The Impact of Clinical Sample Transportation by Unmanned Aerial Systems on the Results of Laboratory Tests

Abstract

Transport by unmanned aerial systems (UASs) (e.g., drones) could save time and personnel. Our study aimed to assess the effect of drone transportation on the clinical laboratory results of biological samples by examining its impact on pre-analytical and analytical processes. We performed a cross-sectional study of healthy volunteers from Sha’ar Menashe Mental Health Center between July and November 2022. Blood and urine samples were transferred to the central laboratory at Hillel Yaffe Medical Center. Overall, 40 healthcare workers aged 21–67 years (57.5% females) with a mean age of 45.8 (SD = 11.3) years from Sha’ar Menashe Mental Health Center were recruited in the study. There were no significant differences between transportation modes in the complete blood count levels. We found a significant difference between the transportation modes for GGT (p = 0.01) and PT (p = 0.04), despite the very similar mean results of these tests. In Bland–Altman plots, GGT and PT samples fell within the 95% limits of agreement and were indicated as not clinically relevant; however, glucose and LDH did not meet the 95% acceptance criterion and showed a potential clinical effect. There was full agreement between the two types of transportation for urine glucose, nitrites, and urine cultures. UAS transport is an appropriate method for maintaining the quality of most routine clinical laboratory specimens, similar to the routine procedure of using a vehicle. For the 34 biochemistry, hematology, and coagulation assay parameters, only glucose and LDH did not meet the 95% acceptance criterion and showed a potential clinical effect.

1. Introduction

Terrestrial vehicles are the most common mode of clinical laboratory specimen transportation [1]. Unmanned aerial systems (UASs) are a rapidly evolving technology with significant applications in various fields [2]. UASs are promising and planned subjects that provide seamless advantages in extensive wireless application scenarios and meet the increasing communication requirements with enhanced network performance through expanded UAS-assisted communications coverage integrated with emerging and potential technologies [3]. UASs can fly to inaccessible areas to collect data and then distribute the collected data to remote access points using their high flexibility and maneuverability [4]. However, the enhanced multi-objective salp swarm algorithm displays advanced performance. It can significantly decrease time and energy costs compared with benchmark strategies that require the UAS to fly frequently, and is valid when given unexpected circumstances [5]. Drones are a safe way to transport blood samples in a hot tropical climate that could serve as a solution to improve access to healthcare, save transportation time during emergencies, and provide much broader healthcare coverage to the population [6]. UASs can be an available, convenient, safe, and economical alternative mode of transportation compared to conventional methods that rely on road infrastructure and are affected by traffic (as with motorized vehicles). Drones have become a routine and preferred means of transportation for goods, foods, and other items in some areas and companies. The Swiss Postal Service started a pioneering program for transporting packages (including biological samples) with drones. Thus far, their drone delivery service has made over 3000 flights with only one crash that ended without casualties or property damage [7,8]. In the USA, some branches of a large fast-food chain have recently advertised new drone-delivery services. Given the expansion of drone technology to commercial use for transporting various items, it has great potential for use in the medical field with appropriate safety measures [9,10].

Aggarwal et al. indicated that there is no environmental influence, such as from humidity, temperature, wind, etc., on drones and no effect of vibrations on the physical integrity and leakage of the dummy vaccine vials [11]. In a prospective pilot study, the transport of simulated blood products by drone was significantly faster than using ground transport. The temperatures of the simulated blood products remained within acceptable ranges throughout the transport period [12]. Rwanda successfully conducted a drone blood delivery program in 2016. It used drones flying up to an altitude of 150 km and carrying up to 1.5 kg of blood. Since August 2019, blood delivery times have decreased from 4 h to 15–45 min over more than 18,000 flights [13]. Drones in Tanzania and Ghana have been used to deliver blood, vaccines, and medications [14,15]. Previous preliminary reports have demonstrated the feasibility of drone-related transportation of human organs [16,17]. The maximum time to transport biological samples was approximately 40 min (equivalent to 40 km) for a fixed-wing drone and 27 min (equivalent to 13–20 km) for a multi-engine drone. One report noted that the overall concordance of all sample types (hematology, chemistry, and coagulation) between airborne and transported samples was approximately 97% [18]. Another study evaluated the effects of longer flight times of up to 3 h and longer distances of up to approximately 258 km [19] and found a tendency for glucose levels to decrease and potassium levels to increase, with biases of 8% and 6.2%, respectively, between parallel samples transported by air or on the road. The researchers believed that this deviation was due to changes in the temperature between the two groups of samples.

