampled from the MC was 685 hours (28.5 days). In 226 cases total, nurses reported that they were not able to sample blood from the MC after some point in time during its dwell due to no blood return when attempting to withdraw using with a syringe. This often happened later in the course of the dwell time of the MC, and its frequency significantly varied.
Focus group results
The investigators held three focus groups of nurses to determine their practices when drawing blood from the MC. Two groups were nurses working on the study units (n=16) and another group with 8 VAT nurses. Three nurse investigators attended each focus group, took field notes, and recorded sessions that were transcribed and reviewed using constant comparative analysis. After discussion and review, consensus was reached by the team. Nurses described general principles they used for safe blood withdrawal as described in the hospital policy and INS standards, such as proper use of hand hygiene, gloves, disinfection of injection surfaces, and infusion interruption. They also followed principles of flushing to remove debris, aspiration for blood, clearance of the line, and avoidance of diluted blood or incompatible fluids or drugs. Key themes emerged from qualitative analysis of these focus groups. Themes from nurses who use MCs were sources of information, methods used, and supplies used. For sources of information, nurses most often learned techniques to withdraw blood from their preceptors or other nurses on the unit. Most nurses indicated they attempt to withdraw blood from the MC first to avoid venipunctures on patients and that many patients request and expect the MC to be used for blood collection when possible. Patients also prompted nurses with “tips” to be successful with blood collection. Methods used by nurses to withdraw blood varied considerably; however most used 10 mL syringes with various flushing efforts. Nurses tended to use saline syringe flushes and varied regarding the number of syringes used. They described using “power” flushes and “pulsed” flushes to ensure line patency. They rarely used a vacutainer device, but rather used a transfer device to inject blood into the collection tube. As a last resort, nurses would, on occasion, use a tourniquet at various locations on the upper arm. They described manipulation of arm and catheter position to “get the kinks out” of the catheter where it interfaced with skin. They described this “angle” to be where the catheter kinks due to its position in the arm and the patients' excess skin and fat. They described techniques used to “straighten out the angle,” such as pulling on the catheter hub or straightening out the arm. Other manipulations used were to take the dressing off the catheter when they deemed it to be “too tight” such that no blood would flow. They also seemed to borrow techniques commonly used with CVCs to achieve blood collection, such as turning the head, coughing, raising the arm, and other such maneuvers. In contrast, different themes arose from VAT nurses who insert MCs: vein preservation and selection, insertion request burden, and nurse education. VAT nurses valued the use of the MC access for therapy over blood collection and felt vein preservation was important. They preferred to use the cephalic vein as it was easier to access, but they stated it was not the vein of choice due to its anatomical position. They indicated they believed that MCs are requested more often since in-patients are admitted with higher acuity, are older, and have comorbidities that lead to poor venous access or depletion. They also stated that the pH of medications impacted dwell time and that nurses should have more education on MC maintenance care.
Discussion
In this exploratory study on MC processes and outcomes for blood sampling, few references exist to compare for previous studies, as none were found in the literature at the time of the study. Interesting findings were that hemolysis incidence was very low. Previous studies on the use of SPCs for blood sampling have shown high incidence of hemolyzed samples.13 In our study, the <1% hemolysis rate was lower than generally acceptable levels of 2% by the American Society for Clinical Pathology17 and, as reported in the literature, ranging from 3% to 32%.14 In the study by Green,15 estimated costs in North American hospitals per preanalytic errors such as hemolysis from blood specimen collection is approximately $208 per specimen. Also reported in this paper was that such costs can add up to 0.23%–1.2% of hospital operating costs, which could represent about $1.2 million per year in a hospital size of 650 beds, and there is a potential of 26% unnecessary diagnostics or treatment. Recommendations from their study and others16 are to develop improvements in blood collection techniques to reduce unnecessary costs associated with preanalytic processes. The ability to use the MC for blood specimen collection with low hemolysis occurrence is one way to potentially reduce costs associated with hemolysis and multiple phlebotomy venipunctures with its associated pain and suffering, particularly for those with difficult IV access. There may be differences in outcomes from other MC brands, as this specific catheter was designed to reduce hemolysis in blood sampling. More studies are needed to test this in other types of MCs and compare results.
Few complications occurred in all patients with the MCs, including those used for blood sampling, which is consistent with findings from other outcomes studies.6 The average and mean dwell time in our sample was less than found in other recent studies.4–6 The fact that most of the MCs in our sample were removed at discharge may have impacted our overall dwell time. Dwell time of catheters was longer in those who had MCs used for blood draws when compared to those who did not. This may indicate that the use of MCs for the indication of blood sampling may not affect dwell time or ability to use the catheter for therapeutic reasons. Additionally, more patients who had their MC used for blood sampling completed therapy than those who did not, indicating that when the MC is used for this purpose it may not impact therapy completion negatively. Overall, when the MC was used for almost any reason, dwell time was higher. It is possible that when the MC was used for infusions of medications and fluids and withdrawal of blood specimens, the MC may have been flushed more frequently (before and after), thus maintaining its patency and usability. Positive benefits were that those patients whose blood tests were sampled from the MC (n=1021) were spared the experience of venipuncture and its associated discomfort and costs.
