Increased hemolysis rate in plasma tubes after implementation of a fully automated sample delivery and acceptance system

Introduction

Hemolysis arises from the pathological cell death of erythrocytes [1]. This process releases hemoglobin and intracellular substances or components to the surrounding plasma that subsequently can cause interference in laboratory testing [2]. Common mechanisms include divergence due to concentration differences, interactions between intracellular substances and extracellular substances or test reagents used for analysis, spectrophotometric interferences due to changes in optical absorption, or cell debris [2]. About 70% of hemolytic samples are due to errors in the preanalytical phase and are related to blood collection or specimen transport into the laboratory [1, 3, 4]. Technical advances in the analytical phase have reduced analytical laboratory errors [1, 3, 4]. Errors arising in the preanalytical phase can affect all subsequent steps of the laboratory process and thus, as a consequence, lead to inaccurate results and protracted patient management [1]. A possible source of hemolytic specimens are pneumatic tube transport systems (PTS), frequently used in modern hospitals. Based on pressure differences generated by pumps, single samples or capsules are delivered through a transport pipe from the areas of patient care to laboratory areas [5]. While PTS have the potential to reduce turnaround times, workload, and costs, their implementation may influence the quality of blood sample analysis [6]. Fully automated sample transport and acceptance systems (PTAS) represent technical progress by combining pneumatic tube transport systems for sample delivery with acceptance systems to allow further processing without conventional interventions. In our institution, an automated PTAS, capable of not only transporting but automatically transferring samples from transport capsules to the centrifuge and inserting them into instrument-specific test tube racks, was recently established. Following implementation, a steep increase in hemolytic plasma samples has been reported from the clinical departments. Thus, our study aims to systematically investigate the influence of PTAS on hemolysis rates in lithium heparin plasma tubes.

Materials and methods

Sample handling

The departments in our study were located on different floors within the same building. Per institutional standards, all blood samples were taken by physicians or nurses and collected into 5 mL lithium heparin separator tubes (13 × 100) (Greiner Bio-One, Kremsmünster, Austria). The standard operating procedures and handling instructions for blood samples were not modified between 2015 and 2017. Before installation of the PTAS, all samples were conventionally delivered by hospital staff via transport cart and were physically handed over to laboratory personnel. Since 2017, samples have been transported in the mother-child surgery center using the PTAS (Sumetzberger, Wien, Austria) in automated load/and unload transport capsules (Sumetzberger, Wien, Austria) as shown in Figure 1. Our automated load and unload capsules use non-padded transport compartments for sample storage during PTAS transport. After insertion of the capsule into the transport station, the capsules are automatically delivered to the laboratory, guided by an integrated RFID chip. The exact length distances of the different PTAS transportation routes were not available for this study and could not be assessed systematically. The transport speed of the capsules within the entire PTAS relied on the manufacturer’s instructions and ranged between 1 and 3 m/s. After arrival in the laboratory, capsules are slowed down due to cushioning O-rings and compressed air and are then opened by air pressure. The capsules are emptied, and the sample tubes fall 30 cm onto a metal chute, from which they are taken over by an automatic pre-analysis system (RSA Pro Roche, Basel, Switzerland) encompassing automated centrifugation (Cobas p 471, Roche, Basel, Switzerland and Rotanta 460 Robotic, Hettich, Tuttlingen, Germany) and at 3500RPM for 10 min at a temperature of 20 °C and with an RCF (=g) of 2,643. Afterwards, samples are sorted into instrument-specific racks.

Figure 1:

Schematic representation of our PTAS with the different possible causes of hemolysis: Lithium-heparin plasma tubes (1) and pneumatic tube systems, transport capsules (2), speed of specimen transport (3), length of transport path (4), G-forces during specimen transport (5) and shock forces during specimen transport (6). Samples are placed into transport capsules. The capsules are inserted into the transport station and automatically delivered to the laboratory, guided by an integrated RFID chip. After arrival in the laboratory, capsules are opened by air pressure, and the sample tubes are ejected into a metal chute, from which they are retrieved by an automated pre-analysis system (RSA Pro, Roche, Basel, Switzerland), encompassing automated centrifugation and sample sorting into instrument-specific racks.