In recent years, several reports have demonstrated that it is possible to transport biological samples using drones without negatively affecting the samples [20,21,22,23]. No significant differences were observed in any measure of hematology parameters between blood samples analyzed pre-flight compared to those transported and recovered from the drone [20]. Perlee et al. indicated that most of the blood parameters were not affected by drone transport. However, glucose, phosphate, potassium, chloride, hemoglobin, platelet count, activated partial thromboplastin time, and lactate dehydrogenase (LDH) showed significant differences between transport by car and drone. They showed a clinically relevant effect only in LDH and glucose values [21]. However, Weekx et al. showed that in 3 of the 30 biochemistry and hematology parameters, a statistically significant difference was found in gamma-glutamyl transferase (GGT), mean corpuscular hemoglobin, and platelet count. A clinically relevant effect was only shown in potassium and LDH [22]. A study from Japan demonstrated that the LDH values were not significantly different between the flight and non-flight samples [23].

The increasing experience with drones as a means of transport and supporting evidence from previous studies regarding their potential for safely transporting biological samples without clinically significant effects on laboratory results led us to investigate their potential in the Mediterranean weather conditions in Israel. Establishing drone technology as a means of transporting biological samples can optimize the time and availability of transportation from hospitals to distant destinations and may even reduce transportation costs. Accordingly, our study aimed to assess the effect of drone transportation on the clinical laboratory results of biological samples by examining its impact on pre-analytical and analytical processes.

2. Materials and Methods

2.1. Study Design and Population

We performed a cross-sectional study among healthy volunteers aged 21–67 years from Sha’ar Menashe Mental Health Center. Blood and urine samples were collected between July and November 2022. Initial samples were made with 47 volunteers of healthcare workers on a total of five test days. On one day, the drone needed to make an emergency landing for technical reasons; thus, the test day was stopped, and the five samples that were performed were excluded from the study. Another flight was canceled due to weather conditions, and two blood samples were also excluded from the study; therefore, a total of 40 samples from eight separate flights on three different days were enrolled. The pre-analytical phase in our study included the identification and selection of an appropriate test, specimen collection, and transport conditions, while the analytical phase mainly explored the effect of the different transportation modes (drones vs. cars) on the clinical laboratory results of biological samples.

2.2. Pre-Analytical Phase and Flight Data

We ensured the same conditions and quality of the sample set in addition to pre-transport controls to prevent bias and compare the transport types (drone versus car) accurately to assess the effect of transport type. The choice for comparison with a car as a control method is because it is the accepted method in Israel and following the literature. For example, Perlee et al. stated that ground transportation by car is the regular method of transporting blood tubes from the blood collection site to the laboratory [21]. Drones are unmanned aerial vehicles, and ensuring the security of the samples during flight and guarding against unauthorized access is paramount. Thus, the packages were locked and the drone was equipped with cutting-edge navigation and real-time tracking system technology in our study. One drone had an unscheduled landing due to technical problems. During another flight, the drone was returned to the base owing to low battery power but was subsequently resent. One flight was canceled because drones are sensitive to adverse weather conditions, including high winds, rain, and fog. Considering this, and until their ability to withstand weather improves, in our study the UASs were used mainly during calm weather between July and November. However, the mean temperature ranges in the Hadera district (the study area) in 2022 were 23.0–31.8 °C in July, 23.6–33.0 °C in August, 21.7–31.3 °C in September, 18.5–29.2 °C in October, and 14.0–25.0 °C in November [24].