The main reason for catheter removal was at discharge, and the average dwell time was similar to the average length of stay (LOS) on the study units. These findings are consistent with those found in the Chopra et al.4 study, where completion of therapy and discharge were common reasons for removal. In acute care, there is an emphasis to reduce unnecessary patient LOS with its associated costs and potential for complications during hospitalization. As older patients and those with chronic and debilitating comorbid diseases are admitted to the hospital, many also have vasculopathy and need hospitalization and administration of medications and blood samples for laboratory analysis. This is particularly concerning in patients with repeated admissions requiring vascular access and loss of vein integrity. While MCs can be used up to 30 days, hospital LOS is usually much shorter, thus there is likely less duration of therapy with MCs in the acute care setting. When multiple venous actions are required, the MC can serve as a vehicle to withdraw blood for sampling safely and reduce unnecessary costs and suffering associated with multiple venipunctures. One caveat and trade-off for this privilege is a potential loss of catheter function over time and catheter dwell time. More studies are needed in this area to determine efficacy under various clinical conditions and circumstances.
The average time over which the MC was used for blood sampling was 2.67 days, with a median of 1.45 days. Anecdotal reports in the literature about withdrawing blood from MCs indicate the time that the catheter could be used for blood sampling was often short and limited to either the first day or two, and this is supported by this data. In this study, the duration of time that the MC was known to be functional for blood sampling was measured, so if blood tests were ordered later during the course of dwell time, the actual endurance of the MC for blood sampling was likely underestimated and may have been functional beyond measures used in this study. More studies are needed to verify this metric. Additionally, patients received a variety of irritating and nonirritating IV medications that may have contributed to chemically induced phlebitis or hyperplasia that may have influenced the failure to aspirate blood for sampling. Finally, only one type of MC was used in this study, and its design was purposed to reduce hemolysis and turbulent blood flow at the catheter tip. This may have affected the ability for blood withdrawal with retrograde flow and the results seen in this study. Thus, studies to evaluate time to effectiveness of blood withdrawal in other types of MCs and with patients who have more frequent and long-term blood sampling may help us to better understand the length of time MCs function for this purpose.
Since little was known about nursing practices and perceptions about the use of MCs for blood specimen collection, findings from the focus groups supported the presumption that variation in practices exist for this procedure. As no guidelines exist to direct nursing actions with regard to blood collection, there are opportunities to test and develop best practices for blood sampling from MCs for safe and effective use.
Conclusions
In this study, findings demonstrate that blood can be withdrawn successfully from this type of MC with very low rates of hemolysis. These data may further support the use of MCs for the purpose of phlebotomy. However, since only one type of MC with structural alterations intended to reduce hemolysis was used in this study, and the study was conducted at one clinical site, it is not clear that similar results would be achieved in other types of MCs or settings. Additionally, withdrawal of blood samples for testing had little impact on dwell time and completion of therapy. Since removal of MCs at the time of discharge was the main reason for removal, complete understanding about catheter longevity and usefulness over longer periods of time is limited, particularly when subjected to repeated blood withdrawal. When withdrawing blood from MCs, nurses tended to use procedures recommended for phlebotomy in other vascular access devices to guide their actions. More information was gained to better understand how nurses perform and use the MC for blood withdrawal and provides opportunities to develop and standardize practices for blood sampling from MCs.
Limitations
To our knowledge, this was the first known exploratory study to evaluate use and outcomes of MCs for blood sampling in a single, acute care hospital. For this reason, findings may not be generalizable to all patient populations and settings. Since only one type of MC was available for use in this study, findings may not reflect outcomes for other types of MCs. Another limitation is associated with manual collection of data for hemolysis; other settings with automated and robust hemolysis reports would lessen complexity of data collection. However, all historical results were available in the EHR for later capture. The investigator was present only 5 days a week and may have missed patients with a short-term LOS or was not able to capture the data in 24 hours. However, the investigator was present in early hours during usual blood specimen collection times to capture subjects, and another investigator did rounds on patients in the afternoons for possible inclusion or for other blood specimens being collected. One limitation in the study was that the investigators did not perform or observe all blood sampling events by the nurses, and some data were collected from the EHR for some of the metrics in the study. Thus, the actual techniques used by nurses likely varied. Another limitation was that the patients' LOS was not collected to compare with dwell time, so other factors may have influenced dwell time beyond LOS. More studies are needed to replicate or substantiate outcomes and uses.
Recommendations for practice
Given that this study is the first to evaluate outcomes from using MCs for blood sampling and few references exist to guide clinicians as to the best procedures to use, evaluation of these practices and outcomes at other settings is recommended. Further research is needed to explore outcomes from MCs other than the one tested in this study to compare results, particularly with regard to hemolysis and catheter performance when multiple actions are applied to MCs. Other recommendations are to develop and test procedures for blood sampling from MCs to contribute to and establish standards of practice in this area.