Results

In 2015, a total of 32.614 samples were sent to the laboratory by conventional transport. Among these, 3.815 (11.7%) were hemolytic based on an HI of 25 and 2.403 (7.4%) based on an HI of 50. In 2017, after installment of the PTAS, 35.899 samples were sent to the laboratory, of which 9.441 samples (26.3%) were hemolytic based on an HI of 25 and 4.044 samples (11.3%) based on an HI of 50.

The observed increase varied in the different departments, as shown in Table 1. The department with the highest absolute increase in hemolysis rate and the highest observed hemolysis rate was the department of surgery, with a hemolysis rate of 15.3% in 2015 and 37.9% in 2017.

Logistic regression was used to analyze the relationship between transportation modality (conventional transport = 0, PTAS = 1) and the odds of obtaining a hemolyzed blood sample. It was found that the odds of obtaining a hemolyzed blood sample increased by 64.1% (95% CI [0.61, 0.68]) for blood samples sent by the PTAS compared to conventional transport (Table 2). After adjustment for the submitting department, the odds of obtaining a hemolyzed blood sample increased to 67.7% (95% CI [0.64, 0.71]) for blood samples sent by the PTAS compared to conventional transport. The odds of obtaining a hemolyzed blood sample varied within the different departments. When compared to the reference department, the odds of significant hemolysis ranged from an increase of 103.3% for the department of dermatology to up to 403.9% for the department of surgery (Table 3).

Discussion

The present study found an association between a new PTAS and an increased rate of hemolytic specimens, particularly an increase in samples with HI values between 25 and 50.

The findings in this study are consistent with and extend those of previous reports [3, 5, 6, 9, 10] and contradict some studies showing no significant influence of PTS on hemolysis rates [11], [12], [13]. The heterogeneity of the effects may be explained by the fact that different PTS have a broad range of configurations and influencing factors, such as type of transport capsules, transport speed, the presence of distribution centers, transport distance, and maximum g-forces applied to the samples. Furthermore, the studies used different cutoffs for the definition of hemolysis. Moreover, some groups investigated other laboratory parameters to test for the presence of hemolysis [3, 5, 6, 9], [10], [11], [12], [13], ranging from visual inspection of the samples to the use of a modular analyzer system [12, 13].

A PTAS with automated load and unload pods has been used in the present study, which necessitates a sliding drop into the pods from a height of roughly 30 cm. Arguably, the abrupt opening and ejection of the samples from the capsules could also increase the risk of hemolysis. Streichert et al. demonstrated that a fall from a height of one meter increases the likelihood of sample hemolysis [14]. However, to the best of our knowledge, there is no evidence that supports attributing the observed increase in hemolysis rates to the sample acceptance part of the PTAS. Other factors that influence PTS/PTAS sample transport are briefly summarized as follows.

Lithium-heparin plasma tubes are more susceptible to hemolysis than is the case for serum tubes [15], [16], [17]. This is possibly due to the increased mobility of erythrocytes in anticoagulated lithium heparin tubes compared to coagulated serum samples [16]. If anticoagulated plasma tubes are incompletely filled and transported, hemolysis might occur due to air bubbles passing through the sample, causing additional hydrodynamic pressure on the erythrocytes [18, 19]. Hemolysis could thus be reduced using smaller or completely filled sample tubes [19, 20].

Reducing the sample movement space during transportation and padding the capsules or samples can decrease the occurrence of hemolysis [9, 21, 22]. Especially cushioning individual blood samples with bubble wrap proved effective but is not possible due to the automatic emptying process of the transport capsules [23]. In contrast, some systems try to avoid the application of transport capsules [24].