The temperature of the samples was documented by a temperature-monitoring device with sensor model T-174, from the Testo company, during each transport, both by car and drone. The drone was equipped with an innovative flight automation system (FlightOps) that maintained cellular network connectivity, even in communication instability, to ensure safe and controlled operations. The drone was also fitted with a parachute to reduce the impact of a fall in the event of malfunction as a precautionary measure for passage over urban areas. These integrated technologies significantly enhance drone safety and reduce the risks associated with these flights. FlightOps seamlessly integrated these advancements into the Sentinel G3 drone, an 11 kg device with a payload limit of 3 kg. The flight path was maintained at an altitude of 70–100 m to avoid potential obstacles, and the drone maintained a speed of 10–12 m/s, resulting in a flight duration of 25–30 min. A potential obstacle that might hinder drone implementation is the strict aviation regulations that limit drone use [25]. We intended to perform flights from another medical center when planning the study, but the air route was not approved by the Civil Aviation Authority Israel (CAAI) because it is located above major transportation routes. All the operations and air routes used in our study were approved in advance by the CAAI. However, the flight distance between the Sha’ar Menashe and Hillel Yaffe hospitals is approximately 14 km, falling under the category of “Beyond Visual Line of Sight”.

2.3. Analytical Phase—Clinical Laboratory Methods

Blood and urine samples were transferred to the central clinical diagnostic laboratory at Hillel Yaffe Medical Center; however, the assays were performed on the same day of the transport on three different days. Five tubes were collected in duplicate from each person, namely a serum gel tube for biochemistry panel that included the 19 most common biochemical parameters: electrolytes (sodium, potassium, and calcium), glucose, lipid profile, kidney and liver function panels, EDTA for complete blood count (CBC), sodium citrate for coagulation assays that included prothrombin time (PT) and partial thromboplastin time (PTT), INR, urinalysis, and urine culture. After collection, the tubes were separated into 2 identical sets. One set was transferred as routine samples by car, whereas the other was transferred by UAS. All samples were packaged according to the national safety regulations in boxes containing ice packs and a temperature-monitoring device with sensors model T-174 from the Testo company. The boxes contained 25 tubes for each shipment. The tubes transferred by the drones were identified by a code number, and the package was locked to maintain confidentiality and limit access to the samples in case of an accident. The code list was then fed to the laboratory. Biochemical assays were performed using a Cobas-8000 instrument (Roche Diagnostics, Mannheim, Germany). CBC was performed using a DxH-800 instrument (Beckman Coulter, Brea, CA, USA). PT and PTT were conducted using CS2100i (Siemens Healthineers, Erlangen, Germany). Urinalysis was conducted using a Cobas-6500 instrument (Roche Diagnostics), and urine cultures were performed manually. Hemolysis levels were detected using the Cobas-8000.

2.4. Statistical Analysis

The Kolmogorov–Smirnov test was performed to examine normal distribution. Continuous variables in this study are described as the mean and standard deviation (SD), while continuous variables with skewed distributions are presented as the median and interquartile range (IQR). We used the McNemar test for dichotomous variables and the Friedman test for categorical variables. The paired t-test was used for normally distributed continuous variables, whereas the Wilcoxon signed-rank test was used for variables with skewed distributions. Bland–Altman plots were used to illustrate the agreement between the two different transport (drones and cars).

All statistical tests were performed using the Statistical Package for the Social Sciences (SPSS) version 28 (IBM, Armonk, New York, NY, USA), and statistical significance was set at p < 0.05.

2.5. Institutional Review Board Statement

This study was conducted following the Declaration of Helsinki, and all procedures were performed following local guidelines and regulations. The Institutional Review Board of Hillel Yaffe Medical Center (protocol number HYMC-0022-22) approved the study protocol. All participants provided informed consent to participate in the study and agreed to the publication of the manuscript by signing an informed consent form.

