Sample transportation – an overview

In-house sample transportation

For in-house samples, transportation can be manually or automated:

Manual transportation by trolley or by hand (e.g. by the phlebotomist or the clinician drawing a STAT sample) are both very reliable as no technical instruments are involved and also, it is an accustomed transportation form. It is however slow, and if emergency samples are carried by hand the time spent with sample transportation is time lost to perform other tasks. Another hurdle is the tendency to gather samples at the wards in order to send them in batches to the laboratory. This perhaps saves transportation labour, but the first sample collected may have been waiting for hours at the ward before it is transported to the lab. Also, if samples are received batch-wise, it is not possible to maintain a “first-in-first-out” (FIFO) process, as the laboratory does not know in which order the samples were collected.

As a consequence of these drawbacks, an increasing number of hospitals use automated transportation systems, which can be pneumatic tube transportation systems (PTS) [15] or, e.g. an electric track vehicle [16]. Both fit nicely into automation of the in-house sample reception at the laboratory; it also tends to make the wards send their samples in more “real time”, enabling the laboratory to maintain the FIFO principle, and as a natural consequence, it facilitates speed, unidirectional flow and high through-put. However, it also poses obstacles to be addressed by the laboratory professionals: It is important to note that the test request information has to be available at the laboratory when the sample arrives, which makes electronic requisition more flow-efficient than paper requisitions. Separated transport of samples and test request information in terms of paper requisitions leads to further delay in the processing of the ordered analysis, which often leads to resampling and is the cause of laborious tracing efforts for both the wards and the laboratory. Thus, if an automated sample reception system is to work, test requests must be electronically handled. A more basic problem is the obvious risk of processing the serum samples too quick, i.e. not allowing time for clot formation before centrifugation. This results in fibrin issues, which can be either interference with a number of analyses or a fibrin clot blocking the sample pipetting needle leading to a false result and possibly “down time” of the analyser – in both instances a new blood sample is needed causing delay in the clinical process. This can be solved by programming a halt in the sample flow for these serum tubes, but one must however first be aware of the problem. Another main issue is the increased risk of haemolysis shown by a number of studies (e.g. [17], [18], [19]) and also demonstrated by use of a smart phone [20]. But as revealed by a systematic review there is no general evidence for the safety of using PTS for blood sample transportation [15]. This is mainly due to the high degree of heterogeneity of the retrieved studies, but also because the local, physical PTS arrangement impacts the sample transported in an unpredictable way: Automated PTS are very individually constructed, and conditions concerning the usability (how easy can the samples be loaded?), the g-force impact (due to speed, twist and turns), the actual physical impact on the sample (which along with the mentioned g impact also includes bumps of the tubes during “pit stops” during transport and at the arrival at the laboratory), and finally the temperature (if the PTS goes underground or outside a building) are all parameters that requires a closer scrutiny to assure sample integrity during automated transportation.

So ideally, laboratories must measure and document the actual acceleration forces in their existing PTS. For this, use of G-loggers to measure acceleration vector sums has been suggested, and the laboratory should accordingly institute quality target thresholds for these values [10]. Another more simple possibility is the use of daily test results from a particular PTS, where, e.g. median potassium values can be used to monitor changes in the impact on blood samples. Single potassium values will thus not have any relevance, while running potassium median values will provide sufficient monitoring; again, the laboratory must establish a target threshold and define what action should be taken if exceeded. Haemolysis index values for each PTS can also be used, but notably, samples can be affected without haemolysis, e.g. if potassium leaks from leukocytes or platelets during transportation [21], a preanalytical error that will remain undetected by the haemolysis index. A significant preanalytical effect on samples transported by PTS has for instance been noted for thromboelastographic analyses. For these analyses, manual transport of samples is therefore recommended [22], and in the case of platelet function testing best practice is to avoid any transportation at all and if possible, perform the blood collection at the test site.

In order to continuously minimize turnaround times in all clinical settings, a forthcoming challenge will be the use of automated PTS for all different kinds of sample materials, e.g. paediatric samples, urine, cerebrospinal fluid [23], free DNA measurement [24], culture media and even pathology specimens [25].

External sample transportation (from general practitioners or external laboratories)

External samples are transported in a variety of ways, e.g. by car [26], in boxes [27], using drones [28], [29], or even by planes and trains. If the samples are properly protected from temperature deviation and agitation, none of these transportation forms should affect the samples significantly [10]. It is however crucial to monitor these conditions – and also a demand according to the ISO15189 accreditation that most laboratories carry.

The transportation of venous blood samples from outside the hospital setting is often costly and logistically challenging. Flaws in tube labelling, packaging and transport cause delays and increased costs with several stakeholders involved. The transportation routines should be described and agreed upon, also including what to do when these routines are deviated from. A number of technical support systems for this, including software [30] and intelligent transportation boxes with GPS and temperature loggers [31], are already on the market, and such features will for certain be an important part of an improved preanalytical quality assurance in the future.

Third party delivery

Security systems must be established that ensures sample transportation by a third party if such are used, e.g. samples send by mail: Although used widely in many countries, the efficiency of postal systems is generally deteriorating due to the now-a-days more IT-dependent communication, and one must therefore assure sample integrity if postal delivery is delayed. This is, e.g. a problem with centralised investigations such as screening for colorectal cancer using immunochemical testing of faecal occult blood (iFOBT), where samples are sent by public postal delivery. A recent study with stability tests of faeces samples stored at 30°C for 14 days showed no change in the distribution of iFOBT tests below and above the cut-off [36], and the significantly delayed sample delivery was therefore not an issue. Such developments can be foreseen for a variety of samples and analytes, and it is of course the responsibility of the laboratory to monitor and if necessary, validate any alterations in a third party transport that could affect sample integrity.