Especially at higher speeds, samples are exposed to substantial g-forces in curvy transport sections of PTS. This causes mechanical stress on erythrocytes and induces hemolysis [25, 26]. A significant difference in hemolysis rates can be observed among samples sent by PTS between 2 m/s and 3 m/s [26]. It is possible that the speed within the PTAS might have contributed to increased hemolysis rates.

Due to the high transport speed and the resulting short transport time, we assume that the temperature, which could alter the sample quality by creating hemolysis, didn’t vary during transport [14, 27]. PTS transport distance length demonstrably influences sample quality [28, 29]. Gils et al. reported a correlation between hemolysis and the length of transport in their PTS with individual sample transport [28]. Wei et al. showed that the longer PTS pathway in their study increased the percentage of hemolytic samples by 56,7% [29]. Therefore, the differences in hemolysis rates across the various departments of Clinic Favoriten may be partly explained by the different transport routes and lengths between sending and receiving locations.

The 3-axis acceleration during sample delivery significantly varies at different transport speeds as compared to delivery hospital staff [14]. During PTS transport, g-forces as high as 15 g for 2.5 m/s, 14 g for 2.0 m/s, and 13 g for 1.5 m/s have been measured in contrast to 9 g for hand-carried transport [14]. Clinically relevant differences have already been observed at a PTS transport speed of 1.5 m/s for the laboratory parameters aspartate aminotransferase and lactate dehydrogenase (LDH) [14].

After arrival, samples are usually centrifuged at a speed of 2,500–3,500 rpm for 10 min with high g-forces of up to 3,000 g without increasing the hemolysis rate. Therefore, g-forces per se most likely do not increase hemolysis rates [14, 24]. Rather the rapid acceleration and abrupt deceleration than the g-forces during sample transport may influence hemolysis rates in PTAS [5, 18].

A significant influence of abrupt acceleration, rapid deceleration, and the number of sudden changes in tube routing on hemolysis has already been demonstrated [9, 18, 22, 30]. The extent of these shock-like forces can cause a higher hemolysis rate within a particular PTS route and may explain our observed differences in hemolysis rates between departments. To reduce acceleration and deceleration effects, compressed air [22] and cushioning O-rings, as used in our system, are utilized to minimize hemolysis rates [9].

HI, lactate dehydrogenase (LDH), and potassium represent established references to monitor PTS-induced hemolysis in daily practice. Based on these indicators, any systemic hemolysis can be quickly noticed [28]. The specifications of the PTS influence the extent of sample hemolysis. It thus has been recommended recently that each laboratory should evaluate the PTS with particular emphasis on g-forces [5, 31, 32]. Smartphones equipped with appropriate apps are a simple and easily accessible option to measure acceleration and g-forces during sample delivery [14, 18, 33]. Moreover, commercial data loggers are suitable for recording the number and magnitude of g-forces within the PTS [14].

Our study has several limitations. First, we analyzed retrospective data of mostly surgical departments and were hence not able to compare two randomly assigned arms as a 1:1 comparison. Second, due to the nature of data retrieval, it was impossible to exclude the influence of the sample collection procedure. Third, due to nature of the available data, a differentiation between hemolysis prior to transport, hemolysis generated by the PTAS, and hemolysis due to the acceptance system was not possible. Fourth detailed information about system configuration and transport routes was not available to the authors. However, since hemolysis rates increased in all departments under investigation, a major impact of blood drawing practice seems unlikely. Nonetheless, additional training regarding sampling techniques has been implemented in response to our findings. The evaluation of the PTAS with a data logger is planned to find further possible sources of hemolysis and continue to improve our system.

Conclusions

Implementation of a PTS in combination with an automated load- and unloading infrastructure is associated with hemolysis affecting clinically relevant hemolysis-sensitive laboratory parameters. Our data confirms the importance of evaluating novel PTAS prior to routine implementation to consider conventional delivery for hemolysis-critical blood panel parameters.