3. Results

3.1. Pre-Analytical Indices

Overall, 40 healthcare workers aged 21–67 years (57.5% females) with a mean age of 45.8 (SD = 11.3) years from Sha’ar Menashe Mental Health Center were recruited in the study. The mean package temperature of the drone was 25.2 (SD = 2.7 °C) compared to 21.4 (SD = 2 °C) by car (p < 0.01). The distance between the Sha’ar Menashe and Hillel Yaffe hospitals was 14 km by flight with drones compared to 17 km by car. However, the mean time by drone was 39 (SD = 12) minutes, similar to the transport time mean of 37 (SD = 12) minutes by car, p = 0.07. Transportation flight time depended on weather conditions, and on one occasion, the drone was returned to the base and resent after adjustments. Although the duration of this flight was an outlier, the samples from this flight were analyzed and the data were included in the study. Hemolysis was rare and occurred in two samples, of which one was transported by drone and one by car.

3.2. Biochemistry and Hematology Panel

There were no significant differences between transportation modes in the levels of creatinine, Blood Urea Nitrogen (BUN), sodium, calcium, phosphorus, uric acid, protein, total bilirubin, LDH, alkaline phosphatase (ALP), alanine transaminase (ALT), aspartate transferase (AST), protein, high-density lipoprotein (HDL), total cholesterol, or triglycerides (

Table 1. Differences in the serum levels of biochemistry markers by transportation type (drones vs. cars).

	Car n = 40	Drone n = 40	p Value a
Glucose (mg/dL), median (IQR)	89.5 (16.5)	89.0 (16.7)	0.8 b
Creatinine (mg/dL), mean (SD)	0.84 (0.14)	0.85 (0.13)	0.5
BUN (mg/dL), median (IQR)	12.3 (5.0)	12.3 (5.5)	0.6 b
Potassium (mmol/L), mean (SD)	4.41 (0.33)	4.44 (0.36)	0.6
Sodium (mmol/L), median (IQR)	139.0 (2.9)	139.0 (1.2)	0.7 b
Calcium (mg/dL), mean (SD)	9.82 (0.49)	9.85 (0.41)	0.4
Phosphor (mg/dL), mean (SD)	3.59 (0.55)	3.59 (0.55)	0.8
LDH (U/L), mean (SD)	358.2 (58.5)	365.7 (62.7)	0.3
Uric acid (mg/dL), mean (SD)	5.0 (1.3)	5.0 (1.3)	0.9
Total bilirubin (mg/dL), median (IQR)	0.43 (0.34)	0.44 (0.32)	0.4 b
ALP (U/L), mean (SD)	75.08 (19.21)	75.05 (19.16)	0.8
ALT (U/L), median (IQR)	17.0 (12.5)	17.0 (13.5)	0.4 b
AST (U/L), median (IQR)	19.0 (6.8)	19.0 (6.5)	0.2 b
GGT (U/L), mean (SD)	23.55 (12.20)	23.18 (12.42)	0.01
Albumin (g/L), mean (SD)	4.88 (0.31)	4.94 (0.30)	0.08
Protein (g/L), mean (SD)	7.40 (0.4)	7.38 (0.4)	0.5
Total cholesterol (mg/dL), mean (SD)	189.30 (40.3)	189.40 (40.5)	0.6
HDL (mg/dL), median (IQR)	51.00 (15.4)	50.85 (15.3)	0.6
Triglyceride (mg/dL), median (IQR)	120.90 (65.9)	121.00 (65.8)	0.9

^a p value was obtained by a paired t-test; ^b Wilcoxon Signed Rank test. ALP: alkaline phosphatase; ALT: alanine transaminase; AST: aspartate transferase; BUN: Blood Urea Nitrogen; dL: deciliter; g: gram; HDL: high-density lipoprotein; IQR: interquartile range; L: liter; LDH: Lactic dehydrogenase; mg: milligram; mmol: millimole; SD: standard deviation.

). The effects of parameters sensitive to pre-analytical conditions, such as potassium and glucose, were carefully evaluated. For potassium, the mean level in aerial samples was 4.44 mmol/L (SD = 0.36) compared to 4.41 mmol/L (SD = 0.33) in samples transported by a car (p = 0.6). The median glucose was 89.5 mg/dL by car and 89.0 mg/dL by drone (p = 0.8).

The mean level of GGT was significantly higher in samples transported by a car {23.55 U/L (SD = 12.20) {than by drone }23.18 U/L (SD = 12.42)}, (p = 0.01). However, the mean level of albumin was slightly lower in samples transported by a car 4.88 g/L (SD = 0.31) than by drone 4.94 g/L (SD = 0.30), p = 0.08 (Table 1).

Similar findings were obtained for CBC; however, there were no significant differences between transportation modes in the levels of neutrophils, monocytes, eosinophils, red blood cells, hemoglobin, hematocrit, mean corpuscular volume, mean corpuscular hemoglobin, mean corpuscular hemoglobin concentration, red cell Distribution Width, and platelet count (

Table 2. Comparison of complete blood count parameters by transportation type (drones vs. cars).

	Car n = 40	Drone n = 40	p Value a
White blood cells (103/μL), mean (SD)	7.3 (1.4)	7.3 (1.4)	0.08
Neutrophils (103/μL), mean (SD)	4.1 (1.1)	4.1 (1.1)	0.9
Lymphocytes (103/μL), mean (SD)	2.4 (0.7)	2.3 (0.7)	0.09
Monocytes (103/μL), mean (SD)	0.5 (0.1)	0.5 (0.1)	0.2
Eosinophils, (103/μL), median (IQR)	0.2 (0.1)	0.2 (0.1)	0.7 b
Red blood cells (103/μL), median (IQR)	4.9 (0.6)	4.9 (0.5)	0.9 b
Hemoglobin (g/dL), mean (SD)	14.2 (1.0)	14.2 (1.1)	0.8
Hematocrit (%), median (IQR)	41.7 (3.1)	42.0 (3.2)	0.4
Mean corpuscular volume (fL), median (IQR)	85.5 (4.9)	85.4 (4.6)	0.1 b
Mean corpuscular hemoglobin (pg), median (IQR)	29.2 (1.9)	29.2 (1.9)	0.9 b
Mean corpuscular hemoglobin concentration (g/dL), mean (SD)	34.4 (0.9)	34.4 (0.9)	0.4
Red cell Distribution Width (%), median (IQR)	13.4 (0.9)	13.5 (0.8)	0.5 b
Platelets (103/μL), mean (SD)	241.0 (48.0)	240.0 (47.2)	0.5

^a p value was obtained by a paired t-test; ^b Wilcoxon Signed Rank test. IQR: interquartile range; SD: standard deviation. dL: deciliter; g: gram; μL: microliter.

). Borderline significances were in the mean level of white blood cells and lymphocytes between the car (7.3 × 10³/μL, SD = 1.4; 2.4 × 10³/μL, SD = 0.7, respectively) and drone transportation (7.3 × 10³/μL, SD = 1.4; 2.3 × 10³/μL, SD = 0.7, respectively), p = 0.08, and p = 0.09, respectively (Table 2).

3.3. Coagulation Assay

All samples showed normal coagulation results. The mean PTT levels were very similar between transportation methods (27.2 s, SD = 2.3 and 27.2 s, SD = 2.4 for car and drone, respectively), p = 0.7. However, a small variation (0.03 s in the mean levels of PT between transportation methods (10.27 s, SD = 0.59 for a car; 10.30 s, SD = 0.58 for a drone) was significant (p = 0.04) (Figure 1).

A sensitivity analysis to describe an agreement between the two different modes of transport (drones and cars) is presented in Figures S1 and S2 by the absolute differences in the results obtained between samples that were transported by drones versus cars. The green lines draw the 95% limits of agreement. The red line shows the mean difference for the two different types of transport (drones and cars). As shown in Figure S1, PT (second), creatinine (mg/dL), BUN (mg/dL), total bilirubin (mg/dL), ALP (U/L), ALT (U/L), AST (U/L), GGT (U/L), LDH (U/L), total cholesterol (mg/dL), and HDL (mg/dL) fell within the 95% limits of agreement. This indicates no clinically relevant difference for these parameters. However, glucose (mg/dL), had 36 out of 40 (90.0%) samples falling within the 95% limits of agreement, and LDH (U/L), had 37 out of 40 (92.5%) samples falling within the 95% limits of agreement. Thus, glucose and LDH did not meet the 95% acceptance criterion and showed a potential clinical effect (Supplementary Figure S1). Figure S2 shows that white blood cells, neutrophils, lymphocytes, monocytes, eosinophils, red blood cells hemoglobin, hematocrit, mean corpuscular volume, mean corpuscular hemoglobin, red cell distribution width, and platelets fell within the 95% limits of agreement and indicated no clinically relevant differences for these parameters (Supplementary Figure S2).

3.4. Urine Tests

There were no significant differences between transportation modes in the leukocyte (p = 0.2) and erythrocyte (p = 0.3) counts; however, a mismatch in the leukocyte count was observed in two samples that were negative by car and low by drone transport. A mismatch in the erythrocyte count was observed in one sample that was high by vehicle and medium by drone transportation (

Table 3. Differences in selected urinalysis parameters and urine culture tests by transportation type (drones vs. cars).

(A). Semi-quantitative leucocytes and erythrocyte’s count		Car n = 40					p value a
Drone n = 40	Urine leucocytes	Negative	Low	Medium	High	Total	0.2
	Negative	24	0	0	0	24
	Low	2	4	0	0	6
	Medium	0	0	4	0	4
	High	0	0	0	6	6
	Total	26	4	4	6	40
	Urine erythrocytes	Negative	Low	Medium	High	Total	0.3
	Negative	26	0	0	0	26
	Low	0	9	0	0	9
	Medium	0	0	3	1	4
	High	0	0	0	1	1
	Total	26	9	3	2	40
(B). Urinalysis parameters and urine culture test		Car n = 40					p value b
Drone n = 40	Urine ketones	Negative		Positive	Total		1.0
	Negative	38 (97.4%)		0 (0.0%)	38
	Positive	1 (2.6%)		1 (100.0%)	2
	Total	39 (100.0%)		1 (100.0%)	40
	Urine glucose						1.0
	Negative	38 (100.0%)		0 (0.0%)	38
	Positive	0 (0.0%)		2 (100.0%)	2
	Total	38 (100.0%)		2 (100.0%)	40
	Urine nitrites						1.0
	Negative	40 (100.0%)		0 (0.0%)	40
	Positive	0 (0.0%)		0 (100.0%)	0
	Total	40 (100.0%)		0 (100.0%)	40
	Urine protein						1.0
	Negative	38 (97.4%)		0 (0.0%)	38
	Low	1 (2.6%)		1 (100.0%)	2
	Total	39 (100.0%)		1 (100.0%)	40
	Urine culture test						1.0
	Negative	31 (100.0%)		0 (0.0%)	31
	Positive	0 (0.0%)		9 (100.0%)	9
	Total	31 (100.0%)		9 (100.0%)	40
(C). Urine PH		Car n = 40			Drone n = 40		p Value c
Urine pH, median (IQR)		6.0 (2.0)			6.0 (2.0)		0.3

^a p value was obtained by the Friedman test; ^b McNemar test; ^c Paired t-test. IQR: interquartile range; SD: standard deviation.

A). There was full agreement between the two types of transportation (car vs. drone) for urine nitrites, glucose, and urine cultures. The assessment of colonies in the positive cultures (cfu/mL) was identical between the cars and drones. For other parameters (protein and ketones), there were minor mismatches (p = 1.0) (Table 3B). For urine PH, the mean level in aerial samples was 6.0 (SD = 2.0) compared to 6.0 (SD = 2.0) in samples transported by car (p = 0.3) (Table 3C).

4. Discussion

Our study explored the potential use of drones as a tool for transporting clinical laboratory specimens compared to car transport in Mediterranean weather conditions mainly during calm weather between July and November in Israel. An interesting finding, that UAS transport is an appropriate method that maintains the quality of most routine clinical laboratory test results, similar to the routine method by vehicle, was consistent with Perlee et al.’s study indicating that most of the blood parameters were not affected by drone transport. However, glucose, phosphate, potassium, chloride, hemoglobin, platelet count, activated partial thromboplastin time, and LDH showed significant differences between transport by car and by drone [21]. We found a significant difference between the transportation modes for GGT and PT despite the very similar mean results of these tests. However, in Bland–Altman plots, GGT and PT samples fell within the 95% limits of agreement and were indicated as not clinically relevant. This finding was similar to that reported by Weekx et al., who showed that in 3 of the 30 biochemistry and hematology parameters, a statistically significant difference was found in GGT, mean corpuscular hemoglobin, and platelet count without a clinically relevant effect for these parameters [22].

We found no significant differences between transportation modes in the following hematology parameters: levels of neutrophils, monocytes, eosinophils, red blood cells, and hemoglobin; hematocrit; mean corpuscular volume; mean corpuscular hemoglobin; mean corpuscular hemoglobin concentration; red cell Distribution Width; and platelet count. However, there was borderline significance for the mean level of white blood cells and lymphocytes between the car and drone transportation without a clinically relevant effect as indicated in Bland–Altman plots, in line with previous studies [20,21,22,23].

We found that glucose and LDH did not meet the 95% acceptance criterion and showed a potential clinical effect, in agreement with Perlee et al.’s report that showed a clinically relevant effect only in LDH and glucose values [21]; however, Weekx et al. showed a clinically relevant effect in potassium and LDH values [22]. A study from Japan demonstrated that the LDH values were not significantly different between the flight and non-flight samples [23].

A surprising finding is that transportation by the UAS did not shorten the mean transport time, despite the ability of the drones to bypass traffic. This finding is inconsistent with other studies [12,26,27]. In our study, the short distance and traveling during hours when traffic is relatively calm led to the absence of a difference in times; however, on one occasion, the drone was returned to the base owing to low battery power and resent after adjustments. These situations and others prolong delivery time and might challenge trust in drones, negatively influencing their implementation in healthcare, particularly in emergencies. We anticipate that the transmission time of drones will become significantly shorter when this technology improves and advances. In our study, the weight and the distance were limited to 3 kg and 30 km, respectively. Improvements increasing the maximum weight and volume capacity and extending flight distances are needed to make UASs practical for carrying clinical laboratory specimens.

We found that the mean package temperature was significantly higher in drone transportation than in a car, which is inconsistent with the results of Amukele et al. [19], who reported lower temperatures in flow samples. We also found that the temperature difference did not affect glucose and potassium levels, which are very sensitive to pre-analytical conditions, in agreement with a study by Amukele et al. [18]. We believe that the effects of higher temperatures should be investigated further in future studies.

Our study has some limitations. The small number of participants requires us to confirm the results in future studies with larger sample sizes. We did not address the cost-effectiveness of flights funded by the Israeli Innovation Authority. Nevertheless, investigating the economic implications of UAS transportation is important, and should be examined in future studies. Our study investigated the effect of transportation modes on samples from healthy volunteers; however, further investigations are needed to examine the potential effects on pathological samples from patients. The small distance which was limited according to CAAI approval might have caused a bias in the estimation of the transport times. Thus, greater distances for UAS transportation should be examined in future studies.

5. Conclusions

Transportations of laboratory specimens by car and drone are equivalent. UAS transport is an appropriate method for maintaining the quality of most routine clinical laboratory specimens. For the 34 biochemistry, hematology, and coagulation assay parameters, only glucose and LDH did not meet the 95% acceptance criterion and showed a potential clinical effect. Technological, logistical, and regulatory obstacles must be overcome before implementation.